Revisiting agrarian questions of capital: examining diversification by capitalist farmers in Punjab, India

Abstract

While economic diversification by capitalist farmers in India is a commonly accepted fact, it is rarely analysed through the lens of agrarian questions of capital. This paper argues that questions about the movement and transformation of agrarian capital continue to be significant in understanding contemporary processes of agrarian change and rural development. However, these need to be studied by looking beyond the agrarian transition debate in contexts where agrarian capitalism has consolidated itself and non-agricultural capitalist development is not fuelled by agrarian capital. Using the case of the state of Punjab and drawing on intensive field research, the paper examines a broad spectrum of non-farm investments and activities of large capitalist farmers, including agriculture-based business, non-agricultural business, education-based diversification and international migration. It shows that both industrial investments and mobility through education have been limited. International migration is a preferred, but also risky, channel for the utilisation of agrarian surplus. The analysis takes seriously experiences of failure of non-agricultural businesses, resulting in the circulation of agrarian capital across the agricultural and non-agricultural sectors. This both further strains accumulation within agriculture and reveals the limits to productive investment in the non-agricultural economy imposed by historically specific social, political and economic conditions.

Keywords:

• Agrarian change
• capital
• economy
• diversification
• Punjab
• India

Introduction

Economic diversification by agricultural households in India has become a commonly accepted fact that we need to grapple with to understand the nature of agrarian and overall capitalist development in the country. Typically, scholars have considered the declining significance of agriculture as representative of structural transformation, desirable in terms of releasing capital and labour for industrial development. In the seven decades since India’s independence, the declining share of agriculture in India’s gross domestic product has not been replaced by a concomitant increase in industry but by an increase in services (Chandrasekhar 2007), even as economic diversification by agriculturists has increased. What this means for rural capital and labour released from an increasingly capitalised agriculture has been studied extensively from the point of view of rural labour, eg through the framework of ‘classes of labour’ (Lerche 2010) and rural non-farm employment (Chandrasekhar 1993; Visaria 1995). However, fewer studies explain this from the vantage point of agrarian capital.

Such extensive attention to labour is not unwarranted: India has a large rural landless population, and over 80% of landholdings are small and marginal (Agriculture Census Division 2011). Moreover, under neoliberalism, these sections of the rural population have suffered acute distress (Reddy and Mishra 2009); this is often the salient problematic in critical agrarian studies debates. However, since a class of capitalist farmers in India continues to accumulate, or make a profitable surplus, through agriculture (Lerche 2013; Ramachandran 2011; Sinha 2020a), it remains important to understand whether and how this class uses its agricultural surplus outside agriculture.

This paper begins with a discussion of the salience and the subsequent decline of the agrarian question of capital within the agrarian transition debate. This debate presents a key framework within which agrarian capital has been studied in order to map the development of agrarian capitalism and the subsequent emergence of industrial capitalism within a nation state. I argue that the framing of the agrarian transition debate does not allow us to study agrarian capital comprehensively in the twenty-first century in those locations of the Global South where agrarian capitalism has been consolidated (as in the Indian Punjab). There is, therefore, a need to study agrarian capital beyond the framework of this debate and to understand developments in rural areas on their own terms rather than in terms of the emergence of national industrial capitalism.

Mapping patterns of economic diversification by capitalist farmers is a productive way of doing this since it helps us consider which areas capitalist farmers seek to invest in, what they consider profitable and whether they can expand the basis for their accumulation through non-agricultural activities. Capital, of course, may be generated and expanded by a whole range of agrarian classes. But here, we focus on capitalist farmers since they are most likely to be able to accumulate capital in a sustained manner and then reinvest it for expansion.

The second and third sections of the paper analyse the evidence and explanations for diversification by capitalist farmers in India and the specific case of Punjab, respectively. I show that there is a gap in this literature in terms of mapping the full range of diversification strategies of agrarian capitalists and drawing links between these strategies and capitalist development outside the agricultural sector.

To address this gap, in the fourth section I discuss the findings from a field study in Punjab on capitalist farmers and patterns of agrarian accumulation conducted in 2014–2015. I examine their non-farm investments in businesses linked to agriculture and those outside agriculture, in education and in international migration. I argue that these investments are rarely industrial and more commonly involve trading and a variety of public and private employment. This preference is in turn a function of the state of the non-agricultural economy in Punjab as well as the relative power of the agrarian caste in the region. At the same time, even these do not always yield adequate returns, a tendency that results in further pressures on capital accumulation through agriculture. The paper, therefore, crucially advances the existing scholarship on agrarian questions of capital, on diversification by capitalist farmers in India and on structural transformation within Punjab and India.

Agrarian questions of capital

Within agrarian political economy, agrarian questions of capital have been studied through the agrarian transition debate. This debate concerns, firstly, the emergence of agrarian capitalism in the countryside and, secondly, its contribution towards the development of industrial capitalism in a nation state (although the former has been more widely discussed than the latter). Byres (1986) identified different paths towards agrarian capitalism. He argued that Prussia represented a case of transition ‘from above’: it was the landlords who ushered in capitalism in the countryside. In the USA, in contrast, family farms led the agrarian transition, representing a transition ‘from below’.

