Training for employment or skilling up from employment? Jobs and skills acquisition in the Tiruppur textile region, India

This paper explores how skills for garment work are acquired in the rural hinterland of Tiruppur, one of India’s largest garment manufacturing clusters. Drawing on a quantitative survey and qualitative interviews with garment workers in Tiruppur’s hinterland, we document the infor- mal pathways of skill acquisition for garment work and advocate a demand-driven approach to vocational training. Such an approach, first, unsettles linear policy assumptions about direct linkages between training, skills acquisition and access to decent and rewarding employment. We show how rather than being formally trained for employment, vil- lagers gained skills from employment and upskilled themselves on the job. Such upskilling took the form of self-directed learning rather than formal training, and involved spatial and job mobility between companies and sectors. Second, a demand-driven perspective reveals how access to more advanced skills and more desirable jobs is shaped by the structural inequalities of gender, age and caste, which curtail the opportunities of women and the elderly in particular. Finally, policy and research would benefit from a demand-driven approach to training and recruitment that prioritises the skill development needs of local popu- lations and that supports those whose participation in training and labour markets remains constrained by gender, age or caste.


Introduction
Over the last two decades skills development has received increased policy attention and investment in India. Skills training courses have been launched to address rural poverty, improve human capital and provide a skilled workforce to industry and services. Underlying such upskilling initiatives, policymakers and training providers assume a linear pathway from skills training to employment and poverty alleviation. This paper builds on the concept of the 'social life of skills' to explore how people in the Tiruppur textile region of Tamil Nadu access garment jobs, obtain relevant skills, and move up within the sector (Carswell and De Neve 2018). By revealing the diverse ways in which people acquire skills and the different pathways of employment and non-employment that follow, we seek to unsettle assumptions embedded in skills development policy about direct linkages between skills training, skills acquisition and access to better, more rewarding and more dignified employment (McGrath 2012). Drawing on a quantitative survey conducted in a village in the hinterland of Tiruppur in Tamil Nadu as well as on a series of interviews with current and past workers, we document the formal and informal ways in which skills are acquired and jobs obtained by this ruralbased labour force. The paper contributes to debates about skills acquisition in India's garment sector by demonstrating how established pathways of learning shape entry and mobility in the industry and how formal skills training initiatives fall short of meeting the requirements of both industry and workers.

Vocational education and training in India
Vocational education and training (VET) policies have long been cast within a productivist paradigm in which vocational training is aimed primarily at enhancing productivity and economic growth, increasing workers' employability and getting young people into jobs (Anderson 2009;McGrath 2012). As Saraf points out, much of the rationale for strengthening vocational training in India's New Education Policy of 2015 was to provide the informal economy with skilled labour so as to enhance its productivity (2016,16). Authors have emphasised the exclusionary effects of such a singular focus on employability as it not only reduces the very aim of education to employability but also enhances inequality by relegating youngsters from already disadvantaged classes and lower castes to vocational training for informal work (Sadgopal 2016;Saraf 2016). As Sadgopal rightly argues, 'this "extremist" focus on employability hardly makes sense against the background of the "jobless economic growth" model pursued since globalisation ' (2016, 35). Saraf similarly argues that most skills training is for work that is low in the employment hierarchy and will ensure trainees 'remain employed in informal blue-collar work where working conditions are severe, unstable and unequal ' (2016, 17). What, indeed, is the point of inculcating India's poorer and less educated classes with low-wage earning skills for labour markets tainted by informality and jobless growth?
Researching India's recent skills development policy, Nambiar similarly observed that skill development was considered 'a magic bullet, which would simultaneously generate employment, reduce poverty, educate and empower youth, and promote rural development ' (2013, 62). Empirical research, however, reveals a different reality. Indeed, vocational training under India's recent skills development drive appears more geared towards providing employers with a steady labour supply than enhancing quality of employment, working standards or wage levels (Nambiar 2013;Ruthven 2016Ruthven , 2018. Under the dominant neoliberal paradigm, serving the demands of industry took the upper hand, and these demands are shaped more by a search for flexible, compliable and cheap labour than by a search for specific skills (Ruthven 2018, 331). Nambiar too emphasised that skills training programmes for garment work are more preoccupied with the mobilisation and recruitment of new labour forces to the industry than with enhancing skill levels, improving the quality of employment or fulfilling youth aspirations for a better life (Nambiar 2013(Nambiar , 2021. In this paper, we argue that a focus on the recruitment of rural youths from distant places to supply Tiruppur's garment factories has trumped the training of local populations who tend to (up)skill themselves outside of any formal training framework.
While vocational and skills training has typically been supply-driven and top-down, Ramasamy and Pilz highlight that a demand-driven approach in VET is gaining momentum (2020,257). however, as they point out, most research has focussed on macro-level needs of industry and on employers' demand for skills (McGrath 2012), rather than the needs of millions of poorer and less educated workers employed in the informal economy. What is needed, therefore, is a demand-driven approach to VET 'from the perspective of individuals, learners and society by keeping them at the centre of the vocational skill development process' (Ramasamy and Pilz 2020, 258). A demand-driven approach prioritises the perspectives of learners themselves, including their existing capabilities, preferences and long-term aspirations (Ramasamy and Pilz 2020, 259). Such an approach pays attention to local skills requirements and constraints, existing pathways of skills acquisition, and the social and cultural contexts that shape skill acquisition and employment trajectories. A demand-driven focus, then, not only provides insight into the training needs of specific labouring populations, but also offers a means to make training more attuned to the needs of target populations and more aligned with existing patterns of skills acquisition and learning (Ramasamy 2016, 31-39).
