Clark University

Caste, Class, and Clientelism: A Political Economy of Everyday Corruption in Rural North India Author(s): Craig Jeffrey Source: Economic Geography, Vol. 78, No. 1 (Jan., 2002), pp. 21-41 Published by: Clark University Stable URL: http://www.jstor.org/stable/4140822 . Accessed: 07/02/2011 01:09
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=clark. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Clark University is collaborating with JSTOR to digitize, preserve and extend access to Economic Geography.

http://www.jstor.org

Caste, Class, and Clientelism: A Political Economy of Everyday Corruption in Rural North India*
Craig Jeffrey
Department of Geography, University of Edinburgh, Drummond Street, Edinburgh EH8 9XP, Scotland cj @geo.ed.ac.uk
Abstract: Corruption has reemerged as an important issue in research on geography and development, but there has been little research on the relationship between corruption and class reproduction in rural areas of poorer countries. This article presents insights into how low-level economic corruption actually works within institutions that are responsible for purchasing sugarcane in rural western Uttar Pradesh, India, and the role of this corruption in perpetuating material inequalities within rural society. The discussion is based on 12 months of intensive field research on the economic and social strategies of a dominant caste of rich farmers in Meerut District, western Uttar Pradesh. In this article, I note periodic ruralprotest againstthe government's mismanagement of sugarcane marketingand corruption and describe everyday, disguised, and discrete forms of corruption that allow rich farmers to obtain privileged access to lucrative marketingopportunities. I also show how discourses surroundingcorruption are politicized along the lines of caste and class. I conclude by relating my empirical material to debates on local state-society relations in India. I stress the need to understand corruptionwith reference to local political economy and the broader distribution of social and economic opportunities in rural society and point to future avenues for geographic research on corruption in the Indian countryside. Key words: corruption, clientelism, Uttar Pradesh, India, agriculture, class, caste.

Recent research on geographyand development has tended to examine corruption with reference to the state and the macroeconomic implications of public officials' abuse of office. The study presented here focused instead on the social context of cor* I would like to thank my research assistant, O. P. Bohra, for his friendship and help in the field and the people whom I interviewed in north India for their time, openness, and hospitality. I am grateful to the Economic and Social Research Council for funding the research on which this article was based and to three anonymous referees for their comments on an earlier version of the article. I am also grateful to Stuart Corbridge, Roger Jeffery, and participants at the working-the-fields' panel of the European Conference of Modern Asian Studies in September 2000 for comments on an earlier draft and to Stephen Clibbery for assistance in analyzing the data. I alone am responsible for any errors that remain.

ruption. I examined the role of corruption in social differentiation and how rural agents construct corruption in discourse, focusing on "low-level corruption": relatively routine forms of malfeasance involving low-ranking officials who act as patrons or brokers in the disbursement of state resources. In this article, I explore an apparent contradiction between critical collective action and private accommodation in rural north India. A vibrant public culture of anticorruption appears to have emerged in rural and smalltown Uttar Pradesh, north India, but some of the key figures in this movement are heavily involved in everyday corruption at the local level.

Rent Seeking in Rural North India

To understand the relationship between collective anticorruption protest and every-

21

22

GEOGRAPHY ECONOMIC

day corrupt practices among local elites, it is necessary to examine low-level corruption as a process that operates through clientelist networks. In this respect, I have been inspired by the work of Khan (1996; 2000), who argued that to understand whether rent seeking is productive or unproductive in a particular setting, it is necessary to examine how far the state is able to maintain autonomy from its "clients," including powerful business interests, professionals, and other rent seekers. Khan maintained that in India, a range of "intermediateclasses"' have come to occupy positions of leadership in civil society and are adept in exerting political pressure on the state. These classes are able to influence the state's disbursement of rents (subsidies, tax breaks, licenses, and the like) by demanding new concessions and blocking the state's efforts to remove unproductive rents. Drawing on the work of Harriss-White (1997) and Wade (1985, 1988) in south India, Khan described how these elites shore up their position through forms of political rent seeking, for example, creating links with the police and politicians, and building cross-class alliances. Khan contrasted rent seeking in India with rent seeking in South Korea, where the relatively weak political and organizational position of the intermediate classes has allowed the state to maintain control over the management of rents. This comparative perspective allowed Khan to distinguish between clientelist patron-client networks, which he viewed as characteristic of most developing countries and where the state is weak relative to its clients, and patrimonial patron-client networks (such as South Korea), characterized by a strong state. Khan further distinguished between
'This category includes an increasinglybroad group of richer landowners, petty industrialists, merchants, professionals, and white-collar workers. The notion of "intermediate"classes is usually traced to the work of Kalecki (1972). The defining feature of these classes is that there is no contradiction between capital and labor (see Harriss-White 1999).

clientelism"of the Indian the "fragmentary subcontinent, where the authority to distribute rents is relatively dispersed through a complex and unruly bureaucracy, and Malaysia,where rent seeking was relatively centralized for much of the last part of the twentieth century. Most of the political economic work on clientelism in poorer countries has pointed to the role of corruption involving local elites in preventing equitable economic growth (see, especially, Herring 1983; Wade 1985; Bayart 1993; Harriss-White 1997; Perry 1997; Harrison 1999; Robbins 2000). Within scholarship on Uttar Pradesh, it is possible to identify three strands of research on corruption and patron-client networks. In reviewing these bodies of work, one can begin to trace some of the links between corruption within patron-client networks, rural-class reproduction, and poverty in the north Indian countryside. First, considerable work has focused on struggles between different rural strata to co-opt or colonize local government bodies (panchayats)that are partly responsible for the disbursementof development resources in ruralIndia. Several authorshave demonstrated that dominant classes in ruralnorth India, who control landownershipand usually belong to an upper- or middle-ranking caste, have been able to convert their economic strength into power over panchayat decisions and administration. monopolizBy ing headship of the panchayat councils, dominant classes are able to obtain privileged access to government contracts, employment, and grants. These dominant classes are then able to act as "brokers" between poorer sections of the ruralpopulation and the state (Lieten 1996; A.Gupta 1999). 1998; Lieten and Srivastava The second strand of research in Uttar Pradesh concerns the relationship between rural people and the police and judiciary. Scholars have demonstrated that access to effective police and judicial assistance in north India is mediated by a person's economic and social status and the local and nonlocal political economy (Brass 1997;

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

23

Jeffrey 2000). This finding has been confirmed by research in other parts of northern and western India (Bonner 1990; Bhatia 1997; Breman 1997). In these accounts, dominant classes are shown to be fairly (but not often wholly) successful in manipulating the local police force and judiciary through bribery, influence, and intimidation. This form of political rent seeking supports other means of protecting or increasing an elite's supply of rents, such as rigging panchayat elections, capturing electoral booths, and using preelection intimidatory tactics in elections for the state assembly and central government (Singh 1992). To obtain assistance from the police or judiciary, nonelites in rural Uttar Pradesh have to rely on one or more brokers, who may charge money in their role as gatekeepers to the local state. Recent research on the operation of panchayats, the police, and the judiciaryin ruralUttar Pradesh supports Khan's(2000) thesis in demonstrating that contemporary forms of fragmented clientelism in India are deeply implicated in local processes of class differentiation and accumulation. This clientelism threatens the well-being and security of the urban and rural poor and prevents the state from obtaining funds for its developmental and revenue functions. In this way, the capacityfor rural elites to appropriate state resources illegally at the local level is connected to wider aspects of the central and state governments' failure to foster equitable economic growth. Within Uttar Pradesh, these failures are manifest in continuing high levels of poverty (Dube 1998; Hasan 1998; Lieten and Srivastava 1999); sluggish growth in formal-sector salaried employment (Hasan 1998); wholly inadequate government credit, educational, health, and infrastructural provision (Drize and Gazdar1996; Jeffery and Jeffery 1997); and low rates of growth in rural off-farm businesses (Sen 1997). The liberalizationof the Indian economy from 1991, provoked, in large measure, by a growing fiscal crisis in the 1980s, does not appear to have reduced opportunities for elites to seek illegal rents