In later years, Bernstein (1996, 2006) questioned the centrality of agrarian capitalism to the emergence of capitalist industrialisation. He argued, for instance, that once capitalist agriculture was established in Prussia east of the Elbe, as against transfer of surplus from agriculture to industry, the terms of trade were favourable to agriculture. Similarly, in the USA, it was the drive of merchant capital that facilitated agriculture’s contribution to industrialisation. Building on this, Bernstein argued that as the capital fuelling structural transformation in developing countries since the late twentieth century has come from non-agricultural sources, the agrarian question of capital (or the emergence of agrarian capitalism) has become redundant for overall capitalist development in such countries.

Scholars working at national levels of analysis compared to Bernstein’s world-level analysis have confirmed or refuted the latter’s proposition to different degrees. In the case of India, Lerche (2013) has argued that the ‘classic’ agrarian transition, ie capitalist agriculture leading to industrial development, has been bypassed, but to the extent that capitalist farmers continue to accumulate, a different resolution of the agrarian question of capital cannot be written off. On the other hand, examining the evidence on land rush in African countries, Oya (2013, 1554) challenged Bernstein’s position by arguing that ‘national capitals’ that ‘straddle rural–urban boundaries’ were working both independently and in collaboration with global capital in effecting changes to rural land ownership and use.

It is important to understand that the agrarian question of capital in the aforementioned debate was concerned with the transformation of ‘predatory landed property’ into productive agrarian capital and its decline as a ‘significant economic and political force’ (Bernstein 2006, 452). The debate, in this sense, no longer applies in parts of the Global South where this has already been achieved. The other sense of the agrarian question of capital, on how agrarian capital contributes to the emergence of industrial capitalism, is now widely accepted as a non-starter. Upadhya (2009, 23) has argued that such a formulation ‘assumes that industrial capitalism is the “highest” form’ of capitalism. It is incapable of explaining the complex combinations of production and circulation activity engaged in by capital under contemporary capitalism.¹ Therefore, while an understanding of agrarian capital is important to assess trends in rural non-agricultural production and employment, the terms of the agrarian transition debate are not sufficient in aiding this understanding in the current historical moment.

One aspect of Bernstein’s argument, however, continues to be relevant for this purpose: that of taking the non-agricultural sector seriously in any account of agrarian capital. This is particularly significant in the twenty-first century when the logic of capital has come to dominate global and national economies. The development of agrarian capital may not contribute to transforming pre-capitalist sectors, but it will have to contend and compete with capitals in an existing capitalist non-agricultural sector. Much of the discussion on agrarian transition still works within the framework of emergent capitalism, but that is hardly applicable to countries like India.

Agrarian capital beyond the farm in India

For India, a country where agrarian capitalism has been consolidated in many regions for a few decades now, we may ask: what can the diversification by capitalist farmers outside the farm tell us about the expansion of agrarian capital and trends in rural development?

Some of the earliest scholarly discussion relevant to this question started in the 1980s when India experienced a marked increase in rural non-farm employment. But the relationship between agricultural production and off-farm employment in India remains complex and contested. Visaria (1995) noted an increase in non-agricultural workers from 14.4% of the rural workforce in 1972–1973 to 21.7% in 1987–1988. Some scholars suggested that the increase in rural non-farm employment represented structural transformation, something anticipated by the capitalist transformation prompted by the Green Revolution of the late 1960s. Vaidyanathan (1986) argued that the increase in rural non-farm employment was explained by a higher incidence of wage labour and agricultural commercialisation. Similarly, Unni (1991) argued that high levels of male non-farm employment correlated positively with higher agricultural productivity, rural incomes and extent of non-food crop cultivation. Chandrasekhar (1993) disputed this, arguing that there can be no monotonic relationship between agricultural change and rural non-farm employment in a country like India. An increase in the latter could be a function of either positive linkages through the flow of capital and generation of rural demand or the inability of backward agriculture to support people adequately.

Through their analysis of the nature of urbanism in a southern Indian town (Arni), Harriss and Harriss (1984) also suggested that the links between agriculture and non-farm employment were contingent and weak. Agricultural expansion in Arni’s hinterland did not generate local non-farm employment and investments; it directed capital flows to metropolitan centres. These tenuous links between agriculture’s contribution to non-agricultural development have been even more firmly established by Chandrasekhar (2007) and Lerche (2013).

In addition to mapping such inter-sectoral links, studies focussed on economic diversification (employment and enterprise) identified differential benefits for better-off households (variously defined as elite, landed, rich, capitalist etc) compared to petty producers and labouring classes. For example, Himanshu, Murgai, and Stern (2016) argued that access to non-farm employment had been differentiated in Palanpur village since the 1970s: better-networked households were able to access regular employment. Over time, as such employment declined and casual wage work and self-employment increased, these households invested in training for small enterprises such as marble-polishing and machine-repair shops, while the poorest households could only access poorly paid non-farm employment. Studying three villages in Tamil Nadu, Colatei and Harriss-White (2004) showed that elite households were more diversified in terms of production, even hiring in non-agricultural labour. The poorest were ‘net sellers of labour’ (139) while the middle peasants also performed non-agricultural labour to a large extent. Ranganathan (2013), in his analysis of National Sample Survey Office (NSSO) data, showed that nationally, while wages constituted 63% of the total income of the smallest landholders, they only accounted for 3% of the largest landholders’ income. The income share of the smallest landholders from non-farm business was also higher, at 10%, compared to 4% for the largest landholders. But in absolute terms, the average non-farm business income of the latter was nearly four times that of the former.

