Interlinkage in Indian Agriculture: Some Findings and Comments (1997)
Particular observations about land hold true for cultivators and scholars alike. Land generates wealth, and struggles for landcontrol connect periods and regions in Indian history. Land in Indian history has been a central tool for the production of wealth and is perceived, in abstracted, ahistorical terms, to be wealth itself. The productive potential of all and any land depends on other capacities to make it productive, including labor, technologies and already existing wealth, so that land constitutes one element in combination with others in the production of wealth. All forces of production available to a society at a given time work as a system, such that a change in the efficacy of one of them effects them all.
If land ahistorically is wealth, then historically it becomes wealth only if complementary forces of production are adequately mobilized, or kept from being immobilized. Thus landcontrol has meant, more specifically, control over access to other forces of rural production, particularly agrarian markets in land, labor, credit and commodities. Differential access to these means and markets is a dominant aspect of life in rural India: understanding the mechanisms of differential control is critical to picturing socio-economic foundations and change.
This essay will examine differential control of the major forces of production and agrarian markets (land, labor, credit and commodity) in contemporary India, as they are extrapolable from the available current literature, with particular attention to coastal Andhra Pradesh. Broadly the task will be to set out the varieties and causes of economic differentiation according to patterns in landholding, employment, labor hiring and selling capacities, credit availability, indebtedness, and power of purchase. A major theoretical objective of this paper will be to define what I wish to call “interlinkage” as a just and comprehending (if not comprehensive) paradigm of rural economic behavior. Interlinkage analysis extends and adjusts traditional Marxian class-based investigations in ways suitable to the Indian, and possibly general postcolonial context. It retains the ethical and historical-scientific power at the core of Marxist insight, while answering critiques of social determinism and ethicistic historical projects.
Socio-economists and historians of India implicitly and explicitly have attempted to explain socio-economic activity in terms of contractual links between markets and productive means forced by stronger parties upon weaker parties (Bharadwaj 1985, Bhaduri 1973, Bardhan and Rudra 1978, Brass 1990, Frankel 1971, Ramachandran 1992). Their efforts probe how the provision of designated goods and services depends on provision of other goods and services, how access to certain markets binds (in senses, determines) accesses to other markets. Interlinkage thus describes how the contracting-in and -out of productive forces makes them vulnerable to one another. This interbinding of market accesses and productive means creates mutually dependent socio-economic relations, complex and changing, articulate and disarticulate at particular times and places. The interknitting is flexible: new arrangements accommodate existing arrangements by perpetuating actively and passively the immiseration, or non-amelioration, of one party (productive force) by another. Durable arrangements are possible because of, not despite, inequities and imperfections in the nexus of arrangements. For rich and poor these durabilities carry significantly different values.
Following Marx, what makes the productive forces of a society (land, labor, capital) productive is their use by particular persons or groups to establish or maintain their own economic leverage or control. Productive arrangements form around efforts to control productive forces of a society. As these forces develop, relations of control change, and forces change again.1 Often forces and relations of production change at different rates, sometimes "corresponding," sometimes clashing. In a nutshell, productive relations are relations of control—at times, ownership—of forces of production, and of the relations themselves. The historical dialectic between material forces and relations designates the ongoing conflict between productive development, and ownership of the means of that development: groups dominant in a particular socio-economic relation counterpose those which aspire to dominate a changed set of relations—based on access or nonaccess to, or identity with, changing productive forces. Marxian socio-economic class, then, describes a polar relationship of differential control of the productive forces of a society, not a frozen historical entity or category of persons (Thompson 1962). Such an understanding of class is hard and yielding both. It refuses to legitimize static reading, or to procrusteanize socio-economies into overdetermined paradigms of rightful domination, necessary historical devolution or revolution. Marxist epistemology names socio-personal dynamics whose economies of movements are class. Class, in short, is the tightening and loosening of socio-economic tensions toward and away from the rupture of polarities which define them. To designate class as encoded in a society is to know persons and groups necessarily to depend on one another for gain and loss of material control.
Theories of interlinkage understand class and sub-class in this dynamic, relational sense to be a primary pattern of social and economic behavior. Socio-economic differentiation precisely means interlinkage, mutually determining connections: ordinated classes straining differential access to benefit themselves, subordinated classes straining it toward rupture along its very lines. At the microlevel (freezing the class movement spatially and temporally), interlinkage designates reciprocal increases and decreases between wages, prices, employment by type, credit, tenancy and holding of land over time between economic classes and subclasses, particularly those comprised of the landless and landpoor (see the discussion on this description of class to follow). This is more than an explanation of supply and demand reciprocities by which neo-classical economists describe spatial and temporal localities: it is a claim that supplies and demands in particular markets cannot be studied as autonomous operations. Markets arise, move in contexts of and reproduce already existing differentiation; persons enter markets advantaged and disadvantaged by previous or simultaneous participation in other markets. Not to account for this phenomenon as structural and not merely coincidence is to imply freedom of economic agency where it does not exist. Such free agency is, of course, often enjoyed precisely by that class to which economists searching for it belong.
It is useful to review briefly the salient history of landreform legislation enacted in Andhra following Independence, and its primary effects. This history is in many ways the chassey upon which interlinkages in the contemporary agrarian socio-economy ride.
Landholding systems before Independence differed between the Hyderabad princely state and the Madras Presidency, the former characterized by jagirdari and the latter by a mixture of zamindari (principally in the Northern Circars, Nellor and Chittoor districts) and ryotwari tenures (Jatkar 1980). 2 Abolition of zamindars and other intermediaries and creation of private, taxpaying landholdings in the coastal areas began in 1948 with the Madras Estates (Conversion into Ryotwari) Act, amended by a string of legislation in the 1950s. 3 The state thereby claimed rights to collect taxes formerly paid to intermediaries by conferring alienable ownership rights directly on farmers. The state agreed to compensate newly freeholding ex-zamindars, previously granted rights to the land by the colonial government "in perpetuity". 4
The Andhra Pradesh (Andhra Area) Tenancy Act, 1956, amended 1974, effective 1980, established former intermediaries' rights of land "resumption" for "personal cultivation" (under the slogan, "Land to the Tiller"), as well as rights of purchase for certain "tenants-at-will" (at least half of the land cultivated by them prior to such resumption), as well as guidelines for land consolidation and rent regulation. 5 The legislation did not attempt to remedy the serious problem of documentation of tenancy by mandating written leases, and did not attempt to register existing tenants. Neither did it confer landtitle except by right-of-purchase, contingent of course on a landlord's right-of-non-sale and on financing. No protection was given to tenants against eviction and coercion to declare membership in non-protected categories, including, importantly, sharecroppers and "farm-servants." Both eviction and coercion routinely accompanied important legislation in all parts of the State (Damodar Reddy 1980, Haque and Sirohi 1986).
In 1961 and 1972 ceiling legislation was passed for the coastal areas, establishing limits on the number and size of landholdings per family according to soil quality and irrigation of land. Leased-out land was exempted from the ceiling, as were state farms, plantations and farming societies. Surplus land was to be redistributed to the landless. However, no effective provisions were made to ensure correct records of landholdings before or after implementation of ceilings, to counter the transfer of lands to exempted categories, check benami transfers, clarify the priority of ceiling legislation over laws of inheritance, or remediate landlords' judicial tie-ups in the courts or their intimidation of land-tribunals (Haque and Parthasarathy 1992).
Table A: Operational Holdings and Area, All-Andhra (Source: Agricultural Situation in India, October, 1981, in Bergmann)
Table B. Operational Holdings and Area, West Godavari (Sources: Government of India, Government of Andhra Pradesh, in Parthasarathy and Pothana)
Table C: Jati-wise Distribution of Land in Gowthami and Krishna Villages, East Godavari and Krishna Districts (Source: Viswanadhan and Swarna Kumari, 1980)
Tables A and B, measuring the Lorenz differential for the entire state and for a district high in the production of cash crops (tobacco, chilis, etc.), both show a fragmentation of the size of holdings over time, with the largest increase in numbers of cultivators holding small and marginal plots. Likewise both show the largest landholders (holding more than four hectares) falling only slightly in numbers and area held. Do the changes in numbers indicate changes in the status or control of the landless and landpoor? Do they indicate meaningful land redistribution? Table C, 6 showing the distribution of landholding by caste over time for two villages in coastal Andhra, shows effective landcontrol changing little since Independence. Viswanadham and Swarna Kumari observe that in Gowthami village "...[a]round 1940, about 74 per cent of the land in Gowthami was owned by Brahmins; of this a little over 75 percent was concentrated in the hands of 5 families...[around 1980, at the time of the study,] 94.3 per cent land has remained in the hands of the traditionally land owning class of Brahmins, Kshatriyas and Setti Balijas" (sic.). The numbers concerning landholding patterns over time point not to redistribution but redivision of land, or redistribution of nomenclature, such that the sons (and grandsons, cousins, dead uncles) of nuclearized landed families now register holdings in their own names and become "cultivators" in their own right (cf. also Patnaik 1986). Tables A and B alone do not indicate the movement of small landholding proliferation: down, indicating divisions of semi-medium and medium holdings into smaller holdings, or up, indicating acquisition of land by the landless. This problem is endemic to the categorizations used for redistribution measurement. Viswanadham and Swarna Kumari's evidence of non-redistribution of caste and family holdings, combined with evidence of substantial increases in agricultural laborers (Bardhan, K. 1977 and see discussion to follow), and evidence of the slow rates of surplus redistribution (Narayana Rao 1980, Haque and Sirohi 1986), demonstrate that the division of former extended-family holdings into nuclear-family holdings extends certainly to the level of semi-medium holdings and perhaps even to small holdings. Tables A and B become more a record of changes in inheritance than land redistribution to the landless and landpoor.
