New Economic Sociology: Mark Granovetter

 

New Economic Sociology: Mark Granovetter

Gaurang R. Sahay

Despite the efforts of Parsons and Smelser in the mid-1950s and the 1960s to revive economic sociology, it attracted little attention, and by the 1970s the field was somewhat stagnant. However, in the early 1980s, a few studies suggested a new stirring of interest in economic sociology (e.g., White 1981; Stinchcombe 1983; Baker 1984; Coleman 1985). If one nonetheless were to choose one single year as the birth-year for New Economic Sociology, it would be 1985 since this was the year when the term 'New Economic Sociology' was born and also the year when an article appeared that was soon to become the most popular article of all in contemporary new economic sociology. The article, which galvanized quite a few sociologists into action as well as supplied them with intellectual legitimation for venturing onto the economist's turf, was Mark Granovetter's brilliant piece of work 'Economic Action and Social Structure: The Problem of Embeddedness', which appeared in the November 1985 issue of American Journal of Sociology. The article clearly contains the best critique up to date of New Institutional Economics as well as the conceptualisation of individual as an atomised rational decision making agency. The article focuses on the failure of economists to incorporate social structure into the analysis. To quote Granovetter,

Critics who have attempted to reform the foundations of economics have mainly been economists themselves. Their attack has typically been on the usual conception of rational action. It is my argument here that there is another fundamental feature of neoclassical economic theory that provides more fertile ground for attack: the assumption that economic actors make decisions in isolation from one another - independent of their social connections: what I will call the assumption of 'atomized' decision-making. (Granovetter 1982:2)

It should also be noted that Granovetter's move made it possible to introduce a new kind of analysis: where the actor is rational but their actions are grounded in social structure, so social structure counts. He argues that a sociological approach to the economy does not mean that the actors are viewed as irrational; sociology and rationality can very well coexist.

There exists an undercurrent of optimism and enthusiasm in this article, which is connected to the idea that if social structure matters in the realm of economy, then sociologists have a whole new world of exciting research opening up to them. He writes that ‘a large and important agenda thus remains nearly virgin territory with abundant fertile land to till' (Granovetter 1982:27). ‘I think there is a huge, untouched territory there, a whole Virgin lake' - again - for anybody who knows some sociology’. (Granovetter 1987:18)

Old economic sociology was here identified, by Granovetter, primairily with industrial sociology and the economy and society perspective of Talcott Parsons, Neil Smelser and Wilbert E. Moore. These two approaches, Granovetter said, had been full of vitality in 1960 - but then 'suddenly died out' (Granovetter 1985a). While stressing that the tradition of Parsons, Smelser and Moore had still much to give, Granovetter nonetheless drew a sharp line between old and new economic sociology:

In general, one of the main differences between the new and the old economic sociology has been precisely that new economic sociology does not hesitate to attack neoclassical arguments in fundamental ways, whereas the old economic sociology kept its criticism rather muted, and almost never constructed alternative models at the same level of detail. Granovetter writes ‘My position is that there is something very basically wrong with microeconomics, and that the new economic sociology should make this argument loud and clear especially in the absolutely core economic areas of market structure, production, pricing, distribution and consumption. What is wrong is that economic actors are not atomized from one another, as the theory requires us to believe, but are involved in interaction and structures of interaction that are theoretically central to outcomes - they are not peripheral frictions’. (Granovetter 1985a; cf. Granovetter 1990a:107, 1990b:95)

During the thirty five years or so of its existence, New Economic Sociology has become quite popular. This is particularly true for the United States, and it deserves to be stressed that New Economic Sociology is basically a North American phenomenon, even if a small number of supporters also can be found in various European countries. It should therefore be noted that a reader in New Economic Sociology appeared in 1992 (Granovetter & Swedberg 1992). Three anthologies have also appeared as well as a huge Handbook of Economic Sociology with some 40 contributors during 1990s (Friedland & Robertson 1990; Zukin & DiMaggio 1990b; Swedberg 1993; Smelser & Swedberg 1994). Every year there are a couple of sessions in ‘economic sociology’ at the American Sociological Association (ASA). But there still is a long distance we have to cover.

New Economic Sociology developed its own theoretical perspective by introducing a few key theoretical concepts; and one of these is obviously 'embeddedness' that comes from Granovetter's 1985 article. The basic thrust of Granovetter's 1985 article was to change the focus of the critique of economics from the assumption of rationality to the use of isolated actors. Embeddedness is then introduced as a counter-concept to atomization, Granovetter writes:

The opposite of atomization is something I want to call 'embeddedness', and I believe that the usefulness of social structural analysis in economic life has to do in crucial ways with recognizing the importance of embeddedness. (Granovetter 1982:11)

Whereas Karl Polanyi had introduced the notion of embeddedness to emphasize that the economy was an organic part of society in pre-capitalist times, Granovetter wanted to substantiate that economic actions are embedded in social structures of capitalist society too. He writes that economic actions are 'embedded in concrete systems of social relations', and these 'social relations' are to be understood in network terms (Granovetter 1985b:487; cf. 1990a: 107-108). Economic behaviour is 'embedded in networks of interpersonal relations' (Granovetter 1985b: 504). Granovetter introduces two related concepts 'relational embeddedness' and 'structural embeddedness' (Granovetter 1990b: 98-10).

Granovetter, following Berger and Luckmann's famous book The Social Construction of Reality (1966), talks about 'the social construction of the economy'. According to Granovetter, the notion of social construction constitutes - together with the idea of embeddedness - a fundamental sociological proposition in new economic sociology (Granovetter 1990b:95). What Granovetter, however, adds to Berger and Luckmann is a networks perspective, more precisely an argument about how networks play a particularly important role in the early stages of the coming into being of an institution. The US electrical utility industry, to use one of Granovetter's favourite examples, was early on fundamentally influenced by the networks of firms, holding companies and regulators. Later 'this [whole] network . . .congealed', as Granovetter phrases it, something which means that personal networks played less of a role after the evolution of an institution (Granovetter 1992a: 9). Granovetter says the electrical utility industry during the course of time acquired a specific institutional form with its own dynamic; it was 'locked in'.

Granovetter’s paper (1978) argues that knowing the norms, preferences, motives, and beliefs of participants in collective behavior can, in most cases, only provide a necessary but not a sufficient condition for the explanation of outcomes; in addition, one needs a model of how these participants in collective behavior or individual preferences interact and aggregate. He terms the model “threshold model”. The model admits that in most cases the decision of an actor can be thought of as having a positive and negative side— deciding to do a thing or not to. The decision of the actor depends in part on how many others make which choice. The cost to an individual of taking a decision declines as the number of people taking the same decision increases, since the probability of being apprehended is smaller the larger the number involved.

The individuals in this model are assumed rational—that is, given their goals and preferences and their perception of their situations, they act so as to maximize their utility. Individual differences are a main focus of the models. Difierent individuals require different levels of safety before they make a decision and also vary in the benefits they derive from decision making. The crucial concept for describing such variation among individuals is that of “threshold.” A person’s threshold for making a decision is defined here as the proportion of the group he would have to see join before he would do so. The threshold is simply that point where the perceived benefits to an individual of doing the thing in question exceed the perceived costs.

Granovetter has cited some examples in favour of threshold analysis. Women in Korean villages were wary of adopting birth control devices and wait to do so until some proportion of their fellow villagers do. Different women had different thresholds, depending upon their education, age, husband’s opinions, position in a hierarchy of informal leadership, or personal tastes. Workers deciding whether to strike will attend carefully to how many others have already committed themselves, since the cost of being one of a small number of strikers is high. One’s decision to vote for a particular candidate depends heavily on how many others have already decided to do so, partly because of social influence, partly because one does not want to waste one’s vote. This is what he calls “bandwagon effects.” The decision to go to college depends in part on what proportion of one’s cohort does so. If the the proportion is greater, there is greater the possibility of joining the college. We have all had the experience of sitting impatiently at a boring lecture, unable to leave because not enough others have yet done so. It is well known that migration decisions depend heavily on those of others, as in “chain migration”.

