Shifting agrarian labour regimes, ecology, and the crisis for Dalit women’s work in India

Abstract

Building on the differentiated analysis of India’s agrarian crisis, this article argues for centring ecology in understanding the crisis faced by agricultural labour. The empirical case is of landless Dalit women in Punjab, India, experiencing the shift from a cotton-dominated labour regime to a paddy-dominated one. It delineates the materiality(s) of commodity, workplace, and body associated with the two regimes and explores its contingent intersections with social reproduction, caste oppression, capital’s strategies, and the state. Overall, the article argues that labour’s agrarian crisis is both produced through ecology and manifests in/as ecology across different moments and levels of analysis.

1. Introduction

India has been in the throes of an agrarian crisis for almost three decades now. The continuous farmer suicides across India since the liberalization of its economy have come to represent the severity of agrarian distress. Simmering under the surface of these tragedies are declining farm incomes, chronic indebtedness, rural outmigration, and rural unemployment (Reddy and Mishra 2009; Swaminathan and Baksi 2017). Several scholars have offered a differentiated perspective on the agrarian crisis, that is, outlined the differentiated impacts on large and small farmers, between regions, communities, and genders (see below). Building on this perspective, this article explores how we may understand labour’s agrarian crisis, a relatively neglected issue, and how ecology can be integrated into this understanding. According to the 2011 Census (the last available), ‘agricultural labourers’ constituted over 50 per cent of farming households, which the Indian government considers a proxy for landless peasant households (Government of India 2013; Dogra 2020). Moreover, there is evidence that their distress is increasing alarmingly (The Hindu 2021).

The agrarian crisis in India is seen primarily as a crisis of subsistence and, to some extent, expanded reproduction within agrarian capitalism and studied through income, profits, and debt. Some argue that the entire sector is plagued by the crisis (e.g. Patnaik 2011) and others that this does not necessarily imply a crisis for large, capitalist farmers (Lerche 2013; Sinha 2020). But there is a broad agreement that the crisis exists for small farmers (or petty commodity producers) and landless workers. Further analytical layers have been added to this by other scholars. Taylor (2011: 484), for example, writes, there is a ‘generalized crisis of social reproduction among land-poor farmers and landless labourers’ (italics mine). A step further, Naidu and Ossome (2016) highlight that the burden of this crisis of social reproduction falls primarily on rural women and is linked to India’s unfinished land reform process where significant concentration of land ownership exists alongside decreasing access to private and common land.

Within this, there exists an important, through relatively smaller, scholarship focused on agricultural labour. Pattenden (2016), for example, demonstrates how forms of labour control, domination, and resistance have changed in a neoliberal economy (see also Shah et al. 2018). Other studies demonstrate how labour, especially women, negotiates debt for their survival, and occasionally, empowerment (Guérin and Venkatasubramanian 2022; Mehrotra 2022). Note, however, that ecology is not a category of analysis in these studies.

Other scholarship does point to the ecological dimensions of the crisis. But much of it is concerned with the environmental degradation and pollution caused by intensive agricultural practices, that is, in terms of water, soil quality, and biodiversity (e.g. Sidhu et al. 2008; Pingali 2012). Here, the environment is seen fundamentally as an external input into agricultural production. Such a view is, at best, a partial way of seeing ecology in relation to agriculture. It misses the deep interconnections between agriculture and ecology developed through Marxist and materialist ecofeminist scholarship (e.g. Moore 2010; Salleh 2010; Barca 2018).

To be sure, some (though not many) studies use a political ecology perspective to understand the crisis, highlighting the intersections between ecological uncertainties and socio-economic differentiation. Patel (2013) analyses the long history of the Green Revolution in India (and Asia overall) to show how production technologies and properties of high-yielding variety (HYV) crops intersected with class relations and geopolitics to produce adverse outcomes for both ecology and the poorest rural populations. His analysis, though at a world-historical level, clearly establishes that ecology is not an externality of agriculture and demonstrates the operation of a nature-society dialectic in agriculture. Similar insights are developed by field-based studies. Taylor (2013) argues that in the southern Indian state of Andhra Pradesh, the region’s geo-hydrology, the use of water-intensive HYV seeds, and the imperatives of commercialized agriculture have together led to continuous drilling of wells to extract groundwater for irrigation, leading to severe stress on the aquifers. Meanwhile, the drilling expenses led to chronic debt for small farmers. Using the case of central India, Matthan (2023) argues that climate risk articulates with capitalist risk through social differentiation, industrialized monocropping, and status-quoist state and non-state interventions to produce ‘climates of uncertainty’: where it is increasingly unviable to be a small farmer but not possible to disengage from the risks (see also Münster 2015 and Flachs 2020). While these studies have been crucial to embedding an ecological lens into the scholarship on agrarian crisis, they are almost entirely focused on small farmers, with scarce mention of labour. To be sure, Taylor and Bhasme (2021) discuss the socially differentiated dynamics of agricultural and non-agricultural labour generated by climate stress as part of their study on climate resilience initiatives in Indian villages. But their focus is on labour by small farming households rather than landless labour, which forms the focus of my study.