from the state and may have increased the vulnerabilityand isolation of the rural poor (S. Bhalla 1997; Sen 1997; Corbridge and Harriss 2000). The third strand of research on clientelism in Uttar Pradesh has considered the relationship between rich peasant farmers and local state institutions that are involved in regulating or assisting in agricultural production and marketing. These local state institutions include government-run cooperative societies, bureaucracies that are responsible for land administration, government institutions that regulate the supply of electricity or maintenance of agricultural equipment, and public agencies that are charged with buying and/or managing the sale of crops. Much of the work on the relationship between farmers and these institutions has focused on the Bharatiya Kisan Union (BKU) farmers' movement, which rose to prominence in western Uttar Pradesh in 1987, fronted by a prosperous stratum of middle-caste and locally dominant cultivators belonging to the Jat caste (Bentall and Corbridge 1996). As both Hasan (1995) and A. Gupta (1998) have shown, BKU activists achieved political visibilityand broadened their rural support base by organizing urban demonstrations, camps, and marches in protest against the deteriorating agriculture-industry terms of trade, disparities in rural and urban standards of living, and the corrupt activitiesof local representativesof the state. The BKU's demand for higher agricultural prices for key cash crops, such as wheat and sugarcane, and cheaper inputs was undergirded by populist verbal and physical assaults on representatives of the state (Hasan 1989, 1994; Lerche 1995; Bentall 1996). A. Gupta (1998) demonstrated that violent attackson government officialswere accompanied by a critique of local-level state functioning. Drawing on the political vocabularyof the politician,intellectual,and peasant-leader,CharanSingh, the BKU created a particulartype of subject position that celebrated the virtue of those who were poor, politicallyhelpless, and exploited. The implicit contrastwas with the alleged high-

24

ECONOMIC GEOGRAPHY

handedness, arrogance, and corruption of the powerful, notably government officials and the urban bourgeoisie (A. Gupta 1998, 84). A rural, legitimate, Gandhian Bharat (the Hindi word for India)was contrastedin BKU discourse with a degraded, illegitimate, and corrupt "urban India." Sartorial signifiers expressing the poverty of the farmer, such as torn clothes or traditional dress, became crucial to the expression of rural solidarity in the face of government officials'malpractice. A. Gupta (1998) linked the BKU movement to everydaycorruptioninvolvingrural people in north India and described how all members of small-town and rural India are brought into contact with government officials who demand bribes and ration state resources according to their own financial and social interests. The BKU drew on a popularfeeling of frustrationat these everyday forms of low-level corruption.Its attack on urban interests and the state was partly channeled through discourses of corruption (brashtachar)that were aimed at criticizing and undermining the functioning of local state institutions.These discourseswere circulated by vernacularlocal newspapers that deployed the Indian state'sown discourse of impartial bureaucratic procedure, strongly influenced by the colonial model (Kaviraj 1991; Chatterjee 1993), to critique the nefarious activities of low-ranking government bureaucrats.The BKU employed this critique of corruption circulating within public culture to downplay class contradictions underlying the union's support base, highlightan issue that affects all members of the rural population, and thereby sustain politicalpressure on the state to protect and extend subsidies on key cash crops (A. Gupta 1995, 1998). As researchon panchayats and the police has demonstrated, the ideal of the "servant state," enshrined in the Indian constitution and championed by JawaharlalNehru, is unraveling in rural Uttar Pradesh. At the same time, however, the ideal of an impartial bureaucracylives large in the political imagination of rural north Indians. These findings question the assumption that the

norms of liberal democracy are alien to the mass of ordinaryIndians (see Kaviraj1991) or that "clientelistic" patron-clientnetworks are uncontested by them. Gupta's analysis also raises questions concerning the relationship between the public culture of anticorruption and classdifferentiated access to state patronage. As A. Gupta (1998) pointed out, the BKU was a movement that involved many sections of the rural population. But richer peasant farmers, often belonging to the Jat caste, appear to have led many of the agitations (Bentall 1996). This observation raises the question of the extent to which rich peasant farmers' opposition to local state institutions is a real reflection of their inability to be effective in manipulating public-sector agencies that are engaged in regulating agriculturalmarketing. Research on the operation of panchayats and the police in western Uttar Pradesh has suggested that the rich farmers who were in the vanguard of the BKU movement have been reasonably successful in manipulating local state bureaucracies. Are marketing institutions different? Assuming for a moment that they are not and that rich farmers are successfully manipulating state marketing institutions, it is necessary to ask why rich farmers seek to circulate anticorruption discourses. The larger issues at stake here are, first, how everyday corruption within marketing institutions sustains or undermines social inequalities within rural society, and, second, how discourses of corruption are politicized with respect to class and caste struggles over resources. This article addresses these questions and issues with reference to a particular form of corruption that is prevalent in western Uttar Pradesh: that related to the marketing of sugarcane. Drawing on field research in Meerut District,2western Uttar

I Throughout, refer to the administrative unitof MeerutDistrictasit existedat the beginning of 1997,beforethe creationof a separate District. Baghpat
2

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

25

Pradesh, I present an account of how and with what success rich farmers monopolize access to state institutions that mediate between farmers and sugar mills. I demonstrate that many rich farmers are routinely involved in corrupt activity that is aimed at improving their access to lucrative marketing channels. I also show how they flexibly deploy different conceptions of "corruption"in an attempt to justify their collusive activity while they also criticize bureaucratic malpractice. This account points to the fungibility of discourses of corruption and anticorruption and the close links between corrupt practice and social exclusion. In the remainder of the article, I first introduce my field research in Meerut District and the pattern of class and caste differentiation in the settlements in which I worked. This discussion provides a basis for examining the links between class and corruption in a particular rural setting. Then, I consider sugar production and marketing in rural western Uttar Pradesh, examining how rich farmers collude with marketing officials to obtain privileged access to sugar mills and therefore higher prices for their sugarcane and discuss how these forms of corruption are legitimated and contested by rich farmers and poorer rural people, respectively. In the conclusion, I consider the implications of my analysis of north Indian corruption for research on agrarian change and local state-society relations in India and examine the relevance of my case study of corruption for future research on corrupt practices.

