In this article, I summarise the basic features and specific components of land distribution and forms of tenancy and mortgage in two villages in Bihar in 2011–12, with a few additional observations from 2018. The article is arranged as follows. The first section introduces the study villages. The second section describes the patterns of land use and land ownership in the villages. The third section discusses the inter-relationship of land and caste structures, and examines the pattern of land holding (owned and operational) based on land-size and caste categories. The fourth section focuses on forms of tenancy and mortgage relations, and the fifth section analyses the persisting inequality in the two villages. The final section makes some concluding observations on the similarities and differences between the two contexts: one in which the older structure of near-total monopoly of the Bhumihars over land and, consequently, over other spheres of life and work persists; and the other where there is a new caste, the Yadavs, gradually taking over land from the erstwhile landed caste, the Bhumihars.

Introduction to the Study Villages

Bihar is one of the most backward regions of India in terms of industrialisation and urbanisation. Historically, the State was divided into two separate socio-cultural regions, north Bihar and south Bihar. After Jharkhand was carved out of the State, the region north of the Ganges is referred to as north Bihar and the region below the Ganges is called south Bihar. The study villages, Katkuian and Nayanagar, are both located in north Bihar: in Bagaha II block of West Champaran district and Hasanpur block of Samastipur district respectively. Both villages fall under Zone I (North-West Alluvial Plain) of the agroclimatic classification (Shankar et al. 2009).

The town (and railway station) nearest Katkuian is Bagaha II, which is about 12 km away. There is no bus service, government or private, to and from the village. The only available mode of transport is a private shared autorickshaw, available every half an hour or more. The nearest post office is in Tarwalia, and the nearest bazaar town, Semra, is two or three km away. Katkuian has no bank branch, but it does have an extension centre of the State Bank of India (SBI). According to the Census of India 2011, 98 per cent of the agricultural land in the village was irrigated through canal irrigation. The major sources of surface irrigation were the Triveni canal and its tributary, the Samarkand canal, part of the Gandak irrigation project. Apart from canal irrigation, there is the Harha, a seasonal stream and tributary of the Gandak river. The main sources of groundwater irrigation were borewells. The main crops cultivated in Katkuian at the time of the survey were sugarcane, paddy, wheat, and pulses. The village is not far from the Bagaha Sugar Factory, located in Bagaha II.

Nayanagar, by comparison, is better connected and more developed. The nearest town to the village is Rosera, which is eight km away. The village has its own railway station, and the nearest bus service is from Sihma Chowk (Begusarai district), which is about three km away. There are private shared autorickshaws and tongas that ply to and from the village. Nayanagar has a post office and a branch of the Bihar State Grameen Bank. According to the Census of India 2011, 89.5 per cent of agricultural land in the village was irrigated by groundwater from tubewells and borewells. The main crops were wheat, maize, sugarcane, and rapeseed/mustard. Nayanagar is prone to flooding, because of which a significant proportion of land is not cultivated in the kharif season. Horticulture, particularly the production of litchi and mango, was a very significant contributor to the village economy. The village is not far from the Hasanpur Sugar Mill in Samastipur.

In May–June 2012, the Project on Agrarian Relations in India (PARI) conducted by the Foundation for Agrarian Studies (FAS) undertook a census-type survey of the 350 households in Katkuian and a sample survey in Nayanagar of 348 households. The data presented and analysed in this article are from these surveys. In June 2018, the FAS team went back to the study villages to conduct small sample surveys and qualitative interviews focused on certain themes, such as land and tenancy, forms of labour, and changes in cropping pattern. All the interviews and case studies included here were conducted in June 2018. There was no fresh house-listing in the 2018 re-survey round. Instead, a purposive sampling method was used to select sample households from the households surveyed in 2012. The 2018 round covered 44 households in Katkuian and 64 households in Nayanagar.

In what follows, I present some basic facts and observations about demography and socio-economic relations in the two study villages.