Broadly then, landreform since Independence succeeded in altering the configuration of rural society, establishing legislative and bureaucratic apparatus for state taxation. Changes in the superfices of the agrarian structure brought by landreform did not make tillers of land out of landlords. Reform did not enjoin more than nominal state reclamation and redistribution of land according to class. It did not remediate absenteeism. In practice, it was by omissions in the legislation, intentional and not, that agrarian reconfiguration proceeded.
If the agrarain structure remained substantially the same before and after reform legislation, the primary changes being superstructural, then the goals of reform can reasonably be called misplaced or false. Rather than serving as a base for comprehensive agrarian reform, land reform co-opted and deflected agrarian reform impulses. A structural economic problem remained legislatively untreated, and was effectively reinforced by adoption of a freshly obfuscating nomenclature and the authorizing voice of new state bureaucracy. The structural problem, I want to show, was the foundation of agricultural productivity precisely on the economic vulnerability of vast sections of the population via the interlinkage of markets and productive forces by those who stood to benefit from differential access to them. Interlinkage itself thus spoke both a class relation, and a mode of production with two basic forms. The active form came to be associated with entrepreneurial, or "capitalist" farming, the passive with rent and debt payment, or "feudal" farming. These modes of production, commonly counterposed, do not, I think, correspond to clear economic epochs in Indian history (presuming they ever existed in clear forms—see discussion below), and in fact share much in common. Perhaps most fundamentally, together they form a compromise of silence over landreform: agricultural productivity (together with industrial productivity), not state assumption and redistribution of land and capital, was given to stimulate economic growth, and perhaps in time bring economic justice. Interlinkage, the fundamental vulnerability of massive sections of the population to markets and productive forces owned by a few, was retained both as the legacy of the past and the beacon of the future.
Interlinkage is rarely explicitly measured statistically. For all that landreform failed to do, it succeeded in authorizing a socio-economic lexicon and the use of categories for the non-study of economic interlinkage. It gave rise to a deflective, rather than a reflective statistical praxis. Reconstructing interlinkage precisely means making tables vivify lives, holding measurements together and looking into the spaces they point toward, the forged-together choices of actual persons.
Table D: Workers by Category, All-Andhra (Source: Census of India, m1981, in Bergmann)
Table D follows the convention for both government (Census and Agricultural Census) and quasi-government (National Sample Survey) statistical analyses—prompted by legislated categories—to consider all cultivators workers. This category confusion is not found in the countryside itself: wealthy "cultivators" do not actually work their land, and poor landholders do not merely work theirs. K. Bardhan's study (1977) ably iterates the mixed composition of the rural labor force: "[the] major portion of rural workers combine self employment with wage employment, subsistence production with market participation, use of family labor with use of wage labor," such that the variety of employer-employee relations consists, in combinations, of capitalist farmers hiring-in all or virtually all their labor, large "feudal" landowners with tenants or sharecroppers, "peasant" farms subsisting on family labor with and without hiring-in and hiring-out labor, and laborers holding no land and hiring-out their labor exclusively.
As the largest number of cultivators, according to Tables A and B, are small and marginal, and those cultivators needing to hire-out labor to survive are small cultivators (see the discussion of size productivity below), fifty to sixty percent of the persons classified in Table D as "cultivators" could just as well have been classified as agricultural laborers, because they are. That the largest landholders do not, and never did till land—despite being termed "cultivators"—hardly bears repeating. Small and marginal landholders are not delineable in any hard way from agricultural laborers, and, I will maintain, share more classwise than do small and marginal landholders with larger landholders—precisely because of interlinkage.
Table E: Indebtedness among Agricultural Labor Households, All-Andhra (Source: Rural Labour Inquiry, Gov't of India, in Haque and Sirohi)
Table F: Employment and Earnings of Agricultural Labor Families in Andhra Pradesh, 1979 (Source: K. C. Alexander, Indian Labor Journal, 5/1980, in Haque and Sirohi)
* Annual per capita expense necessary to live above the poverty line at 1960-1 prices, Rs. 246, multiplied by consumer price index for agricultural laborers, 8/1979, 301, and the average size of agricultural labor family, 4.1.
Table G: Inputs and Returns by Size and Caste, Jonnalagadda Village, Guntur District, 1977 (Source: Reddy, 1985)
By Table E, laborers both with and without land receive credit, indicating that loans go for production and consumption both. This is not surprising. That indebtedness is extremely high among both groups suggests that such loans are necessary for survival, and are not merely calculated risks. We can deduce, then, that credit provision both determines and is determined by land use. Ramachandran (1990), in his study of agricultural labor in Tamil Nadu, likewise reports that in 1975, eleven percent of debts incurred by agricultural laborers went to production, while sixty-eight percent went to household consumption. Bhende reports (1986) in a comparative study of credit in villages in Maharashtra and Telangana that as much as one hundred percent of credit assumed by agricultural labor households went to consumption needs. Table E indicates further that indebtedness among laboring households with land—tenants or marginal-small cultivators—exceeded that among households without land. This finding indicates that landholding itself constituted a burden for the landpoor, variously in the form of rent payment (for tenants), credit-taken for consumption and production, wages not received as a result of labor not hired-out, or low prices received for produce. At best, landholding for the landpoor was an unrealized asset against the need for credit. At worst, self-employment caused debt for subsistence farmers due to interlinkages in farm size, labor, credit and prices.
Table G indicates that a higher proportion of landpoor farmers received loans from private moneylenders-cum-financier brokers. We can assume that landless laborers, without even the collateral of a small piece of land, prima facie received their loans from moneylenders. The exorbitant rates of interest of these moneylenders—from twenty four percent to thirty three percent for Backward Caste small farmers—indicates that wages and/or prices received for the selling of labor or produce went mainly for the repayment of interest. This conclusion is corroborated by the terrific gap between net earnings and total debt in Table F.
That a greater proportion of institutional credit went to larger landholders according to Table G (corroborated by Bhende 1986, Ramachandran 1990, Parthasarathy 1971, Frankel 1971), 7 and that debt among small landholders was chronic as shown above, points clearly to a higher demand for (consumption) credit among poorer laborers and cultivators. The demand follows both from the lack of availability of institutional credit, which charges lower interest rates and therefore in theory obviates the need to receive credit to pay off credit previously taken, and, among small landholders, from the inability to hire-in labor or make productive investment in land which would also, in theory, lessen or relieve debt burden in the long run (see more on size productivity below). Small cultivators absorb family labor in order to pay off debt incurred, or sell some labor for wages (in fact both), but according to the data neither wage-labor nor small farm self-employment suffice to relieve chronic debt. This abiding vulnerability of landless and landpoor persons to usurious credit can also be inversely explained: wages possibly earned, and prices possibly received from land are sacrificed in the form of increased debt incurred, or increased credit extended. Interlinkage suggests perpetual debt not simply by discreet contractual arrangements between debtors and creditors, but the nexus of credit-related occlusions an individual or family enters into even by a single loan. Debt acts as a disincentive for small farmers to hire-in wage labor to increase production, and to hire-out their own labor for supplementary income. The burden of debt effectively means the depression of the exchange value of labor. Such depression results alternately in the non-sale of labor, and the sale of labor for non-living wages (see below), precisely by those cultivators whom greater commerce in labor would stand to benefit: the landpoor whose small holdings are unprofitable assets, and the landless with no assets other than their labor days.
Productivity of farm by size has been an issue of much discussion among agricultural economists, particularly since the introduction of bio-technical innovations such as high yielding varieties of crops, and mechanization, especially tractor use (Sen 1962, Frankel 1971, Parthasarathy 1971, Bhaduri 1973, Chattopadhyay and Rudra 1976, Patnaik 1976, Bardhan, K. 1977, Bardhan and Rudra 1978, Byers 1981, Parthasarathy and Pothana, 1983, Bhardawaj 1985, Reddy 1985, Patnaik 1986, Upadhya 1988). Sen's influential 1962 thesis held family farm size and productivity to be inversely correlated, such that increase in family farm acreage was complemented by diminishing outputs and returns. An important corollary to this thesis held that large non-family farms not only were more productive, but more profitable, because the ratio of debt and labor costs to price received, was higher than for small farmers.