Field Research in Meerut District

Western Uttar Pradesh is often grouped with Haryana and Punjab as part of the heartland of India's Green Revolution (see Fig. 1). In the mid1970s, the western region accounted for 64 percent of the total cultivated area of sugarcane in Uttar Pradesh and pro-

duced 58 percent of the state's sugar (Hasan 1995, 173). Meerut District is one of the most agriculturally advanced and fertile districts in western Uttar Pradesh. In 1980-83, it achieved the second-highest value of crops per hectare in Uttar Pradesh (G. S. Bhalla and Tyagi 1989). In 1990, roughly 79 percent of the land in Meerut District was under cultivation, 73 percent of which was irrigated (28 percent by canals and 70 percent by tube-wells) (Malik 1993). Sugarcane and wheat are the most important crops by value and area. Virtually all land is double cropped, and many farmers cultivate three or four crops a year (Sharma and Poleman 1994). I conducted 12 months of field research on the economic and social strategies of rich Jat farmers in the sugar township of Daurala and two villages of Khanpur and Masuri in Meerut District between December 1996 and December 1997 (see Fig. 2). The three settlements, located within 25 kilometers of Meerut City, are relatively large, wealthy, and accessible by the standards of the district, and western Uttar Pradesh. Jats comprise between a third and a half of the total population of these settlements and dominate local landownership and opportunities to obtain lucrative off-farm incomes. The total populations of the three fieldwork areas were Daurala town, 10,025 inhabitants; Masuri, 4,373; and Khanpur, 4,913, according to the 1991 census (Office of the Registrar General 1991). During the course of the research, I interviewed 250 male heads of households belonging to four Jat lineages within the three settlements. The genealogical tables for these groups, which stretch back at least ten generations, were constructed from the memory of respondents and the records of village genealogists. The large kinship groups contain households that are broadly representative of the wider Jat population in each settlement. I also interviewed 30 male Muslim, randomly selected

26

ECONOMIC GEOGRAPHY

NORTH-WESTSI JAMMU-KASHMIR FRONTWIER PESHAWR LADAKHN

PRADESH HIMACHAL

" QUETTA PAKI TAN

LAHO RE R/ PUNJAB

\CHANDIGARH . PUNJAB HARYANA TTARANCHAIX%... S ELNEPAL R ?HpTDESH * NAGALAND PATNA BEHAR MEGHALAYA BANGLADESH TRIPURA MANIPUR

BALUCHISTAN JAPUR SIND KARACHI KRACHI RAJASTHAN

UT

PRADESHASSAM LU KNOW

BHOPAL MADHYAPRADESH UARMADABAD

ARMADABAD HOPAL

JHARKHAND W AL

CALCUTTA

DL\

DACOARAM
MAYANMAR

ZR

A MUMBAI
MAHARASHTRA SHYDERABAD

ORISSA
BHUBA ESWAR RANGO N0

C ANDHRA

KARNATAKA BANGALORE CHENAI

LAKSHADWEEP

C TAMIL

o? o o o
0 0

KRALA
0 500 kms
m 300 miles 300 SRI LA KA COLOMB

A B C

DIUTRIVANDRURN GOA, HAVELI DADRA NAGAR & PONDICHERRY DAcMAN

0 0

0S

Figure 1. Map of South Asia.

Scheduled Caste (SC), or Most Backward Caste (MBC)3 heads of households and 40 SC, MBC, and Jat women in Daurala, Khanpur,and Masuri (combined). In addi-

of "untouchable" jatis 3 SCs are a collection (castes) identified by the British in the 1930s as requiring special government assistance and listed on "schedules." MBC refers to former service castes, including Kumhars, Dhimars, Goswamis, Lohars, Telis, and Julai (Pai and Singh 1997). MBCs are just above SCs in terms of caste status.

tion, I conducted numerous interviews with lawyers, politicians, policemen, and state officials who were responsible for agricultural marketing and regulation and spent time observing the interaction between officials and members of the public in offices, tea stalls, and the homes of rich farmers. The semistructured interviews, conducted with Jat, lower-caste, and Muslim household heads, focused on five aspects of their economic and social status and activity: agricultural assets and production; off-

IN CASTE, CLASS,AND CLIENTELISM NORTHINDIA

27

SSaharanpur

SAHARANPUR'

Muzaffarnagar

I
Bunor

MUZAFFARNAGAR

Daurala ?Baraut
Baghpat

Sardhana,

aghpat

MEERUT
-

cerut

Masuri

Mawana

Khanpur*

GHAZIABAD
Delhi Ghaziabad ,-

'--

.Bulandshahr

BULANDSHAHR
e'

20 km

Figure 2. Mapof the upperdoabof western

Uttar Pradesh.

farm business and employment; education and fertility;relationshipwith in-laws, marriage, and dowry; and political activity and affiliations. These interviews generated quantitativeand qualitativedata that I analyzed using a relational database developed and customized for my work. I also conducted flexibly structured interviews that centered on how and when access to local state concessions and resources becomes significant for rural people and the social strategies that they employ in pursuit of state patronage. These interviews were generally conducted in Hindi. My research on the social and economic strategies of rich farmerswas centrallyconcerned with differences in status, security,

and wealth that are emerging within the Jat caste. For the purposes of analysis, I distinguished between a class of "rich farmers," possessing more than 5 hectares (12 acres) of agricultural land, and a class of "poor farmers,"possessing less than this amount, as a basis for examining social differentiation within the caste. This distinction corresponds with local definitions of what constitutes a "large"farmer, as opposed to a "medium"or "small"farmer. Furthermore, farmers who possess more than 5 hectares (12 acres) are much more likely to hire laborers on long-term (six- to nine-month) contracts. Analyses of the patwari's (local land revenue officer's) records for each settlement, supplemented by numerous efforts to cross-check this information through fieldbased interviews, suggested that Jats possess more than 90 percent of the agricultural land in Daurala, Khanpur, and Masuri. The concentration of land among the Jats was confirmed by the household surveys. Of the 250 households within the four Jat lineages, 66 percent own 2.5 hectares (6 acres) of agricultural land or less, 22 percent own 2.5 to 5 hectares (6 to 12 acres), and 12 percent own more than 5 hectares (12 acres). Of the 73 SC households in the sample, only 33 percent own agriculturalland, and none of these households possesses more than 2.5 hectares (6 acres) (Jeffrey and Lerche 2000). My research demonstrated that agrarian and nonagrariandifferences in wealth tend to reinforce one another. Rich Jat farmers have invested in the schooling of their children, high dowries, and modern technology in the pursuit of secure off-farm incomes. Public-sector employment is particularly valued owing to its long-term security, perks, and potential to obtain illegal uparli (on-the-top) incomes. Jat farmers who possess more than 5 hectares (12 acres) of land have been considerably more successful than poorer members of their caste and SCs, MBCs, and Muslims in obtaining valued public-sector employment and thereby avoiding recourse to laboring work. Of the 128 low-caste and

28

ECONOMIC GEOGRAPHY

Muslim men in the sample, 69 percent were employed as unskilled or semiskilled wage laborers, compared to just 8 percent of the 413 Jat men in the sample (Jeffrey and Lerche 2000). Social inequalities in access to land and public-sector employment within the Jat caste are reflected in and sustained by inequalities in physical mobility and ownership of consumer goods. Of the 26 Jat households in a sample of 202 who owned more than 5 hectares (12 acres) of land, 85 percent (22) owned motorcycles and 46 percent (12) had telephones. Equivalent figures for the 176 in the sample who possessed 5 or fewer hectares were 44 percent (82) and 7 percent (3). This analysisof class and caste differentiation in rural Meerut District shows that the power and wealth of different sections of the ruralpopulation are reflected in both agriculturalassets and access to nonagricultural economic and social resources. Two processes of class polarizationappear to be occurring: first, the emergence of rural classes within the dominant Jat caste based on access to land, lucrative employment, and social contacts, and second, increased differentiation between Jats and the lower castes in the wake of the Green Revolution. The following section explores-through ethnography and survey evidence-how these processes of class differentiation are reinforced through differentialaccess to the state officials who are responsible for regulating the marketingof sugarcane.