Katkuian is a multi-caste village, now dominated by Yadav and Koeri castes. In Bihar, in general, the backward castes are divided into two categories: Extremely Backward Classes (EBCs), also referred to as Backward Classes-1 or BC-1, following the listing of caste names in Annexure 1 of the Mungerilal Commission report (see Kumar 2013, p. 17); and Backward Classes-2 or BC-2, including the relatively less backward caste groups (see Table 1).

In 2012, EBCs constituted the largest caste group in Katkuian. Out of a total of 350 households, 163 households or 47 per cent belonged to EBCs. The Mallah caste (also known as Sahni/Been/Bind/Nishad/Machhuara) alone accounted for 12 per cent of all households. The BC-2 castes within the category of Other Backward Classes (OBCs), mostly Yadavs, Koeris, and Kurmis, constituted 35 per cent of all households (121 households out of 350). The Yadavs were numerically dominant with 70 households (20 per cent of the total). There were 42 (12 per cent) Scheduled Caste households, including Chamar and Musahar. Chamar households constituted about 11 per cent of total households in the village. There were nine Scheduled Tribe households, including Gond and Lohar households, and 15 other households, covering Brahman, Bhumihar, and Sheikh families.

An important feature of Katkuian is its pattern of migration of labour. Since the 1960s, with the beginning of the Green Revolution, male workers from Katkuian have migrated to Tehang in Phillaur tehsil, Jalandhar district, Punjab. During the survey, 73 workers reported seasonal migration to Tehang. A small number of worker households have bought land and settled there, but most workers follow a circulatory pattern of migration – going to Tehang for a few months during the harvesting season and coming back to Katkuian once the season is over. Almost every household in the village has at least one member who has gone or still goes to Tehang for work. The younger generation, however, prefers migrating for non-agricultural work to cities such as Bengaluru, Mumbai, and Kochi.

There are differences between the populations covered by the Census of India 2011 and the FAS survey in 2012 (see Table 2). This is mainly because the FAS survey covered a settlement that, while in the physical area of the village, turned out to be within the revenue boundaries of the neighbouring village. It is the part of the village where a large part of the Dalit work force of Katkuian lives.

The Census of India 2011 covered 246 households in Katkuian, and the FAS survey covered 350 households. The most striking aspect of this difference between the Census of India and PARI data is in respect of the shares of Dalit and Adivasi households in the village population. As mentioned above, the difference can in part be explained by the fact that the village as studied in the PARI survey included a Dalit hamlet that was not part of the village in Census 2011.¹

The other study village, Nayanagar, is much bigger than Katkuian; for this reason, a sample survey was conducted in 2012 (see Table 3).

Nayanagar, in both 2012 and 2018, was a Bhumihar-dominated village. Bhumihars alone constituted 22 per cent of the village households. Bhumihars and Brahmans together made up one-fourth of the total households in the village. At the other end of the caste hierarchy, Paswans/Dusadhs constituted 21 per cent of the village households, while the share of Scheduled Caste households in total households was 34 per cent. The EBC caste group accounted for 31 per cent of total households, within which 11 per cent households belonged to the Mallah caste and 8 per cent belonged to the Dhanuk/Mahato castes. If we compare data pertaining to the demographic features of the village from the Census of India 2011 and the FAS survey 2012, again we notice a lot of differences (see Table 4).

While the Census of India 2011 reported 1,946 households in Nayanagar, the FAS survey of 2012 covered 1,205 households. The most notable aspect of the variance between the Census and the FAS survey is the considerable difference in the shares of Dalit and Adivasi households in the village population. While the Census of India 2011 reported that only 19.2 per cent of total households were Dalit households, the FAS survey found that 35 per cent of village households belonged to Dalit castes.²