This evidence suggests that despite weak links between agriculture and industry, agrarian capitalists were using capital to expand their portfolio in different ways. Given the considerable overlaps in the caste-class hierarchy of rural India, it is not surprising that many of these studies also find successful non-agricultural diversification to be linked to the households’ dominant caste position. Harriss-White (2003) labels caste, religious, kinship and political networks as ‘social structures of accumulation’ in India’s largely informal economy.

Yet the motivation for such diversification has been debated. Djurfeldt et al. (2008) suggest that in Tamil Nadu, the old landed elite left farming due to the difficulties of controlling labour and rent and because of more attractive opportunities in the non-agricultural sector. In contrast, Colatei and Harriss-White (2004) and Upadhya (1988a, 1988b) have argued that diversification by such households was a risk-minimising strategy. Village studies in Andhra Pradesh and Karnataka have argued the same while adding that for the richest capitalists, diversification also served to create new assets and networks through which they could assert their power (Ramachandran, Rawal, and Swaminathan 2010; Swaminathan and Das 2017). Ranganathan (2013) presents yet another view in arguing that the returns to investment in non-agricultural businesses were so low that diversification, even by larger farmers, must be seen as distress-driven. The motivations of capitalist farmers in diversifying are, therefore, not straightforward and may reflect the perceived relative profitability across different sectors.

Despite this rich tapestry of insights into diversification by capitalist farmers, rarely do these studies consider the links between such differentiated patterns of diversification and capitalist transformation outside agriculture. Djurfeldt et al. (2008) and Harriss-White and Janakarajan (1997) are exceptions, having shown that it was the non-farm sector that lent dynamism to the agricultural sector in Tamil Nadu by drawing capital and labour, although the latter also argued that no classic industrial transition was in sight. Moreover, there are hardly any studies that examine the entire spectrum of contemporary non-agricultural diversification undertaken by capitalist farmers in India, even though there are those that do this historically (Damodaran 2008) and those that examine such diversification across specific areas like agricultural trade and industry (eg Harriss-White 2008, 2016). But if we are to persist in our ‘investigation into the actual processes of agrarian change that clearly are taking place in India’ (Lerche 2013, 400, italics original), then a more thorough examination of this phenomenon is imperative. The next section contributes to this agenda through a focus on capitalist farmers in Punjab, with particular attention to the trajectory of both its agricultural and non-agricultural sectors.

The case of Punjab

The Green Revolution in the late 1960s brought fundamental changes to social relations in Punjab’s countryside, the most significant from our point of view being the emergence of capitalist farmers who reinvested their surplus into agriculture for expanded reproduction (Frankel 1971).² This raised expectations for a transition to industrial development through surplus capital and labour flows (see Chadha 1986). A modern capitalist class of farmers was expected to progressively buy up land from less competitive households, pushing such households towards proletarianisation and urban migration. At the same time, agricultural modernisation was expected to energise upstream and downstream agro-enterprises.

Green Revolution technology – characterised by high-yielding seed varieties that needed high levels of irrigation, fertiliser and crop chemical application – certainly generated demand for different kinds of agricultural inputs, leading to the emergence of agro-engineering and implement units. Growth in rural incomes led to a surge in consumer goods industries. Punjab also housed some of the country’s leading textile and hosiery, cycle and sports, steel, machine tools, and liquor and malt beverage industries (Damodaran 2008; Jain 2016). Alongside this, there was a considerable increase in rural employment in manufacturing through the 1970s and 1980s (Chandrasekhar 1993; Visaria 1995). In 1993–1994, Punjab was one of only four states (theothers being Haryana, Kerala and West Bengal) where non-farm employment was the principal work done by a majority of the rural and urban workforce (Bhalla 1999). This was partly a consequence of agriculture’s declining labour-absorptive capacity and partly due to the emergence of new rural manufacturing units (Chandrasekhar 1993).

Despite these gains, Punjab’s leading status in agriculture did not translate to a leadership position in industry. In 2009–2010, the level of manufacturing employment in Punjab was only just above the national average (Lerche 2014). Jain (2016, 265) has argued that after 2000–2001, ‘output growth has been negative in almost all the industries’. He also pointed to the relatively much better industrial performance of its neighbouring states, Himachal Pradesh and Haryana, over the same period, and the flight of industry from Punjab to them.

Partly, this is seen as a consequence of government neglect. A rural bias in Punjab’s politics since Independence, Chadha (1986) argued, led to resources being channelled disproportionately towards agriculture. Pritam Singh (2009, 2016), however, blamed the centrist bias in India’s federalism for Punjab’s lacklustre industrial development. As food self-sufficiency became a crucial pillar of planned development and therefore of economic and political independence and stability, Punjab’s agricultural strengths were harvested at the expense of its industry. The decline of planning in recent decades has led states to compete with each other for industrial investment, also disadvantaging ‘industrially backward states such as Punjab’ (2016, 385).³

A more socially rooted explanation comes from Murali (2017). She argued that the ‘wide-capitalist coalitions and competitive populism’ (151) that inform Punjab’s political fabric are responsible for its weak industrial development. Both the major parties/party coalitions in the state – namely, the Indian National Congress party and the Bharatiya Janta Party–Shiromani Akali Dal coalition – rely on a spectrum of voters to gain power. While this includes entrepreneurial castes, their numerical strength pales in comparison to that of the agrarian castes, forcing governments to make populist decisions in favour of agriculture and at the expense of industry.