Sugar and Corruption in Western Uttar Pradesh

In 1996, India overtook Brazil as the largest producer of industrialcane sugar in the world (Godbole 2000, 421). In Brazil, sugar production is organized around factory-owned plantations that evolved to overcome technical difficulties associated with coordinating cane-cultivation and sugar-production plantations (ScheperHughes 1992, 37ff). Since the sucrose content of cane declines rapidly after it is har-

vested, factories are unable to stockpile the crop and must seek a regular supply of fresh cane. The plantation system binds field and factory within a closely coordinated industrial system of sugar production. In India, the colonial government lacked the political strength required to displace large numbers of peasants from the land to create a plantation system. Nearly all India's sugar is produced on peasant holdings of 8 or fewer hectares (Attwood 1992, 68). Historically,this situation has created problems in coordinating the cultivation of cane and the production of sugar. Maharashtran peasant-managed sugar cooperatives have addressed these problems quite effectively, according to Attwood (1992), who located the move toward the cooperative management of with sugarcane production in Maharashtra reference to the particularsocial, political, and ecological conditions prevalent within the Bombay Deccan in the late nineteenth and early twentieth centuries. Attwood showed that the availability of irrigated land, combined with the in-migrationof an enterprising and relatively egalitarian caste, stimulated peasants' entrepreneurial activity in the cultivation, and later processing, of sugarcane. By 1990, Maharashtra's cooperatives, run by and 92 for peasant cultivators, produced 30 percent of India's white sugar.These cooperatives maintain efficiency through a centrally organized system of harvest and transport "planned like a military operation" (Attwood 263). The expression of structural contradictions and tensions among different classes of cultivators are largely avoided as a result of the common caste background of most farmers and the shared feeling of having a stake in the efficiency of the cooperatives. The transparent,effective, and cooperative system of cane cultivation and sugar production said to be characteristic of Maharashtrais frequently contrasted with the endemic corruption and inefficiency of sugarcane production and processing in north India (Batra 1988; Attwood 1992).

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

29

Scholars who have paid attention to the dynamics of sugarcane marketing in and around western Uttar Pradesh have stressed the competitive nature of cane supply and malpractice within the District Cane Societies (henceforward referred to as Cane Societies) that act as intermediaries between farmers and the sugar factories (Amin 1984; Batra 1988; Attwood 1992; see also Baru 1990). In Uttar Pradesh, the high level of government regulation associated with the management of sugarcane marketing is said to have encouraged widespread malfeasance within the Cane Societies and generated antagonism between cane growers and the societies' employees (Hirsch 1961, 111-25; Amin 1984, 272-77; Batra 1988; D. Gupta

1997,100-01).

Sugarcane Cultivation and Marketing in Meerut District

Sugar and corruption are intimately linked in the rural economy of rural Meerut District. Rich farmers in this region, the vast majority of whom are male, view the efficient and profitable cultivation of sugarcane as an important economic objective and source of social status. Between 1965 and 1985, rich Jat farmers in Daurala, Khanpur, and Masuri invested in tractors, electric tube-wells, chemical fertilizers, and high-yielding varieties of sugarcane and wheat. In combination with the high government-support prices offered for sugarcane and wheat, associated with the policies of Charan Singh (see Byres 1988), these investments substantially increased agricultural profits. In 1997, almost all Jat farmers in the three settlements who cultivated more than 0.8 hectares (2 acres) of land grew rotations of sugarcane and wheat or sugarcane and potatoes for the market. The bulk of the farmers' profits were derived from the sale of cane (Jeffrey and Lerche 2000). Sugarcane is sown in February, March, or April and is harvested between

November and May, depending on the variety and efficiency of the sugar factories in which much of the crop is processed. Following the harvest of the first crop of cane, the shoots are left to regrow. A second crop is usually ready after 10 to 12 months (from October to April). Thus, the "supply season" for sugar mills and private cane crushers usually runs from October to May. After the harvest of the second crop, the shoots may be left to regrow once more or grubbed out to allow the soil to be prepared for another crop. Assuming that the soil is ready in time, a wheat crop is planted between the middle of November and the middle of January. This crop is harvested between March and the end of April, and cane is planted. Jats who own more than 5 hectares usually hire farm servants on 6- to 9-month contracts. These servants are responsible for a variety of tasks, including the supervision of temporary laborers and the irrigation of crops. Jats who cultivate more than 2.5 hectares (6 acres) of land tend to employ temporary workers for the most labor-intensive agriculturalactivities, such as tying and weeding sugarcane. Most of the temporary laborers and farm servants are local SC or MBC men or women. Farmers in Meerut District may choose between marketing their cane privately to small crushing enterprises that produce low-grade sugar or gur (a crudely refined raw-sugarproduct) or selling their crop to government-run sugar mills at a price fixed by the state government. In 1997, three of the seven sugar mills in Meerut District were run directly by the government (Sakauti, Muliana, and Mohinddhinpur), two were cooperatives but were nevertheless managed by the state government (Ramala and Baghpat), and two were privately owned (Daurala and Mawana). Meerut District does not have sugar mill cooperatives that are run along the lines of those in Maharashtra,in which farmers are responsible for managing the production units (Attwood 1992). Throughout Uttar Pradesh, Cane Societies act as intermediaries between the farmers and the sugar mills. The Uttar

30

GEOGRAPHY ECONOMIC

Pradesh government established these organizations in the late 1930s to reduce the large cut taken by middlemen, encourage technical investment by cultivators, and improve the scheduling and coordination of cane delivery (interview with the president of the Sugar Factory Union 8 November 1997; Attwood 1992, 73). Each sugar mill has a corresponding Cane Society that is charged with organizing the controlled and fair supply of sugarcane to the unit during the supply season. The Cane Societies also provide loans to farmers to buy equipment, seed, and fertilizer. Each Cane Society has local elected officials and permanent staff who are recruited by the government. Farmers who supply sugarcane to government mills through the Cane Society are all members of the society. Membership entitles a farmer to vote in elections for a delegate, although farmers who are in debt to the society are not allowed to vote. One Cane Society delegate is elected from a circle (usually containing four or five villages), and there are between 400 and 500 delegates for each Cane Society. The delegates are then responsible for electing directors, who, in turn, elect a single chairperson for each Cane Society, and there are about 60 to 70 directors in most Cane Societies. The elected officials are responsible for overseeing the supply of sugarcane and decisions regarding the disbursement of loans to cultivators. Members of the Cane Society's staff, who are, on the whole, formally employed by the government, are charged with more specific tasks, such as accounting, weighing cane, and sending out supply slips. They are formally committed to acting impartially in their dealings with the sugar mills and the farmers (interview with the secretary of the Cane Society, Mohinddhinpur, 17 February 1998). Each year, the Cane Society arrives at a basic quota for each farmer. Cane Society officials calculate this quota on the basis of the needs of the sugar mill, each farmer's supply during the two previous years, and the amount of cane being grown by the farmer. A farmer may also apply for an

additional bond over and above his or her basic quota. A clerk in the Cane Society then draws up a schedule for the delivery of cane and sends out supply slips (parchi) to farmers when it is their turn to deliver cane. Each supply slip entitles a farmer to deliver 42 quintals (1680 kg) of cane. The supply slips are administered throughout the supply season until a farmer's basic quota and additional bond are exhausted. Having received a supply slip, a farmer transports the cane by buffalo-drawn cart or tractor-drawntrolley to the sugar mill or a collection point known as a "center." There the farmer is given a token and proceeds to a weighing bridge, where the gross weight of the cane and vehicle is recorded on a weighing slip and checked by factory officials. The cane is then unloaded, the empty vehicle is weighed, and a receipt is issued. Until recently, the farmer or his or her representative would then take the receipt to the Cane Society, where it would be checked by a clerk and exchanged for a token. A cashier would process this token and send payment not less than ten days later.