Chattopadhyay and Rudra (1976) observed that larger farms used a higher proportion of hired to family labor, and that labor productivity for family based (less- or non-hiring, small) farms was higher than for hired-labor based farms. They also found large (labor-hiring) farms to be more productive overall, and concluded that farms using more hired labor also used more non-labor inputs per acre, so that hired labor on large farms was less productive than family labor on small farms. Their findings indicate that the cost of labor and other inputs, not the size of holding, principally determined productivity, and that the productivity of the family farm was better understood as a function of distress—the necessity of intensive cultivation with no hired help—than of efficiency. Their findings indicate further that large, labor-hiring based farms were more efficient—due not to more intensive labor use, but to the use of labor-displacing capital intensive inputs. The distress factor for labor was, however, much the same, only based on predictably low wages rather than compulsory self-employment (see discussion below on wages). In this connection, Patnaik's (1976) thesis of labor-hiring as the indicator of economic class relates paradoxically that as farmers can afford to hire-in labor, they can also afford not to hire-in labor, or alternately can afford to pay their hired labor poorly. The capital which permits labor-intensification represents a capacity to depress labor's exchange value, overtly so when capital intensification displaces labor (particularly with tractorization—bio-technologicization alone has been shown to increase labor demand). If wages on large, labor-hiring farms are low or stagnant, small and marginal cultivators do not stand to benefit from hiring-out their labor to meet the non-increase in productivity of their small farms. Likewise they do not stand to benefit from not hiring-out their labor and remaining self-employed, due to low prices and high credit. The upshot is that economic improvement for the landpoor and landless does not follow the productivity of their labor irrespective of farm size.
Kalpana Bardhan (1977) and Haque (1987), corroborating Table A and the non-profitability of self-employment noted above, show the proportion of agricultural laborers rising in all parts of India. Bardhan and Rudra's (1980) study from West Bengal shows casual laborers to comprise between seventy-seven and ninety-six percent of total agricultural labor families, and Rudra (1987) shows that between thirty and fifty percent of the laborers in his West Bengal sample villages were casual, with the greater number found in the more prosperous areas. Why the increase in casual and not other types of labor? Presuming that the increase comes from marginal and small cultivators, it becomes useful, following K. Bardhan, to try to identify the allocation of labor supply across kinds of employment at the farm level, in light of the preceding discussion on the differential demand between family labor and types of hired-in labor.
Bardhan and Rudra (1980) schematize agricultural labor attachment as follows:
•Totally unattached, or "casual," free in principle to sell their labor freely
•Totally attached, or "farm servant"
•Type 1 semi-attached, or "free" to work for other employers despite attachment to one employer for part of the year
•Type 2 semi-attached, or "obliged" to work for an employer for a stipulated number of days in a stipulated period
•Type 3 semi- attached, or "obliged" to work for an unstipulated number of days over an indefinite period
Labor contracts and arrangements are sensitive to changes in every agrarian market and vary according to duration (single days, harvest months, full years or more), frequency of work (regular, intermittent), method of payment (cash, kind), timeliness of payment (regular, irregular), task, and rates of pay. Rudra (1984) demonstrates that labor attachment, or the ties between employers and employees, including varieties of compensation and noncompensated services, and the ability of a laborer to sell his labor to multiple employers—is significantly determined on a regional, even village by village basis (though see discussion of transregional factors below). Even so, both the First and Second Agricultural Labor Enquiries (1950-1, 1956-7) distinguished attachment by lower pay, continuity of employment and restricted freedom to seek alternative work while debts were owed to a creditor-employer (debt bondage, sometimes called "tie-in allotment") (Brass 1990).
Ramachandran (1990) and Bardhan and Rudra (1978 and 1980) show that the overwhelmingly decisive factor in determining attachment for casual laborers, tenants, attached and semi-attached laborers alike—those contracting-in wages or renting-in land for cash or kind, or a mixture—is the availability of credit for consumption. Thus the landlord, perhaps an entrepreneurial farmer, will allot land to laborers on the basis of commitment of labor for a stipulated part of the season, for a stipulated wage or service, or payment of rent, and cash or kind advances to meet household consumption needs. The landlord thereby assures the labor necessary to plant and harvest the holding.
In Bardhan and Rudra's sample (1980), the number of laborers who accepted credit both against a commitment of labor and against a wage lower than the market rate was universally reported. Further, this tendency was highest among landless casual laborers, next highest among fully attached laborers, and lowest among landed casual laborers. From this finding we can deduce that landholding provided some security against the need to take out loans (repayable in labor, low wages or low prices for produce-sold), but landrental did not. It is important to reiterate that landholding as security depends on the agrarian structural context. As mentioned above, landholding constitutes a burden to the extent that consumption loans are anyway necessary, credit-taken is repaid in low prices for produce, and subsistence farming absorbs family labor, depressing its exchange value. If not an outright fetter, landholding under interlinkage (i.e. poor land and agrarian reform) does not constitute the asset it should be. Labor migration patterns corroborate this conclusion, albeit elliptically, indicating that landpossession, rather than income, is inversely correlated to the search for wage employment (Parthasarathy and Dasaradha Rama Rao, 1973). Rudra's figures (1987) report longer contracts and fewer new consumption loans among laborers in advanced areas. Unremarked upon by him, however, respondents in the advanced area reporting longer contracts together with new consumption loans equaled in attachment (the total obligation of debt plus the commitment of labor time) the number reporting consumption loans alone in the poor area. This correlation suggests that longer contracts did not bring prosperity, but that reported decreases in new loans in the advanced area were matched by decreases in real wages and/or net income. Duration of contract, whether for attached or casual labor, functioned primarily as a form of labor control, tying-in employment to low wages and aggravation of arrears.
Bardhan and Rudra (1980) report that virtually one hundred percent of both their fully attached and fully casual respondents did not prefer to change their contracts, and that attached families seemed "to have no special reason for their attachment." How does this non-preference relate to wage and attachment differentials? Rudra (1987) finds wages to be more uniform and stable in the backward areas than advanced areas. In microstudies from around India, Parthasarathy and Adiseshu (1982), Bardhan (1970), K. Bardhan (1977) and Ramachandran (1990) all report that real wages either fell or remained stagnant with increases in productivity. Together these findings indicate that differences in labor attachment are not due to universal trends of higher incomes for wage labor, though of course small increases or decreases spatially may motivate some workers to try to seek certain kinds of employment, or motivate employers to avoid hiring on particular terms. They show that there is little qualitative difference between attachment and non-attachment: that they share something fundamentally in common.
The underlying condition is an encompassing, totalizing, not-quite-annihilating unfreedom. Interdependences of labor commitment and credit provision, of tenancy and credit provision, of productivity and credit provision (as discussed above) and of prices and credit provision (see below), make a self-reinforcing cycle of poverty and non-possibility of opportunity. Benefits gained from contract employment are lost in direct consumption loans; benefits gained by security of employment in attachment are lost in bargaining power and fixed, noncompetitive remuneration (in cash or kind). Thus traditional relationships of landed employers extending credit, tenancy or wages to landless or landpoor employees should not be mistaken as patron-client relations, fostering congenial familiarity or understanding between parties, or the bestowal of advantages (as do both Bardhan and Rudra 1980 and Rudra 1987). This, following Brass, is to “downgrade...or...discard...altogether” debt bondage by another name. It is to assume complicity with the non-recognition of chronic indebtedness for the self-employed, casual and attached laborer alike, of rent-payment and stagnant, non-living wages despite productivity increases on labor hiring farms—as bound together in a singular denial.
Table G illustrates that inputs and outputs of small farmers come from and stay within the village itself. This indicates, perhaps, an infrastructural problem, the inaccessibility of markets due to poor roads, transportation, etc. (Bardhan 1985). Alternately it suggests that prices did not justify transportation to central markets (and government subsidies). Finally it suggests interlinkage: that debt owed could be repaid only to local buyers, often landlords and moneylenders. The effective convergence of landedness and merchant-broker activity is well known. Reddy (1985) relates that "...since the cotton buyers in the villages happen to be the village moneylenders-cum-traders in fertilizers and pesticides...the small and medium sized cultivators are forced to sell their produce to these brokers/traders in the villages, i.e., the credit and output markets are interlocked." Satyanarayana (1991) observes that the marketing of agricultural produce needs, or cannot prevent, mediation by merchant capital. This mediation extends beyond the buying and selling of produce, as Bardhan and Rudra (1978) duly observe: "the majority of villages where professional moneylenders lease out land...[are]regarded as technologically advanced." Bharadwaj (1985) remarks that access to commodities markets both for buying and selling is constrained by the initial resource position of the buyer or seller in relation to the merchant: when sale or consumption of commodities depends intimately on provision of credit from the buyer or seller, the exchange relation becomes a social relation. By implication, production relations themselves become socio-economic interrelations of class.