Before presenting a detailed analysis of land relations, I present below some general facts about the social and economic life of the study villages, identifying some historical factors and contemporary features of the dominant households and castes in the villages. In the 1950s and till the 1980s, Katkuian was a typical north Bihar village showing a dominance of Bhumihar landlords prominent in its social and economic life. By 2012 and 2018, our survey years, the village was largely dominated by Yadavs and Koeris. At the same time, our survey data show that Katkuian had a high share of crop lands under absentee ownership.³ The most prominent household, headed by YPJ, the sarpanch of the village panchayat, was among those that had wrested village-level political power from the Bhumihars. The absentee landlords control their crop lands through managers. One of the managers, JN, a Maithil Brahman whose family had migrated from the Mithila region to Katkuian, managed a large parcel of about 73 acres. A few Koeri families controlled the Primary Agricultural Credit Society (PACS) and public distribution system (PDS) in the village.⁴ The president of the PACS, RK, headed a committee of 12 members. The procurement process through the PACS has become more complex in recent times,⁵ but it continues to be a source of social and political power in the villages of Bihar. Given the continuing concentration of land ownership, it is not surprising that in 2012, in terms of the socio-economic class structure of the village, more than 60 per cent of the population depended on manual work either in agriculture or non-agriculture.

Katkuian was in a state of flux as far as the nature of socio-political dominance in the village was concerned. As pointed out earlier, the Bhumihar landlords still controlled a considerable amount of land in the village. Further, and as we will see later, the inter- and intra-caste differentiation among the relatively powerful Backward Castes (BC-2), such as Yadavs and Koeris, was high. Except for a few households at the top of the scale, most Yadav and Koeri households lived in thatched houses with meagre amenities. Inter-caste relations in everyday social interaction, and even in land-leasing and tenancy, were much more open in Katkuian, despite and perhaps because Backward Castes constituted 81 per cent of all households in the village. Katkuian did not have a single dominant caste in the sense proposed by M. N. Srinivas (1959); there were at least two major castes, namely the Yadavs and the Koeris, who competed for political power (for a discussion on the idea of “major castes,” see Sahay 2009).

A very different mode of control was observed in Nayanagar, where a single household – the historically dominant landlord family – continued to be at the helm of the socio-economic and political structures of the village. This family alone owned 33 per cent (self-reported) of all cultivated land in the village – more than 350 acres in 2012. As pointed out earlier, the village had a substantial population of Dalits but almost all these households were landless, to the extent that they did not have even homestead rights. Bhumihars led by the big landlord family controlled politics and society, and new technology in agriculture, in the village. The levels of socio-economic inequality in Nayanagar were extremely high (see the article by Swaminathan and Nagbhushan in this issue). In almost a textbook case, the Dalits of the village, though constituting 34 per cent of all households, lived a life of almost total separation both physically and socially. The Bhumihars cultivated export-quality litchi and possessed a state-of-the-art technological support system for transporting the produce to Mumbai and then abroad. Many of the Bhumihar households in the village included highly educated individuals with a variety of professional achievements – we came across filmmakers, marine engineers, doctors, professors, and corporate professionals. At the other end of the social spectrum was the Dalit settlement where most of the households had no homestead rights. Some of them kept their important documents such as ration cards in the custody of the landlord household, often referred to as malik.⁶ The level of education and amenities in the Dalit settlement were very poor, more so in comparison with the Bhumihars. Many Dalit children worked in litchi cultivation, hired for their “nimble fingers,” and often without being paid cash wages. In other words, social and economic inequality in Nayanagar was far more deep-rooted and structural than in Katkuian.⁷

The following sections delve more deeply into aspects of social and economic inequality through empirical data collected in 2012 and 2018.

Land Use and Land Ownership in the Study Villages

In 2011–12, the land-use pattern of the study villages displayed characteristics typical of regions almost completely dependent on agriculture (see Table 5).⁸ The total acreage was 809 acres and 2,142 acres respectively, in Katkuian and Nayanagar.

In Katkuian about 66 per cent and in Nayanagar about 82 per cent of all land was used for farming. Interestingly, while more than 12 per cent (266 acres) of all land was used for cultivation of litchi and other horticultural crops in Nayanagar, about 30 per cent of all land (240 acres) in Katkuian was “other types of land,” most of which was land outside the village owned by one landlord household.