Therefore, building on the existing scholarship on agricultural labour in and the political ecology of India’s agrarian crisis, I ask: how can we centre ecology in our understanding of the crisis for agricultural labour? And what other layers of the agrarian crisis—and struggle—are opened through this lens? I argue that the starting point of answering these questions must be to recognize that agricultural production is a socio-ecological process and agricultural labour is ecological labour. This work is mediated through ‘materiality’, understood as the biophysical quality of things, but which I argue is itself historically contingent. I borrow Baglioni et al. (2022b)’s delineation of materiality into that of commodity, workplace, and body within a labour regime framework to build this argument. I combine it with Mitchell’s (1996, 2010) relational understanding of landscapes and Rioux’s (2022) emphasis on the corporeality of labour regimes to develop the cross-scalar dimensions of agrarian labour’s predicament.

The article uses the case of the northern state of India, Punjab, where a sub-region is witnessing a shift from a cotton-dominated labour regime to a paddy-dominated one, with profound implications for the local Dalit (or Scheduled Caste, the constitutional category for ‘lower’ caste groups) female agricultural workers. Using Raj’s (2020) concept of ‘categorical oppression’, I argue that the situated category of caste is crucial to how the materiality of the labour regime is structured, experienced, and negotiated by the workers in this context. It can appear disingenuous to examine the predicament of landless, lower caste workers through the lens of ‘crisis’ given the sustained, historical forms of oppression they experience. Yet, in a context where agrarian crisis is the dominant political and policy narrative, it is important that the structural conditions of their exploitation are explained and not allowed to be ‘naturalized’ (Raj 2022).

The next section develops the analytical agenda of the article and is followed by a description of the site and methods. The remaining article presents the empirical material and analysis, comparing the two labour regimes and discussing how the crisis manifests for labour through and beyond wage work. A conclusion, including theoretical reflections, follows.

2. Agrarian labour regimes and ecology: An analytical agenda

Agricultural labour confronts nature directly, whether in the cultivation of crops, rearing of cattle or farming of fish. It is a form of labour that directly evidences Marx’s (1976) idea of social-ecological metabolism, that is, the relation between humans and nature, including human control of nature, is established through the labour process. In other words, it is through the application of labour’s body and mind that elements of nature are rendered useful for human existence (see also Foster 1999). This section argues that moving from this theoretical understanding of labour’s role in shaping human-nature relations to a context-specific, empirical analysis of how ecology shapes the predicament of agricultural labour requires unpacking the specific biophysical properties and processes, or forms or materiality, shaping labour relations.

Studies in agrarian change have long recognized the importance of ecology (Bernstein 2010). Building on the Mann and Dickinson thesis (1978), Boyd et al. (2001) argued that firms in nature-based industries like agriculture face the ‘problem of nature’, where nature could serve as ‘obstacle, opportunity or surprise’ (560) in obtaining competitive advantage. Recently, Banoub et al. (2021) have argued that capital’s strategies in nature-based industries include expansion and intensification of commodity production à la Moore (2010) but also commodity transformation, e.g. through genetic modification (1551). While these perspectives shape this article’s approach, they are primarily a theorization of capitalist penetration and capital’s strategies for accumulation in sectors that confront nature, even if labour processes are discussed. Instead, Baglioni and Campling (2017) theorize the role of ecology from the vantage point of labour by arguing that the labour process is both socially indeterminate (due to the persistent, overt or covert, conflict between labour and capital; see also Smith 2006) and ecologically indeterminate (owing to uncontrolled/uncontrollable environmental conditions). Arguably, in agriculture at least, ecological indeterminacy exists not only in the labour process but production overall because crops’ production times are greater than labour time. This exposes crops to environmental uncertainties even when labour is not working (Section 4 discusses how this could still impact labour relations).

Operationalizing the implications of this ‘problem of nature’, in other words, of ecological indeterminacy, for labour brings me to labour regime analysis. A vastly rich and varied body of scholarship exists on labour regimes, mapping which lies beyond the scope of this article (see Baglioni et al. (2022a) for an overview). However, it should be noted that labour regime is an analytical framework (2022a: 82), a heuristic device of sorts, that enables contextual and granular examination of labour-capital relations. Some key elements of labour regime analysis include a multi-scalar analysis of the economic, social, and political forces shaping the labour process in a given spatial-historical context, the role of social difference in the reproduction of labour’s exploitation and spaces for labour’s agency. Baglioni et al.’s (2022b) emphasis on ecology as a ‘constitutive part’ (88) of labour regime analysis alongside production, circulation, and social reproduction is relatively new and one sought to be developed by this article in conversation with the aforementioned elements.