Inflation suffered by the landpoor and landless can be described from a variety of approaches. First, from Table G it is clear that the lowering rate of return per rupee-spent by caste and size of holding effectively constitutes inflation for the landpoor. Second, given the evidence that real wages tend to stagnate in areas of high productivity and high incidence of labor hiring-in, lower prices obtainable by the landpoor in Table G mean that wages lag behind prices. Third, given evidence that wages tend to be more stable in non-advanced areas, in a region of high productivity like Guntur, money wages are vulnerable and cannot be relied upon as store against fluctuation in wages or prices. Not only real wage decreases and stagnations, but deviations from pricing trends seriously effect the incidence of poverty (Bliss 1986). In this connection, Ahluwalia (1986) records that "...[a sharp rise in food prices] could take place without any change in production per head in the rural economy because of a general inflation in which food prices move up with other commodity prices. Food prices could also rise owing to developments outside the rural economy such as a rise in export demand or reduced imports of food. ...Even if the rural poor are self-employed peasant producers who produce goods other than those they consume, a rise in food prices could accentuate rural poverty if it is not matched by a rise in the prices they receive." Patnaik (1985) explicitly links inflation at the local level to extralocal developments, showing that increases in state expenditures not matched either by sufficient increases in agricultural or industrial productivity, or increased taxation on property or luxury consumption of the wealthy, are financed out of the savings—or increased demand for credit—of wage and salary earners. Either way, landlords, capitalist farmers and manufacturing capitalists enjoy increased profits and/or interest payments.
From the foregoing discussion and Tables A-G, the following general theses on interlinkage emerge:
1) Poor tenants pay exorbitant interest rates on consumption loans in higher rents or lower prices for their produce; contracted workers pay exorbitant interest rates on consumption loans in lower wages or unremunerated labor. These interlinkages constitute debt bondage.
2) The high rent or usufruct tenants pay to landlords, the usurious interest rates debtors pay to creditors, and the low wages employers pay to employees form variations of a singular class differential, explicitly so when landlord, creditor or employer are combined in the same person, institution, family or nexus of families.
3) The principle security available to the landless against loans is sale of labor. For the landpoor securities may consist in the price- obtainable for output (varying seasonally—generally lower after the harvest), sale or lease-out of land accompanied by pledge to work that land for a wage or for the price of sale of its proceeds, sale of labor for wages, donation of labor for rights to lease-in land, or possibly small savings. These securities are only securities to the extent that they can be used as leverage against efforts by the agrarian elite to tie land, labor, credit and commodities markets together. To whatever extent the tying-together of markets succeeds, these securities are fetters.
4) Survival consumption, in the form of credit taken and debt borne, becomes an economic liability, limiting the amount of capital a family can purchase, the amount of land it can lease-in or purchase, and indeed the food it can afford: in this sense the family itself is an economic liability.
5) For the landpoor, landholding becomes a burden where it should be an asset: family cultivation in the context of near-total dependence on consumption loans at usurious interest rates acts as a disincentive for hiring-out of family labor and for hiring-in of outside labor where it might be productive: "self-employment" blurs into self-exploitation.
6) Family labor effectively encourages noncapitalization of wages, i.e. nonproletarianization, down to a certain limit, below which the family is forced to abandon small farm cultivation and sell or lease-out their land.
7) The uncapitalized wages a laborer loses in exchanging labor for terms of lease, price, personal services or credit may be more than the value of land worked, price paid, services unbought or interest accrued. The very absence of markets to retrieve liquid wages, saturated into the structures of production, is proof that labor-value flows recurrently in invisible payments.
8) Rents and prices likewise may bear little correlation to the real or potential productivity of land.
9) Lower wages and prices suffered by small cultivators result from higher rates of interest on loans, which is to say smaller farmers pay the interest on their loans in lower wages and prices.
10) Types of crops to be grown are stipulated as part of the provision of credit, and as effectively presold, preclude crop differentiation which may be to a cultivator's or tenant's economic or nutritional benefit.
11) Greater usury in the way of credit and un- or underpaid services is more common in more productive areas: although production loans are more important for both rich and poor in productive areas, the landpoor will be more restricted to noninstitutional credit, precisely because a larger share of institutional credit is taken by the landrich, who promise to deliver guaranteed returns—although, ironically the landrich have a poorer record of repayment than the landpoor.
12) Wealth does not trickle down with rising productivity: depressed and compromised wages derive from usurious credit and a higher incidence of wage-labor in capital intensive farming.
13) Higher debt borne by the rural poor as a result of higher interest rates, themselves a result of greater productivity, will depress or stagnate wages and prices obtainable by the poor. This in turn encourages attachments formal and informal to employers and creditors, and ensures the availability of labor necessary to maintain productivity.
14) Capitalist farmers are as likely to become creditors as are landlords or moneylenders.
15) Eviction and insecurity of tenure are higher in more productive areas, not necessarily because of fear of ceilings legislation, but because of vulnerability of the landpoor to landowners-cum-creditors.
16) Though the majority of lessors may be small cultivators, tenancy is profitable for middle cultivators, who tend to lease-in from the landrich.
17) With increased production and profitability comes a decline in willingness of large landholders to lease-out land, and a complementary increase in the leasing-out of land by small or marginal holders to large holders: likewise an incentive for poor landholders to rent out their land and work on it as casual laborers.
18) Pure tenants with no assets other than landtenure invite greater exploitative capacity of landlords to demand rent in labor or kind.
Interlinkage, as I have presented, it describes a condition of endemic pauperization of the primary producers of agricultural wealth. As an analytical tool interlinkage names, on the one hand, the network of economic exchanges that forces cultivators into chronic debt and systemically ensures their vulnerability. On the other hand it names the modes of ruling class surplus appropriation and their attendant socio-political forms. I have tried to show interlinkage is not a peripheral phenomenon in Indian agriculture. On the contrary, agricultural productivity at many levels is founded precisely on the economic vulnerability of the tillers of the fields (i.e. a dire want of agrarian reform), in combination with the deliberate frustration of effective landreform. Moreover, I have argued that this political-economy of vulnerability designates a class relation. All of this I have been presenting in light of data from the post-Independence period, with attention specifically to the paddy-producing coastal areas of Andhra Pradesh. At this point I wish to turn to the origins of interlinkage. To what extent is interlinkage an historical phenomenon that predates the colonial period? In what ways did it acquire decisive forms during the colonial period? Can interlinkage be discerned in a strong form in the Mughal and other early modern/medieval states? What aspects of it are evident in antiquity? Though a thorough historical explication is beyond the range of this essay, I wish to trace, skeletally, the genealogy of structural, vulnerability-based agricultural production.
It will be helpful to begin by considering briefly the ongoing debate over trends in the “mode of production” in the Indian rural economy. The discussion has attempted to evaluate the changes in agricultural production and agrarian society in the subcontinent in terms of the well known European historical-economic development articulated by Marx and Engels, describing the historical transitions between primitive communism, the so-called slave mode of production, feudalism, capitalism and finally, socialism (Hobsbawm, 1980). In light of the preceding discussion, the debate raises questions of whether commercialization and capital intensification in Indian agriculture constitute a transformation to agrarian capitalism from feudalism; what feudalism thereby means in the Indian context (regardless of whether the question is answered positively or negatively); whether Indian agriculture is better described in terms of some hybrid of capitalism and feudalism, or an altogether exceptional mode of production; what the nature of class formation is in Indian agriculture; what generally the impacts of colonialism and imperialism have been on Indian agriculture; how state formation has historically functioned in relation to agrarian political-economy.
Following Hamza Alavi (1981), a mode of production refers to the logical and mutual coordination of the following:
a) a determinate type of ownership of means of production
b) a determinate form of appropriation of economic surplus
c) a determinate degree of development of the division of labor
d) a determinate level of development of productive forces.
A mode of production refers to the structure of a dialectic between social relations of production and economic forces of production, which need not and often do not correspond in a one-to-one fashion. Mode of production can be distinguished from the concept of "social formation," (Rudra 1978; Alavi 1981; Chattopadhyay 1972) which names a particular, geographically and temporally bounded society, its specific political-economic organization and cultural features. Social formation is a descriptive concept. Mode of production naturally includes social formations, but is an analytical concept. As such mode of production does not correspond to otherwise designated historical epochs. The feudal mode of production does not refer simply to the medieval period, the capitalist mode to the colonial or postcolonial period, etc., but one or several modes of production may be evident in any particular geo-historical situation. The discussion on the mode of production has tended to focus empirically on the post-Indepedence period. Of course there is no reason not to extend the empirical consideration into the past as well.
A. Bhaduri (1973) argues in a seminal article that the mode of agricultural production in India can rightly be called feudal (or semi-feudal) as long as the following characteristics are widely evidenced:
1) Extensive, non-legalized sharecropping
2) Perpetual indebtedness of small tenants
3) A rural “ruling” class acting both as landowners and creditors to small tenants (i.e. landlords investing in usury, financier-brokers investing in land)
4) Small tenants having incomplete access to the market forced into involuntary exchange through the peculiar organization of these “markets.”
It is well to consider these characteristics point by point. Sharecropping, legal and non-legal, and both registered and unregistered tenancy are widely associated with feudal economic formations (see also Alavi 1981) because they constitute systemic insecurity for cultivators. They entail, Bhaduri argues, the constant possibility of eviction, and also the extreme personalization of landlord economic power over tenants. The claim for the possibility of eviction depends to some extent on the collection of information on actual evictions, but the threat of eviction is not necessarily quantifiable, or if it is, as during the period of widespread agrarian reform, must be measured inversely. Neither threat of eviction nor tenancy (rent payment) can themselves be considered exclusively feudal: both eviction and for-profit land rental have accompanied the so-called green revolution (and even for-profit sharecropping is reported—see P. Bardhan 1985). Personalization of relations, sometimes in the form of political coercion, sometimes in the form of terror, does seem to accompany what can be called feudal forms. Whether this personalization as extra-economic is foundational to a feudal mode is something I will contest below.