Antistate Dissent
There are certain irregularities associated with this process of sugarcane marketing that have a negative impact on all farmers and provide slogans and issues for periodic protests and more sustained efforts at political mobilization in rural western Uttar Pradesh. For example, at the beginning of 1997, the most pressing issue was the nonpayment of farmers following a dispute between sugar mill representatives and the Uttar Pradesh government. Some farmers had been waiting as long as nine months to receive remuneration from the Cane Society. Resentment focused on the fact that many farmers had been driven to sell wheat early at a lower price to finance the planting and irrigation of their next cane crop. The Cane Society, however, had refused to pass on the interest it received from factory owners as compensation for the delay. Angered by such apparent greed

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

31

and continued delays in payment, the BKU protested on 24 January1997 by blocking a railwayline leading to Simbhavlisugar mill, Ghaziabad.This agitation ended in a clash with the police in which 2 farmers were killed and 50 were injured (Trouble at Simbhavli Sugar Mill 1997). Although BKU activity became much less intense after the early 1990s (Madsen 1998), such clashes continue to be a feature of the economic and political geography of western Uttar Pradesh. In addition to the issue of late payment, the BKU voices other criticisms of sugar marketing. BKU leaders complain that, unlike in Maharashtra, farmers in Uttar Pradesh receive no share in the sale of the by-products of cane (molasses sold to distilleries and a dry residue used in local paper factories). They also complain about technical faults within the sugar mills that delay the processing of sugarcane. Farmers within and outside the BKU are frustrated by the long waits to deliver cane and a subsequent increase in the amount that the factorydeducts from the price of sugarcane for the drying of the crop in transit (D. Gupta 1997, 100). These criticisms mesh with the BKU's calls for higher government sugarcane prices and the ending of interstate restrictions on the trade in agricultural products. In February 1997, I interviewed farmers who claimed to have been waiting three days to deliver sugarcane to the Mohinddhinpursugar mill near Meerut. These men are reluctant to leave their carts or tractor-trolleysfor fear of losing their place in the line. Cut off from their families, without any facilities for washing or relieving themselves, these men become bitter and frustrated. The long lines become fertile ground for the growth of dissatisfactionand the circulation of gossip and rumors against the Cane Societies and the corruption and inefficiency of state bureaucracies more generally. Farmers' statements resonated with A. Gupta's description of agrarianpopulism in western Uttar Pradesh. They emphasized their powerlessness ("What can we do but wait?")and frustrationat the government's

inefficiency and corruption: "The factory and officials are sick to the core (Mohinddhinpur sugar mill, field notes, April 1997).

Cane Societies and Corruption

Popular discontent regarding the marketing of sugarcane should not obscure the covert and discrete efforts of rich farmers, and some poor farmers, to influence local officials. To understand the economic rationale behind corruption in the delivery of sugarcane to the mills, it is necessary to appreciate the dilemmas that farmers face in marketing this crop. During most years, the government price offered by the sugar mill is far enough above the private price to encourage farmers to risk administrative and technical delays and to supply the government-run sugar mills. For example, in the 1996-97 supply season, farmers who supplied sugar to the local kolhus received about 40 rupees per 100 kg at the beginning of the supply season and 65 rupees by the end of season (in 1997, 40 rupees was roughly equivalent to one U.S. dollar). During this period, the price paid at the government mill remained at about 70 rupees. Many farmers claimed that this price differential is typical. But occasionally, as in the 1997-98 supply season, the private price rises above the government's price. When it does, particularlyif it occurs at the beginning of the supply season, farmers may decide to supply most of their cane privately.The sugar mills, through the Cane Societies, seek to guard against this defection by using the actual amount of cane supplied by a farmer as one consideration in setting a basic quota in the two subsequent years. If a farmer has arranged an additional bond and fails to deliver the required amount of cane, the Cane Society institutes a more direct punishment and reduces the rate paid to the farmer for the cane that he or she has supplied. In light of these checks, when private prices significantly exceed the government's price, farmers face a dilemma: They may choose to supply privately, in which case they can

32

ECONOMIC GEOGRAPHY

expect that their quota will be reduced during subsequent years, or they may "play safe" and continue to supply to the government mill.4 Many farmers opt to supply privately. As a result, there is often particularlyintense competition to reestablish a sizable basic quota after a year of high private prices. Competition among farmers to increase or maintainquotas or to seek supply slips is played out through interactions with Cane Society officials. In 1998, the Uttar Pradesh government announced that "misappropriation and financial irregularities" amounting to 240 million rupees had been identified in various Cane Societies of the State (Rs 24 Crore Cane Scam Unearthed 1998). In Meerut District, the illegal activity of Cane Society officers is manifest in their abilityto promote a shadow market in supply slips. In any one year, a small number of farmers have an oversupply of slips. They may choose to give these slips to close relatives or marketing officials, in the latter case as a means of staying on good terms with these officers. Alternatively,they may sell the slips to those who are seeking to increase the amount of cane they can sell at the government's price. To avoid criticism from other farmers,those who sell slips disguise this activity by using Cane Society clerks as brokers. "In this way we can occasionally make some extra money and still appear as good Jats"(interview with a rich farmer, Meerut, 1 October 1997). Unless a farmer has influence with the Cane Society, the society's officers may delay issuing a slip that rightfullybelongs to a farmer until the farmer pays a bribe. The purchase of a supply slip or the guarantee of the rapid delivery of one's own slips fre4 Farmers' to indebtedness CaneSocietiesdid not appearto be a strongfactorin decisions aboutwhere to marketcane in the villagesin whichI worked.It is possible,though,thatthe CaneSocietiesmayuse a farmer's indebtedness to tie the farmerto providing to sugarcane the mills when the private price is government abovethe government's pricefor cane.

quently required the payment of 200 to 1,000 rupees per slip in Dauralain 1997. In a similar vein, clerks who are responsible for remuneratingfarmers for the cane supplied often seek rents from their posts by withholding payments to farmers until a bribe is paid. Batra (1988, 108) maintained that a number of cane growers in Haryana have resorted to mortgaging cane receipts to local moneylenders "who in turn mainwhom they later recover payments." I did not encounter references to similar forms of brokerage in Meerut District. Cane Society clerks who are responsible for weighing sugarcane at the point of delivery also extract large illegal rents, understood as "on-the-top" incomes. Farmers recognize that these officials can have a significant impact on the level of payment they receive for their sugarcane, either negatively, by underestimating the weight of the crop, or positively, by falsely declaring it to be a particularly valuable variety of cane. In this case, the connection between local corruption and the wider bureaucratic system of collecting "on-thetop" incomes is clear. One Cane Society official maintained that to receive a 15-day posting as a weighment clerk, it is necessary to pay the secretary of the society 2,500 rupees (interview with a Jat Cane Society official, Daurala, 8 November 1997).5 Each Jat household head attempts to exert influence over the weighment clerks. Until recently, farmers were also able to bribe or influence the Cane Society officials who were responsible for estimating the number of hectares planted with sugarcane on individual farms. In return for a bribe, the surveyor would overestimate the amount of cane being grown, "and in this way some larger farmers in 5 As Wade(1985,470) argues his research on corruption "a officials, noramongirrigation malpricetendsto emergeforeachpost,around which bargainingin specific instances takes
place." tain regular links with the . . . [Cane Society] . . . and factory officials through

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

33

[could] increase their basic quota" (interview with a rich farmer, Daurala,21 August 1997).