In both the villages, almost all the crop land received irrigation from either public or private sources, and therefore the share of unirrigated cultivated land was negligible (see Table 6). Tubewells/borewells were the most important sources of irrigation in both the villages. Nayanagar was almost completely dependent on tubewells/borewells for irrigation, whereas in Katkuian canal irrigation was a significant additional source of irrigation. This is captured by the proportion of acreage receiving irrigation from multiple sources in Katkuian (32 per cent) as compared to Nayanagar (two per cent).⁹

As part of the survey, we collected data on landholdings – operational and owned. Table 7 gives basic descriptive statistics about landholdings in the two study villages. The average ownership holding per household was 2.3 acres in Katkuian and 1.8 acres in Nayanagar. The average operational holdings were slightly lower than ownership holdings in both the villages.

The largest ownership holding in Nayanagar was 377 acres, whereas it was 246 acres in Katkuian. The coefficient of variation for ownership holdings was 6.7 in both the villages, signifying high levels of inequality in access to land.

The most striking feature of the agrarian structures in both study villages was the extent of landlessness. In 2012, about 51 per cent of households in Katkuian and about 69 per cent in Nayanagar were landless (see Table 8). The extent of land inequality can also be understood through another indicator. In 2012, the top 5 per cent of households owned about 74 per cent and 72 per cent of total owned area, respectively, in Katkuian and Nayanagar. The bottom 50 per cent of families owned no land at all in both the villages.

In Katkuian, there were four families owning more than 25 acres of land each, and together they owned more than 56 per cent of all cultivable land. In Nayanagar, 13 families owned more than 25 acres of land each, and together they owned about 41 per cent of the total land owned in the village. The usual tendency towards fragmentation of landholding with a rise in population and division of families does not seem to be at work in both villages. Let us try to understand the distribution of landholdings according to the size of holdings. Table 8 shows that about 94 per cent of all households in Katkuian belonged to landless or small/marginal farmer categories in 2012. Similarly, in Nayanagar, about 93 per cent of all households possessed less than 5 acres of land in 2012. Both the study villages therefore displayed a very high concentration of landholdings, while more than 90 per cent of the households were either landless or belonged to small/marginal categories.

Land and Caste Structure in the Study Villages

I have shown in the previous section that the landholding structure in the two villages was highly skewed, with very few families owning a disproportionate share of the total land. These few families came from particular castes, and there was an intrinsic connection between agrarian class and caste.

Katkuian was dominated by castes referred to administratively as Other Backward Classes (OBC). The picture was different in Nayanagar, which was a more traditional Bhumihar-dominated village. The main features of the landholding and caste structure in Katkuian are shown in Tables 9a and 9b. Of the total 350 households, about 81 per cent – that is, 284 households – belonged to the OBC category, and these castes alone accounted for about 97 per cent of all the land owned by households in the village. Each of the landed OBC households, on an average, owned about 5 acres. About 15 per cent of all village households belonged to Scheduled Caste (SC) and Scheduled Tribe (ST) categories. However, these households together owned less than 1 per cent of land in the village. About 86 per cent of SC households and 57 per cent of ST households were landless.

There were only 15 “other” caste households in the village, of which seven were landless. Each of the remaining eight landed households in this category owned, on an average, about 2 acres. These households mainly belonged to the Bhumihar, Brahman, and Shaikh castes.