To study how ecology shapes a given labour regime, Baglioni et al. (2022b) suggest delineating the materiality (or biophysical qualities) of commodity, workplace, and body, with materiality being the mediating category through which dynamic intersections of ecology and social relations can be explored. I consider such a delineation within a labour regime framework to be instructive as it has room for social and historical contingency in the way ecology matters to labour. Put differently, the issue is not just whether materiality matters but which aspects of it are activated in which context, why and how, and with what effect on differentiated social groups. For instance, the materiality of the same commodity can change depending on the technology being used to produce, process or distribute it. Similarly, the environmental conditions required for a crop’s production (materiality of workplace) may imply difficult working conditions for labour (e.g. hot, humid weather during paddy transplanting, as we will see below).

The contingent understanding of the role of ecology in labour regimes being developed here is complemented by Mitchell’s (1996, 2010) relational understanding of landscapes, especially with respect to agriculture where the landscape is the workplace. Using the case of 20th century Californian countryside, Mitchell argues that the morphology of landscapes (including ‘natural’ and built elements) shapes labour-capital relations (as discerned through processes of production, circulation and social production) and vice versa. For him, the labour process is key to understanding this dialectical relationship: while labour produces the morphology of the landscape, the latter then “determines the rational, ‘natural,’ or necessary form that labor demands, labor relations, and labor processes must take” (2010: 148). One could infer, in other words, that the landscape embodies and encompasses a labour regime(s). However, he goes on to demonstrate that this dialectical relation ‘creates a sort of path-dependency’ (2010: 149) in, or inertia against, the transformation of landscapes and their attendant social relations. This raises questions of change and continuity in the context of shifting agrarian labour regimes in the present study as well. I will reflect on this briefly in the conclusion, while identifying how the historically produced landscape works into the materialities of commodity and workplace in the preceding sections.

Finally, towards analysing the materiality of the body, I note the argument of Rioux (2022; see also Taylor and Rioux 2017) that the labour in labour regime analysis is not abstract but embodied labour, that is, situated in historically and spatially constituted social relations and marked by particularities and social differences like gender, race, ethnicity, migrant status, etc. For example, Baglioni (2022) demonstrates how rural Senegalese women are disciplined ‘as women’ through domestic responsibilities and marriage structures in the household, which in turn made them more pliant in the workplace. While the question of how the labouring body is controlled is not new in labour regime analysis (e.g. Burawoy 1983), the specific emphasis on what Rioux terms the ‘corporeality’ of labour, that is, the (assumed or assigned) ability or capacity of specific bodies to perform specific labours, is an important intervention.

In the Indian context, and especially in my case study, such a corporeal perspective demands that we factor the role of caste in structuring the labour regime. A vast literature in agrarian political economy (and beyond) has demonstrated that caste relations are deeply entangled with class (e.g. Breman 1996; Byres et al. 1999), and even ‘co-constitutive’ (Lerche and Shah 2018). Raj (2020, 2022) has argued, however, that caste operates not merely to reproduce capitalist production relations but also to keep Dalits stigmatized in wider social life, a phenomenon he terms ‘categorical oppression’. He writes, ‘the Dalits have to struggle not only for economic security, but also for social dignity within the caste society’ (2022: 12). The empirical analysis below will thus also reflect on how Dalit women workers negotiate this double-edged struggle.

3. Site and methods

This study is based in the north Indian state of Punjab. The state experienced spectacular agricultural growth through Green Revolution technologies introduced in the late 1960s and early 1970s. In the following decades, the state shifted decisively towards growing paddy in the monsoon season and wheat in the winter. A small south-western pocket, however, continued to predominantly produce cotton in the monsoon season till the 1990s. Since then though cotton farmers in this pocket too started shifting towards paddy production as cotton costs were increasing, markets prices were volatile, and the yields were impacted due to weather and pest attacks. The repeated pink bollworm attacks in recent years have shifted the weight decisively towards paddy (Nitnaware 2023). Paddy, on the other hand, is more expensive to produce than cotton and less suitable ecologically for the area. But its purchase is guaranteed by the central government at somewhat remunerative prices in state-regulated wholesale markets. The simultaneous expansion of free electricity connections for tube well irrigation by the state government also meant that the cost of producing water-intensive paddy is subsidized. These conditions, elements of what Mitchell (2010: 149) would term the ‘actually existing landscape’, have incentivized farmers to produce paddy rather than cotton (see Sinha 2022 for details). The shift from cotton to paddy is then a function of farmers mitigating their own crisis. How this crisis-mitigation by farmers amplifies the crisis for a specific group of agricultural labour is this article’s core interest.

In Punjab, agricultural land is owned primarily by the dominant Jat caste, while the landless labour is usually Dalit or Scheduled Caste. This is also true for my research site in Bathinda district. The landless labouring castes include different sub-castes such as Mazhabi and Boria Sikhs, but for the purpose of my research, I used the combined political category of Dalit as there was no discernible difference in work done by the women of the two sub-castes.