For Granovetter, there are mainly three approaches in New Economic Sociology: networks approach, cultural sociological approach and organization approach. There are a number of concrete studies in new economic sociology which are inspired by networks approach, cultural sociological approach and organization approach.

Networks approach

The relationship between new economic sociology and network theory is divided into two stages. During the first stage, which began in the late 1960s and lasted to something like the mid-1980s, network theory led to the study of corporate interlocks, that is, a social structure that is created when an individual sits on two or more corporate boards. A number of studies of this type were produced, and some of the results were the following: there exists regular patterns of interlocks that persist over time, and banks plus insurance companies tend to be the most central corporations in these networks (Mizruchi 1966). Donadd Palmer (1983) made a study of the situation when an interlock is broken, through an individual's death or retirement. Palmer concludes that only a minority of the interlocks are replaced. Michael Useem shows through a comparative study of top managers in the United States and England that executives in huge corporations tend to develop an overview picture of the economy when they sit on several boards (Useem 1984). Granovetter's early studies The Strength of Weak Ties (1973) and Getting a Job (1974) are in a class of their own among early network studies. He argues that people get jobs depending very much on their connections. People with many casual contacts ('weak ties') tend to find jobs much easier than those who have only regular contacts ('strong ties'), and the basic reason is that people with many casual contacts or weak ties have access to much more information. If strong ties are those between groups of close friends, weak ties are those of acquaintances. The importance of the latter is that they may form a ‘bridge’ between two clusters of people and therefore will be able to communicate novel information from one cluster to another. Strong ties are less important because information will circulate more readily within a group. For example, everyone in an office knows about the new vacancy in it, but not about the vacancy in the offices across town where you friend’s sister’s friends work. All ‘bridges’ between clusters should be weak ties. Similarly, strong ties should be ‘transitive’ – friends of friends should also be your friends – whereas weak ties should not – acquaintances of acquaintances should not be your acquaintances. Unique information often travels between clusters through a weak tie. It has been suggested by many studies that people with comparatively more weak ties than others have higher self-reported happiness, empathy, and feelings of belonging. Furthermore, new opportunities often arise through weak ties – so perhaps pushing through those awkward conversations with your hairdresser is worth it. This basic insight by Granovetter spawned a vast literature on social network analysis which continues to this day.

From the mid-1980s and onwards, network studies in economic sociology changed course and became more interesting as a whole. One reason for this was that some new economic phenomena emerged in the meantime, which fit the networks perspective very well. The rise of new industrial regions, held together by networks of smaller firms, is one of these; and it has inspired a number of fine networks studies (e.g. Lazerson 1993; Perrow 1993; Powell & Smith-Doerr 1994; see adso Piore & Sabel 1984). Important single monographs have appeared as well, such as Ronald Burt's Structural Holes (1992), which contains an imaginative theory of entrepreneurship. It is here suggested that entrepreneurship consists of tying together two isolated parts of one's networks that can benefit from one another (more precisely: two non-redundant contacts). Burt refers to Simmel's idea of 'tertius gaudens' - 'the third who benefits' - and one can say that he has helped to flesh out this idea with the help of networks theory.

Finally, a very significant recent development in networks theory is represented by the introduction of the concept of ‘business groups’ - a concept which has been sociological content and status by Granovetter in his contribution to The Handbook of Economic Sociology (1994). A ‘business group', according to Granovetter's definition, 'is a collection of firms bound together in some formal and/or informal ways', and it differs from the average conglomerate by displaying 'social solidarity'

(Granovetter 1994:454, 462-463). The concept of business groups fits perfectly such phenomena as the Korean chaebol, the Japanese keiretsu and many other, less-known formations of firms from all over the world (see e.g. Gerlach 1992; Biggart et al., forthcoming). The idea of business groups is one of the most fertile concepts generated by New Economic Sociology.

Cultural Sociological approach

The relationship between cultural sociology and New Economic Sociology is mainly discerned in the writings by Viviana Zelizer and Paul DiMaggio. Viviana Zelizer’s work ‘Beyond the Polemics of the Market: Establishing a Theoretical and Empirical Agenda’ (1988) provides programmatic statements on the need to integrate a cultural perspective into New Economic Sociology. This essay contains a sharp critique of the tendency in contemporary economic sociology to reduce everything to social relations and networks, a position she terms ‘social structural absolutism’ (Zelizer 1988: 629). But she also rejects the tendency to analyse economic phenomena exclusively in cultural terms, as if they consist only of meanings — a tendency she labels ‘cultural absolutism’. A well-balanced analysis, she concludes, would simultaneously take structural, economic and cultural factors into account. The goal should be ‘to plot a middle course between cultural and social structural absolutism’ (Zelizer 1988:629).

A similar argument can be found in an interesting discussion of the concept of embeddedness by Sharon Zukin and Paul DiMaggio (1990a). ‘Structural embeddedness’ or embeddedness in networks and social structures, they argue, is of great importance — but there also exist other types of embeddedness. There is, for example, ‘political embeddedness’ or the fact that economic action is always set in a specific context of political struggle. Then there is ’cognitive embeddedness’, which has to do with factors that limit the human mind in its mental processes. And, finally, there in ’cultural embeddedness’ or the embeddedness of economic action in culture. Culture affects the economy through ‘beliefs and ideologies, taken for granted assumptions, or formal rule systems’ (Zukin & DiMaggio 1990a:17). Culture can be either “constitutive,” referring to categories, scripts, and conceptions of agency, or “regulative,” referring to norms, values, and routines.

Political embeddedness has been well explored by Bruce Carruthers (1996) in new economic sociology. He has shown that not only do economic interests influence politics, but also the opposite: “political interests influence economic action” (Carruthers 1996, 7). Using primary material on the trade in shares in the East India Company in the early 1700s, he establishes that political ambitions clearly influenced the choices of buyers and sellers. The critique of Alfred Chandler’s key idea—that recent advances in technology had made it necessary around the turn of the last century to reorganize the large corporation – has similarly emphasized the state’s role in the emergence of the large industrial corporation (e.g. Fligstein 1990; Roy 1990, 1997; Freeland 1996, 2001).

As to empirical studies in economic sociology from a cultural perspective, there is first and foremost Viviana Zelizer’s trilogy: Morals and Markets (1979), Pricing the Priceless Child (1985) and The Social Meaning of Money (1994). In all of these works, the emphasis is on ‘the social construction’ of economic phenomena: on the social construction of life insurance, of the economic value of children, and of different kinds of money. The focus is on the cultural response or resistance to life insurance in 19th-century United States, when the sacred value attached to life clashed with the secular tendency of setting a price on someone’s death. In Pricing the Priceless Child, Zelizer studies nearly the reverse movement: children, in the American perception, went from having an economic value in the 19th century to being economically ‘worthless’ but emotionally ‘priceless’ in the 20th century. And in the last volume of her trilogy — the Social Meaning of Money - Zelizer shows that money is not a neutral, non-social kind of substance, but actually appears in a variety of culturally influenced shapes (‘multiple monies’).

Organization approach

The third approach in contemporary sociology that has contributed to economic sociology in a decisive manner is organization theory. New economic sociology has been very successful in using organization approach to explore a number of important topics, such as the structure of firms and the links between corporations and their environments. One fine example is Nicole Woolsey Biggart’s Charismatic Capitalism (1989), which deals with a very special type of organization: direct selling organizations, such as Tupperware and Mary Kay Cosmetics. Three theoretical approaches in organization theory have been especially important for the development of new economic sociology: resource dependency, population ecology, and new institutionalism.