Perpetual indebtedness cannot be considered exclusively feudal. Perpetual indebtedness is rightly considered a condition of unfreedom insofar as it means the inexorable dependence of the cultivating classes on the rural ruling class (which include large and middle farmers cultivating with hired labor, large and middle farmers leasing out to tenants, large and middle farmers leasing in land, large and middle farmers who personally participate in cultivation and who do not, and moneylenders and traders) for subsistence. Indebtedness becomes unfreedom when basic survival for the majority of actual workers is tied into a whole complex of economic and superstructural relations of dependence. Moreover, chronic indebtedness may occur systemically with or without wage-labor, either if wages are sufficiently chronically low, and/or if rent either in labor, cash or kind is sufficiently high, and in either case if the need for consumption credit is ineliminable.
Bhaduri's third and fourth criteria, the collapsing of landowners and moneylenders, and involuntary exchange resulting from incomplete access to markets—are essentially statements of interlinkage. These form the most important part of his explication. Bhaduri asserts that the feudal or semi-feudal landowner "exploits the tenant through his traditional right of land as well as through usury and the economic basis of semi-feudalism is the combined operation of these two modes of exploitation.." This is to say that throughout the year tenants are typically dependent on the landowner for advances to pay landless laborers, and tenant and laborer alike are dependent on the landowner for consumption loans for food, clothes, medecine, family emergencies and special expenditures like marriages. At the end of the crop season, the advances are deducted from the tenant's or sharecropper's keep, which may leave the tenanat/sharecropper as little as 12%-15% of the total output. This does not include payment on back loans at exorbitant rates of interest. Cultivators in such conditions survive only by taking additional, often larger loans, and in this way the cycle of permanent indebtedness from chronic poverty replicates itself. The efficient cause of the cycle is the necessity that the tenant lease in land from the same individual to whom he is permanently indebted—or if not the same individual, the same small group of individuals.
What Bhaduri calls involuntary exchange also follows from a land-credit interlinkage. Because of the constant compulsion to settle past debt, the cultivator produces prinicpally for landowners and moneylenders, who offer depressed prices against the debt outstanding. Cultivators do not produce crops for sale as commodities on an open market, except to the extent that they can maintain some leverage against the forced, often early sale to landlord-creditors. Subsistence farming is ensured by the personalization of markets for indebted cultivators' crops and the fragmentation of access to open markets (the impediments to large, regional rural markets may be literal: roads and transport may be kept deliberately inadequate if the political power of one or several landlords is great enough). The lack of access to better organized money markets (banks, cooperatives, etc.) and those markets' clear preferences for large-scale investment in entrepreneurial farming over subsistence agriculture, certify the small cultivator's vulnerability to usury and to distress sale of crops (i.e. selling early). In short, what Bhaduri wishes to call a feudal mode of production is characterized by a linkage of credit, labor and prices, such that the cultivator unremittingly earns lower wages, receives lower prices, and/or pays higher rent—against debt which is perpetual because it primarily covers consumption costs. The cultivators' loss in income though lower wages, lower prices or higher rent, or a combination, "more than outweigh[s] the gain from...productivity." Low productivity often results, but not necessarily. As Bhaduri and Patnaik (1971) pointedly note, the level of productivity may be of secondary importance to rate of return on the existing system. Higher productivity by capital intensification may be less than income from usury. It is necessary to view the development of capital intensification of agriculture not in terms of a metahistorical destiny, but in terms of the real considerations entertained by the very persons who would elect to change. The system that Bhaduri describes as feudal has the advantages of securing control of labor and political efficacy for its elite, and these are not to be considered insignificant.
Rudra (1978) offers an expanded set of essential characteristics of a feudal mode of production in response to the well known Kautsky-Lenin Laws of Development of Capitalism in Agriculture. These include the law of increasing returns—large farms are more productive than small; differentiation of the peasantry—land is increasingly concentrated into fewer and fewer hands and as a result there is an increase in the agricultural proletariat; decline in sharecropping; production for market rather than for subsistence. The following are Rudra's criteria:
a) extraeconomic coercion by landlords, including both political control and direct physical intimidation
b) direct control of the means of production by the primary producer, i.e. the nonreduction of labor to a commodity bought and employed by capital along with other forces of production
c) petty and localized production, i.e. production for local consumption rather than for transregional, in some cases regional, and possibly standardized markets
d) production for use rather than for exchange, i.e. non- or minimally commodified production (commodities representing principally a heightened exchange value) ;
e) labor or kind as opposed to money rent; in many shorthand feudalism, monetization is taken as the decisive index of capitalization—for the problems with this view, see below
f) low rate of technological progress, sometimes characterized as an extraeconomic or cultural adversity to technology or preference for labor-oriented production
g) landlords and/or moneylenders using rent for personal consumption rather than productive investment in land.
Against Kautsky-Lenin, Rudra notes first that capitalism may generally be characterized by a disappearance of petty production, but some kinds of mechanical improvements are scale-neutral, such that small capital-intensive farms may be sustainable by scale-neutral capital inputs. Correspondingly, land values may increase while average farmsize does not. Insofar as there is a concentration of capital without a concentration of landholding, i.e. the persistent fragmentation of landholdings, and rigidity of the land market restricts the accumulation of land—productivity may still increase due to the input of certain productive capital technologies. Size productivity cannot straightforwardly be measured simply by surface area of land owned, but must measure the value of land possessed. Second, as noted above, tenancy and sharecropping may be adaptive. Moreover, while rent payments may be burdensome for subsistence farmers, they may be profitable for capitalist farmers who find renting-in land more attractive than purchasing land from small or large landowners (see also Patnaik 1976, Haque and Parthasarathy 1992). Tenancy may likewise "adapt" to spatial particularities, replacing compulsion and coercion with written contracts (Chattopadhyay, 1972, Bardhan 1985). Finally, Rudra asserts that cashrent payment by capitalist farmers (“reverse tenancy”) may not result in capital accumulation for small rentier-farmers, or a rise in their standard of living above the subsistence level. Capitalist farmers may themselves do a business in credit, so that usury may assume an aggressive, monetized form, and indices of monetization may not indicate a reduction of economic distress. Indeed, Rudra argues, the prevalence of moneylending is not a straightforward index of feudalism insofar as loans may be taken for capitalist production as well as for consumption, and so need not result in perpetual indebtedness.
Clearly to attribute any characteristics to a "feudal," "semi-feudal or "pre-capitalist" mode of production is to claim that they appear exclusive of developments otherwise associated with capitalist agricultural production, particularly decline in tenancy, increase in market rather than subsistence production, monetization, increase in large farms following capital accumulation, and increased social differentiation (i.e. increased land concentration and increased numbers of agricultural laborers). For Rudra, only his condition c), the necessity of petty and localized production under feudalism, analytically accords with Kautsky and Lenin, who assert capitalist agriculture necessarily to be oriented toward the market rather than toward the cultivator's self-consumption. Rudra argues, in short, that tenancy, cash rent, nonconcentration (even fragmentation) of holdings, usufruct and usury are compatible with, and occur within capitalist development. Rudra is loath to call an economy in which capitalist elements survive and indeed flourish, "feudal."
While it is true that these characteristics may separately or together materialize in capitalist production, that is may be assimilated into capitalist production, Rudra does not answer the implications of the systemic interlocking of these elements as a function of the destitution of labor—as elucidated both by Bhaduri and by his most vocal critic, Utsa Patnaik. Rudra's argument is of a specious type. It is rather like saying that because a hammer can be used as a toothpick it is not essentially a hammer. Rudra does not challenge the fundamental interlinkage of labor, credit, commodity that sustains economic vulnerability and class differential. Instead he articulates alternative, capitalist, applications for it. Nor does he recognize that his applications are formalist in nature, and represent a disaggregate, distended capitalism. Rudra's emoluments would exist as features of capitalism only in the context of their non-capitalist application. His exceptions to prove the rule.
Rudra's oversights reach their limit when he argues the following:
[i]t seems to us that the one common denominator among the essential and characteristic features of feudalism which have been recognized by different scholars is the extra-economic coercion that it involves. 'A typical expression of such coercion would lie in the landlord possessing the legal power to compel him to work for him gratis or to serve him in many other ways, curtailing thereby his individual liberty.' (Rudra 1978, p. 963). This presumably means that Rudra understands his condition c), petty/localized production, to occur as a result not of economic constraints on primary producers but political-judicial coercion, physical threat or patron-client relations. He does not treat the interlinkage of artificially low wages and prices due to the tying-together of credit, labor, land and commodities markets as a form of coercion. As such he does not explain how or why it should be treated as extra-economic—when it seems precisely economic. Nor does he offer any argument for why interlinkage should be viewed as in any primary way extra-economic because it may be accomodated by modes of production specified as distinct social-relationally and otherwise. In Rudra’s contribution are the seeds of an insight that the mode of production in Indian agriculture is not neatly predicted by the historical material dialectic patterned by Marxists for the west. From this insight Rudra concludes, however, that modes of production are idealities superimposed upon spatially and temporally specific relations of production which are admixtures: landlords sometimes pay wages, sometimes receive groundrent; exploitation is sometimes visited upon the landless and landpoor by landlords, sometimes by capitalists, sometimes by rich peasants. In Rudra’s view, we can describe only the relations of production in the Marxist dialectic between productive forces and relations. As such we can neither pattern the past nor predict the future but only describe spatial and temporal particularities with greater or lesser accuracy.