The logical next question, then, is why agrarian castes have not developed industrial interests in Punjab as in some other states, an issue that is often overlooked in the literature on non-farm employment and sectoral shifts in the state. The most credible existing explanation about this comes from Damodaran (2008). He shows that capitalist farmers, predominantly of the Jat caste, could never control leading industries in Punjab since trade and industry were monopolised by the mercantile castes, the Hindu Aroras, Banias and Khatris, collectively known as the Mahajans, leading to insurmountable entry barriers for the farmers. This was in contrast with the experience of dominant agrarian castes of other states, for example the Kammas in coastal Andhra Pradesh and the Patidars in western India. Damodaran did point to new industrial investments by Jat farmers but added caveats about their scale and sustainability.

Therefore, a combination of socio-economic and political factors restricted the transformation of agrarian capital to industrial capital in Punjab. However, if economic diversification is an ‘objective necessity for the rural capitalist’ (Damodaran 2008, 274) when he has exhausted possibilities of agrarian expansion, and industrial investment is limited, a different approach to studying agrarian capital is warranted. Taking a cue from Upadhya (2009)’s exhortation to stop privileging industrial capitalism in studying the trajectories of capital’s evolution, this calls for exploring the full range of avenues towards which agrarian capital is drawn. In other words, it demands a study of where ordinary Jat farmers invest, why, their experiences of diversification and its consequences for the region’s development.

This paper draws on data collected in 2014–2015 for a study on agrarian accumulation in Punjab, India. Undertaken over four villages around a single market town, Khanna, in Ludhiana district, patterns of economic diversification formed one axis of investigation. The study includes a household survey of 93 agricultural households across these villages. The sample was selected purposively to over-represent large capitalist farmers (56 out of 93), who constitute the focus of this paper. The research was not designed as a village study nor was it designed to decisively identify existing classes (done far more competently by others, most recently and notably by the Project on Agrarian Relations in India (PARI)). As discussed elsewhere (Sinha 2020a), the class categorisation was done as a heuristic device through which the functioning of agrarian capital may be represented. The survey took, ex ante, operational holdings of 2 hectares or 5 acres or above to represent capitalist farmers and those below as petty commodity producers. At the same time, the survey itself recorded not just details of land, but also the extent and patterns of labour employment and capital intensification (using investment in fixed capital as a proxy). The research found that capitalist farmers were more likely than petty producers to hire attached labour and invest in farm machinery and land. Within the class of capitalist farmers, those with 4 hectares or 10 acres or above were distinguished as large-capitalist farmers since they were more likely to have a secure resource base to accumulate and to be able to overcome losses.⁴

The study also involved detailed semi-structured interviews with some farmers included in the survey and some outside the survey, identified purposively through snowballing. In addition, I conducted interviews with different kinds of agricultural traders in and around Khanna. I draw on these to demonstrate experiences of different kinds of diversification and to support my overall analysis. Insights from a brief revisit in August 2018 are also included. I must also note that the data were not collected longitudinally and, therefore, I am unable to draw out causal relationships between different kinds of diversification over time. However, the survey recorded in which decade a particular form of diversification took place for each household, and the paper makes tentative arguments from this information.

The primarily qualitative data presented here aims to identify broad patterns of investment and document households’ strategies and experiences. This approach, while ill suited to generalisation, presents a rich landscape of diversification strategies and experiences that allows us to identify tendencies and move the conversation on agrarian questions of capital in India forward.

Patterns of diversification

Overview of the data

Table 1 shows the distribution of large capitalist farmers from the survey across four types of diversification strategies: agriculture-based business, non-agricultural business, education-based diversification and international migration. Households with more than one diversification strategy have been counted twice, while households with multiple members engaged in the same type of strategy have been counted only once. Therefore, the total of these figures is not meaningful. Eliminating duplication, 32 out of 59 large capitalist households, or 54%, had attempted some form of diversification.

Table 1. Types of diversification.

	No. of large capitalist households
Agriculture-based business	9
Non-agricultural business	11
Education or skills-based diversification	11
International migration	9
Total sample	59

Source: Own household survey, 2014–2015.

These figures include both diversification efforts current at the time of the survey and those attempted by a household member in the past. Thus, they span a long period – from the 1960s to the 2010s – and the subsequent sections signpost timelines for different kinds of diversification. Here, it is worth noting that the earliest diversification was in civil or military government jobs, with some agriculture-based businesses starting in the 1980s. All other businesses and private employment were reported as having started in subsequent decades. The evidence does not suggest that one form of diversification necessarily led to another. In many cases, military or civil employment of one generation did lead to diversification by the next generation. But this did not appear to lead uniformly to government jobs or lucrative businesses or migration, and indeed some only diversified into petty businesses.