Resource dependency, as its name suggests, rests on the postulate that organizations are dependent on their environments to survive. An example of this approach is work by Burt (1982, 1983), who suggests that three important factors that affect profits are the number of suppliers, competitors, and customers. The more “structural autonomy” a firm has, the higher its profits; that is, a firm with many suppliers, few competitors, and many customers will be in a position to buy cheaply and sell expensively.

Although resource dependency has lost some of its popularity during the past few years, some new approaches in organization theory such as population ecology and new institutionalism have emerged in the meantime; and these have resulted in a number of studies that are of great interest to economic sociology.

In population ecology the main driving force of organizations is survival. It has been shown that the diffusion of an organizational form typically passes through several distinct stages: a very slow beginning, then explosive growth, and finally a slow settling down (e.g., Hannan and Freeman 1989). Individual studies of this process in various industries, such as railroads, banks, and telephone companies, fill a void in economic sociology (e.g., Carroll and Hannan 1995).

New institutionalism is strongly influenced by the ideas of John W. Meyer and is centered around what may be called cultural and cognitive aspects of organizations (see Powell and DiMaggio 1991). Meyer argues that organizations seem much more rational than they actually are, and that specific models for organizing activities may be applied widely—including to circumstances they do not fit. It has been argued that the strength of new institutionalism is its exploration of “factors that make actors unlikely to recognize or to act on their interests” and its focus on “circumstances that cause actors who do recognize and try to act on their interests to be unable to do so” (DiMaggio 1988, 4–5). John W. Meyer has inspired a number of excellent empirical studies in economic sociology. Two of these deserve a special mention: Neil Fligstein’s Transformation of Corporate Control (1990) and Frank Dobbin’s Forging Industrial Policy (1994). The former is a very sharp and skilful attempt to recast the history of the huge US corporation from a sociological perspective. Dobbin’s work is a comparative study of the evolution of railroad policy in France, the United States and England during the 19th century. The strength of Forging Industrial Policy resides primarily in the way that the author demolishes the idea that there exists only one (rational) way of doing things.

Agrarian Class Structure in India: A Note

 

Agrarian Class Structure in India: A Note

Gaurang R. Sahay

The issue of agrarian class was not discussed in social science till late nineteenth century. Even Marx and Weber did not at all pay attention to classes in an agrarian society because they believed that the development of capitalism in the countryside will lead to a countryside reflecting the class profile of industrial society. However, when agrarian society differentiated itself and continued to survive along with the industrial society, it attracted attention from the scholars particularly from those ones who were involved in concrete Marxist politics such as Lenin and Mao. Lenin and Mao are the first scholars who developed the schema of agrarian classification. Land and Labour have been the main basis of the classification of agrarian society. Using these criteria it has been observed that some own land but do not work on it, some work on land but do not own it, some own land and work on it, and some own land and work on it too but they also used hired labour to work on land. This observation has resulted into categorisation of classes in agrarian structure.

Lenin used three related criteria to classify peasants into classes in the Preliminary draft Thesis (1920). The criteria are: 1. The possession of land, 2. Labour hiring vs. self employment, and 3.Subsistence vs. surplus above subsistence that can be invested and reinvested and can lead to capital formation in agriculture. On the basis of these criteria he differentiated agrarian society in Russia into six classes: 1. Agricultural proletariat or wage labourers, 2. The semi-proletariat and dwarf peasants, 3. The small peasantry, 4. The middle peasants, 5. The big peasants, and 6. The big landowners.

Mao Zedong did not use the term agricultural proletariat because, given the colonial character of China, the creation of a class of agricultural proletariat was substantially the result of pauperization of poor peasants under colonial oppression, rather than the proletarianization of poor peasants and workers accompanying the growth of capitalist production. Mao also argued that because of the low level of capitalist development in rural China rent exploitation and loan interest (usury) also played an important role in the formation of agrarian classes along with possession of land and labour hiring. Thus, Mao, on the basis of mainly labour power exploitation, rent-based exploitation and loan interest, talked about five agrarian classes in How to Differentiate Classes (1933): 1. The Landlord, 2. The Rich Peasants, 3. The Middle Peasants, 4. The Poor Peasants, and 5. The Workers.

There are scholars such as K. N. Raj argued that the concept of class is irrelevant in India because India is a complex society characterized by the coexistence of both capitalist and pre-capitalist features. It is also irrelevant because the institution of caste has been the primary institution everywhere in India. Besides caste is not coterminous with class, and more important social structure to explain the existing structural situations and their dynamics.

This position has been criticized quite a lot in India social science. It has been argued that class and caste can coexist in the same social formation. They reflect on the two different realities. Caste exist in the realm of ideology whereas class exist in the realm of economy. This apart, historians like D. D. Kosambi, Irfan Habib and sociologists like Dipankar Gupta argue that it is not the caste but class which has been the primary institution in India, and therefore class reality has always affected or shaped the nature of caste in India.

Before the debate started on the nature of mode of production in agriculture in India scholars, there were a few social scientists who had reflected on the nature of agrarian classes. Daniel Thorner on the basis of his surveys of rural India in 1950s argued that agrarian India is characterised by three agrarian classes: Maliks (owners), Kisans (peasants) and Majdurs (wage labourer). Maliks are those landlords who do not cultivate land but lease out to kisans and derive rent from them. Kisans are the peasant who cultivate land employing majdurs. Majdurs are the agricultural labourers. Ram Krishna Mukherjee in his study of six villages in Bengal observed that there are agrarian classes in rural Bengal. He conceptualised agrarian classes on the basis of household’s income level and amount of land. The criteria should therefore be one which has traditional recognition in the society. For him, there are three agrarian classes in rural Bengal and they are 1. Landlords, 2. Peasant-proprietors, and 3. Bargadars (sharecroppers) and labourers. Peasant-proprietors is a heterogeneous category consisting of Jotedars, Rich farmers, and Big Ryot.

Understanding of agrarian classes started systematically with debate on the nature of mode of production in agriculture in India. There is a number of scholars such as Joan Mencher, Nirmal Chandra, Utsa Patnaik, Ashok Rudra, Pranab Bardhan, Pradhan Prasad and John Harris who presented a description of agrarian classes in rural India during the course of the debate. For Joan Mencher, an important reason for understand the nature of agrarian classes in rural India is to try to find an explanation of why peasant organisations have developed in certain regions of India, but not elsewhere, "why people in one area are involved in continual revolt, while those in another area are relatively quiescent". (Mencher 1974) Looking at village India in recent years, Mencher concludes that movements have occurred "where there is a strong polarisation between landless and all others". In effect, these are areas with a large agricultural labour class, although peasant organisations are not strong in all such areas. In the South Indian context, Mencher asserts, Eric Wolf's hypothesis that it is middle peasants who constitute the pivotal groups for peasant uprisings does not hold. Rather it has been the possible ties to the land" who have been the main agitators or strikers.