As is clear from Table 9a, about 46 per cent of the OBC households in Katkuian were landless. Table 9b shows the same figures for two distinct categories within the OBCs, i.e. EBC and BC-2. The village was dominated by Yadavs, Koeris, and Kurmis, all belonging to the BC-2 category. The EBC category was larger and more heterogeneous, as has been pointed out earlier. The divergence within the OBC caste category is apparent when we look at Table 9b closely. While EBCs formed about 47 per cent of all households, their share in total area owned was only 12 per cent. On the other hand, the BC-2 category controlled about 86 per cent of all land in the village even though only 35 per cent of the total number of households came from these castes. To put this in another way, a BC-2 category household, on an average, owned about 8 acres, whereas an EBC household only had about 1 acre of land. This differentiation within the OBC category in Katkuian stood in stark contrast with the land and caste structure of Nayanagar (see Table 10).

Table 10 shows that “other” caste households in Nayanagar (mostly Bhumihars) constituted only 25 per cent of all households, but they owned about 98 per cent of all cultivable land in the village. On an average, a Bhumihar household owned about 8 acres of land. The OBCs and SCs, even though they accounted for about 75 per cent of the village households, owned only about 2.5 per cent of land. About 86 per cent of OBC households and 92 per cent of SC households were landless. Even those who were landed within these categories had negligible amounts of land under their ownership and control.

The two villages demonstrated two different types of class–caste control. While both were numerically dominated by OBCs and SCs, or, in other words, the historically oppressed castes, Katkuian displayed a more caste-differentiated landholding pattern where OBC castes were dominant in terms of landholding and access to other resources. On the other hand, Nayanagar was an example of the longstanding and almost complete monopoly exercised by Bhumihars and other “upper castes” over land and, as we will see, other resources.

Another way to understand this difference within and across the two study villages is to look at the patterns of tenancy, and of land sales or purchases. These are the two possible pathways for the historically landless to access land. The relative opening up of the land market for castes with little historical claim over resources has allowed some of those households to enter into cultivation.

Forms and Patterns of Tenancy and Mortgage

The landless, mostly from the oppressed castes, have always sought to lease in land in a variety of ways, in order to secure access to land. Has this been the case in the two study villages? The extent and pattern of tenancy may provide more evidence on the nature of control exercised by the major castes and sections in the two villages (see Table 11).

As was expected from earlier analyses, the incidence of tenancy in Katkuian was higher than in Nayanagar. In Katkuian, more than one-third of the total operated area was under lease, whereas in Nayanagar the share of the same was about 14 per cent. Similarly, about 59 per cent of households with operational land holdings were tenants in Katkuian – of 298 identified tenancy contracts, 151 were on fixed rent contracts and 147 on share rent contracts – whereas their share in Nayanagar was 43 per cent. Table 12 reinforces the observation of deeper concentration of land in Nayanagar. About 85 per cent of all operated land was owned and self-operated in this village, whereas the same share for Katkuian was 60 per cent.

Distorted access to land in the study villages can also be gauged from the distribution of operational landholdings, as shown in Table 13. This table read in conjunction with Table 8 highlights the extent to which tenancy allowed access to land in the two villages.

Overall landlessness in terms of operational holdings came down to 43 per cent (as compared to 51 per cent in terms of ownership holdings) in Katkuian, and 62 per cent (as against 70 per cent in terms of ownership holdings) in Nayanagar. In both villages, the major lessors were the larger landholders (owning more than 25 acres). In Katkuian, these households owned more than 56 per cent but operated only 34 per cent of total land. Similarly, in Nayanagar, large landholders owned more than 41 per cent but operated only 27 per cent of total land. However, there was a key difference in how tenancy arrangements benefited other households between the two villages. In Katkuian, the medium and semi-medium landholders owned about 20 per cent but operated some 35 per cent of total land. In Nayanagar, the same figures were about 38 per cent and 45 per cent. The small and marginal landholders owned about 24 per cent but operated 32 per cent of total land in Katkuian, whereas the same figures for Nayanagar were 21 per cent and 27 per cent.

Looking at Table 13 alongside Tables 11 and 12, we can deduce that tenancy was more widely prevalent in Katkuian, with small cultivators as tenants, whereas in Nayanagar this form of tenancy was less common (see Table 14 and Table 15).