The fieldwork for this research was conducted between August and December 2019 and a short revisit in August 2022. Three villages near Bathinda city were chosen to represent different production patterns or social organization in 2019: Village A had predominantly cotton cultivation, while Village B had seen a decisive shift from cotton to paddy over the past twenty years. Village C, on the other hand, was an entirely Dalit settlement where all the land, albeit of much poorer quality, was owned in relatively small plots by Dalits. At the same time, not all the households owned land. Here, both paddy and cotton were cultivated as the monsoon crop. Notably, by the time of my revisit in 2022, both A and C had moved towards cultivating much more paddy and there was hardly any cotton being produced in any of the three villages.

Data were collected using semi-structured interviews with thirty-eight women across the three villages in 2019. This was complemented by data from interviews with farmers, traders, and male workers. In the revisit, sixteen of the same women were re-interviewed along with one new respondent.

4. Labour regimes of cotton and paddy

4.1 Production process

To understand the different materiality of the cotton-dominated and paddy-dominated labour regimes, I first outline the production processes of the two crops, as described by farmer-respondents.

The cotton crop in Punjab is sown in April or May. Cotton seeds, primarily genetically modified Bt cotton, are sown mechanically using a seed drill attached to a tractor, where one person drives and another clears the drill of the mud. After sowing, fertilizers and pesticides are applied once every 15 days. This is done usually by a local Dalit male labour on a daily wage or farmers themselves using a backpack sprayer (recently, some large farmers have started using tractor sprayers that are faster). Once the plant has sprouted, weeds are removed both mechanically and, where that is not possible, using local daily wage labour (male or female). There are about three rounds of weeding, with six workers needed to complete one acre in one day. From sowing to harvest, Bt cotton also needs four to six rounds of irrigation, which farmers typically manage on their own or with the help of one male worker if the landholding is large.

While only a few labour days are required through most of the production cycle of cotton, the harvest, starting in late September, is labour intensive. Cotton fibre bolls are picked primarily by local Dalit female labour: the first is wet with low yields, while the next two have higher yields. Wages are paid by the amount of cotton picked by each worker on the days they have worked (see more below). Following this, any fibre bolls that are not fully ripened are broken at the stem manually and left under the sun to allow the fibre to break out of the shell. This could be paid by weight or by daily wage. Finally, the cotton plant stalks are cut down to be dried as fuelwood; this work is considered ‘heavy’ and done primarily by local Dalit male labour, with women accompanying them occasionally. This could be paid by daily wages or on the condition that the labour can take home some portion of the stalk.

Paddy, which competes with cotton as a monsoon crop, is sown later. Starting in mid-May, the seeds are first grown into seedlings in a nursery, created in a part of the landholding, a process that takes about a month. The seedlings are picked and bunched together to be transplanted manually in puddled farmland by wage labour, which is the most labour-intensive part of the production process. This work is done usually by groups of migrant male labour from eastern India and sometimes by groups of local Dalit women. Wages are fixed per acre of transplanting done and paid to the group as a whole.

Fertilizers, pesticides, and weedicides are applied by male labour or the farmers themselves, scattered over a few days in the production cycle. Fertilizers are applied to the crop in two rounds within forty days of transplanting, although some farmers apply it once while preparing the land for transplanting. Pesticides are applied using backpack sprays, and the number of rounds and types can vary from year to year, depending on the prevalent disease risks. Paddy is famously water intensive, with some farmers claiming that it takes about four times as much water as cotton. But like cotton, this can be managed by the farmers themselves, with large farmers sometimes employing labour to help with this.¹ Most paddy varieties are harvested in October, while some varieties may be harvested from late September. This is done entirely mechanically using a combine harvester, an exorbitantly expensive machine which is rented by most large and small farmers from a handful of large farmers who own it.

To unpack the varying materiality(s) of the two labour regimes, I focus below on cotton harvest and paddy transplanting as these are the only parts of the two crops’ production processes, respectively, that employ labour on a significant scale and constitute the chief source of agricultural wage work for local Dalit women.

4.2 Materiality of the commodity

The materiality of the labour regime is firstly and most obviously reflected in the commodity being produced. Here, I demonstrate how different elements of the landscape, namely, the labour process of the commodity’s production in the given context, ecological indeterminacy, and market structures confronted by farmers, impact the amount of work available to local Dalit women, the risks they face, and the space for negotiation.

In the case of cotton, the primary demand for labour emerges during the harvest period. Assuming there is a good harvest, wage work in cotton is available for around two and a half months (although Section 4.3 will show that some women do not or cannot work this full length). No migrant labour is employed for this work in the study region. This contrasts with paddy where the main labour demand is for paddy transplanting. This process lasts only up to a month in the area (although this could extend if farmers have sown late maturing varieties, a recent development in the study area). Therefore, growing paddy production has nearly halved the workdays available. The loss of income this entails is amplified by the fact that, as mentioned above, most farmers prefer to hire migrant male labour for paddy transplanting from poorer eastern Indian states like Uttar Pradesh, Bihar, Jharkhand, and West Bengal.²

In interviews, women often stated that if it was not for the migrant men, they would have had work for around two months in paddy transplanting itself. But this is unlikely since, by law, farmers cannot start paddy transplanting before June 15 but need to finish it by mid- to late July to get a reasonable yield. This means there is a short period of intense labour demand which migrants play a crucial role in fulfilling. These men work longer hours at lesser pay, while many women do not do paddy transplanting work due to other aspects of materiality, that is, of the workplace and body (Sections 4.3 and 4.4).