Mencher contrasts developments in two regions known for successful organisation of landless labourers and sharp agrarian conflicts – Thanjavur in Tamil Nadu and Kuttanad in Kerala - with Chingleput district, also in Tamil Nadu. In Chingleput, Mencher points out, there is a large proportion of agricultural labourers, but in any given year a handful of these landless families may become share-croppers "on a 50/50 basis if they have bullocks; if not, on a 1-to-6 basis". Competition for obtaining land on crop share acts to inhibit unity among the landless, as also between peasants with small plots of their own who equally hope to rent in additional fields from the same land-owners. The rural bigwigs, for their part, are quite capable of juggling tenancies not only to prevent the actual cultivators from being able to claim customary rights but precisely to keep the poor divided and leaderless. ‘Thus, one way of handling a potential organiser of the landless or of the small peasants has consistently been to threaten that he might not get any share-cropping land for the next season or for the next year’. (Mencher 1974)

Mencher calls attention to the use of caste allegiances "by people in the system, as well as by outside observers. to mask class differences". For example, well-to-do members of middle-ranking castes may give land preferentially- to their own poor caste mates, who are thus led to identify with the village landed. On the basis of detailed information which she gathered in 10 villages studies in 1966-67 and 1970-71, Mencher proposes "a very rough socio-economic classification of the rural population of Chingleptit District". She talked about six agrarian classes which are as follows:

(1) The landless - "those who primarily derive their livelihood from working in agriculture either as day-labourers, as attached permanent labourers for particular landlords, or as ... a kind of share-cropper, receiving one portion for every six retained by the landowner".

(2) Poor peasants - "those who own small pieces of land, between 1 and 2.5 acres, small enough to require that on occasion some of the members of a household do day-labour."

(3) Middle peasants - "those who are clearly self-sufficient and able to sustain themselves without ever doing coolie-work…. Households with over 2.5 acres of land are employers of labour, and rarely go out as manual labourers themselves."

(4) Rich farmers - "those having between about 7.5 and 15" acres, "not only self-sufficient they are also able to store surplus for a bad year, and still have enough grain to sell to obtain cash for the purchase of cocisumer goods. (Most have a transistor set or a regular radio, and if possible, electricity)."

(5) Rich farmers, capitalist farmers, and traditional landlords - "households owning between 15 acres and approximately 30 acres of land. In this category there are three types of agriculturists:

(a) "rich farmers who, apart from giving small parcels of land to share-croppers, cultivate most of the land themselves with the help of coolies and actually go into their fields and do some of their manual work...."

(b) "capitalist farmers ... who do not do physical labour themselves...."

(c) "traditional landlords of the old school. I hesitate to call them 'feudal' because this is an area that has been subjected to capitalist penetration for a long period.… Mostly this category consists of landowners who give their land on various kinds of tenancy to labourers who look after their land for them."

(6) Indeterminate class of large landholders - "a few households in the over-3S-acre category.... I frankly question whether it would even be useful to decide if they are 'capitalist farmers' or 'feudal landlords'." (Mencher 1974).

Putting together Census of 1971 figures for Chingleput District and her own survey data, Mencher provides an idea of the relative strength of the six classes. According to the Census, 43 per cent of all working males in the rural population of the district were recorded as agricultural labourers. Another 32 per cent were returned as cultivators. The vast majority of these latter, Mencher judges on the basis her survey villages, belong to her categories 2 and 3, So far as concerns the three upper classes, "Those owning more than 7.5 acres, even of dry land are quite rare...". It is these same well-to-do households in which more family members tend to be employed outside of the village or to have additional sources of income "which add to the households' resource base and serve to raise them further in the socio-economic sphere". (Mencher 1974)

In the Chingleput villages studied by Mencher, class and caste hierarchies overlap to a considerable degree. Landless households (those owning less than one acre) were mostly Vanniyars (low-caste) and Paraiyans (untouchables). Proportionately fewer untouchables were to be found in the second, poor peasant, category. Large landowners (over 30 acres) belonged almost exclusively to Brahman, Reddiar and Mudaliar castes (these last two considered locally to be agriculturalist castes). Although majority of households holds in each of these castes had smaller holdings.

Caste loyalties tend to blur class boundaries, as does the fact that owners of even tiny plots become employers of labour at peak moments for transplanting and harvesting rice. Thus "on the whole the well-to-do Vanniyars have managed to keep the Vanniyar poor politically isolated and segregated from the untouchable poor". On the other hand, "families in category 3 (middle peasants), and even many in category 2 (poor peasants), do not see a commonality of interest with the landless - not even with their own landless relatives". (Mencher 1974)

Nirmal Chandra, in the Frontier articles to which we have already referred, delineates the rural classes in his Burdwan (West Bengal) villages in somewhat different fashion, and also considers the implications of the class structure for political action. He defines as "upper classes" landed families which do not depend to any significant extent on income from agricultural wage-labour. These classes include landlords "who depend mainly on their rental income,", jotedars "those funectioning in a capitalist manner", rich peasants who are dependent upon non-family labour although engaging personally in some major field operations, and middle peasants who cultivate with only marginal help from workers outside of the family. He proposes two separate estimates of class strengths which we can set out in tabular form. (a) taking into account only household agricultural income, and (b) according to households income from all sources.

In other words, the villages are split almost into equal halves between those with unearned incomes, i.e., all who lease out land or hire in worker the one side, and the poor peasants and agricultural labourers on the other. Given this quasi-equality, a struggle launched by the exploited half against the exploiting half "would never get off the ground". (Nirmal Chandra 1975b)

An alternative political approach implicit in the two giant waves of peasant struggle, for the reform of the tenancy system in undivided Bengal in 1946-47, and for the recovery and distribution among the poor of surplus land in West Bengal in 1967-70, was to concentrate on a struggle against "one particular feudal remnant". The difficulty encountered was that in both cases "the exploiters were sometimes men with very small means, and had close friends and supporters among sections of the middle and poorer peasantry". Too many enemies were created. This enabled the "most powerful sections in rural society" to create divisions among the ranks of the militants and their followers, and eventually to defeat the movements. Nirmal Chandra proposes instead a two-stage approach with left-wing political hegemony as the first goal. Once this has been achieved, the main task becomes the elimination of all forms of exploitation. He foresees the possibility of a number of sub-stages in the course of the movement "when the lines between 'friendly' and 'hostile' elements may have to be redrawn". (Nirmal (handra 1975b)

Utsa Patnaik returns to the centre of the debate on the nature of mode of production in agriculture in 1976 with an article on class differentiation among the peasantry. Citing evidence from successive censuses of landholdings, she emphasizes the extent of "concentration of the means of prodcuction". This high degree of concentration, Patnaik reasons, implies "a correspondingly high degree of economic differentiation within the cultivating population". Thus, there is no single representative type of holding, but rather a series of qualitatively distinct types, "which differ in the way their production activity is organised". (Patnaik 1976)

At one end of the scale, Patnaik continues, a small minority of households have resources so great in relation to family size that they must rely primarily on labour from outside the family. At the other pole, a large proportion of households "which may be the majority" have so few resources that in order to meet their family consumpion needs they must rely primarily on working for others whether as labourers or as tenants. In between these extremes we may expect to find a middle category of petty producers neither employing others nor employed by others. Taking together the National Sample Survey figures on landholding and the results of various Farm Management Surveys, Patnaik finds that "the majority of holdings in most regions do not fall into this category". The bulk of agricultural holdings, she argues, are so small that peasant families must hire themselves out or take in land at high rents in order to make ends meet. (Patnaik 1976)

Reiterating her earlier contention that the size of landholding is insufficient as an indicator of class status among the peasantry (see Patnaik 1971b), Patnaik elaborates a composite "labour-exploitation criterion", This ratio, to which she assigns the letter E, takes into account for each household hiring in, hiring out, renting in, renting out, and use of family labour. Her E ratio, Patnaik stresses, "has been formulated as an empirical, and therefore descriptive approximation to the analytical concept of economic class". (Patnaik 1976) In much the same manner as Nirmal Chandra, Patnaik distinguishes between the exploiting classes -landlords and rich peasants - and the exploited classes - poor peasants and labourers.