Political Strategies of Rich Jat Farmers

The efforts of farmers to influence Cane Society officials may be broadly described as either proactive or reactive. After failing to receive a supply slip or payment, farmers often attempt to take up the matter with an official in the Cane Society office. This reactive strategytypicallyinvolves negotiating a scrum of complainants in a crowded, hot, and disorganized office. The speed with which a person's complaint is addressed and the likelihood of success depends on the person's capacity to bribe an official, sifaareesh (influence or recommendation) within the Cane Society, and perceptions of the person's political clout. It also depends on the person's performative skill, including his or her physical stature, dress, and language. Most complainants attempt to ingratiate themselves with an official before they address a particular complaint. Smiles and fawning occasionally give way to a more intimidatory approach. There have been several examples of rich farmers threatening Cane Society officials with rifles to obtain supply slips or payment for cane delivered. During an interview (in Daurala, 15 November 1997), one of the most renowned Jat bullies in the area raised a gun to my head and stated simply, "This is a good way to persuade those bastards" (referring to Cane Society officials). Since they comprise 30 to 40 percent of the population in Daurala, Khanpur, and Masuri, Jats have relatively good access to a local supply of muscle men, and many own unlicensed handguns or rifles. To guard against falling victim to Cane Society malpractice, rich farmers also attempt to foster relationships of trust and friendship with the government officials who are employed in the Cane Society in anticipation of the need for future help.

These proactive measures frequently involve visiting offices to drink tea and gossip about politics and village affairs with Cane Society officials. Dress and demeanor are especially important in these social encounters. On several occasions, I observed rich farmers deliberately changing out of "cityclothes" (shirts and trousers) into the more traditional kurta pyjama in visits to officials.6 Less frequently, rich farmers invite officials to their rural homes for food and entertainment. As a rich Jat put it, "Thisis how we market our cane and become rich: through playing politics with these men" (the Cane Society staff) (interview, Masuri, 16 October 1997). Such "social investment" forms part of a wider effort by Jats to protect and enhance their wealth, security, and standing by maintaining strategic political contacts in urban areas. In particular, efforts to maintain social bonds with Cane Society officials are bolstered by rich Jats' attempts to establish "friendlyrelations"with local politicians. A link with a politician acts as a kind of trump card in encounters with local officials, who are keenly aware of a politician's ability to punish or reward them (see Jeffrey 2000, 2001). The capacity to raise money for bribes, make routine and timely visits to Cane Societies, and master the performative dimensions of corruption depends on one's material resources. There are class- and caste-based inequalities in the capacity of rural people to guarantee a flow of supply slips and receive remuneration for cane delivered, partly because of rich farmers' superior ability to obtain government posts in the Cane Societies. Five Jat men in the sample of 415, 4 of whom were raised by guardians possessing more than 5 hectares (12 acres) of land, obtained clerical positions in the Cane Society; in contrast, no low-caste men in the sample of 128 did so.
6 A. Gupta(1998)noted the symbolicsignificance of traditional ruraldress and its connection to a cultureof ruralprotest.

34

ECONOMIC GEOGRAPHY

Local influential Jat men have also tended to monopolize the top elected posts in the Cane Society. I found it difficult to obtain clear statements about the relationship between elected officials and the Cane Society's government employees. But close contact with a delegate, director, or (still better) the Cane Society chairperson was widely regarded as improving a person's capacity to obtain supply slips and maintain or increase one's basic quota. With regard to the colonization of key posts, many farmers in and around Daurala referred to the activities of a Jat named Satender. In the early 1990s, Satender used a link with the politician Mulayam Singh Yadavand forms of intimidation that appear to be his trademark,' to install his wife as chairperson of a Cane Society. According to several informants, he has been able to earmark other positions in this Cane Society for friends and relatives by intimidating or petitioning delegates and directors. I lack quantitative data on the proportion of Jats who occupy elected posts in Cane Societies in rural Meerut District, but Jats and members of other castes claimed that the vast majority of delegates and directors in Cane Societies are either Jats or members of other middle or upper castes. Proactive collusion instigated by Jat farmers builds upon kinship or caste solidarities. While visiting a farmer in his rural home, Jat Cane Society officials are frequently encouraged to join in a smoke on a hookah (pipe), a form of mutual consumption specifically associated with Jat identity (D. Gupta 1997). On other occasions, rich Jats seek to build on masculine bonds with the officials and a shared sense of common Jat or upper- or middle-caste heritage by sharing liquor, cigarettes, and meat. These forms of consumption, facilitated by Green Revolution wealth and originating, in part,

from Jats' historical contact with the culture of the British army, are becoming increasingly associated with the rich male middle- and upper-caste lifestyles in this part of India (Chowdhry 1994; Dutta 1997).

Class Differentiation and Corruption

A survey of 87 Jat farmers that I conducted as part of my research also pointed to the class-differentiated nature of influence within the Cane Societies (see Table 1). Table 1 suggests that in the 1996-97 supply season, when prices at the sugar mills were much higher than those available at private crushers, only 56 percent of the farmers who possessed fewer than 2.5 hectares (6 acres) of land supplied cane to the mills versus 82 percent and 78 percent of the medium and rich farmers, respectively." I also heard far fewer complaints about the operation of sugar mills from rich farmers than from poor farmers. The case of Surrendra,a Jat farmerwho owned 1.3 hectares (3 acres), is typical of the predicament of poor farmers. Surrendra's basic quota was inexplicably reduced from 60 metric tons for the 1995-96 supply season to just 5 metric tons for 1996-97. Surrendracould not afford to follow up this issue, since he was unable to pay for replacement agricultural labor while he was awayfrom the farm and found bus fares a significant drain on his household resources. On the one occasion when he visited the Cane Society, he was told that no one knew where the slips had gone. Surrendra shrugged, saying, "to raste men ghumne,"literally, "so [the slips] wandered on the road"(interview with Surrendra,12
8 The smallsize of the samplepreventedme firmconclusions fromdrawing fromthesedata. Nor did it allow me to establishwhetherthe to capacity deliverto the millin a yearof higher

During an interview that my research assistant and I conducted with Satender, he placed a shotgun on his lap and slowly turned it over as we asked him (increasingly innocuous) questions.

mill prices is normally distributed or whether there is a sharp break in one's capacity at a crit-

ical level of landownership.

IN CASTE, CLASS,AND CLIENTELISM NORTH INDIA

35

Table 1 Percentage of Sugarcane Marketed to Sugar Mills and Complaints Regarding the Misallocationof Supply Slips for a Sample of Jat Farmers
Number of Jat Farmers in the > 0-2.5 hectares (0-6 acres) 2.6-5 hectares (6.01-12 acres) > 5 hectares (12 acres) Total Sample 11 13 30 54 Number Sending a Portion of Their Sugarcane to Sugar Mills 8 10 29 47 Mean Percentage of Cane Sent to Sugar Factory in 1996-97 Supply Season 56 82 78 76 Number Who Complained About SupplySlip Allocations 5 1 1 7

Source: Interviews with rich Jat farmers in Daurala, Khanpur, and Masuri, conducted between August and December 1997.

October 1997). As Surrendrastated later in the interview, glumly, "Thiswould not have happened to a large farmer [bare bare kisan], but is the common experience for small farmers and the poorer communities." Surendra'sgloomy assessment of the social location of marketing privileges is supported by the interviews I conducted with Muslims and low castes who market sugarcane in Daurala, Masuri, and Khanpur. Of the five low-caste or Muslim households that occasionally cultivated sugarcane for the market, only one was able to obtain supply slips at the sugar mill. Poorer famers' preference for marketing cane to private crushers is partly a result of their frequent need for immediate cash to cover household consumption needs and occasional additional expenses (especially dowries). But in explainingtheir sale to private crushers, usually at lower rates, the heads of SC, MBC, and Muslim households also cited the fact that they lack influence in the Cane Society. It is necessary, though, to guard against arguing that poor farmers are always excluded from clientelistic networks involving Cane Society officials or that rich Jats are always successful in their efforts to collude with officials. Poor farmers pointed to a small number of low-caste poor farmers who were routinely successful in their

attempts to bribe or influence government officials in the Cane Society. For example, Surrendra'scousin, Brijpal, who owns 1.3 hectares (3 acres) of land, has been able to obtain some political leverage in the Cane Society. Surrendra and Brijpal explained the latter's success by noting that Brijpalis confident, a "good talker," and has acquired a habit of traveling to urban areas to meet people. Even so, these two lowcaste men claimed that Brijpal and other poor farmers are successful only in exerting pressure on officials who have not delivered sugarcane on time; they find it much more difficult to build links proactively. Low castes and many Jats in rural Meerut District emphasized that there are some government officials in Cane Societies who abide by official rules, refuse bribes, and do not engage in corrupt practices. Rich Jats have obtained a degree of control over the activities of Cane Societies, but they have not comprehensively captured these institutions.