Rudra errs perhaps in not recognizing interlinkage, which he explicitly describes in his 1978 article co-authored with Pranab Bardhan, as fundamentally a condition of coercion of any kind. Beyond that he errs in representing it as essentially extraeconomic in character. It is, as I have tried to show, precisely economic: it constitutes a traditional economic status quo into which capitalist elements may be introduced and assimilated, or may perhaps struggle to overcome. This is not to say that the interlinking of markets may not occur by means of personal coercion by stronger over weaker parties, assuming a superstructural dimension. The existence of superstructural (read: political, ideological) support for interlinkage was never at issue. However, it is imperative to remember that this support is supporting something—an economic structure of debt-compulsion which operates by its own force.
Against Rudra, Patnaik (1971, 1976), after Daniel Thorner, emphasizes the ground rent “barrier" as the critical element in the transition from feudal to capitalist modes of production. Patnaik argues that ground rent constitutes an impediment to profitability which capital intensification and/or capitalist relations must surmount. The profitability of the traditional (what Bhaduri calls feudal) mode of production is precisely its success in controlling labor to frustrate the formation of a free proletariat. The traditional mode operates by maximizing destitution, particularly by maintaining endemic underemployment (“a vast underemployed reserve army”) to ensure bare subsistence income. The traditional landowner may employ wage-labor, but does not do so for proft—rather uses wage-payment to bind the cultivator into the pattern of dependence already established. It must be stressed that the traditional mode of production is neither a relic nor a failure of economic rationality. In Patnaik’s words (1971, p. 42), “[t]he choice between operating with hired labor and leasing out to tenants may represent for such landowners a purely contingent, reversible decision taken on the basis of current circumstances—e.g. the comparative terms on which labourers and tenants are available in that locality.” This is to say that wage-labor is a necessary but not a sufficient condition of capitalist production. By itself it is not an indicator of the mode of production, but requires a determination of its use. Likewise production for the market is only a necessary condition: commercialization is not by itself an indicator of the absence of vulnerability-based production if, for example, it is generated by family labor.
For Patnaik, the fact that capitalist elements exist and thrive within traditional modes of labor-expropriation, with or without decisively replacing them, is no basis for concluding that those traditional modes of production are not essentially feudal. As Patnaik (1971) rightly observes, with Rudra (1978), Banaji (1972), Alavi (1975, 1981), the operative characteristic of capitalism is not merely the appropriation of surplus value, but the accumulation and reinvestment of that surplus on an ever-expanding scale. Precapitalist methods of surplus use do not involve capital investment and intensification, but rather moneylending, letting-out of land to tenants, and trade. The rent barrier thus might be better termed the non-capital investment barrier, or the destitute-labor barrier, because its profitability depends on the sedulous debt of primary producers under the weight of hunger-rent obligations, consumption and production loans, artificially low prices obtained, and pervasive underemployment. In short, what Patnaik calls the rent barrier is better termed an interlinkage barrier, dependent as it is on the pauperization of the cultivators. Insofar as feudalism describes a mode of labor control based on the economic, and perhaps socio-political obligations of cultivators to the demands of a surplus-appropriating class of intermediaries, the traditional mode of production in India can rightly be called feudal.
This appellation is only a matter of convenience, given not to procrusteanize Indian agrarian forms into western categories, but to expand those very categories. At this point it is important, following Marx (1980) and Byers (1985), to note that feudalism as it existed in Europe was transformed into capitalism not simply out of its own internal dynamics, including land enclosures, migrations to towns, formation of an urban proletariat (free labor) from guild control, etc. The critical element of an ever-expanding market was provided by the opening of overseas trade which developed colonial possessions and eventually, imperial possessions. Political control of overseas territories meant not only market expansion, but also the wholesale drain of wealth and natural resources from countries around the world to Europe, providing an indispensible capital base for new capitalist operations. In short, capitalism developed hand in hand with colonialism, and as such developed not merely as a European phenomenon, but as a global phenomenon.
The rise of capitalism-colonialism as a global phenomenon means that changes in the mode of production in colonized countries likewise must be analyzed globally. David Washbrook (1990, 1994) has pioneered studies of the shifts in Indian agriculture under the impact of colonialism. Broadly, the penetration and eventually domination of Indian trade by the European trading companies, with the attendant political struggles and attainment of supremacy by the British, destroyed the thriving precolonial manufacturing activity. By the early nineteenth century, a massive migration of labor back to the countryside was underway, along with the emmigration of more than a million Indian laborers to southeast Asia, Africa, and the Caribbean as indentured servants. At the same time, as noted in the in the foregoing discussion, the colonial government radically hardened the agrarian structure in forming their alliances with the classes of subinfeudated intermediaries in the Permanent Settlements. Though the British did not introduce rackrenting and subinfeudation into the agrarian situation, they manipulated production toward the imperial market and destroyed subsistence in food production in India. The South, by the end of the century, saw a steady decline in yields of jowar (millet), the staple grain of the majority of the population, and a corresponding rise in prices. Perhaps the worst legacy of the colonial regime (and among the least studied) was the virtually constant presence of famine in one or another part of the colonial dominion from the mid-nineteenth through the mid-twentieth centuries (Bhatia 1963), amidst the increasing drain of revenue. Revenue drain was foundational not only to the extravagance of the colonial government itself, but to industrial development in Britain and especially to the financing of the imperial armies and the Empire’s martial demands. The cumulative result in India was casualization of wages from increases in consumption loans, deficient supplies of basic goods, and the proliferation of a hungry, landless peasantry. In short, colonialism effected a feudalization of India. Capital intensification in agriculture in the form of an overall increase in wage-labor simply aggravated the coercive power of economic interlinkages and reinforced extraeconomic coercion by calcifying the rural class differentiation. Whether this feudalization amounted to a refeudalization or a genetive feudalization is in certain respects immaterial. The point is that the development of capitalism in Europe (particularly Britain) and the relative capitalization of agriculture and manufacturing in India during the colonial period buttressed material interlinkages which are rightly called feudal. 8
Certain scholars (Banaji 1972, Alavi 1975 and 1981, P. Chattopadhyay 1972) have proposed the notion of a “colonial mode of production,” by which imperial capital disarticulates the internal economy of the colony and integrates the disarticulated segments into the metropolitan economy. The colonial mode represents an effort to specify the nature of capital intensification during the colonial period—to account for the external source of investment and the appropriation of surplus not just by local or regional elites, but the exportation of that surplus to Europe and other parts of the British Empire. Thus Alavi offers the following comparative view of modes of production:
Table H: Structural Conditions of Various Modes of Production
(Source: Alavi 1981)
Alavi’s scheme may be taken as clear and relatively comprehensive. He avoids the common problem of equating capitalization with monetization and wage-labor by speaking rather of the freedom of labor from the direct possession of the means of production—such that laborers earning a wage but not alienated from the means of production are still considered feudal. He correctly separates economic from extra-economic coercion, but would also be right to call the extraction of surplus in the feudal mode economic. He emphasizes that the capitalist mode involves more than just the expansion of a localized power structure to a translocal structure, but the creation of a bourgeois state with metropolitan and peripheral (various colonial) forms. Likewise, the expansion of the economy from self-sufficient localized production to generalized commodity production precisely involves the alienation-cum- commodification of labor within particular peripheral forms. In some sense this might also be understood as the overcoming of the personalization of production articulated by Bhaduri. Finally, Alavi includes Patnaik’s point (following Lenin) on production for static versus ever-expanding markets, again within particular colonial forms.
While Alavi’s contextualization of capitalist production in India in terms of the prerogatives of imperial capital is doubtless necessary, Alavi may not be contextualizing enough. Are the characteristics of the feudal and capitalist modes of production uniquely functions of colonialism?