Within each of the first three categories she further specifies two different strata or divisions on the basis of the predominant form of exploitation, whether wages or rent. The resulting array of classes and divisions is as follows:

(1) Landlord

(a) Capitalist (Labour hiring greater than rent)

(b) Feudal (Labour hiring at most as high as rent)

(2) Rich peasant

(a) Proto-bourgeois (Labour hiring greater than rent)

(b) Proto-feudal (Labour hiring at most as high as rent)

(3) The Middle Peasants (No hiring in or hiring out labour)

(4) Poor peasant

(a) Agricultural labourer operating land-owners (hiring out greater than rent payment)

(b) Petty tenant (hiring out at most as high as rent payment)

(5) Full-time labourer (hiring out only form, no rent payment)

Distinctions among the five main classes are those familiar to the Marxist classics. Thus in the case of big landowners, whether feudal or capitalist, family members do not per- form manual labour in major farm operations. Supervision or operating machinery, Patnaik specifies, is not considered manual labour. Rich peasants do participate in manual work; however their resource position is such that appropriation of others' labour is at least as important as use of family labour. The middle peasantry is primarily self-employed since on the average the resources per capita just suffice to employ adequately the supply of family labour and to provide a living "at a customary subsistence level". The poor peasant family must hire out its members for wages or lease in land no matter how high the rent, or combine these two expedients. Typically these families "cannot make ends meet and have to depress consumption standards below customary levels". The same is true of full-time labour families; some of these may own small strips of land which they do not cultivate, but lease out. But the labour equivalent of the rent received is not large enough to balance, let alone outweigh, the amount of family labour hired out. (Patnaik 1976)

Patnaik explains that her labour-exploitation criterion is designed to bring out the necessity for the different classes within the peasantry to enter into relations with each other in the process of production. Ashok Rudra makes a very similar point in the first of three 1978 articles on class relations in Indian agriculture. Classes, he understands, "are defined by class contradictions". The relations between classes are relations of production, but (here Rudra diverges) "not all relations of production define classes". They define various "social groups", but "only some social groups are classes" (Ashok Rudra 1978a)

Having thus ruled out any theoretical obligation to fit the whole of the agricultural population into one or another class category, Rudra proceeds to argue that there exists in Indian agriculture today two, and only two, classes. These are "a class of big land- owners and a class of agricultural labourers". The latter include landless labourers, landed labourers, and poor tenants who do not hire any labourers. (Rudra 1978c) So far as the big landowners are concerned, Rudra can discover no contradictions between those with capitalistic features and those who operate along feudal lines. There may be co-existence of more or less feudal and more or less capitalist farmers "in the same region, or in the same village, or even in the same family". There may in fact be co-existence of "some traits typical of capitalists and some other traits typical of feudal land-owners in the same farmer". Rudra also rejects the classical distinction between 'landlords' and 'rich peasants' on the basis of participation in the manual work of cultivation. In India, he maintains, this criterion is negated by the caste factor. In some cases even very small and impoverished landholders will not take to the plough because they belong to upper castes. On the other hand, with the introduction of mechanisation, one may find, for example in the Punjab, women members of families owning several hundreds of acres who do not hesitate to drive their own tractors. The class of big landowners, in Rudra's view, is "a single class" and also "a hybrid class: part feudal, part capitalist". He refers to it as the "ruling class in Indian agriculture". Apart from the big landowners and the agricultural labourers, the rest of the population may be disregarded: "they do not constitute or belong to any class or classes". This classlessness results from the fact that, while they have contradictions among themselves, they do not have clear contradictions with the two principal classes. Or such contradictions "are of a subsidiary nature". Only the struggle between the two main classes "can provide the motive force for any changes in the agrarian structure". (Rudra 1978c)

In his conclusion Rudra spells out the political implications of his class analysis. Since he lumps together landlords revealing a preponderance of capitalistic traits and those displaying a more feudalistic prose, Rudra sees no justification for "those who believe in progress" to support "the assumed capitalist forces against the assumed feudal forces in an assumed struggle between the two". He rejects scornfully the concept of "an alliance of the entire peasantry from landless labourers up to capitalist farmers against the feudal landlords". Such a political line, he pronounces, "cannot but objectively betray the interests of the peasantry not belonging to the ruling class", and in particular the interests of the agricultural labourers. In point of fact, Rudra tells us, this is what has happened in previous peasant movements, "led by the political parties of the country" which have "by and large benefited the middle and rich peasants, but not the landless or the landed labourers". By contrast, the line of political action which would follow from Rudra's thesis "is one of struggle by the class of agricultural labourers against the class of big land' owners, without making any reservation on account of some members of the ruling class revealing more capitalisic traits than some others". (Rudra 1978c)

Rudra's rather drastic disposal of commonly held notions of class structure draws a comment from Pranab Bardhan, who had previously worked together with him in a largescale survey of land, labour and credit relations in West Bengal, Bihar and eastern Uttar Pradesh. (Bardhan and Rudra 1978). While listing five main points on which he wishes to state his agreement with Rudra, he specifies two major disagreements. Essentially, Bardhan approves Rudra's proposition that the most important contradiction in Indian agricultture is that between big landowners (including rich peasants) and labourers (landed or landless), and Rudra's criticism of the political line adopted by Left parties. But he takes Rudra to task for denying the significance of the middle peasants - who do not hire themselves out very often, or hire in much labour of others, as a separate class. For this purpose he cites data compiled by his wife, Kalpana Bardhan, from Farm Management Surveys carried out during the years from 1967 to 1972 in four states: Punjab, Uttar Pradesh, West Bengal and Tamil Nadu. Observing that the numerical strength of the middle peasantry varies sharply from one part of India to another, he calls for a more extensive investigation of the phenomenon. Bardhan also takes exception to Rudra's assertion that within the hybrid class of big landowners the feudal elements do not have any contradictions with the capitalist elements. (Bardhan 1979)

Pradhan Prasad, writing in 1979 and 1980 about Bihar in particular and, by extension practically the whole of the North-Indian Hindi-speaking belt (Rajasthan, Madhya Pradesh, Uttar Pradesh), provides yet another array of agrarian classes. As Mencher did for South Indian, but with less precision, Prasad indicates which castes tend to be found in which classes.. His three categories are as follows: (1) Top peasantry, including land- lords, who deem physical labour even on their own lands below their dignity - upper castes. (2) Middle and poor-middle peasantry, who do manual work on their own farms but do not labour for others. The middle peasants hire in agricuttural labourers; the poor-middle do not - these are essentially “middle-caste Hindus (i e, backward castes other than scheduled tribes)". (3) Agricultural labourers, "a sizeable number of whom have small operational holdings"; these are drawn "mostly from scheduled castes, scheduled tribes and some middle caste Hindus". Prasad points to sharp contradictions between the middle peasants, whose landholdings have increased and whose overall economic position has become stronger over the years, and the top peasants - doubled and made more acute by conflict between the "rising" middle castes and the "traditionally dominant" upper castes. He also speaks of an "emerging contradiction" between the "landlords, cultivators and big peasantry on the one side and the poor peasantry on the other". This antagonistic relationship arises "out of semi-feudal "bondage", and is destined to become less important "as the semi-feudal set-up disintegrates". It will be replaced by "another contradiction between new upper caste Hindu kulaks and the poor peasantry". At this stage, Prasad predicts, the landlords and big peasants will retract their earlier resistance to modernisation and "will take steps to dynamicise their cultivation". In his words, “The fanning of caste passions which at one time led to a diffusion of class contradictions, and thwarted agricultural growth, now turns out to be a factor which may sharpen the contradiction and cause the disintegration of 'semi-feudal' production relations in Bihar. (Prasad 1979, 1980)