In this paper, we build on this research and take a learners' perspective to explore the ways in which garment skills are acquired by villagers in Tiruppur's hinterland. We show that only a tiny minority of garment workers received any formal training, but that this did not prevent them from becoming highly skilled workers and moving up within the industry. Rather than formal training leading to skills acquisition and subsequent employment, we show how the vast majority of garment workers gain largely unskilled employment firstprimarily as helpers to tailors -and then learn more advanced skills from friends, co-workers and contractors on the shop floor. King (2012), in a critical study of India's skills policy agenda, similarly observed that across South Asia there is a widespread preference by employers for training workers on the job. A 'shadow training system' -referring to the unorganised and often largely unacknowledged forms of training within India's informal economy -is what provides industry with the skills it needs (King 2012, 668). yet we still lack detailed analyses of how such training takes place despite it being the dominant mode of skill development in India's informal economy. Moreover, such informal processes of skilling 'would better be termed skill acquisition …, as there may well be no explicit commitment to train, but just a process of learning on the job' (King 2012, 672). This observation is echoed by Chea and huijsmans (2018), who emphasised the role of highly informal apprenticeships as a route towards acquiring skills in Cambodia and Laos, which involved self-directed learning rather than explicit training (see also Regel and Pilz 2019). Our research similarly analyses the self-directed ways of learning that underpin skill acquisition processes in rural Tamil Nadu. Central to both skill and job acquisition processes are informal ways of learning that are located within social networks and relations of social reproduction. This is what we have termed 'the social life of skills': that is, the social processes, relationships, and ideologies that enable (or constrain) people's access to skills, and subsequently to employment, wages, satisfaction and dignity (Carswell and De Neve 2018). We argue for an approach to skills that shifts the focus from formal training programmes to the social life of skills in society, where the acquisition of skills is mediated by social institutions, such as gender, caste and age, which enable or constrain both skill development and access to employment (see also McGrath et al. 2020, 7-9). We suggest that such a demand-driven approach to skill, which acknowledges the informality of both skill acquisition and employment, can offer new opportunities for designing formal training initiatives that meet the gendered needs of youth from marginalised backgrounds more effectively.

Skilling India's garment workforce
Since the early 2000s, the government of India has launched a series of policies and programmes to promote skill development aimed at the country's less educated, economically deprived rural youth. The launch of the National Skill Development Policy (NSDP) in 2009 formed a landmark moment that started more than a decade of subsequent investment in skills training. This commitment was boosted by a new National Policy on Skill Development and Entrepreneurship (NPSDE) introduced in 2015. Skills training initiatives were started up across sectors from textiles to manufacturing and from services to agriculture (Brown 2020;Gooptu 2018;Nambiar 2013;Ruthven 2018). The discursive goals were clearly articulated in the 2009 policy: skills training would lead not only to enhanced productivity, but also to more inclusive growth, more regular forms of employment, and more 'decent work' (NSDP 2009, 1;NPSDE 2015, 2). The overall aim was to supply industry and services with the skilled labour force required to stimulate national economic growth, while simultaneously enhancing livelihoods and lifting people out of poverty.
These policies shifted attention towards shorter skills training courses, spanning anything from a few days to a few months in length, with industry placements on completion. Funds were channelled through public-private initiatives, which formed another cornerstone of the policy, with local training institutions responsible for recruiting trainees, delivering training and arranging placements (Mehrotra 2018). In the garment sector, the focus of this paper, skills training initiatives were promoted through different ministries and different schemes over time. The Ministry of Textiles initially introduced the Integrated Skill Development Scheme (ISDS), which channelled funds through local training providers across the country.
In Tiruppur, the Skills Training Institute (STI), established in the late 1990s, is one such training provider. 1 In addition to offering more established vocational degree courses, they ran a training centre that offers an array of short-term garment skills training courses in tailoring, pattern making, embroidery, etc. Funded through ISDS until 2017, they trained 15,000 students, of whom 12,500 were subsequently placed in garment factories. While some of these were trained in Tiruppur itself, most received their training in one of the many local centres with which STI collaborates across Tamil Nadu. The main aim of such training, as stated by the training centre, was to provide the Tiruppur garment industry with the skilled labour force it needs to continue to grow. From 2018, ISDS was replaced with Samarth (Capacity), a new government scheme for capacity building in the textile sector, again under the aegis of the Ministry of Textiles. In addition, STI also provided training funded under the DDU-GKy (Deen Dayal Upadhyaya Grameen Kaushalya Yojana) scheme, an initiative of the Ministry of Rural Development aimed at skilling up rural youth of poor and marginalised backgrounds. Under the banner 'Skill Up Rural India, Lead the World' , DDU-GKy sought to reduce poverty and enhance livelihoods by providing the rural poor with skills and job placements for sustainable employment. 2 STI recruited rural men and women from across Tamil Nadu, offered them 3-4 months of free training, and then placed them in a garment company. Placing at least 75% of trainees in industry jobs was a key target of the policy and reflects the emphasis on labour recruitment to the industry. Driven by the need to place trainees in jobs, skills providers effectively acted as recruitment agencies supplying migrant labour to the industry rather than (up)skilling local labour whose long-term livelihoods are more likely to depend on garment work (Nambiar 2021;Ruthven 2016).