In cotton, on the other hand, once the crop is ready for the first harvest, the farmers need an almost continuous supply of labour to pick the fibre bolls, lending the women some structural power to negotiate the wage rates and make sure the farmers arrange for their pickup and drop-off from their fields if they are far away (see also Singh 2017). Farmers often give an advance cash amount (shahi) to some labour, within their own or other villages, to work throughout the season in their farm (terms of work described as pakka or fixed). This provides crucial cash flow for these women and their households.

It is important to note here that unlike countries like China, USA, Brazil, and Turkey, mechanical cotton-picking has not been introduced in India yet. Therefore, the low production acreage is the key impediment to wage work in cotton. But in paddy too, the limited paddy transplanting work that is available is under threat. Interviews in village B in 2022 revealed that some large farmers had started mechanical transplanting leading to even less wage work available. The level and type of mechanization are thus also features of the landscape configuring the materiality of the commodity’s production and, through that, the nature of labour relations.

While cotton’s labour process in this region then affords some leverage to the labour, its intersection with the ecological indeterminacy of agricultural production also creates risks. As work is available during harvest, or at the end of the production cycle, any damage to the crop during the growing period (e.g. due to disease, weather fluctuations) can reduce the yield and affect the work available. As mentioned above, this has happened every year recently, deterring farmers from cotton production. Interviews revealed that after the pink bollworm attack in 2021, some farmers wanted to retrieve the undamaged bolls and plant stalks. Since this would take a lot of work but would not yield enough volume of ripened or unripe bolls, women refused payment by weight and were successful in getting a daily wage. But many farmers also simply chose to mow down the crop and grow a late paddy variety instead of paying labour. Contrarily, in paddy, work is available at the start of the production cycle and the subsequent crop risks do not impact labour.

Farmers’ (representative here of capital) constraints and strategies also impact how promptly labour is paid in paddy and cotton. Farmers make relatively more prompt payments for paddy transplanting either from their savings or from formal or informal credit they have secured for paddy cultivation. This is also conducive to rehiring migrant labour who leave as soon they are paid and has become the norm even when local Dalit women do the work. For cotton-picking though, farmers often delay payments by saying they can only do so once they have sold the crop. This is true to an extent as most, if not all, farmers cultivate on credit and the cotton markets are volatile (Sinha 2022). However, this means the women must continuously chase the farmers to pay their due wages, reinforcing the indignities of their lower caste status. The difference in the ecological indeterminacy in the production/labour process and payment schedules for the two crops suggest that, in principle, the growth of paddy cultivation is less risky for the labour involved. However, this is a moot point for most local Dalit women, given they do not get or are unable to do much of this work.

4.3 Materiality of the workplace

Baglioni et al. (2022b) argue that the ‘physical environment of the workplace’ can shape the form of labour’s exploitation. Given that the workplace here is open farmland, it should be noted that cotton-picking takes place during autumn and winter, which is considered pleasant for the most part and so more women feel able to work in these conditions. Paddy transplanting, on the other hand, happens in the peak of summer and several women stated that they simply cannot do work in the open fields in that much heat and humidity. This links directly to bodily capacities, a part of the body’s materiality, to which we turn our attention in Section 4.4.

However, especially in the case of women, the materiality of the workplace should be seen in terms of not only the physical environment but also the specific entanglements it has with issues of social reproduction. For paddy transplanting, the workplace for local Dalit women is the same village in which they reside, while for cotton, it could also be neighbouring villages or occasionally distant villages in other states. The workplace is thus typically embedded in their relatively immediate surroundings, the landscape of their wider social life. The varying levels of proximity between their workplace and homes have implications for the nature of the labour regime as they impact how women negotiate wage work with domestic work. In fact, women consider both types of work as part of a continuum of things they do to fulfil their social role; in their own words, everything they do is part of ‘kabeeldaari’, that is, to shoulder family or household responsibilities (note that this term applies to men as well). They considered wage work ‘as a supplement to and break from their household activities’ (Baglioni 2022: 451). Another recent study in rural Punjab has similarly argued that within a household, the boundaries and balance between productive and reproductive work (both waged and unwaged) is blurred and can shift over time (Rao et al. 2024).