Although Marx never explicitly took up the analytical tool of the mode of production, he did ascribe to India what he termed the Asiatic mode of production, which he took to operate in conjuction with what the British colonial historians, particularly Alexander Dow (1770, 1772), called Oriental Despotism. These theories, criticized roundly for their fautly empirical foundations (for a summary see Thapar 1992), posited an all-powerful, despotic Oriental potentate who creamed agricultural surplus off the peasantry through an effective bureaucratic machinery; the absence of private property in land; self sufficient, isolated village communities, politically autonomous except for the revenue collection activity; a state-monopolized irrigation system; the absence of important urban centers and of trade; and a king invested with powers of divine. While Marx, who never visited India himself, was wrong in asserting the absence of private property, the undifferentiation and self-sufficiency of villages, the absence of cities, the unchallenged supremacy of the Emperor and singular effectiveness of the Mughal bureacratic machinery, nonetheless certain corrolaries of his assertions remain valuable. Most important of these, as Irfan Habib points out (1985), is the recognition that agrarian subjection in the Mughal state did not proceed along the lines of European serfdom. Rather the village community was the object of subjection. This is not to say that collective subjection was not accompanied by individual subjection in various forms—by village strongmen who exercised political control, lent money, hired laborers, etc., and by internal divisions of jati (such that certain jatis formed the bulk of the landless population)—but that even individual subjection proceeded by way of the individual’s position in a collective schema.
Insofar as surplus appropriation either by the state (through the jagirdars) or hereditary claimants (zamindars) 9 was successful, and insofar as state and hereditary claimants to agricultural surplus were relatively cohesive as against the claims of the primary producers themselves, rent payment and tax collection coincided in the Mughal dominions. Of course, the payment of rent/tax to superiors, whether zamindari or jagirdari, constituted a major burden on primary producers, and may be taken as the fundamental class differential, beyond which caste stratification and landlessness constituted additional burdens. Although jagirdars and zamindars did make separate and at times competing demands on agricultural surplus, together they formed an appropriative class of intermediary assessees which could, and did, convert the surplus into commodity production and circulation both inside and outside the village under the aegis of the state, or perhaps more precisely, the particular spatial-temporal coalition of ruling elements.10 The state, then, should be considered not just in terms of centralized political control, but in terms of a centralized fiscal structure. This fiscal structure supported both the immiseration of primary agricultural producers (which facilitated both jagirdari and zamindari surplus extraction) and extensive commodity circulation (which was accompanied by extensive banking and insurance operations, and the growth of urban centers specializing in the manufacture of commodities for distant markets). Although commodity production was directly dependent on state-sponsored agrarian exploitation, i.e. effectively centralized fiscal control, this exploitation did not destroy the self-sufficiency of the local economy. Indeed as Habib shows (1982), there was practically no change (either up or down) in the yield of foodgrains per unit of area for Agra and Delhi between 1540-45 and 1870.
Habib notes that centralized control was not always effective, in that the insecurity of jagirdari holdings increased fiscal pressure on the primary cultivators (jagirdars extracting as much as possible), which led sometimes to agrarian revolts. During breakdowns of imperial power, zamindars commonly reasserted their authority, becoming more “feudal” in character. Habib claims that this increase in local power stands in contradistinction to Europe, which saw in the breakdown of feudal power in the countryside the development of a regime of private landowners and a rent-paying free peasantry. Out of the contradictions of this system grew the expansion of commerce, the universalization of petty production, and eventually the growth of nationstates and capitalism. This development, he says, did not occur in India until the second half of the nineteenth century, and then only as a result of colonial rule. What remains unclear in his explanation is just how the rent-paying primary producers in Europe were "freed" from direct control over the forces of production to form a rural and then urban proletariat, but rent/tax paying primary producers in India continued to be obligated to the surplus-appropriating classes. Why, in other words, did landless British cultivators and British tenants not depend (if indeed they did not) on credit from their former feudal overlords in such a way as to prevent their capital accumulation? Why, on the other hand, did not Indian rent-paying landless cultivators and tenants transform themselves into a free, petty commodity producing class?
Some answers may lie, as already suggested, in the parallel developments of colonialism and capitalism. Broadly, the British peasantry did not develop into a proletariat until the development of an international market and an influx of bootied capital from overseas. A proletariat in Britain did not develop, that is, based simply on itself as a market for its own goods. On the other hand, Tapan Raychaudhuri (1982) argues that a sizeable middle-income group—including the lower ranks of the bureaucracy, professionals, holders of rent-free tenures, etc.—constituted a proportionately large internal market for comfort and luxury goods during the Mughal period. The grossly inequitable distribution of surplus “both inhibited and stimulated” its demand, such that it did not develop into an economy of scale, and at the same time was sufficient to prevent the accumulation of surplus either by primary agricultural producers or artisans (which often overlapped in the countryside) and thus its transformation into productive capital. While there is evidence of a vast market for basic commodities, long-distance trade in luxury goods provided the most lucrative markets. Such markets were sufficiently liquid to support a certain amount of commercialization, but were not ever-expanding as would have been necessary for the development of a full-scale capitalism. As a result, both Indian agriculture and manufacturing remained labor-intensive, and necessary divisions of labor did not seed awareness of workers as “free” proletarians, but rather proliferated jatis.
As such, the reasons why the Indian domestic market did not become a market for Indian manufacturing but did become one for European manufacturing cannot simply be the arrival of European trading companies and their usurpation of markets, though such competition was undoubtedly significant. Rather, to the extent that petty manufacturing was not commercialized, standardized or centralized, but was dispersed, local and family based, it was, like agricultural production, immiseration-based, and so proved a “barrier” to state-merchant-landed interests. It is not, as Raychaudhuri claims, that Indian manufacturing or agricultural production could not develop past a “hoarding” mentality, but that the conversion of savings into productive investment was simply not sufficiently profitable. For state and local elites, surplus appropriation from the land, savings and investment in luxury good production was lucrative enough. Such a productive-investment barrier can rightly be considered an indicator of interlinkage. Class exploitation in the Mughal period was sufficiently effective both in agricultural production and manufacturing to prevent the growth of mass purchasing power, so that the production of food and petty commodities was minimally but basically satisfied, without need for mass productive investment by the coalition of state-merchant-landed interests.
The Mughal system of generalized, minimal self-sufficiency in food and commodities based essentially on peasant-artisan self-exploitation was undone during the colonial period by the colonial government’s destruction of Indian manufacturing, the forced importation of basic commodities which precipitated a slow exodus from the towns to the countryside, and the consequent overabundance of rural labor. The economic harships unleashed and legitimated by the Permanent Settlements further caused an overall decline in agricultural productivity, hence vulnerability to ecological crises and the widespread incidence of famine. Interlinkage made the Mughal political-economy vulnerable to the colonial threat, and the British secured their markets and their rural power base by destroying the Mughal equalibrium—not by breaking interlinkage apart (thus precipitating proletarianization), but by tightening it.
If interlinkage characterized the Mughal political-economy, it was also present, we can gather, in medieval south India. Alaev (1982) stresses the importance of the landless population, which constituted as much as 25% of the total agricultural population, as well as the non-expansion of agricultural productivity (in fact the overall decline) despite the widespread availability of unoccupied lands. R. S. Sharma’s study (1985) of the inscriptional and literary evidence of state sanctioned land grant charters establishing usufructory and taxation rights, and of various classes of subtenants and leaseholders—suggests that hierarchical control over land and large-scale subinfeudation existed in the north at least from the eighth century onwards. Sharma believes that the beneficiaries of such titles generally were able to make their economic claims and administrative powers sufficiently comprehensive to convert their rights into possessions. In this way they controlled agricultural production by the efficacy of their appropriation of surplus, and so controlled labor. He suggests that such intermediaries did not constitute such a burden on primary producers to impede agricultural production. Likewise he maintains that the agrarian base was strong enough to support state expenses—priests, military, etc. It is perhaps difficult to ascertain, but it seems possible that the Mughal structure of interlinkage may have inherited much from preceding systems.
D. D. Kosambi’s insights (1965) into the growth of the Mauryan state push, perhaps, the origins of interlinkage into antiquity. According to Kosambi, the sixth century B.C.E. saw the rise of new classes in the Gangetic basin, in particular a propertied class not bound by tribal regulations and obligations. The first evidence of sreshthis (lit. superior, pre-eminent), wealthy traders and financiers who received the respect even of despotic kings, dates from this period. More important is the changed significance of the word gahapati (Skt: grihapati, lit. lord of the house) from non-royal host of a Vedic sacrifice to independent patriarch of any caste whose wealth was gained in trade, manufacture or farming, but in any case not measured merely in cattle. The gahapatis germinated a political-economic sensibility that would build the Mauryan empire: as agriculturalists and traders they suffered from the constant warfare that preceded the yajna fire sacrifice, and the burden of animals requisitioned for the sacrifices without payment. Their livelihood depended on political peace in the countryside and on the preservation of safe trade routes. This led, in time, to the ideal and in large part the realization of “universal monarchy,” a single state that could end petty warfare, police the countryside and regulate trade. Gahapatis and sreshthis formed the elite of the newly emerging free peasantry, “Aryan” homesteaders who had begun to clear their own land, and whose land, crops and animals were not the property of the tribe or subject to redistribution by tribal councils. As independent agriculturalists and traders they developed an interest in private control over their surplus and means of production, in contrast to the interests of Yajurvedic kingship whose demands inhibited trade and production. On the strength of Kosambi’s insight that agricultural and commodity production was controlled by a cohesive class which formed the foundation of state power—we may speculate that Ashoka’s true legacy was the development of an ideal of political-economic interlinkage.