Within this web of skills training provisions, it appears that at STI the majority of recruits had completed 10 or 12 years of education, with some having completed further education. Most were recruited from rural areas outside the Tiruppur region, primarily from poorer and less industrial districts of Tamil Nadu, which suggests that here also skills training was at least in part about recruiting rural labour to the industry. In our study village, called Allapuram and located in the hinterland of Tiruppur, 3 very few men and women had received any formal training at all, whether in garments or other industrial work. This raises the questions as to why so few people in Tiruppur's immediate hinterland accessed formal garment training and how they obtained the skills required for such work. To answer these questions, we present data based on interviews and a worker survey conducted in Allapuram in 2020 that included all types of work, but with a focus on garment jobs. We describe the different pathways of skill acquisition among our rural informants and consider the gendered, castebased and age-specific patterns of inclusion and exclusion from employment. We argue that rather than skills leading to jobs, the reality of skills acquisition within the garment sector is exactly the other way around: jobs enable the acquisition of skills. Men and women in Tiruppur's hinterland obtain garment jobs first and in the course of employment acquire the skills needed to become more skilled workers and to attain higher levels of competence, speed and experience. Such on-the-job learning is seen by workers themselves as the most effective way of obtaining the skills required to access more desirable and rewarding garment jobs. This type of learning also explains why so few villagers in Tiruppur's direct hinterland come forward to enrol themselves in formal training courses -they consider these inadequate to help them acquire the more advanced skills needed to access better job opportunities and fulfil their aspirations for upward mobility within the industry. This fits with the recruitment priorities of training initiatives too. Motivated to attract new labour supply to the industry, training centres prefer to recruit rural men and women from farther afield over local labour forces, whom they know train themselves on the job anyway.
We start by briefly describing our methodology and research context, and we then present the occupational profile of Allapuram, the skills required for different jobs, and the pathways through which different workers acquired the skills they need. We then reflect on the gender-, age-and caste-based nature of skill acquisition and employment, and the interconnectedness of productive and reproductive spheres in the skilling of rural labour.

Methodology and context
The evidence presented in this paper draws on long-term fieldwork conducted in Allapuram, a village located about 20 km to the south-east of Tiruppur that is dominated by the textile industry. The authors have conducted fieldwork in this village and a second village in the region since 2008, with a focus on livelihood strategies around textile employment De Neve 2013, 2014). The material presented here is primarily based on research carried out between July 2019 and March 2020. We conducted a worker survey and a household survey and applied a range of qualitative research methods.
Firstly, the worker survey was undertaken among a sample of 165 informants, who were either currently doing paid work or had been in paid employment in the past. Whilst this included work across different sectors, the focus of this paper is on garment employment.
The worker survey contained questions about skills, work experience, job arrangements and social networks with reference to employment trajectories. In particular, we focussed on patterns of skill acquisition and the ways in which villagers acquired skills, engaged with training and accessed different labour markets. The worker survey was purposively sampled to capture the village's mix of employment sectors as well as the gender, caste and age profile of the workforce as identified in the household survey. This sampling was based on our longer-term knowledge of the socio-economic profile of the villagers and the different labour markets people were involved in. Current workers provided details on their current and previous employment, while those who had worked in the past provided details on their last and second-last jobs. Thus, having conducted a worker survey with 165 informants, we obtained data on 241 jobs. Secondly, we undertook a household survey which was a repeat of a survey conducted in 2008-2009, collecting data on occupations, incomes, migration patterns, caste, education and asset ownership. In 2008-2009, all 240 households living in the village were surveyed, whilst in 2019-2020, all 224 households were surveyed. Finally, our qualitative research included in-depth interviews with over 50 informants, collecting life histories with a wide range of villagers to contextualise the survey data and cross-check our quantitative findings. In earlier research we also undertook participant observation in garment factories and at family functions and village celebrations, which provided us with contextual knowledge of local livelihoods.
In Allapuram, the landowning Gounders (Backward Caste) make up the dominant caste in terms of numbers, economic wealth and political power. While many of them set up garment units and textile businesses in Tiruppur from the 1990s, Gounders continue to be the main landowners and agriculture is still reasonably viable in Allapuram thanks to canal irrigation. Members of intermediate castes, such as Nadar, Boyyar, Vannan (dhobi) and Navithar (barber), are involved in a range of jobs, but also primarily in the garment industry. Among the Dalits (Scheduled Caste), there are about equal numbers of Adi Dravida Christians and Matharis (also known as Arunthathiyars), and historically both groups used to be primarily involved in agricultural coolie work. 4 Some still are -especially older villagers and women -but a significant number of them now commute to Tiruppur's garment factories, where both men and women work as helpers, tailors, ironing masters, cutting masters, and so on. Ever more Adi Dravidas and Matharis now also find employment in local garment workshops set up more recently by Gounders and members of intermediate castes in Allapuram itself.