As mentioned earlier, the three study villages had different production patterns and conditions in 2019. The women in village A, where cotton production dominated, worked in their own village. When asked about whether they go outside the village, they unequivocally stated that there was so much cotton being produced in the village itself so there was no need; in fact, women from other villages were being brought to the village by farmers to pick cotton (also corroborated by multiple farmers). In villages B and C, on the other hand, most women went to nearby villages to work as there was little cotton being cultivated inside their own, although some worked within the village too. In 2022, the year of my revisit, there was hardly any cotton being produced in any of the three villages and most women had no cotton-picking work either inside or outside the village.

Nevertheless, my data show that working in other villages versus in one’s own has implications for women’s domestic responsibilities. This is partly due to the time it takes to travel to other villages as well as related terms of work. As mentioned earlier, cotton-picking is paid by weight. Even though some women may go as a group to a farm (as it allows them to negotiate wages), the payment is made individually, according to what each person has picked. Moreover, they may either work casually or they can become fixed (pakka) with a farmer for the season. If they decide to work casually, then they can come and go whenever they want and even skip days. Usually, it is women picking in their own village that work in this way. When they work on a fixed basis, skipping days is still an option but less so—the farmers expect the women to come to work early in the morning (according to one respondent, by 8 am) and to come daily during the harvest period. When they work in other villages, they work on a fixed basis with one or a few farmers as a group because they do not have networks with other farmers and because they want to be sure that travelling all the way will be worth their while. As a result, women who work on a fixed basis either do it at the expense of the domestic work, if they have older women relatives or slightly older children who can manage on their own. Evidently, some women simply do not have such support, or if they have infants, they necessarily choose to do what they can locally and/or to work only casually. But this still allows women in different situations to earn some wage income.

In the case of paddy, there is much less flexibility because payment is made by acreage transplanted. Groups of labour are contracted to complete transplanting on a given number of acres at a fixed rate per acre. The payment is then shared between all group members. There is thus considerable peer pressure on every member to work regularly with the rest of the group (see also Section VII). This applies to the migrant male labour as well. However, as they do not have comparable domestic work, they can work longer hours, transplanting the same number of acres in lesser time than the women.

The overall decline in cotton production in favour of paddy in the study area has also meant that a handful of women were migrating to more distant parts of Punjab or to neighbouring states like Haryana or Rajasthan where more cotton was produced. I interviewed five such women in 2019, three of whom went with other family members, and each of them clearly stated that this migration was crucial if they are to earn a lump sum amount of cash in the year. As migrant labour themselves, they too prioritize working longer hours over all else.

4.4 Materiality of the body

This section focuses primarily on the physical capacities and skills involved in cotton-picking and paddy transplanting. While both types of work are physically strenuous, all the women I interviewed agreed that paddy transplanting was more difficult than cotton-picking. Cotton-picking is done standing—women hang a cloth from their heads or less frequently on their side in a way that it can hold cotton as they are picking it. They make a pile of this cotton (dheri) at one spot through the day, or the hours they are working. This work involves some difficulties: some women reported how they wear long sleeves on their upper garment to avoid getting scratched from the thorns on the side of the plant. Moreover, the cloth with the cotton filled up can become heavy and hurt their back. However, to address this, they can take breaks or empty out the cloth as frequently as possible. This means that old women and children can also do this work, even if they cannot pick as much as some other women. By the time of the final harvest, the winter sets in fully and some women stop picking as they cannot withstand the cold. But even such women would have been able to earn some wages in the warmer weeks.

Similarly, picking the unripened bolls is hard work as they have to break it from the thick plant stem and branches. But women can choose not to do it if they are not working on a fixed basis. Here again, the money is paid individually (either daily wage or by weight, sometimes in kind), so there is an incentive to do the work. They say ‘we have greed so we push ourselves in doing cotton work’.

Yet, paddy transplanting is considered significantly more challenging. Respondents described it vividly as one of the most difficult tasks they ever did: they must stay in wet fields for hours which affects their skin and work continuously in a squatted position, moving backwards to plan the seedlings in straight columns. One woman reported that after each season of paddy transplanting, she has an illness, a fever of some kind or skin infections (a good share of their income is then spent on obtaining treatment). As a result, women who are old or have health problems and young children do not do this work. Also, as discussed in the previous section, paddy transplanting is done in groups whereby they need to work together to cover an agreed acreage. Each member is accountable to the wider group, and so it is harder to take breaks as per individual need. Women who do paddy transplanting work clearly state that they only make or join groups with women who have similar bodily capacity to do long hours of such strenuous work.

It is also the case that several women do not know how to do this work. Most women who do it have done it earlier in the village where they were born; many others have only ever lived in cotton-growing villages and so find it difficult to learn. Meanwhile, the migrant workers who come from the rice-growing areas of east India are said to work extremely fast and the women concede that they cannot match their speed.