Several respondents claimed that economic diversification (‘side business’) was necessary due to limited or declining returns from agriculture. But even such side businesses may be successful only when the investments are large, which only some capitalist farmers could manage. To be sure, it is not that capitalist farmers in Punjab do not profit from agriculture. Kannan (2015) argues that Punjab was the only state where farm incomes increased through the 2000s. However, profits through agriculture have become both limited and riskier in recent decades (Damodaran 2008; S. S. Gill 2016; Sinha 2020a). Diversification is, then, a function not simply of distress per se but of the frustration of capitalist farmers regarding expanding accumulation. Similar limits to accumulation within agriculture have been shown to be the main incentive to diversify in other contexts too, eg in Andhra Pradesh (Upadhya 1988a) and Tamil Nadu (Harriss-White and Janakarajan 1997) and more recently by the PARI village surveys in Andhra Pradesh and Karnataka (Ramachandran, Rawal, and Swaminathan 2010; Swaminathan and Das 2017).

Patriarchal kinship relations structured patterns of diversification significantly as well (also demonstrated by other studies, eg Rutten 1995 and Djurfeldt et al. 2008). As diversification strategies are almost always undertaken by the male household members, typically only joint-family arrangements were found to enable multiple strategies within a household.⁵ In some cases, businesses were initiated in partnership with a relative or another known person. Not only did these strategies allow households to absorb risks of diversification, they also provided an avenue to engage younger male members unwilling to do farming. These kinship engagements are signposted below through reference to sons, brothers, etc. Meanwhile, patriarchal norms deterred women from either operating businesses or doing ‘unsafe’ jobs.⁶ Daughters often had college education but were more likely to work, if at all, after marriage and as part of their husband’s household.

A dozen large capitalists in the survey also occupied positions of political power locally, but not at a higher level, either at the time of the survey or before: as members of village panchayats, cooperative societies or the village gurdwara management. In this, they were also members of either Congress or Shiromani Akali Dal. These positions can be reasonably expected to have enabled these farmers to benefit disproportionately from state programmes and political networks but, as in the case of intergenerational effects of diversification, the impact appeared to be uneven. Many did diversify into other businesses, but while some of these included government jobs and transport businesses, others could only foray into petty businesses. Examining the impact of these political positions on accumulation beyond this cursory level remains outside the scope of this research.

Having noted the limits of the field data and mapped broadly the key features of diversification by large capitalist farmers in the study villages, I examine each type of diversification in some depth below.

Agriculture-based business

Agriculture-based businesses include off-farm businesses that are linked to agricultural production and require substantial investment. The most common is the commission agent (arhtia) business. Arhtias are licensed mediators of sale of farmers’ produce to any public or private purchaser in the regulated wholesale market or mandi. They are also the chief source of informal credit for the farmers and are seen as highly exploitative due to their control of interlocked product and credit markets (Gill 2004; Sinha 2020b).

Historically, the arhtia business was dominated by the mercantile castes or Mahajans across Punjab, but Jat farmers have been known to be diversifying into this field since the 1980s. In 2005, 105 out of 290 arhtias in the Khanna grain mandi were Jats (Damodaran 2008). In my survey sample, there were three households that had diversified into this, in the 1980s, 1990s and 2000s, respectively. In 2014–2015, the head of the Arhtia Association of Khanna was also Jat. Mahajans deeply resented the breach of their monopoly. Jat arhtias, however, argued that, as farmers who had to sell their produce in these mandis regularly, it was only natural that they would establish themselves in this business. They claimed that Mahajans resented them since they could tap into their kinship and community networks to attract clients. Moreover, by channelling agricultural loans to the business, Jat farmers were also able to generate a corpus to lend money to farmers.

While farmers have become arhtias in the grain mandi, this has not happened to the same extent in the fruit and vegetable mandi. This was because, firstly, fruit and vegetable trade had no state support or guaranteed wholesale prices as in the case of paddy and wheat, making it a riskier business to invest in. More significantly, due to the absence of state support, this business relied much more on business networks in larger mandis in other parts of the country. Only one Jat arhtia had newly ventured into the fruit business in Khanna mandi at the time of this study. The handful of Jat arhtias here were those who also grew vegetables and, therefore, had some links with traders. Outside mandis, a few extremely well-off potato-growing Jat households also invested in potato trade by leveraging these links (one in my household survey sample started this in 2011). They worked on a commission of 1% to gather potato supplies from villages for traders in larger cities like Delhi and Agra.

Retail shops for agricultural inputs in villages and market towns were another type of agriculture-related business where Jat farmers had invested. Building relations with farmers is crucial in sustaining this business and therefore, unsurprisingly, some arhtias also ventured into it. But in one case in my survey sample, the shop was also opened with the motivation of creating a business for a young male member unwilling to farm.⁷

Dairy farming is yet another such non-farm agricultural business. To be clear, all agricultural households have some milch animals, often buffaloes. The milk is used mainly for home consumption; any surplus is sold to private or government dairies for supplementary income. But dairy farming may also be a full-fledged business enterprise with 15 or more animals. Depending on the quality and species, a single animal may cost Rs 20,000–100,000. Though many have suggested dairy farming as a diversification strategy with great potential, farmers’ experiences were not unequivocally positive. One respondent had set up a dairy farm with 15 buffaloes in Khanna city in 2000. He incurred losses due to high feed and labour costs and inadequate returns. Other farmers argued that dairy can be successful if household labour is used, but several households do not have such labour or members are unwilling to do this work. More recently, studies have shown how dairy farming is under stress (for farmers) across India (eg Srikrupa et al. 2016), and any possibility of Punjab’s farmers increasing investment into this appears bleak.