At the other end of the subcontinent, John Harriss provides a version of the rural class structure on the basis of his field work in the dry districts of Tamil Nadu. He defines his classes according to two criteria - size of production resources (including land) in relation to household livelihood requirements, and labour relations. This gives four categories as follows: (1) Capitalist farmers, with assets capable of realising more than four times basic livelihood requirements, employing a permanent labour force, not contribulting personally more than a very little family labour; (2) Rich peasants, with assets yielding 2-4 times household requirements, possibly employing permanent labourers,, but substan- tially dependent upon family labour; (3) Independent middle peasants, whose assets yield 1-2 times household needs, employing principally family labour, may sometimes do wage labour for others; (4) Poor peasants, whose assets do not cover their livelihood requirements, so that they must depend primarily upon wage labour, this group includes marginal farmers and agricultural labourers. In the North Arcot village which Harriss studied intensively, he found evidence of all the features of a well- established capitalist mode of production. He emphasises the dominant position, consolidated since the end of nineteenth century, of a class of "landowning moneylender merchants", belonging preponderantly in this region to the Agamudaiyan Mudaliar caste. These landowners operate in classical capitalist fashion. "Money has been invested in agricultural production and profits reinvested; and farmers sold a large portion of their output; and farmers employ wage labour...." (Harriss 1979) Despite the fact that some 80 per cent of the households of the village may be indebted, Harriss insists that the "dominant mode of appropriation of surplus in Randam is capitalist." (Harriss 1979) The distinctive element in Harriss' contribution to the debate is his characterisation of the local dominant class as merchants as well as landowners and moneylenders.

An analysis of agrarian classes was not limited to the scholars who participated in the debate on the nature of mode of production in agriculture, a number of scholars such as D. N. Dhanagare, P. K. Bose, Carol Upadhyay, Jan Breman, T. Byres and Gaurang Sahay who did not participate in the debate have also presented the nature of agrarian classes in their studies.

Mode of Production in Indian Agriculture Debate

 

Mode of Production in Indian Agriculture Debate

Gaurang R Sahay

Agricultural economy in India had stagnated throughout the latter part of the colonial period and the rate of productivity was much lower compared to other places in Asia (Harriss 2013). The basic structure of agrarian economy remained almost the same during a couple of decades after Independence. Daniel Thorner characterised this situation as 'built-in depressor’, which referred to the concentration of agricultural land in the hands of very few households who gave out maximum amounts of their land for cultivation by share-croppers. The share-croppers depended on the same landlords for loans at high rates of interest for the purpose of cultivating that land. The landlords accounted, too, for almost the total share of all the sales of agriculture produce from the village. In other words, the term ‘the depressor’ refers to an agrarian relation dominated by landlords who lived by appropriating rents, usurious interest and speculative trading profits from the impoverished mass of the peasantry. Thus, the landlords had no incentive to invest in agriculture and the impoverished peasant masses were left with no means of investment. Consequently, agriculture remained stagnant. The then ruling Congress Party had recognized this condition in a Report of the ‘Congress Agrarian Reforms Committee of 1949’. Accordingly, land reform laws were enacted to change the situation.

During the early 1950s, various states brought about major changes in the pattern of land ownership through the abolition of the zamindary system or intermediary rights and tenures controlled by the tax-farming zamindars and conferring rights over land to the royats or tenants. But a review of land reform laws clearly indicates that these reforms did not bring about the end of ‘landlordism’. In practice, it was the larger occupancy tenants who benefited most from zamindari abolition by becoming real owners of land. They developed vested interests in the existing pattern of unequal landholding and being a powerful group in the ruling Congress Party, they thwarted further attempts at agrarian or land reform. Thorner’s ‘depressor’ continued to remain in place.

However, after Independence, India’s agriculture benefited somewhat from investments in the development of irrigation and the expansion of the cultivated area, and total food grains production grew at 2.75 per cent between 1952–53 and 1964–65. But the growth in food grains production was not sufficient and could not solve food shortage. In the mid-1960s, this problem became extremely serious. It was in this context the government adopted and introduced ‘green revolution’.

The Green Revolution was introduced to make agriculture more productive by adopting new higher-yielding varieties of wheat and rice, investments in agricultural technology and agricultural education and price incentives. Although the ‘green revolution’ had a positive impact on agrarian structure in terms of the productivity of food grains, increasing the level of income in the villages, increasing real wages in agriculture and through lowering food prices (Lipton with Longhurst 1989), the major beneficiaries were the rich peasants or big landlords. They were in the best position to take advantage of the Green Revolution because the ‘green revolution’ was though technically scale neutral but not ‘resource neutral’. Such a development in Indian agriculture led to a debate among social scientists on the nature of agrarian economy or on the idea of ‘built-in depressor’ or landlordism which is widely known as ‘mode of production in Indian agriculture debate’.

Though the debate properly openrd up with a report on a sample survey of big farmers in the Punjab by Ashok Rudra, A. Majid and B. D. Talib, there are three main precursors to the debate: Sulekh Chand Gupta, G G Kotovsky and Daniel Thorner. Sulekh Chand Gupta (1962) offered an estimate of capitalist farming in agriculture in India in 1953-54. Taking as his point of departure the concentration of hired labourers on the large acrege farm units as shown in the Farm Management Studies, he noted that in the state of Uttar Pradesh the element of hired labour exceeded that of family labour on farms with 20 or more acres. Extrapolating in a conservative fashion to India as a whole, he proposed to take the number of households operating 20 acres or more as a rough approximation of the number of capitalist farms. Applying this procedure to the figures from the 1953-1954 Census of Landholding, he came up with an estimate of one-third of the total area under capitalist cultivation.

A Similar estimate was proposed by G G Kotovsky (1964). In Kotovsky's judgment the area cuiltivated wholly or mainly by hired labour in 1953-1954 amounted to 2 in Indian agriculture, but that it did5 or 30 per cent of the total for India. His conclusion was that the capitalist sector represented the leading tendency in Indian agriculture, but that it did not yet dominate.

Daniel Thorner agreed with these estimates in his articles published in The Statesmant on four successive days in November 1967. In these lively journalistic pieces Thorner related what he had just seen in a village tour which had taken him through seven states in tle Northwest, North, West, South and East of India. Whereas fifteen years previously he had been impressed by rural stagnation (Thorner 1956), this time he found alert, enterprising cultivators, eager to experiment with new scientific methods; quick to switch to power for traction and pumping; ready to invest in improvements; preferring to cultivate themselves with hired labour rather than, or in addition to, giving out their land on rent in small parcels; and able to obtain substantial increases in output. He wrote: What we are witnessing in India is the emergence of an advanced sector in agriculture that is broadly comparable to the advanced sector in modern industry. This new agriculture has been tested, has proved profitable, and is rapidly expanding. (Thorner 1980). Thorner termed new-style farming capitalist farming and cultivators gentleman farmers. . He concluded that the unprecedented wave of capitalist farnming foreshadowed new social and political as well as economic developments in India.

Rudra, Majid and Talib denied the evidence of the existence of capitalist farmers in Punjab (Rudra et al. 1969). They say that farmers in Punjab do not tend to cultivate their land themselves. They do not tend to use hired labour in a much greater proportion than family labour. They do not tend tend to use farm machines in an increasingly greater way. They do not market an important share of their agriculture produce. They do not organise their production as to yield a high rate of return on their investmens.

Utsa Patnaik demolished the findings of the report by Rudra, Majid and Talib terming their analysis of data "unhistorical" (Patnaik 1971a). Patnaik (1971a, 1971b, 1972a, 1972b) observed on the basis of data collected from that though the basic nature of agrarian economy is still non-capitalist, it has started exhibiting some features of capitalism and constitutes a small but growing class of capitalist farmers. Citing the results of her own field survey carried out in 1969 and covering 66 big farm-ers in five states Orissa, Andhra, Mysore, Madras and Gujarat - Patrnaik reported her conviction that a new class of capitalist farmers was indeed emerging. She was persuaded after going through her data that capitalist development was on the way, albeit in varying degrees, throughout the regions which she had studied.