Mezzadri (2016) has written about the depletion of women’s bodies after working in garment sweatshops in India. I contend that the bodies of women in my study context are already depleted due to chronic poverty, food insecurity, and lack of healthcare. Paddy transplanting is strenuous in a way that many of these depleted bodies cannot do; even those who can do it agree that it cannot be sustained endlessly and, in fact, will deplete them further. In a sense, the ‘body discipline’ (Baglioni 2022; Rioux 2022) required by cotton-picking meant that different types of bodies could be accommodated as wage labour. On the other hand, paddy transplanting demands a kind of standardization of the labouring body that excludes many from wage work.

4.5 Crisis beyond wages

My intention in framing the discussion on the agrarian crisis through the shift in agrarian labour regimes is not to argue that the crisis for labour has emerged singularly through this shift. Even in a cotton-dominated labour regime, local Dalit women only had 2–3 months of wage work and their situation was dire. However, unpacking the materiality of the two labour regimes through a focus on the labour process of their main waged work and their intersection with issues of social reproduction, circulation of migrant labour and capital’s strategies have revealed how this shift has intensified the crisis for these women by excluding many of them from agricultural work. In the section, I explain how the materiality of the paddy-dominated regime also intensifies their crisis in terms of social reproduction and their struggle for dignity, and how these are mediated by the state’s interventions and inefficiencies.

First, cotton is an important source of fuelwood. As mentioned above, some proportion of the dried cotton stalks may be taken by labour if they cut the stalks themselves. Admittedly, many cannot do this and so they nevertheless need to buy from the open market. But the overall reduction in cotton production means fuelwood has become quite expensive to purchase. More than ever, women need to spend time gathering fuelwood from common lands and from the peripheries of canals and fields. This means a growing burden of social reproduction.

Secondly, agricultural wage work exists in a context of other possible wage work. Both due to growing agricultural mechanisation and increasing caste assertion vis-à-vis Jat landlords, rural Dalits, especially men, have sought non-agricultural work within and outside villages (Ram 2007). Over two-thirds of the women interviewed reported either a husband or a son in some type of non-farm employment, e.g. in construction, drivers of trucks or construction machines in and around the neighbouring city of Bathinda and selling vegetables bought from Bathinda in villages. However, this work can be irregular and not sufficient for households needs. To be sure, a handful of women I interviewed reported working in non-farm jobs too (e.g. seasonal labour in a ginning mill, cleaner in the government school, labour in a chemical factory, in construction with their husband) but these were exceptions. This is in line with state-wide trends of declining overall employment, poor absorption of agricultural labour into non-agricultural sectors and a male-dominated labour market (Singh and Singh 2022). Any agricultural wage work is thus a crucial income source for the household in such a context.

With the loss of cotton-picking work, women seek more employment from the state in the form of the Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA). Under MGNREGA, people in rural areas have the right to a hundred days of waged employment on demand, but it has almost never implemented properly (Aggarwal 2016). Previously, when cotton harvest was more and better quality, many women would choose to pick cotton over doing any available MGNREGA work. They argued that this was partly because there is the possibility of earning more in cotton by exerting themselves to pick more, and partly because the latter would only be available for a week or two and this would disrupt any ongoing picking commitments they made to a farmer. However, in the new regime, they hope for more MGNREGA work to replace their wage losses from cotton. Here again, the materiality of MGNREGA work—with fixed hours and ‘not heavy’ work—would allow even older women (and men) to earn wages. However, the implementation and funding are both getting progressively worse across India (Sivakumar 2022) and so it offers little recourse.

Thirdly, since the cotton-picking season is spread over a longer period, Dalit women could earn a lump sum amount to contribute to household expenses like house repairs and weddings.³ Moreover, they could take small advances from the farmer they were working for throughout the season as and when the need for cash arose. It was also on the promise of such work that they could take some wheat from the farmers (on credit). The loss of picking work has meant a loss of this informal credit source as farmers are not keen to lend money to those who do not work for them. But neither are women keen to borrow without work: certainly, this type of informal credit reflects a perverse patronage relationship in a caste society, but according to the women, since this was attached to the work they did, it was somewhat dignified. Negotiating this fine line between their desperate economic needs and their caste-based (categorical) oppression, some women do domestic work for these farmers (like cleaning, washing utensils) or occasional daily wage work (like collecting fodder for their cattle) to earn cash but also to maintain some basis of a relationship that allows them to borrow higher sums of cash in times of need.

Additionally, women report an increased dependence on small, often exploitative, micro-finance loans, some describing it as their key source of lump sum cash now, as well as informal credit from grocery stores. The new labour regime thus seems to have led to an increased dependence on usurious capital for their survival. But this too is not devoid of contradictions. All women reported having at least one micro-finance loan with high interest rates and difficulties in paying the fortnightly or sometimes weekly instalments. One woman reported she once had to sell utensils to pay the instalments. However, several women also expressed that micro-finance loans allow them to avoid the humiliation of asking Jat farmers for money.