Table 14 gives the distribution of tenant households in Katkuian in terms of their caste status. There seems to be a pattern in the caste distribution of the modes of tenancy arrangements. First, the Scheduled Caste and Scheduled Tribe households were marginalized in the tenancy arrangements in the village. Scheduled Caste households as a whole leased in just about two per cent of the total land under sharecropping or rent-in-kind tenancy arrangement. Sixty per cent of all households under rent-free tenancy arrangements were BC-2 households. Another 20 per cent were Scheduled Caste households; most were long-term workers or, in other words, in a relationship of labour servitude. In return for leasing land from the dominant households, these long-term worker households were at the beck and call of the former. Secondly, and at the other extreme, even though the non-OBC, non-SC/ST households constituted a tiny minority in the village, one household dominated as far as rent-in-cash tenancy was concerned. This household leased in 90 per cent of the total area under a rent-in-cash arrangement, with the remaining 10 per cent under this arrangement being leased in by the dominant Yadav households. Thirdly, tenancy arrangements were seen to be determined by caste considerations, as a significant share of the land leased in and leased out was within non-EBC households, except for the rent-in-cash arrangement where a very small number of Brahman and Bhumihar households dominated.

In Nayanagar, too, there was a discernible pattern between caste status and tenancy arrangements. First, about 87 per cent of land leased in on cash rent tenancy in this village was by Bhumihar households, and almost 98 per cent of all land leased out was also by Bhumihar households. However, it is to be noted that out of 297 acres leased out in total, 196 acres were leased out by the largest landlord family. Secondly, about 34 per cent of land under sharecropping or rent-in-kind, and 75 per cent of land under rent-free tenancy, were leased in by SC households. Given the overall social structure of the village as outlined earlier, such arrangements turned out to be more exploitative because it was difficult to quantify the labour they had to put in, in return for the land. This often became a vicious cycle, leading to further exploitation of the already landless households. Thirdly, and broadly speaking, the OBC households mostly leased in land on sharecropping or rent-in-kind tenancy arrangements, but they also leased in some areas under rent-in-cash arrangements. Some members of OBC households also worked as long-term workers, signifying that the agrarian social structure in Nayanagar was quite different from that in Katkuian.

A further way to illustrate the difference between the two study villages is to look at the patterns of mortgage among classes and caste categories.

Some features of the system of mortgage prevalent in the study villages are outlined below. A system of mortgage in agriculture, generally speaking, is one where the borrower mortgages out his/her land in return for a sum of money from the lender who mortgages in the plot of land. Usually, the borrower is a small/marginal farmer or a landless worker, and the lender is a bigger landholder. However, as is clear from Table 16 and Table 17, this typical pattern of mortgaging in by large farmers and/or landlord households was absent in Katkuian, but strongly present in Nayanagar. This points to another significant process of change in Katkuian, to which I turn now.

It is interesting to note that about 24 per cent of all mortgaged land in Katkuian was mortgaged in by landless households, whereas the same figure was only eight per cent in Nayanagar. In fact, most of the mortgaged-in land (about 78 per cent of the total) was with landless, small, and marginal households in Katkuian. In caste terms as well, about 80 per cent of all the mortgaged-in land in Nayanagar was with the Bhumihars, whereas in Katkuian, the picture was quite different. The dominant OBC households (BC-2) had mortgaged in about 59 per cent of the total, but EBCs, SCs and STs too had mortgaged-in land in Katkuian.

The pattern of mortgaging out can be gauged from Table 18 and Table 19. About 81 per cent of all mortgaged-out land in Katkuian belonged to small and marginal landholders, though in caste terms it was more varied. On the other hand, almost the entire mortgaged-out land in Nayanagar belonged to the Bhumihars alone.