The characteristic of the genuine capitalist, Patnaik proposes, is not appropriation of surplus value generated by wage labour nor the sale on the market of a high proportion of produce, but also - and indispensably - accumulation and reinvestment of surplus value in order to generate more surplus value on an even-expanding scale. The capitalist in agriculture can be recognised by the "degree of capital intensification": i e, growth of outlay on both constant and variable capital with respect to a given land area and, over time, a tendency towards a higher than average organic composition of capital, leading to higher productivity of land and labour.

Following the debate, scholars like A. G. Frank (1973), Gail Omvedt (1981), Dipankar Gupta (1980), Jairus Banaji (1972, 1973, 1977), Jan Breman (1985), Joan Mencher (1974), John Harriss (1979, 1982), Kathleen Gough (1980) and Paresh Chattopadhyay (1972a, 1972b) argued highlighting the existence of capitalist agrarian economy characterized by self-cultivation, monetization, mechanization, productivity orientation, propensity of profit, free wage labour, accumulation and investment in agriculture, increase in extraction of surplus value, generalized commodity production and the existence of a free market. They argued that, though the capitalist agrarian economy consisted of big farmers, there was no bonded labour or begar and bonded or attached labour-based sharecropping which are the hallmarks of landlordism. Dipankar Gupta even writes that the existence of sharecropping does not indicate the prevalence of feudalism in agriculture. It is, in fact, the sign of early stage of capitalism (Gupta 1980). Differing somewhat from this position, Krishna Bharadwaj (1974) argued that, in the post Green Revolution phase, though smallholding peasant producers became part of the process of commercialization, it was a ‘compulsive involvement in markets’. This economically weak section of the peasantry had an extremely weak bargaining position in the markets. Yet they could not avoid involving themselves in market operations. This reflects on the forced commercialisation and conditions of their distress (Bharadwaj 1985).

Contrary to the position favouring the development of capitalism in Indian agriculture, there were some scholars, namely, Amit Bhaduri (1973), Pradhan Prasad (1973, 1974), Nirmal Chandra (1974) and Ranjit Sau (1973, 1975, 1976), who argued that agrarian economy in India was semi-feudal, that is, an economic system characterized by landlordism and which is more in common with classic feudalism of the master-serf type than with industrial capitalism. They argued that agrarian economy is semi-feudal because it is characterized by landlordism and the associated features like sharecropping, perpetual indebtedness of the small tenants, concentration of two modes of exploitation, namely, usury and land ownership in the hands of the same economic class, lack of accessibility to the market for the small tenant, pervasiveness of attached workers and begar (in- or underpaid labour services), consumption loans at exorbitant rates of interest, non-monetized wage and market relations, massive underemployment, determination of small peasants to continue cultivation no matter how meagre is the returns, power oriented attitudes of the landlords, non-utilisation of resources for agricultural development and lack of accumulation of capital for investment in agriculture. Thus, the proponents of semi-feudalism thesis provided a more rigorous shape to ‘the built-in depressor’ idea of Thorner.

The idea of semi-feudalism and compulsive involvement in markets indicate the failure of land reform. Following the semi-feudalism thesis, an international group of scholars, namely, Donald Attwood, David Ludden, B. B. Chaudhuri, Meghnad Desai, Ronald Herring, Ashok Rudra, Llyod Rudolph, Sussane Rudolph and Sukhamoy Chakravarty, debated the issue of agrarian power and agricultural productivity (Desai, Rudolph, and Rudra 1984).They demonstrated the commercialization of agriculture and its unevenness across different regions of India and the variable relationship between power structures and productivity (Rudolph 1984, 12-13). Many of them advanced empirical evidence refuting the semi-feudal thesis. A critical argument for Herring was that it is the deployment of the surplus that is crucial for agricultural productivity, and that the power structure determining the form of appropriation of surplus does not determine its deployment. Herring, and many other contributors to the debate, agreed that the concentration of scarce resources in the hands of powerful agriculturalists encourage entrepreneurial risk-taking in agriculture. In the context of a modernizing agricultural sector, Herring believed that the classic inverse relationship between farm size and productivity, so often referred to as constituting an essential reason for advocating redistributive land reform, no longer holds.

Some of the contributors to the ‘mode of production’ debate who argued for the emergence of capitalist agrarian economy believed that the peasantry will be ultimately differentiated into the establishment of an agrarian capitalist class and a substantial rural proletariat. But, for Terry Byres, Indian economy is witnessing ‘partial proletarianisation’ with small and marginal producers who continue to reproduce themselves consistently. This is because, to a certain extent, many of them work outside agriculture (Byres 1981). Thus, the agricultural economy is still characterized by extensive small-scale household labour farm and household labour-based production. The ‘marginal’ operated holdings (of one hectare, or less, in extent) or household labour farm now account for around 70 per cent of the total farms. Such marginal holdings are unlikely to provide ‘enough work or income to be the main livelihood of the household’ (Hazell et al. 2007, 1).

This point is also proven in the findings of the Foundation for Agrarian Studies from village surveys in Andhra Pradesh, Uttar Pradesh and Maharashtra. In fact, the net annual incomes from crop production of very many such households were actually negative (Ramachandran and Rawal 2010; see also Basole and Basu 2011, 75). These households reproduce themselves now as a result of migration and of the remittances associated with it, and of increasing employment in non-farm sector but still within the rural economy.

The significance of labour migration in this respect is highlighted, for example, by Gerry and Janine Rodgers in their study of two Bihar villages (Rodgers and Rodgers 2011), in Jan Breman’s studies of the ‘footloose proletariat’ of south Gujarat (Breman2007) and in a re-study of a Tamil village by John Harriss, Jeyaranjan, and Nagaraj (2010). The National Commission on Enterprises in the Unorganized Sector indicates that the number of seasonal migrants is of the order of 30 million, while Jan Breman suggests that it is likely to be closer to 60 million. Some estimates suggest that there may be as many as 100 million migrant workers in the country (Harriss 2013). Therefore, Henry Bernstein has argued that the dominance of capital today is no longer expressed in ‘classic’ capital–labour relations. He suggests that there are ‘classes of labour’ pursuing their reproduction ‘through insecure and oppressive – and typically increasingly scarce – wage employment and/or a range of likewise precarious small-scale and insecure “informal sector” (“survival”) activity, including farming’ (Bernstein 2008, cited in Harriss 2013, 358). As a result of these processes, the differentiation and polarization of peasant classes has nearly frozen.

However, Vikas Rawal (2008), using data from the 59th round of the National Sample Survey (NSS) for 2003–04, shows that 31 per cent of rural households are landless, and another 30 per cent own less than 0.4 hectare or about one acre of land, while only a little over 5 per cent of households own more than three hectares and just 0.52 per cent own more than 10 hectares. He argues that the absolute numbers and the relative share in the rural population of landless households have been increasing, so differentiation has not frozen entirely. Comparing the data from 2003–04 and 1992 NSS rounds he suggests that there has been an increase of as much as six percentage points in landlessness, while inequality in landownership also increased. There is an argument that the increase in landlessness and inequality in landownership have further gone ahead with economic reforms in 1991. This is because in the phase of economic reforms the costs of cultivation has phenomenally gone up along with a drastic decline in the public investment in agricultural infrastructure and research. Besides, economic reforms have facilitated commercialisation of agriculture and contract farming.