Starting in the 1990s, some survey respondents also invested in machinery such as combine-harvesters that they rent out to other farmers when not in personal use. Combines have become indispensable for harvesting most of the wheat and paddy in Punjab. Farmers send their combines to other farms and villages within and outside Punjab. One respondent had two combines permanently lodged in a village in Uttar Pradesh. However, Punjab is now overpopulated with combines (see Bera 2015). Renting them out, therefore, also yields lower rents than before. In 2018, one respondent reported that he had invested instead in eight square hay balers meant for collecting paddy straw after harvest, several hundred thousand rupees. The burning of paddy straw has emerged as a major environmental issue in recent years (Haq 2018) and has been banned by the Punjab government. Large capitalists such as the aforementioned respondent now rent out the balers to smaller farmers at high rates as an additional income source (see Singh 2018).

There are some other major agriculture-related businesses that capitalist farmers in the field diversify into. While not reflected in the household survey sample, it is worth noting trends in them. Rice mills presented, for example, an important agro-industry that farmers invested in. In 2014–2015, 11 out of 55 rice mills in Khanna were Jat owned. In Punjab, paddy is almost entirely procured by the state. Privately owned rice mills then process this paddy against a commission but can use the by-products privately. Therefore, there are no reccurring expenses in the purchase of raw materials that are guaranteed by the state. However, interviews with arhtias and mill owners revealed that setting up a rice mill requires an initial investment of at least Rs 15–20 million, a scale of investment accessible to very few capitalist farmers. Investment in other kinds of agro-industries was uncommon. Flour mills, feed mills and solvex plants⁸ in Khanna hardly had any Jat presence. For example, only one of eight flour mills was owned by a Jat, but that too was leased to another trader.

Another kind of agro-industrial business into which a handful of potato-cultivating farmers had diversified is potato cold stores. This trend was started by a large Jat household with five brothers (in a non-survey village). Pioneers in potato cultivation through contract farming in the area, three went on to open cold stores in the 2000s. Some other farmers followed their lead. Investments for these stores were generated from their own surplus but also using government subsidies. To recover investment costs, store owners leased out a part of their store to corporate firms dealing in seed or processing potato varieties. The lease rates given by the firms were considered sub-optimal but assured the store of a stable income.

Non-agricultural business

Non-agricultural business is a key area through which we can gauge the movement of agrarian capital to other sectors. One such business is transport, where Punjab’s Jat farmers began diversifying in the 1980s as business owners and/or as drivers (eg Kaur 1986). One farmer I interviewed outside the survey, now in his sixties, learnt to drive a truck with the help of relatives in Madhya Pradesh who had a truck business. He spent much of the 1980s driving across north and north-east India. All three cases in my survey sample, though, diversified into this area in the 2000s.

However, this business has lost purchase in recent years, again due to low returns. One respondent was in the business for 10 years but quit. He said that, after accounting for the cost of diesel, toll taxes and maintenance, he was left with a profit of only Rs 10,000–20,000 from a single trip, which, according to him, was inadequate. ‘If I saved Rs 50,000, then it would be different’, he added, an indication of how farmers assess the profitability of agriculture against other options (more on this later).

Two farmers in the survey also invested in tractor dealerships, with one doing this in partnership with other relatives or friends. According to one who has been in the business since 2006, however, even the tractor dealership is not as profitable as before. He said that the business model has changed from sale on commission to purchasing tractors from the company and selling at a profit; this reduced his profit margins, and the fact that he no longer had any employees at his shop was testament to this.

Two respondents claimed to work as property dealers since the 2000s, but were not forthcoming about details, given that much of this business is informal (in ‘black’). However, as land prices have decreased in recent years (see Nandy 2015), such businesses have also been adversely impacted. Other kinds of major non-farm investments in the area included private school and colleges and venues for marriage and social events, popularly known as ‘marriage palaces’, but none of these featured in the survey.

Apart from these large businesses, three large capitalist farmers were also found to own petty businesses such as grocery and garment stores since the 2000s. These businesses were a source of supplementary income, but also served as an employment opportunity for young sons averse to agriculture. Overall, non-agricultural investments by Jat farmers were fairly limited in the area, even more so in terms of manufacturing industry.

Education-based diversification

Capitalist farmers’ investment in the education of their children has been noted by several studies in India (eg Jeffrey, Jeffery, and Jeffery 2008; Heyer 2016). The study villages in Punjab were no exception. Education-based diversification in the form of civil or military government jobs were the earliest, starting in the 1960s, but no military jobs were recorded after the 1980s and only two civil government jobs in the 2000s. Private salaried employment was reported only from the 2010s.