The relatively more diverse pattern of mortgaging-in land in Katkuian was the outcome of a variety of factors. Based on our conversations with village elders and circular migrants to Tehang village in Punjab, we learnt that mortgaging in (rehan/bandhak) had emerged as a popular means to access cultivable land among landless, small, and marginal landholders. The mortgage system operated in the following manner: the bigger farmers in the village mortgaged out their land to the landless and/or small/marginal farmers. The relatively well-off or bigger farmers did not enter into such mortgages for want of money alone. It benefited them because, in this way, they received a cash payment (equivalent to the expected income from the plot of land) in full at the beginning of the agricultural season. Sometimes urgent money needs were met through this arrangement, and in fact, this had emerged as an alternative to bank credit. The going rate in May 2018 for mortgaging out land was Rs 14,000–15,000/kattha (1 kattha – or “cottah” – is equal to 0.078 acre) per annum.

From the perspective of the landless, small/marginal, and EBC/SC households, this arrangement provided an opportunity to access cultivable land. Traditional bataai (sharecropping tenancy) was formerly limited to only khetihar (farming) castes. Landless or smaller farmers used to lease land from the bigger farmers to take care of their basic household food needs. In Katkuian, this form of tenancy was in decline for two reasons. One, a growing number of absentee landholders were not interested in agriculture at all, so they preferred either rehan (mortgaging out) or hunda (leasing out on fixed rent in terms of crop produce). Two, bataai also required a lot of supervision and hence was avoided by large landholders.

Our interviews with a variety of participants in Katkuian revealed that more among the labouring/landless castes entered rehan (mortgaging in of land) transactions in the post-1990 period. This was also made possible because of the migration linkage with Punjab, which provided many of the landless/small farmer and EBC/SC households with ready cash when they came back to their village after the harvesting season. It was important for them to cultivate foodgrains for their family’s annual needs.

The bigger farmers found numerous ways to create trouble for smaller farmers who mortgaged in land for cultivation. As part of these mortgage arrangements, the borrower, often a large farmer, mortgaged out a piece of land for a certain amount, and the lender, often a landless or small/marginal farmer, had to return the land whenever the larger farmer or borrower returned the amount. Sometimes, if the land had standing crop – sugarcane, for instance – the lender or small farmer/landless person paid an additional amount incorporating the cost of cultivation to keep the plot under his/her control till the harvest.

In some other cases, the land was mortgaged in by the lender and then leased out to the borrower at a fixed rent of about six quintals of sugarcane per kattha. From the second year of cultivation (khunti crop), the lender cultivated and harvested the crop until the borrower returned the principal amount to the lender. In cases where the lender was not interested in cultivation because many of them were landless migrants, they mortgaged in and then leased out the land to the borrower/larger farmer in return for a fixed rent in terms of crop produce.

Generally speaking, hunda was preferred by the landed households. Rent-in-cash was not very popular and there were fixed rates for different crops, such as six quintals per kattha for sugarcane and one quintal per kattha for paddy (the precise amounts vary with the relative class positions of lessor and lessee). According to the small farmers we interviewed, in the present times, the lessors (hundedars) did not benefit from hundedari because the landed households demanded a share in every crop that was cultivated by the hundedars. Also, hundedari was a preferred practice among own castes, implying that the farmers leasing out land under hunda wanted their own caste-persons as lessees; no such preference was found in rehan transactions.

Persistence of High Levels of Inequality

Another way to understand the depth of overall inequality is to analyse the composition of asset values in different caste categories (see Table 20 and Table 21). In this analysis, assets include landholding, buildings of all types (such as animal shelters), means of transport, means of production and agricultural tools (manual and machine-based), furniture, livestock, domestic durables including electronics, and food stock.

As stated earlier, the BC-2 caste category owned the largest share of land in Katkuian. The asset profile of BC-2 and EBC castes in the village was constituted very largely by land ownership (91–92 per cent in both cases). Within their asset portfolios, the contribution of value of land was 84 per cent for “other” castes, and the corresponding figure for SCs and STs was about 85–86 per cent. This highlights the importance of land as the most valuable asset in the village economy and its disproportionate distribution across caste groups.