Resistance, Counterresistance, and the Discursive Construction of Corruption

As Khan (1996, 2000) argued, the capacity of elites to control the activities of local officials depends on the wider political

36

ECONOMIC GEOGRAPHY

context in which they seek rental incomes. The emergence of the Bahujan Samaj Party (BSP), which held power at the state level three times in the 1990s in Uttar Pradesh, has posed a challenge to the strategies of ruralelites. The BSP, founded in 1984, has actively sought to increase the wealth, visibility, status, and security of SCs in Uttar Pradesh (see Duncan 1999; Lerche 1999). It has attempted to reduce the class and caste bias of state bureaucracies by installing officers who are sympathetic to the interests of SCs in key posts; enforce legislation to reduce caste-based discrimination; and construct parks, statues, and librariesrepresenting or dedicated to the SC hero, Dr. Ambedkar. These initiatives appear to have increased the confidence of SCs in the villages in which I worked. In the realm of sugarcane marketing, in the 1990s the BSP proposed introducing legislation that would prevent farmers who possess more than 5 hectares (12 acres) of land from supplying more than 70 percent of their sugarcanecrop to governmentmills. Several poor farmers with whom I spoke welcomed this proposal.Some went as far as to taunt rich farmersby frequentlyreferring to the proposed restrictionand the threat it would pose to the elite's monopoly over sending sugarcaneto the government mills. But the brevity of the BSP's time in power in 1997 (just six months) appears to have prevented the passage and implementation of the proposal.To my knowledge, no Cane Societies in Meerut District restrictthe percentage of cane that rich farmers may supply to government mills. The BSP has been more effective in encouraging the emergence of a transparent system for the disbursement of supply slips. This is part of a longer-run effort by the Uttar Pradesh government to reduce corruptionwithin Cane Societies. By 1999, a computerized system of supply-slip allocation had been installed in most Cane Societies in Meerut District, in an attempt to reduce administrative error and the deliberate misdirection of slips. In addition, the system of organizing cane pay-

ment, based on face-to-face contact, has been replaced by the direct crediting of farmers' bank accounts. These changes may lead to a reduction in malpractice. I encountered sporadic references to increased transparencyin the allocation of sugar-factory supply slips. Most farmers also claimed that payment through the banking system had reduced delays in receiving remuneration for the delivery of cane.9 Nevertheless, these efforts to increase transparencyhave been only partiallysuccessful. For example, a number of poor farmers, including one of the few low-caste men to cultivate sugarcane for the market, complained that they continue to be cheated in the distribution of supply slips. It is possible that the shifts in administration are simply redistributingopportunities to extract on-the-top incomes within the Meerut District economy. For example, bank managers may be replacing Cane Society clerks as rent-seeking intermediaries between authorities who issue payment for cane and the farmers.Three poor farmers complained of the belligerence of bank managers who falsely claimed not to have received money "or shut up the bank early so that we cannot get our money for days on end" (interview with a poor farmer, Khanpur, 10 October 1997). Following A. Gupta's (1995, 1998) analysis of rural opposition to the state, one might expect rich farmers to be voicing similar dissatisfaction. Nevertheless, and in spite of persistent questioning, I did not encounter a single farmer who owned more than 5 hectares (12 acres) who complained about banking in connection with obtaining remuneration for the sugarcane supplied.
9 During more-recentfield research,conducted from 2000 to 2001 in BijnorDistrict, with RogerJefferyand westernUttarPradesh, PatriciaJeffery,I encounteredreferencesto in and similar improvements the efficiency fairof ness of the allocation sugarcane supplyslips the systemof following shiftto a computerized disbursement.

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

37

The discourses of rich and poor farmers on corruptionmust be viewed in relation to the newfound confidence of SCs and these recent administrative changes in the marketing of sugarcane. Rich farmers' assessments of corruption associated with sugarcane marketing varied a great deal and often contradicted wider oppositional rural discourses circulated by the BKU. Interpretations of Cane Society corruption also varied according to whether rich farmers were interviewed privately or in large groups. Some rich Jats appear to be aware of the contradictionbetween periodically becoming involved in instigating protests against the corruption and inefficiency of government bureaucracies while developing collusive and friendly relations with Cane Society officials. In discussions of corruption involving Cane Society officials, several rich farmers recalled a popular proverb of the area: "sometimes you must call a donkey your 'father"' ("father"is a term of respect and donkey is a term of abuse). This proverb is consistent with the distinction commonly made by rich farmers between "moral" political action, which was said to include participation in the BKU movement, and "practical"political practices, such as colluding with local officials. The English words "moral" and were used. This usage is consis"practical" tent with A. Gupta's (1995; 1998) accounts of a popular culture of antistate dissent. A pragmaticacceptance that it is occasionally necessary to sweet-talk state officials does not imply that farmers are content with local corrupt practices. Nevertheless, in private, rich farmers more commonly defended the actions of Cane Society officials and celebrated the forms of malpractice that these officials assist in perpetuating. "This system [influencing Cane Society officials] is very good for us, and we would not want it to change. Anyway, many of those in the Cane Societies are our good friends or relations" (interview with a rich Jat farmer, Masuri,4 October 1997). Or, as another rich Jat farmer put it, "In our community, we think

we must help each other in our economic affairs.... Yes, this includes getting supply slips out of turn when we need them" (interview, Masuri, 2 October 1997). The positive value that some rich Jat farmers attach to local corruption in sugarcane marketing reflects the siege mentality of members of this stratum, who fear that their newfound wealth is under threat from the political rise of SCs. Many rich farmers justified their own role and the role of Cane Society officials in corruption by referring to the support that low castes receive from the Indian state in obtaining This tendency public-sector employment.10 is linked to the rise of the BSP. Rich Jats frequently justify local acts of corruption by citing the alleged privileges directed at SCs. Referring to Cane Society officials, one rich farmer maintained that "we must protect our friendships; after all we are not pampered by the government like these Scheduled Castes" (interview, Daurala, 10 August 1997). Picking at his kurta pyjama, another rich farmer claimed, "If this supply-slip system works to our advantage,that is good. You can see how poor we are. Look at my clothes compared to those Chamar [Scheduled Caste] bastards" (interview, 2 October 1997). Among the Jats, assessments of government bureaucratsare occasionally conditioned by the caste and class status of those who are engaged in collusive activity. In Jat accounts, whether an action was labeled a "bribe" (riswat) or a "gift" often depended on the caste of the person who was responsible for initiating the corrupt exchange. A Jat who bribes a Cane Society official to increase his or her basic quota may be regarded by other Jats as having offered a "gift" to the official, whereas an SC who attempts to do so
10 Legislationentitlingformeruntouchables to reservedplaces in government employment institutions passedin 1950; was andeducational for similar legislation a widergroupof economicallyor sociallydeprivedcastes-the so-called Castes-was passedin 1991. OtherBackward

38

GEOGRAPHY ECONOMIC

would be regarded as corrupt (brashtachar) and guilty of bribery. Similarly,when SCs, MBCs, Muslims, or poorer Jats are able to obtain political purchase within Cane Societies, they tend to celebrate their capacity to engage in corrupt activity. Examples of low-caste involvement in corrupt practices are considered to be indicative of their refusal to "sit still" or ability to "stand on their own two feet." Jeffrey and Lerche (2000, 873) noted similar examples of SCs boasting of their ability to cheat or deceive higher castes and state officials in research in eastern Uttar Pradesh. The periodic emergence or espousal of a much broader public culture of dissent should not obscure these important class- or caste-specific discourses of corruption.