A question I asked several women is why they do not simply reduce their asking rate for paddy transplanting to perhaps get more wage work. Their response is that partly, it still would not matter because the migrant men can do more transplanting daily because they do not need to do household work. So, their gender works as a structural limit to the leverage they have as labour. At the same time, they say that they cannot accept a lower wage because they also need to carry food from home that costs money and are not valued, whereas migrants are given some food ration and treated better (as ‘paraune’ or guests). Moreover, they already have some food to eat anyway—through free food rations from the state (lasting them a few months depending on the household size), credit, and/or through income from male members’ work. This shows, again, how women are exercising agency to preserve their dignity as local Dalit workers vis-à-vis local Jat farmers in times of crisis.

Farmers are, of course, acutely aware of this. They commonly argue that the free rations received by Dalit households by the government have made them ‘lazy’ and ‘unwilling to do hard work’ and contrast them with migrant men who are seen as ‘motivated’ to earn more and fast before returning to their villages. Thus, state welfare measures implicitly mediate the local labour market, which is resented by the farmers. But this should not be overstated. As noted above, MGNREGA does not work well, and the food rations do not last long. Moreover, my fieldwork revealed how rural Dalits face the incredible challenge of navigating the state bureaucracy even to create identity cards or documents that would ‘prove’ their age, economic, or caste status for benefits like free household electricity, pensions, and disability allowance.

5. Conclusion

In Introduction, this article asked how we may understand labour’s agrarian crisis by centring ecology and what we may find out about the crisis itself through this. Centring ecology was critical as agricultural labour is a type of ecological labour, embodying the dialectical relation between nature and human society, and any crisis of/for such labour would likely involve ruptures at this interface. I then used and unpacked ‘materiality’ as a mediating category to study this question by focusing on the case of local Dalit female agricultural labour in a sub-region of the Indian state of Punjab where cotton is giving way to paddy as the main monsoon crop, which I describe as a shift in the agrarian labour regime.

The article demonstrates that labour’s agrarian crisis is both produced through ecology and manifests in/as ecology, across different moments and levels of analysis. It is through the shift from cotton to paddy cultivation, an ecological change that alters the landscape’s morphology that the crisis for local Dalit women is intensified. As the delineation of materiality into that of commodity, workplace, and body shows, the change in the crop also led to a change in the amount and type of wage work available. In a testament to the socially and historically contingent nature of materiality, it was not this change by itself that was leading to the women’s exclusion from wage work but how it intersected with factors at other scales: farmers’ strategies (local economy) most obviously but also the women’s physical abilities (body) and social reproduction responsibilities (household) and the circulation of migrant labour (inter-regional).

The new paddy-dominated labour regime also represents a new type of ecology, one where women continue to shoulder burdens of social reproduction through waged and unwaged work but under even more difficult conditions—thus demonstrating how crisis is produced in/as ecology. Negotiating these changes involves a different rhythm of labour—spending more time looking for fuelwood, trying to do more non-farm work, doing less agricultural work. In Rioux’s (2022) terms, this could be understood as a corporeal regime marked by the non-utilization of the Dalit woman’s labouring body as a waged worker. These women are now having to wage the quiet struggle of reproducing themselves by renegotiating their labour and debt, state support, and caste relations in a manner that does not compromise their quest for social dignity.

Finally, this new ecology of the sub-region is swiftly aligning with that of the rest of the Punjab state by producing paddy. In this, it is also assuming the groundwater crisis of the region caused by the water-guzzling paddy crop, which threatens the reproduction of the physical landscape itself in the long term. Equally, the shift to paddy has allowed the preservation of the existing regime of agrarian accumulation by large Jat farmers, at least in the short term (even if it has intensified the class differentiation between farmers; see Sinha 2022). This underlying contradiction attests to Baglioni et al.’s (2022a) articulation of labour regimes as an ‘antagonistic relative stability’.

However, it also reveals that even when labour regimes change, they may do so towards perpetuating more of the same. Much like the changing Californian landscape of the 20th century, Punjab’s landscape remains ‘anti-revolutionary’ (Mitchell 2013), that is, the reproduction of labour’s exploitation remains stable across the two labour regimes, reflecting an inertia against progressive transformation. To be sure, rural, landless Dalit politics in the state is robust and has evolved towards an agenda combining the politics of economic security and social dignity⁴. But so far, it has been unable to present a formidable challenge to the capitalist hegemony and caste and patriarchal oppression that structures the labour and lives of so many.

Footnotes

Acknowledgements

I thank the editors of JoEG and two anonymous reviewers for their constructive feedback. I am especially grateful to Elena Baglioni for her generous comments on multiple drafts and Liam Campling for an early discussion on labour regimes. Feedback from participants of workshops on labour and ecology at CLaSP (QMUL) and Nottingham Trent University and a seminar at QMUL Geography shaped this work.

Funding

The fieldwork in 2019 was funded by the TIGR2ESS project through the Biotechnology and Biological Sciences Research Council (BB/P027970/1) and in 2022 by the School of Business and Management, QMUL.

Conflict of interest statement. None declared.

References