In Nayanagar, the picture was different. Land value figured prominently in the asset profile of the Bhumihars (about 93 per cent), but not so much in other caste groups. As we have seen earlier, landholding among EBC, BC-2, and SC was minimal, resulting in relative diversification of asset portfolios among these categories. This was made possible also because of the more urbanised nature of the village and its surroundings.

Such diversification, however, did not result in dilution of the stranglehold of the major caste groups in either of the two villages. If we analyse the share of caste categories in land and non-land asset values in the villages, the overall inequality displays a strong association with caste (see Table 21). While the proportion of the BC-2 caste category in Katkuian was only 35 per cent, they controlled about 80 per cent of total value of land and 77 per cent of total value of non-land assets in the village. The scenario in Nayanagar was even worse, with the Bhumihars constituting only about 25 per cent of total households, but controlling 95 per cent of total land assets and 92 per cent of total non-land assets.

On the other hand, Dalits who formed 12 per cent of all village households in Katkuian only commanded 0.8 per cent of all land assets and 1.4 per cent of all non-land assets in the village. Similarly, the EBCs constituted 47 per cent of all village households, but controlled only 16 per cent of total land assets and 16 per cent of total non-land assets in the village.

In Nayanagar, the situation was worse. The EBCs constituted about 31 per cent of the population, but their share in value of land was only about 3 per cent and in value of non-land assets only about 4 per cent. The BC-2 category formed 10 per cent of all village households, but owned only 0.6 per cent of the total land assets and 1 per cent of the total non-land assets. Strikingly, the SCs constituted about 34 per cent of all village households, but their share in total value of land assets was 1.5 per cent and about 3 per cent in total value of non-land assets. This strengthens our argument of centralization of land in a few hands while the large mass of households remained landless or held poor-quality land. It is also noteworthy that the share of the BC-2 caste category and the Bhumihars in total non-land assets was lower than their share in total land assets, reflecting their higher control over land in the village. Land, it seems, continued to be the source of economic power for these two groups, but more so for the Bhumihars in Nayanagar.

Concluding Observations

Present-day Katkuian and Nayanagar show evidence of two different types of landholding and caste structures. Many recent studies, such as Sahay (2020), argue that there is no landlordism in Bihar any more. Our findings from Nayanagar challenge such an understanding and show the persistence of landlordism, now transformed into a system of capitalist landlordism but still rooted in almost complete land monopoly by a few dominant households from a certain caste – in this case, Bhumihars. On the other hand, Katkuian shows the supplanting of the hitherto dominant Bhumihars by a few from the middle castes.

The present-day land and caste structure in Katkuian is a result of the upheaval of the 1990s. The class of Bhumihar landlords who had controlled almost all the land left the village in those years for nearby towns such as Bettiah and/or even Patna. However, they continue to own hundreds of bighas (one bigha is 1.5 acres) in the village even today, as reported by their managers in the village. In Katkuian, recent improvements in transportation and connectivity have allowed a large section of the landless, small, and marginal farming households to build upon older migratory routes towards rural Punjab, and this has given rise to newer arrangements of tenancy that provide access to land to these sections of the population. These arrangements, however, have not yet changed the ownership patterns of land.

Landlessness continues to be very high in both the study villages. The problem of landlessness is more severe in Nayanagar, but the situation is not all that different in Katkuian. The prevalence of tenancy enables landless groups to access land in Katkuian, but it appears that much of the leased-out land was leased in by dominant households and only a small proportion of landless households were lessees. Dalits, in particular, were almost completely outside tenancy arrangements in both the villages.

Overall, the partial destruction of older socio-political relations in Katkuian aided by the OBC resurgence in the political sphere in Bihar during the 1990s has provided social and economic mobility to a larger section of the village population across castes, but has not led to a fundamental change in the structure of land ownership. In strong contrast, Nayanagar continues to be a Bhumihar-dominated village with little social freedom for its residents, and a strengthening of the hold of a few families over practically all resources in the village.