Today, the residents of Allapuram -especially young men and, to some extent, young women -have some choice about how to earn a living. Paid employment in Allapuram centres around garments, agriculture, services and construction. In our 2020 worker survey, these jobs made up 44.8%, 25.7%, 14.10% and 8.7% of all reported jobs, respectively, clearly indicating the village's high reliance on garment work. The remaining 6.8% was made up of a variety of jobs, including mechanics, government employment, self-employment and other textile work. Among Adi Dravida Christians in particular, garment work constituted the primary or secondary source of income for 62% of households in 2019-2020, while it was a primary or secondary source of income for 45% of households among Matharis. Of all garment jobs, a minority of 9.3% was made up of homeworkers, while the vast majority (90.7%) were jobs in garment factories, either factories in Tiruppur or the smaller units set up more recently in the village itself. Exactly equal numbers of men and women appeared to be employed in factories, but 70% of the homeworkers were women. While those working in factories are engaged in a range of jobs -tailors, cutting masters, ironers, trimmers, checkers and helpers -those working at home are exclusively doing trimming and checking work. Although our sample of homeworkers is small (8), we were able to confirm its validity through cross-checking with our household survey, observations within the village and interviews with homeworkers.
There was a distinct pattern of age distribution of workers across different jobs too (Table 1). While we found garment workers of almost all ages, the vast majority tended to be much younger than any other group of workers. In our 2020 sample, 87% of garment workers were 45 or under, and all but three workers were under the age of 55. In contrast, the age profile of agricultural workers had the opposite pattern, with 90% of current agricultural labourers aged over 45. younger villagers are thus far more likely to be employed in garment and service sector work, while older villagers are more likely to be involved in agricultural labour and to some extent construction (50% of construction workers were over the age of 45).

What skills do garment jobs require?
Before exploring how skills are acquired and jobs learned, we briefly consider what skills workers reported their jobs required. We asked respondents 'what is needed for your current job' and then let them rank the following options: experience, specific skills or knowledge, good education, good networks, personal reputation, nothing needed, and other (specify). We deliberately wanted to broaden the scope beyond a narrow set of technical skills and allow respondents to consider other factors such as experience, reputation and social networks. Out of 165 workers that we surveyed, 126 were currently in paid work, and of these we obtained valid answers about required skills from 123 respondents (Table 2).
A number of observations are worth making. Forty-six percent of our respondents ranked 'experience' as the main thing required in their job and 36% ranked 'specific skills or knowledge' the highest. Only very few ranked good education, personal reputation or good networks highest, but several respondents did rank these second, third or fourth. Thirteen percent of our respondents said that nothing is needed to do their work. 5 Amongst garment workers, of the 67 respondents working in factories, 48% (32) ranked 'experience' highest and 43% (29) ranked 'specific skills or knowledge' highest. Only 6% (four) put 'nothing is needed' at the top and only 3% (two) ranked 'good education' highest. Eighteen of the 32 who mentioned experience first also ranked specific skills or knowledge in second, third or fourth place, and 21 of the 29 who mentioned specific skills or knowledge first also ranked experience farther down their list. As such, the vast majority of factory garment workers (91%) mentioned 'experience' and/or 'specific skills or knowledge' -in one order or the other -as key to their job. This matches interviews with garment workers in which both qualities are routinely discussed. Garment workers talk openly about the specific skills involved in tailoring and discuss the different types of machines, stitches and patterns they master. Some tailors can only operate flatlock machines, while others proudly state that they can also work on overlock and singer machines. Apart from a range of practical and technical skills, tailors also build up extensive knowledge of fabrics, patterns, sizes, accessories, dyes, etc. as well as of the wider production process. Discussions of specific skills and forms of knowledge are rarely separated from reflections on experience, and old-hand garment workers are eager to mention the number of years they worked in the sector. Tailors explain how their skills range expanded with time and the number of companies they worked in. Skill and experience are thus seen to go hand in hand, and are considered equally essential to the job. Of the six male informants who indicated that nothing is needed for their job, four were older teens who had only recently started as helpers and who considered their job unskilled, especially compared with the skilled tailors they assisted. They too therefore seemed to associate skill with experience and time of service, while seeing themselves as still at the bottom of the ladder. While the responses of homeworkers were similar to factory workers, they did tend to emphasise experience over skills or knowledge. This again reflects that the checking and trimming work they do is widely seen as unskilled and not requiring any form of instruction. It also points to the highly gendered nature of this work, with women having internalised the perception that homework is largely unskilled (see also Rajendra 2022).
like? In order to gain a better insight into the diverse pathways through which skills are acquired, our survey first asked how informants acquired the skills they needed in their employment (formally or informally), and subsequently we asked those who reported informal learning what that consisted of.