The issue of the nature of agrarian economy – whether it is characterized by semi-feudalism or landlordism and associated features like big landlords, bonded or attached labour and attached labour-based sharecropping or by capitalist farmers indulging in self-cultivation – has not yet been settled. There are many works which suggest that the introduction of land reforms and New Agricultural Strategies from Green Revolution to contract farming and the phenomenal growth in population have stamped out landlordism from rural India. To quote Gupta, ‘With the abolition of landlordism and the introduction of adult franchise (the two must necessarily go hand in hand), old social relations that dominated the country side are today in a highly emaciated form, when not actually dead. Roughly 85 per cent of landholdings are below five acres and about 63 per cent are below even three acres. What land reforms and land redistribution could not do, demography and subdivision of holdings have done to land ownership. Where are the big landlords? There are some, but they are few and far between’ (Gupta 2005, 752). According to Harriss, the share of leased-in land in the total operated area, according to the NSS, declined from only 10.7 per cent in 1960–1 to just 6.5 per cent in the kharif (summer) season of 2002–3 (Harriss 2013). This apart, it is being also argued that the decline of landlordism can also be substantiated by the decreasing importance of land as a factor of dominance in the villages. To quote S. B. Singh, ‘The importance of land as a factor of livelihood and dominance is decreasing and other factors of production are becoming more important . . . more and more people are losing interest in village affairs. The urge to dominate over the lower castes always had a political-economic angle, and once the locus of the economy has partially shifted away from the village, the tendency to dominate is beginning to wither away’ (Singh 2005, 3173). Jeffrey Witsoe, from another village in Bihar, reports that Rajputs were/are the biggest landholding caste in the village. ‘Over the last two decades, Rajput dominance was replaced by the emergence of multiple power centres’ (Witsoe 2011, 625). In this case, the decline of Rajput dominance has had to do with the income earned outside agriculture and outside the village. Alongside these developments, there has been the emergence of a new and very different generation of local leaders, as Krishna (2003) has explained with regard to rural Rajasthan, from amongst the educated but often unemployed younger men, including some from a Dalit background.

However, there are some studies that argue otherwise. Rawal, using data from NSS for 2003–4, shows that there are a good number of landlords in rural India who own substantial portion of land in the villages and the rural India is witnessing an increasing inequality in landownership (Rawal 2008). V. K. Ramachandran et al. (2010) and Vamsi Vakulabharanam et al. (2011) have reported that in coastal Andhra Pradesh villages absentee landlordism and tenancy have been, in fact, increasing. Ramachandran et.al show not only the resurgence of landlordism in coastal Andhra Pradesh, but also the persistence of the landed wealth and power of Reddys, Kammas and Kapus in different parts of the state (Ramachandran et al. 2010).

Breman has made a similar point from his observations in a village in south Gujarat. He opines that in the village of Gandevigam, Anavil Brahmans, historically the landlords of the locality, ‘told me that the current generation of Kolis, who formerly were their share-cropping tenants, would not be willing to subordinate themselves as sharecroppers. But in confidential conversations the Kolis clearly expressed their frustrations about the undiminished hegemony and obstinacy of the Anavil Brahmans, large farmers. They admit that there is little else they can do in the shadow of this higher caste than to try to find adequate space to promote their own interests and autonomy’ (Breman 2007, 37).

According to Harriss (2006), traces of classic landlordism have remained in Sahajapur, a West Bengal village. Surinder Jodhka reports that in Punjab there are new ‘political entrepreneurs’, though not necessarily rich, but they are ‘are invariably from upper or dominant caste groups’ (2011, 16). The dominant castes’, according to the definition of this term given by M.N. Srinivas, are the castes who have huge control over land in the village. In northern Karnataka, according to Jonathan Pattenden, the Karnataka Farmers’ Movement was in fact ‘usually . . . controlled by dominant caste men often engaged in perpetuating caste and gender-based forms of domination in their villages’ (Pattenden 2005, 1979). Pattenden argues that ‘whilst traditional forms of control over the labouring class [inherent in the caste hierarchy: JH] have been eroded, gatekeeping increasingly allows the dominant class to exert more subtle forms of political control, which in turn facilitates processes of accumulation’ (Pattenden 2011, 164). Thus, on the basis of these studies, one can say that in rural India those who have historically been poor have loosened ties of dependence on landlords, but overall they still exercise very little leverage over the political space. Besides, there will be huge difficulties with generalizations about the decline of dominance based on land control.

The issue of landlordism or semi-feudalism in Indian countryside is also politically significant because all major Left-wing political parties of India – the Communist Party of India (Marxist) (CPI (M)), the Communist Party of India (CPI) and the Communist Party of India (Maoist) (CPI (Maoist)) – continue to believe that rural India is still significantly characterized by ‘landlordism’. This belief is the basis of their understanding of the agrarian question. They also claim that the successive central and provincial governments have largely overlooked the issue of landlordism and, therefore, have failed to resolve the agrarian question. Such ideas are widely prevalent in their programmes related documents. The Programme of the CPI (M) illustrates this point by observing:

3.15 The agrarian question continues to be the foremost national question before the people of India. Its resolution requires revolutionary change, including radical and thoroughgoing agrarian reforms that target abolition of landlordism, moneylender-merchant exploitation and caste and gender oppression in the countryside.

3.16 After independence, instead of abolishing landlordism, the Congress rulers adopted agrarian policies to transform the semi-feudal landlords into capitalist landlords and develop a stratum of rich peasants. The legislative measures for abolishing the old statutory landlordism permitted them to get huge compensation and retain big amounts of land …. Land ceiling laws provided sufficient loopholes to maintain large holdings intact. The record of the Congress party is one of monumental betrayal of the historic opportunity for rural transformation.

3.19 If the development of capitalist relations in agriculture is clearly the major all India trend, it is equally evident that … there are regions where old forms of landlordism and tenancy and archaic forms of labour service, servitude and bondage still play an important part in agrarian relations …. Capitalist development in Indian agriculture is not based on a resolute destruction of older forms, but has been superimposed on a swamp of pre-capitalist production relations and forms of social organisation. (CPI (M) Party Programme: http://www.cpim.org/party-programme#VI)

The CPI Party Programme also reveals that landlordism is still extensively widespread in rural or agrarian India. Some of the lines in its Programme are as follows:

6.3 Congress party continued its alliance with the landlords as well as the dislodged rulers of princely states…. The landlords were also given an opportunity to become capitalist (farmer) landlords and evict thousands of “tenants-at-will” in the name of resuming self-cultivation.

6.4 Land reforms were also sabotaged due to bourgeois government’s compromising with landlord elements in the country.

6.14 Absentee landlords continue to exist extensively in Bihar and in some other parts, especially in the Hindi belt. The Bihar Land Reform Commission constituted in 2006 prepared a list of big landlords clandestinely holding thousands of acres of land and running a shadow Zamindary system. (CPI Party Programme: http://www.communistparty.in/p/party-programme.html)

The position of CPI (Maoist) on this issue is well established. The Party believes that India is a semi-feudal country with widespread landlordism. Its Party Programme states this point in the following sentences:

14. … the comprador big capitalist and feudal ruling classes of India, introduced some changes in land relations in the name of abolishing feudalism after 1947 …. But the monopoly over land by big landlords continues to remain as such while the vast population of the landless and poor peasants remains deprived of the land.

…. Extreme forms of semi-feudal exploitation are still prevalent in the countryside. The major prevailing forms of such exploitation are extortion of their produce through share cropping, which is robbing them of their produce up to 50%, bonded labour, usurious and merchant capital and other forms of extra-economic coercion. The most vicious form of extracting surplus through extra-economic coercion was through the caste system. Here the artisan and service castes had to serve the landed gentry and priests for a nominal fee. The scheduled castes continued to be treated as near slaves contributing even free labour and services in the name of "begar" etc.

…. The countryside is dominated by landlords, usurers, merchants and religious institutions. These exploiting sections are the mainstay of the semi-feudal relations of production in the country.

15. …. Despite some changes in the areas of the ‘Green Revolution’, no significant change has occurred in the semi-feudal relations of India as a whole. (CPI (Maoist) Party Programme: http://www.satp.org/satporgtp/countries/india/maoist/documents/papers/partyprogram.htm).