Overall, military service accounted for most of the salaried employment. This is unsurprising since the Punjabi region has a long and continuous history of military service dating back to the colonial period. All but two of my respondents who reported experience of military service were at the level of the foot soldier, the minimum qualification for which is the completion of secondary education. These households benefitted from a modest, stable salary and, after retirement, a pension. Only three households had a member in non-military government service, in the form of local revenue and records officers or government school teachers. Several households also had members with jobs that were considered reasonably respectable in private establishments, for example as accountants or supervisors in factories, or teachers in private schools or colleges (both of the latter were daughters-in-law of the farmer respondent). None of the large capitalist households had members who were hired for private labouring jobs such as factory workers or drivers.

Income from the aforementioned jobs could provide some upward socio-economic mobility for the household across generations. So, for instance, households where some older members held such jobs also had sons and nephews who were in salaried employment. But this did not transform them into agents of productive capital outside agriculture. Moreover, hardly anyone, especially the youth, was doing jobs commensurate with their qualifications – a matter of serious consternation for the households given that 61% among them had at least one household member with a graduate and/or postgraduate qualification.

To a large extent, this reflects the wider problem of jobless growth under liberalisation (Raveendran and Kannan 2009). In other parts of the country, this has resulted in political mobilisation by dominant agrarian castes for reservations in government jobs (Deshpande 2017). This has not happened in Punjab. Instead, youth has become prone to ‘timepass’ (Jeffrey 2010), drugs and, last but not least, trying to migrate abroad, to which we now turn our attention.

International migration

In the absence of remunerative diversification options within Punjab, the Jat capitalist farmer is investing heavily in migrating abroad. To be sure, international migration from Punjab has been ongoing since the colonial period, especially from the north-eastern Doaba region (Thandi 2006). In the study area, the push for migration started during the political crisis of the 1980s (see Jodhka 2001). However, in the survey, this was reported only from the 2000s onwards. This push is accompanied by a fetishisation of the developed countries of the ‘West’, namely the UK, the USA, Canada and even Australia and New Zealand.

Nine households had a member who was abroad at the time of the survey or had been abroad earlier. The migrant members did a range of different jobs: labouring jobs in factories and farms in Italy, driving in the Middle East, working in businesses like tractor stores and grocery stores in the USA, working as building contractors in the UK. Yet migration had not been ‘successful’ in some cases, and the migrant member had returned. This included two men who returned from labouring jobs in Italy, another man who had worked as a driver in Dubai and a young man who was sent to Kenya but could not get a job there. The first three returned because the work they did was unacceptable to them; one said, ‘one is simply a servant of the owner (of the business)’. In fact, all but one of the four men who ‘failed’ in migration were attempting to migrate to a ‘better’ country or with the prospect of non-labouring work.

At the same time as immigration through family links or for petty jobs has become more difficult in ‘Western’ countries, another route has emerged, one colloquially termed ‘study basis’. Young men and women from farming households take the International English Language Testing System (IELTS) exam. Based on their score, they can secure admission to a private university in one of these countries. Strikingly, it is often young women taking these exams. If they pass them satisfactorily, their husbands or prospective husbands plan to migrate with them. Studying nursing is another route through which young women hope to be able to migrate with their families.

The gendered nature of this process calls for further study. For our purpose, however, it is important to note that ‘trying to migrate’ is expensive. According to one respondent, the cost of migrating on a ‘study basis’ is at least 1.5 million rupees. A similar amount was paid, unsuccessfully, by at least two respondents for a visa to Canada. Farmers often take loans from arhtias to meet these requirements. One arhtia reported that in 2018 alone nine farmers had taken loans worth hundreds of thousands from him for migrating to Canada.

Thus, migration-related aspirations seem to be a major channel soaking up the region’s agrarian surplus. On questioning why despite making agricultural profits, they were trying to migrate abroad, young farmers expressed disillusionment with agriculture and the general state of affairs in Punjab. Prima facie, there does not appear to be substantial reinvestment into productive development of the region through remittances from abroad.⁹ Although respondents were reluctant to discuss remittances, some claimed to use them for building imposing houses while two explicitly stated that they did not receive any. One farmer actively discouraged his two sons in the US from investing in property in Punjab since he felt the risks were too high.

Failures matter

An underlying current in the stories of diversification is the failure of households in sustaining their investments outside farming – dairy, transport, migration and tractor dealerships.¹⁰

In the literature on economic diversification, structural change and questions of capital, there is hardly any discussion of such failures. To be fair, Djurfeldt et al. (2008) and Upadhya (1988a; 1988b) explore how diversification matters both to the people who leave agriculture and those who remain. However, even they do not account for how consistent setbacks could have wider implications for processes of agrarian change.

In the foregoing sections, we have shown how non-farm economic diversification may be unreliable. Some avenues like petty businesses or leasing out of machines may allow for supplementary income, but it is extremely difficult to move out of agriculture on the back of a different economic activity that is not as remunerative as agriculture. This point needs consideration.

For instance, in each failed attempt at diversification discussed above, the farmers turned their entire attention back to agriculture, often expanding their scale of production through leasing of land. These farmers considered agriculture, grudgingly, a potentially more secure form of remuneration than other activities. Diversifying capitalist households thus retain their landholdings and they are reluctant and/or unable to exit farming altogether. Moreover, continuing to farm may even help an agriculture-related business. For example, agricultural credit can be used to finance the commission agent business, marketing links from cultivating vegetables can be used to trade in the same commodity,¹¹ and rent may be earned from leasing out expensive farm machinery. This demonstrates that diversifying or remaining in agriculture involves continuous decision-making by farmers, and to draw a binary between whether diversification was prompted by push or by pull factors is somewhat misleading. The decisions are, of course, constrained by the diversification options available to them. Most capitalist farmers hope to exit agriculture in the medium term, most commonly through their children whose education they invest in, through migration, or both. Yet this is a choice compared to smaller farmers who are forced to leave agriculture due to their inability to meet subsistence needs through agriculture.