Formal training
A first finding was that very few informants had received formal training in any sector. Across all 241 jobs surveyed, just 10 (4%) had received any formal training. These included three professional occupations, four service sector jobs, two tailors and one mechanic. Focussing on the garment sector, out of 108 respondents currently or previously involved in garment work, only two tailors reported having received formal training. They were both women who had previously completed post-secondary education. One of them was a 28-year-old Boyyar (MBC) woman who had completed a diploma in teacher education and later took a short (10-day) tailoring course in a private training centre. The other was a 31-year-old Adi Dravida woman with a BA in English who had subsequently completed a three-month tailoring course. She had only worked for three months as a tailor before stopping paid work altogether when she got pregnant in 2016. Based on our survey and in-depth knowledge of the village, hardly any people had received formal skills training for garment work.

Informal pathways of skill acquisition
With a tiny proportion of garment workers (2%) having received any formal training, the vast majority of garment factory workers had learned the required skills on the job itself.
Respondents offered a range of answers that we clustered into five main pathways of learning, based on the main process mentioned by each informant: as helper/self-taught, learnt from co-workers, guided by an employer or supervisor, learnt from a relative, and learnt from neighbours and friends (Table 3). While we identified five main pathways across all the jobs, these were not mutually exclusive and several overlap quite substantially. For example, some respondents who mentioned being instructed by a shop-floor supervisor also referred to the guidance of co-workers. And, some of those who mentioned being taught by a relative also listed a friend or neighbour. The pathway most likely to have overlap with other pathways is the one we call 'as helper/ self-taught' . When mentioning 'as helper' , workers were effectively indicating that they started in the role of helper to a main job, for example tailoring, and then learned that job by observing the work and giving it a try themselves. Some of them described this as being 'self-taught' , and hence we consider 'as helper/self-taught' as one pathway. however, many of those who mentioned this pathway probably received guidance from others too, especially from co-workers on the shopfloor, even though they may not have stated this explicitly. There is therefore likely to be a considerable degree of overlap between some of the pathways, such as between the 'as helper/self-taught' and 'co-workers' pathways.
Within the garment industry, skill acquisition is almost entirely informal and the most commonly reported pathway was 'as helper/self-taught' , reported by 55% of those employed in factories. We explore this single most important skill acquisition pathway in greater depth below. The second most important pathway -learning from 'co-workers' -was mentioned by 24% of workers in such jobs and usually referred to the tailors who they assisted as helpers.
Just 12% of workers said they were instructed by an employer or supervisor. In the smaller units, this is likely to be the employer or manager, while in larger factories this tends to be the shop floor supervisor or the contractor, who usually recruits and manages the workers on his team. Observation in factories both in the village and in Tiruppur revealed that labour contractors in particular play a key role in upskilling workers, especially those team members who they identify as hard-working, committed and dependable (Carswell and De Neve 2014;De Neve 2014). Some of this upskilling occurs between different job categories, such as from helper to tailor or from tailor to cutting master, but other upskilling is within tailoring itself and involves improving one's speed or working with new machines, stitches, fabrics, patterns and so on.
In only a very small number of cases did workers mention being guided by a neighbour/ friend (2%) or a relative (4%). This is indicative of the fact that, while workers might initially be introduced to a company or a contractor through a friend or relative, they tend to quickly move between factories and follow different contractors. hence, they end up being employed in very mixed workplaces -in terms of gender, caste or place of origin -where they no longer work alongside old friends or neighbours. This is also indicative of the informal, temporary and irregular nature of employment in this industry. Few workers remain regularly employed in the same company. As such, they primarily learn on their own initiative and/or with the help of co-workers employed alongside them. Friends and relatives thus play a relatively minor role in informal training in factory-based garment work in Tiruppur.
In terms of home-based work, the lion's share of such work is taken up by women (70%) in Allapuram, but it involves some older men too (30%). While our sample of home-based workers is small, we were able to confirm the related findings through extensive observations in the village and interviews with homeworkers. home-based work consists of checking and trimming garments produced in local workshops and has been growing in importance over the last 10 years in line with the mushrooming of village-based garment units. While also here the 'as helper/self-taught' (30%) and the 'co-worker' (20%) pathways were significant, 50% of women mentioned that they learned from either 'relatives' (30%) or 'neighbours/friends' (20%) -very different from factory-based work. None learned from an employer or contractor. This reflects the fact that women work either on their own at home or together in small neighbourhood-based groups on the porch of someone's house. Typically, a daughter-in-law learns from her mother-in-law, while neighbours, often belonging to the same or similar castes, instruct each other as the network of subcontracting spreads from home to home. We strongly expect that the 'as helper/self-taught' and 'co-worker' pathways included a good number of neighbours/friends and relatives too, and so the latter clearly constitute the two dominant pathways of learning among homeworkers. But the high levels of learning through such localised relations is also revealing of how local social networks of caste and neighbourhood reproduce patterns of inclusion and exclusion in home-based work. Even though social networks were hardly mentioned in the survey as something needed for the job, it became clear from interviews with homeworkers that access to home-based garment work and to learning the skills it requires was heavily dependent on membership in localised social networks. Seven out of the 10 homeworkers in our survey belonged to BC castes, one belonged to an MBC caste and two were Christian Adi Dravidas (SC). No Matharis (SC) were involved in home-based garment work. This pattern is in line with the ownership of garment units in the village, which are almost entirely in the hands of members of BC and MBC castes. Apart from the Mathari ex-president of the village, who had been running a garment unit for a few years, only one other Mathari family started a small garment unit in early 2020, largely run with family labour and a couple of workers recruited from the Dalit colony. Similarly, one Christian Adi Dravida family had run a small stitching unit for a couple of years, but it had closed down by the time we interviewed them in late 2019. Apart from them, to date neither the Adi Dravidas nor the Matharis have been able to set up their own units and subcontract work to homeworkers. As a result, home-based work is put out by the higher-caste owners of garment units to their own relatives and caste members before including others in their network. Preferring to give socially acceptable home-based work to the women of their own caste, they keep Dalit homeworkers as a reserve labour force to be temporarily recruited during peaks in the production process. This goes some way to explain the recruitment of two Adi Dravida homeworkers by a local Gounder garment workshop owner. A clear pattern of job hoarding among the higher and middle-ranking castes thus emerges around home-based garment jobs that offer an attractive and socially acceptable form of employment to the women of these castes in the village.