Though this historical survey is strictly preliminary and a just presentation of the argument demands far more extensive research than possible here, I hope to have indicated some possible roots for the problem of the mode of production. My chief suggestion is that the mode of production may effectively be understood as a function of interlinkage historically.
Sen Gupta (1977) remarks that the overlapping of landowning and moneylending in precapitalist production indicates that landowners-cum-financiers have “multiple channels” for production and reproduction of wealth, some of them entrepreneurial. Entrepreneurial channels do not necessarily come to dominate production, but operate to a great extent within the traditional interlocking of markets. Capital intensification introduces more than a semantical problem over modes of production. If he is right, it becomes unclear whether the absence of competitive labor markets, the widespread underdevelopment in marketing, low capital accumulation by the poor, and the persistence of subsistence farming result in compatibility (rather than competition) between traditional labor-expropriative and capital intensive agrarian systems. Such compatibility suggests that the real problem is the mode of political economy that permits the cohabitation of traditional landlordism on the one hand, and on the other capital intensification, development of falsely competitive markets, and uneven “freeing” of labor from the means of production and for wage-earning (Byers 1981). In asking how and why different modes of production persist and develop at different rates, we are asking about the rural ruling class structure: how it appropriates various and methods of economic organization to maximize its own control.
If we accept Sen Gupta's suggestion (which both Rudra and Patnaik accept) that modes of production need not be considered hypostatized categories into which Indian agriculture must fit, then the following conclusion suggests itself. Interlinkage, evident in both traditional and quasi-capitalist modes of production, itself constitutes an indigenous Indian political economy, a mode of production that accommodates patterns of political-economy called "feudal" and "capitalistic" in the West. The political economy of interlinkage—not merely ground rent—ought to be considered the fundamental barrier to the development of capitalist relations according to recognizable western models. By the logic of a political economy of interlinkage, parasitical landlordism (passive labor expropriation by interlinkage of rent and credit) remains sufficiently profitable until some outer stimulus forces change—not toward free access to markets, but toward interlinkage of wages, prices and credit usury articulated in terms of capitalist labor expropriation. What has been called “capitalist” and “feudal” modes of production can better be termed active and passive forms of interlinkage. The operative question within an interlinked political economy is not, “why does not the more active form of labor expropriation-cum-profitmaking arise (as if it should necessarily)?” but “what is the path of least resistance to the securing of labor by those who already have power?”
Following Balagopal (1988), the symbiosis of capitalist and noncapitalist elites must be attributed in the post-colonial period to the mediating role of the state, which provides institutional credit and commodities subsidies for the middle peasantry’s productive investment, and presides over the non-redistribution of the land monopoly—at once safeguarding the traditional rent-credit-labor landeconomy of the landed classes, and the facilitation of landpurchase precisely by ascendant rich tenants and middle peasants. Both passive landlordism and active entrepreneurialism come to depend on the state for protection, engendering a calculating policy of state aid and state non-intervention (through false landreform) simultaneously. The duality of the state's role betrays its comprador nature: it desperately solicits foreign interests, and differentially neglects domestic needs to buttress it own power. The comprador state functions as a bureaucratic repository-cum-intermediary for both investment and noninvestment of foreign and domestic productive capital, according to a balance of delicate expediency. It cements its alliance with ruling class interests not by favoring old as against new wealth, but by making all dominant groups dependent on it actively (by providing them a capital base) or tacitly (by not acting to redistribute land or capital concentration). Perhaps most importantly, the state secures class advantage for both old and new wealth by using institutional powers of coercion—army and police power, administrative use of caste and divisive social categories to allocate jobs and benefits—against class (-for-itself) political organization by subordinate groups. I have tried to suggest in the preceding section that the comprador nature of the Indian state is not unlike the mediating activity of the Mughal state, nor perhaps that of other medieval and even ancient Indian states. What seems to connect them is the effective concentration of the power of surplus appropriation and hence the abiding vulnerability of primary producers.
Broadly speaking, Marxist economics attends to the social-relational constraints on economic behavior, articulate and disarticulate both, caused by structural economic contradictions between forces and relations of production. These contradictions recur despite changes in supply and demand of goods and services, demographic changes including population growth, and the peregrinations of enlightened rational self-interest. Marx posits clearly the social consequences of capital accumulation as he understood them in his own time and culture: the alienation of consciousness in workers' alienation from one another and themselves by extensive commercialization and over-division of labor; the moral degradation in the victory of self- over other-oriented values. Whether the structural contradictions attendant upon rule by capital form universal laws by which material history proceeds across cultures and over time—or at least universal patterns for analysis and critical prediction—are beyond the reach of this essay. Recognizing that Marxism is not a dogma but a critical tool, a broad Marxist reading of agrarian change in India since Independence suggests the following:
1) Historical-dialectical class relations, in the context of Indian interlinkage, cannot be described merely in terms of bourgeoisification and proletarianization with the introduction of capital-intensive agricultural inputs.
2) Capitalist relations in India develop not from a clear strain of feudalism, but from interlinkage as a mode of production.
3) The interlinkage mode evidences and assimilates both feudal and capitalist relations. In India, both modes are characterized by interlinkage: just as credit may force up rent in a precapitalist mode, it may force down wages in a capitalist mode.
4) If capitalist relations are to develop along lines similar to non-Indian societies, the operative "barrier" to be overcome is the mediating comprador role played by the state, whose function is to solicit global capital, play internal elites against each other, render them dependent upon it and facilitate repression of class-based political resistance.
5) The class dialectic as articulated in the interlinkage mode does not necessarily oppose elites from different (Western) historical epochs in the development of capital: the class dialectic in India manifests over time via the perpetuation of differential access to markets and productive means by de-facto cooperation among various elites who control surplus value and expropriate labor with and without the capital intensification of production.
Within interlinkage, labor's unfreedom variously means debilitative and usurious rent payment, chronic debt, compromised contracts, the drain of surplus by the rich into luxury consumption or speculation, and the ability of an agricultural proletarian to reject one, two or three employers—but not the class of employers as a whole. Interlinkage is the adaptive structural denial of economic opportunity, sometimes by profit maximization, sometimes at the expense of profit, sometimes with the aid of violence, sometimes relatively peacefully. Elite groups' maximization of differential socio-economic advantage may mean landlordism, entrepreneurial investment in agriculture, or the monetization and commercialization of agriculture for reinvestment outside of agriculture (Upadhya 1988). Chattopadhyay (1972) may ultimately be right in claiming that the increase in agricultural labor represent the formation of a reserve army of proletarians as per the classical Marxian description. Certainly the movement of agricultural labor toward class-in-itself awareness proceeds in fits and jerks—set back perhaps by increases in attachment in areas of greater productivity (Brass 1990), or pushed forward by increased privatization (Byers 1981, Haque and Parthasarathy 1992), and still possibly reversible by implementation of meaningful land redistribution and provision of working capital by the state (Harriss 1992). However, agricultural labor’s historical destiny may be articulable by something other than teleological proletarianization, insofar as agrarian capitalism as developed in the West may not be a necessary development in the material history of India. If Indian material history is not proceeding through successive epochs toward a known end, mirroring or miming non-Indian material histories, linear bias in Marxist debate over Indian productive development misleads. If the fruitful study is not simply bourgeoisification and proletarianization, but the assimilation of varieties of material relations in India, the epistemic key to these sources may be that they do not necessarily follow prescribed patterns. The ongoing occlusion of free access to markets and productive means by dominant groups occurs with and without the state, by whatever means available. The material history of interlinkage describes the movements of a unique socio-economy in which markets are connected like the toothed gears of a watch: turning one turns all the others; turning one slowly turns all the others slowly; freezing one stops the entire mechanism.
Insofar as subordinate groups (so defined by their unfree access resulting from interlinkage) create value by their own labor, they constitute a motile productive force (a self-turning gear by the above metaphor). Their historical task becomes challenging the structural interlocks that constrain them. This may mean a landpoor family’s effort to retain land, or to search for decent wage labor, or both, depending on the perceived efficacy of either to remediate relational interlinkages. Proletarianization occurs to the extent that it breaks the capitalist-cum-landlord-cum-financier-creditor class’ ability to subject labor’s submission to interlinkage. Thus class relations are the tensions in the movments of markets and the grindings of the social urges into which every human being is born. Material polarities are the ease and difficulty with which groups move in socio-economic connections. This model, rather than a pyramidal-hierarchical reconstruction of classes as entities, suggests that Indian material history may not necessarily evolve toward the inversion of the pyramid, or involve the teleological usurpation of dominance by one and then another socio-economic group, according to scientific laws of structural constraints and contradictions.
If interlinkage is an Indian historical-material episteme, liberative praxis begins in the realistic recognition that classes—in freedom and suffering—are bound always to one another. Just as an interlinked socio-economy continuously adapts productive technologies to perpetuate the vulnerability of sections of the population, retributive violence against such interlinkage likewise merely adapts the violence already within the system of productive relations. It dissipates nothing. Justice begins as the continuous, vigilant analysis of causes and conditions. The first order of action is interpolative analytical balance.
Essay: Madison, Wisconsin, 1993; revised Stanford, California, 1997
Photographs: Nandigama, Andhra Pradesh, 1994-1995