The need to continue farming – either because they fail at or because it benefits a non-farm business – has implications for processes of accumulation within agriculture as well. Djurfeldt et al. (2008) argued that when the older elite left agriculture in their study villages in Tamil Nadu, the middle rung consolidated their position. In these villages in Punjab, on the other hand, the inertia of agrarian capital adds to the strains on expanding surplus through agriculture. It does this by keeping up the pressure on land and driving up land-lease rates that then cut into profits. This, in turn, limits the base of accumulation. The difficulty in expanding either within or outside agriculture in the region thus explains the rush to emigrate abroad.

Conclusion

Agrarian questions of capital have been increasingly marginalised in recent debates on agrarian transitions in the Global South. In this paper I have argued that this is justified at the global scale and in some national cases like India because of the predominance of non-agrarian capital in the capitalist development of the industrial and service sectors. But more importantly, there is a case to look beyond the terms of this debate itself in those regions where agrarian capitalism has consolidated itself. Agrarian capital – its nature, development and failures – needs to be continuously studied on its own terms to understand contemporary processes of agrarian change and rural development in specific contexts.

This paper suggests mapping the non-farm investments and activities of capitalist farmers as one way of studying agrarian capital. Doing this in the case of Punjab, India, the research finds that large capitalist farmers, constituted by the agrarian Jat caste, did not succeed in transforming themselves into industrial capitalists. Existing scholarship has explained this ‘failed’ industrial transformation of Jat farmers through their inability to compete with mercantile castes. However, one cannot help but wonder whether the strained, even declining, industrial sector of Punjab is also to blame, inasmuch as it lacks the dynamism to support any new entrants.

More significant than the fact that diversification is not converting farmers to industrialists per se is that the range of activities that farmers are investing in is not substantively generative for the economy in terms of driving either new production or employment. Business in the form of agriculture-based trading, and to some extent rice mills and cold stores, has been their only success. There appears to have been a period when the transport business and dealerships in tractors were lucrative, but this is no longer the case. Some other petty businesses provide only supplementary income, while well-paying salaried employment opportunities are scant.

The limited success in non-farm economic avenues tends to push farmers into bringing capital back into agriculture, further straining processes of agrarian accumulation. This also implies that there is something akin to a circular movement of agrarian capital across the agricultural and non-agriculture sectors, as capital is unable to expand within the region or the nation. Through circuits of international migration, this capital is now seeking global avenues. Here too, success is checkered and there is no substantial evidence of remittances being reinvested into productive development within Punjab. Punjab’s agrarian capital, much like its overall economy, thus appears to be at an impasse, one that may have significant political implications if left unresolved. Explorations of similar issues and trends in non-farm diversification in other parts of India and the Global South could open up new ways of understanding agrarian capital and accumulation.

Acknowledgements

I am grateful to the four anonymous reviewers and the editor for their engaged and constructive feedback. I would also like to thank Shray Mehta and Neha Margosa for comments on earlier drafts. The usual disclaimers apply.

Disclosure statement

There is no conflict of interest.

Additional information

Notes on contributors

Shreya Sinha

Shreya Sinha is a postdoctoral Research Associate in the Department of Geography, University of Cambridge. She works on questions of agrarian political economy, political ecology and regional development in India. She is currently involved in a GCRF-funded research project, TIGR2ESS, on the political economy challenges of building sustainable agricultural and food systems in India across different agro-ecological regions. She completed her PhD in development studies development studies at SOAS University of London. She is also the Reviews Editor for the Journal of Agrarian Change.

Notes

1 Banaji’s (2020) work suggests that this has been a feature of capitalism historically too.

2 Other changes included the emergence of wage labour and increasing class differentiation (Frankel 1971).

3 Some also point to the political considerations of the party in power at the centre (see Rajshekhar 2015; Jain 2016).

4 An operational holding of 10 acres was also the marker for a ‘tagda zimidar’ (strong farmer) in the field.

5 The continuous shift towards nuclear families in rural Punjab (Wasal and Singh 2018) is likely to constrain this possibility further.

6 Raju (2015) has shown that Punjab has one of the lowest rural female work-participation rates in the country.

7 Aga (2018), in his study of Maratha input retailers in Maharashtra, argued that gains from the business were too unstable to allow consolidation.

8 Solvex plants are mills that extract oil from grains and their by-products.

9 This is supported by other studies on international migrants from the Doaba region of Indian Punjab, which argue that remittances were largely invested in displays of consumerism (eg Taylor et al. 2007).

10 Six survey respondents reported such failure across different types of diversification, while two reported consistently declining profits in their tractor dealership and transport business, respectively.

11 Rutten (1995) described how a grasp of such links was crucial to the transformation of potato farmers into rural entrepreneurs in Gujarat.