'As helper/self-taught' pathway: gendering skill, mobility and work
Given that the 'as helper/self-taught' pathway of skill acquisition was so predominant in our survey, we further unpack this pathway with some observations from the garment sector. We highlight how particular intersections of productive and reproductive labour make this a highly gendered skills acquisition pathway, with radically different labour market outcomes for men and women. First, male garment workers typically explained that they started to work as a helper to a tailor and then learned tailoring skills on the job itself -padi padi or step by step. They then gradually work their way up to tailor or cutting master, and some even become contractors over time (Carswell and De Neve 2014). This pathway of learning reveals a significant feature of the learning process: that employment is obtained first and skills are acquired afterwards. Second, when we asked tailors to describe what being selftaught or learning 'as helper' meant, many emphasised the importance of personal interest, initiative-taking and drive to learn new skills. Self-reliance and initiative were similarly identified as key factors in apprenticeships in tailoring shops elsewhere in Tamil Nadu (Regel and Pilz 2019). These personal qualities were also mentioned when garment workers were asked how long it takes to acquire tailoring skills. Respondents struggled to answer this question or give a clear timeline. Rather, they would explain that the time to become a tailor entirely depends on 'personal interest' and 'motivation' . They would typically illustrate this with an example of one colleague who learned it within one month, while others remained helpers after six months or even a year. Tailors emphasised that those with an interest in learning will take initiative to observe co-workers, give it a try themselves and ask advice from those working around them. The point here is that a great deal of skills acquisition clearly occurs through self-directed learning, rather than informal training (Chea and huijsmans 2018).
Men and women in their late teens almost always start working in a 'domestic' company -that is, a factory that produces for the Indian market. There they observe the tailors they assist as helpers and start stitching during break times or after working hours. Learning in such companies has a number of advantages. Much less stringent quality inspections take place and tailors can get away with minor mistakes during their first weeks or months at work. Standard patterns, usually for underwear, night wear or T-shirts, allow tailors to familiarise themselves with the most common stitches, materials and machines. Finally, producing for domestic markets, these companies tend to have regular orders for the same patterns over time. This gives novice tailors the opportunity to practise and refine the same skills over a longer period without having to adjust to constantly changing materials and designs.
With most garment workers being casually employed, one feature of such employment is that internal promotion within a company to a more skilled and better paid job is very rare. When young helpers want to become tailors, they tend to move to another company once they have the basic sewing skills under their belt. There, they will be recruited as tailors and be able to further refine their tailoring skills and -crucially -enhance their sewing speed. While some keep shifting between domestic companies, most are keen to move on to better-paid export companies as soon as they feel confident in their abilities. In export companies, they usually need to enhance their sewing speed, improve the quality and consistency of their tailoring work, and diversify the range of skills they master so as to be able to meet the flexible and high-quality demands placed on them. In such firms, designs are more complex, products may change on a weekly basis, and quality inspections can be ruthless.
Personal interest in learning and taking initiative are clearly central to garment workers' skill development trajectories, their access to employment and, particularly, their ability to obtain more skilled and better paid jobs in export firms. And yet job progression -such as from helper to skilled tailor -is rarely straightforward or linear. It is not unusual for helpers to try and start work as tailors only to realise that they have not yet fully mastered the necessary skills and are out of their depth or unable to keep up the speed. Some will return to work as a helper and try to improve their skills before having another go at tailoring. Others end up realising sewing does not suit them and choose a different line of work, such as cutting, ironing or packing, or leave the industry altogether. Most workers recognise that while their own initiative, drive and focus matter greatly, a degree of luck is essential too. Some may be lucky enough to have been identified by a contractor who is impressed with their skills and work ethic and therefore keen to recruit them as a member of their team. This then enables them to further upskill themselves under the contractor's guidance. Or they may end up being recruited by a company urgently in need of tailors and therefore willing to recruit less experienced young men and women keen to upgrade their skills. Alternatively, some may have bad luck and end up in a company that runs out of orders or that shows little tolerance towards novice tailors, and hence dismisses them before long.
What is clear, however, is that skill acquisition goes hand in hand with movement between companies, roles and sectors of the industry. While the move from domestic to export firm is a common one, tailors move between companies for a variety of reasons.
Anand, a 24-year-old Mathari man, illustrates a typical trajectory that straddles jobs, companies and learning processes. Although Anand did take a short (10-day) tailoring class, he entered the industry with very few usable skills. he first worked as a helper to a cutting master before moving to a domestic firm where he was given a tailoring job and learned the required sewing skills on the job. he said he mainly learned from the tailors around him but also had to concentrate and take initiative himself. After one year, he moved to an export company, because, he said, that previous firm had 'a very strict manager who didn't give us much freedom' . In this third company, he worked for one and a half years, but after some time, the work didn't suit him anymore 'as there were too many rules and too much compulsory overtime, even till 2, 3 or 4am' . As he got married around that time, he decided to leave and over the next year he worked in several different companies, for a few months each. When we interviewed him in March 2020, he had just left a company as they had announced that there would be no work for the next month. he had moved to a different export firm where he joined a contractor who was an ex-co-worker of his. here, he said, he could start and stop work as he liked as he was employed through the contractor and paid piece rates.
This brief trajectory of one man illustrates how male tailors acquire skills on the job and how this learning is accompanied by frequent shifts between different types of firms and employment arrangements (domestic vs export, direct payroll vs contract work, shift rates vs piece rates, etc.). Once tailors master the basic skills, shifting to a different factory is often the only way to be upgraded from helper to tailor and to have one's skills and experience recognised, both socially and financially. It is skilled tailors who can demand piece rates and boost their daily wages by working long hours. Constant mobility between firms is therefore a key part of workers' search for a better combination of pay and employment conditions. While pursuing pay in line with their skills and experience is certainly a main driver for moving between factories, high levels of labour circulation between firms are driven by many causes, including a search for regular work, convenient locations, amenable contractors and supervisors, and a desire to gain new skills (Carswell and De Neve 2013).
But spatial and job mobility are highly gendered and pathways of skill acquisition and spatial labour circulation work out quite differently for women garment workers, whose reproductive work intersects with their paid employment (Mezzadri 2016(Mezzadri , 2021. While women -like men -predominantly learn garment work on the job, their skill development and their movements between jobs and factories are prompted by different forces. While men's careers are driven by the ideology of the breadwinner, by spatial mobility and by masculine ideas about freedom, women's job trajectories are shaped by the changing reproductive demands placed on them during different stages of the life course and by their heavily restricted spatial mobility -in this case their ability to commute to the city for work. Over their life course, women tend to move in and out of paid employment, but this often results in downward career trajectories, such as when skilled women tailors end up working as helpers or checkers when returning to factories after a break or as home-based workers De Neve 2013, 2018;Carswell 2016).
Over the last couple of decades, it has become acceptable for young women across castes -except for women of wealthier, higher-caste Gounder households -to commute to Tiruppur for garment work before marriage. From the age of 16 or 17 onwards, many unmarried women follow a parent or older sibling to Tiruppur to start as a helper in a company, or they join friends and neighbours on one of the company buses that collect workers from Allapuram. While women also learn a variety of tailoring skills on the job and work their way up to tailor, their learning tends to be circumscribed in a number of ways. Their families expect them to work no longer than an eight-hour shift and to return home by 6pm. As a result, they not only have less time for informal learning after working hours, but also struggle to be recruited by companies that expect workers to complete a 12-hour shift or do flexible overtime, a standard requirement in export companies. Women's limited ability to do overtime disadvantages them in terms of opportunities for learning and for accessing subsequent jobs. This lack of flexibility is one of the main reasons why many women continue as helpers or checkers or return to checking work even after they acquire tailoring skills, as such work rarely requires them to stay late. Their lack of flexibility in terms of working hours is further affected by their restricted physical mobility. Women rely on company buses or public transport to commute to work, and are therefore tied by fixed bus schedules and the geographical areas that public buses cover. Men are at a huge advantage in this respect, as most adult working men in Allapuram own a motorbike nowadays. They travel home whenever a shift ends, even if this is late at night, and can therefore offer the mobility and flexibility valued by employers. Being able to travel to factories farther away or off the main roads also means they can select the most convenient or rewarding jobs. While women too shift companies in search of better wages and more skilled work, their restricted spatio-temporal flexibility limits their options quite considerably.
After marriage, women's flexibility is curbed even further. While it has become acceptable for women to commute to the city before marriage, husbands or in-laws often prevent women from working in Tiruppur after marriage or childbirth. yet given the need for double incomes and the rising levels of household debt, ever more families are keen for their daughters-in-law to bring in extra wages -as long as this does not interfere with their domestic responsibilities. The birth of a first child usually spells the end to a woman's paid employment in the city. At that point women tend to give up garment work for anything from a couple of years to a 10-year period, depending on the number of children they have and the availability of help from in-laws and natal families. Once the youngest children start attending the local balwadi (nursery) or school, women often return to work. Only few of them, however, will return to commuting to the city. Even an eight-hour shift in Tiruppur means being away from home for at least 10 hours, which few women can afford given the ongoing domestic demands on their time.
Until about 10-15 years ago, having children meant the end to a woman's paid employment or a return to agricultural wage labour in Allapuram. Today, women have more options available thanks to the new garment workshops that have been established inside the village. Many now return to garment work in the local workshops, even though this is often accompanied by a downward career trajectory. Given the ongoing constraints on their time, many women who previously worked as tailors now take up lower paid checking or helping roles in the village. This work allows them to do shorter shifts, does not require overtime, and can be combined with domestic work. In 2020, several women complained that even the 45-minute lunch break in the local workshops was inadequate for them to go home, prepare food and eat, even if they had cooked in the early morning hours. Several women, therefore, resort to home-based work, which affords even more flexibility to fit paid employment around domestic work, rather than the other way around. This usually entails a further step down the skill and employment ladder, and makes their salaries increasingly unpredictable, irregular, and marginal to their household's income. In sum, rural women's employment trajectories need to be considered in the context of highly gendered reproductive roles and responsibilities that shift across the life course. Women's trajectories appear to be far less conditioned by the types and levels of skill they acquire than by their reproductive roles. The latter place particular temporal and spatial constraints on their ability to upskill themselves, engage in paid employment, and enhance their earning capacity. highly gendered spatio-temporal constraints, thus, produce very different skill acquisition and employment pathways for the men and women garment workers of Allapuram.

Conclusion: training for employment or skilling up from employment?
The Indian government's interest in promoting skills development to enhance productivity and economic growth, increase employability and get young people into jobs has led to a proliferation of skills training policies and programmes across India. yet evidence from a village in the hinterland of the Tiruppur garment hub in Tamil Nadu shows that most local rural residents employed in the Tiruppur industry did not receive any formal training at all. Instead, they got a job first and through their employment acquired the skills necessary to gradually move into more convenient, desirable and better paid jobs. Our findings thus suggest that rather than skills training and skill acquisition being an essential precursor to employment, the very opposite is true amongst this rural workforce. Rather than being trained for employment, the rural population gains skills from employment (Chea and huijsmans 2018, 43).
In this paper, we identified five pathways of on-the-job learning among garment workers, with the 'as helper/self-taught' pathway being the most frequently mentioned one. It is clear that garment skills are routinely acquired through self-directed learning rather than informal training and that much of this learning depends on an individual's personal interest, initiative-taking and drive to learn new skills (Regel and Pilz 2019). Informants expressed a preference for such learning as it enabled them to obtain the most appropriate skills, access jobs quickly, move freely between companies, and rise up the employment ladder through spatial and job mobility. They saw introductory skills training courses as wholly inadequate for what the labour market required and, as a result, only rarely signed up for them.
A demand-driven approach to the study of skill development (Ramasamy and Pilz 2020) reveals that such self-directed learning -and the employment opportunities that it may or may not translate into -is also shaped by structural socio-economic, gendered and spatial inequalities. In particular, the intersections of productive and reproductive labour generate highly gendered pathways of skill acquisition and labour market access (Mezzadri 2016(Mezzadri , 2021Carswell 2016). Indeed, demand-driven research and policy would do well to pay more attention to the 'social life of skills' or the informal processes and relationships that shape people's access to skills and employment opportunities (Carswell and De Neve 2018). In India, these processes -very much like the informal economy in which they are embeddedare mediated by the social institutions of gender, caste, age and space, and by the intersections of productive and reproductive work (harriss-White 2003; Vera-Sanso 2012). In the Tiruppur region, the constraints of gendered reproductive roles leave women -and elderly women in particular -with specific disadvantages in both the skills and labour market. Age and stage in the life course play a key role too as married women and women with children find it challenging to commute for work, upgrade their skills and access better job opportunities, irrespective of what training is on offer (Ramasamy and Pilz 2020). Finally, while garment workers widely agreed that urban employers look at 'skill rather than caste' , within the village it was clear that caste membership shaped labour market and entrepreneurial opportunities alike. Only members of middle and higher-ranking castes were able to establish garment workshops in the village and recruit Adi Dravidas and Matharis as workers. home-based work was similarly contracted out along caste lines, with Dalit women largely functioning as a reserve labour force.
As identified elsewhere, also in Tiruppur skill development programmes focus on recruiting migrant workers to the industry rather than supporting the gender and age-specific skill and employment needs, and upward job mobility aspirations, of the local rural population (Ruthven 2013(Ruthven , 2018Nambiar 2013Nambiar , 2021. As a result, training initiatives' primary focus was to equip novice migrant workers with entry-level garment skills in an attempt to boost the supply of labour to the industry. however, they did little to support the needs of local populations, who acquired and upgraded their garment skills on-the-job, while pursuing employment opportunities through spatial and job mobility. In order to better understand and address gender-, age-and caste-specific training needs, vocational training research and policy would benefit from a demand-driven approach that pays close attention to the needs of individuals and social groups as well as the social and economic landscapes that shape exclusion from both skill development and job markets (Ramasamy and Pilz 2020). Such an approach to skills training would make skills initiatives more suited to the needs and employment aspirations of rural populations. It would also better equip providers with the means to address some of the most poignant forms of exclusion and discrimination, and to contribute to the sustainable development and policy alleviation goals of national skill development policy. Until we better understand the skilling and labour market dynamics of specific localities, skills training initiatives are likely to misfire: failing to impart relevant skill opportunities, continuing to ignore the gendered political economy of labour markets, and persisting in excluding those who might gain most from targeted training initiatives.