Conclusion
I have demonstrated that many rich farmersin rural Meerut District are able to obtain an inside track on sugarcane marketing by colluding with state officials. Through discrete encounters with Cane Society officers, this stratum recycles profits derived from the Green Revolution into creating social contacts and reducing financial risks associated with capitalist agriculture. These microtactics have the cumulative effect of isolating and further impoverishing the rural poor. I have suggested, though, that the rural poor also attempt to collude with officials and circulate discourses of corruption. Rather than think in dyadic terms-of "corrupter"and "corrupted"-it is necessary to be attentive to how differently positioned agents attempt to insert themselves into patronclient networks. The capacity of rural agents to bribe or influence officials varies along a continuum, from the usually successful and well-connected elite to the poorly connected and usually unsuccessful ruralpoor. The account of corruption presented in this article provides an occasion for commenting on the geography of sugar production in India. Attwood (1992, 275ff)

explained the absence of Maharashtranstyle cooperatives in north India in relation to the lack of an organized harvest and transport system and north Indian cane growers' noninvolvement in the management of the delivery and processing of sugarcane. He also cited the climate of protectionism in the north, which leads to little pressure being placed on cooperatives to operate more efficiently. My analysis suggests two further explanations for the lack of cooperative activitywithin ruralwestern Uttar Pradesh. First, the rich farmers who have the social links, education, and time to lead and sustain cooperatives are precisely those who benefit the most from the present system. Their best interest is served by publicly calling for more protectionism (subsidies for agriculture and higher prices) while privately maintaining influence within the Cane Societies. Second, social tensions within the leading cultivating caste and between farmers and agricultural laborers are far more acute in western Uttar Pradesh and many other parts of north India than they are in Maharashtra. The emergence of the BSP has exacerbated these tensions. Intercaste and interclass suspicion and jealousy in western Uttar Pradesh discourage the type of widespread cooperation that is characteristicof historical experience. rural Maharashtra's My analysis has drawn inspiration from A. Gupta's (1995, 1998) research on the local discourses of corruption circulated by rural people. I have demonstrated that the cane fields of rural western Uttar Pradesh provide fertile ground for the growth of a ruraloppositional culture that is concerned with critiquing the corruption associated with the state's regulation and management of sugarcane marketing. The BKU, led by richer farmers but involving large sections of the rural population, has been an important vehicle for the expression of this protest. Rather more than A. Gupta, though, I have emphasized how classbased practices and discourses of corruption coexist with a public culture of antistate dissent. Farmers'efforts to co-opt local marketinginstitutions are comparable

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

39

with their successful attempts to colonize or influence local panchayats, the local police force, and the judiciary in rural Uttar Pradesh. Forms of open protest against the state provide a convenient screen for a creamy layer of rich farmers, whose everyday, disguised, and discrete acts of collusion prevent the emergence of more efficient and fairer processes of sugarcane marketing. The BSP has emerged partlyin response to the class- and caste-differentiating nature of patron-client networks involving local government officials. This political partyhas improved the confidence of many poor SC people in rural Meerut District and contributed to the emergence of a slightly fairer and more efficient means of allocating marketing certificates for sugarcane. The BSP has also sharpened classspecific discourses of corruption and anticorruption. Nevertheless, it has not radicallyimproved the capacity of the rural poor to compete effectively with richer sections of agrarian society in colluding with government officials in the Cane Societies. The economic, political, and geographic implications of this analysisrequire further consideration. How far can institutional reform of local-level state institutions reduce corruptionin the absence of a more profound redistribution of economic and social resources in rural areas of India? How do opportunities to colonize and coopt different local state institutions vary among districts and states in India, and what factors are important in determining these differences? As Khan (2000) pointed out, the study of corruption and rent seeking has not paid adequate attention to the costs of rent-seeking behavior and its differentiating effects. As a result, the power and organization of vested interests in civil society are routinely either underestimated or ignored. Political economic accounts of corruption provide a solid basis for understanding how corrupt practices are implicated in social reproduction and resistance to this reproduction. Answers to the questions posed earlier are likely to require further research that examines the social con-

text of corruptpractices and the role of corruption in reproducing class and other inequalities.

References
Amin, S. 1984. Sugarcane and sugar in An Gorakhpur: inquiryinto peasantproductionfor capitalist in enterprise colonialIndia. U.K.:Oxford Press. Oxford, University Attwood,D. W. 1992.Raisingcane:Thepolitical economyof sugarin westernIndia. New Delhi:Oxford Press. University Baru,S. 1990. Thepoliticaleconomyof Indian Press. change.New Delhi:Oxford University relations sugar and Batra,S. M. 1988.Agrarian in cooperatives northIndia.In Whoshares? and ed. Cooperatives ruraldevelopment, D. W. Attwood B. S. Baviskar, and 91-111. New Delhi:Oxford Press. University Bayart, J.-F. 1993.Thestatein Africa:Thepolitics of the belly.London: Longman. versusIndia": Peasant Bentall,J. 1996. "Bharat relationsin northpolitics and urban-rural west India. Unpublished Ph.D. thesis, of U.K. University Cambridge, Cambridge, S. Bentall,J., and Corbridge, E. 1996. Urbanruralrelations, demandpoliticsandthe "new in agrarianism" NW India: The Bharatiya 21(1):27-48. Bhalla,G. S., andTyagi,D.S. 1989.Spatial pattern of agricultural developmentin India, Economic and Political Weekly 24(25):A46-A56. Bhalla,S. 1997. The rise and fall of workforce diversification processes in rural India. In and poverty:Change Growth,employment and continuity in rural India, ed. G. K. Chadhaand A. N. Sharma,145-83. New Delhi:Vikas. on Bhatia,B. 1997. Massacre the banksof the Sone. Economic and Political Weekly 32(51):3242-45. Bonner, A. 1990. Averting the apocalypse: Socialmovements India today. Durham, in Press. N.C.:Duke University P. Brass, 1997.Theft an idol:Textand context of in the representation collectiveviolence. of Press. Princeton, N.J.:Princeton University Breman, 1997. Silencingthe voice of agriculJ. tural labourersin south Gujarat.Kingsley Martin Memorial Lecture, November, of U.K. University Cambridge, Cambridge,
Kisan Union. Transactions of the Institute of New Series British Geographers, sugar: State intervention and structural

40

GEOGRAPHY ECONOMIC

Byres, T. J. 1988. Charan Singh (1902-1987): An assessment. Journal of Peasant Studies 15:139-89. Chatterjee, P. 1993. The nation and its fragments: Colonial and postcolonial histories. Princeton, N.J.: Princeton University Press. Chowdhry, P. 1994. The veiled women: Shifting gender equations in rural Haryana 1880-1990. New Delhi: Oxford University Press. Corbridge, S., and Harriss,J. 2000. Reinventing India: Liberalization, Hindu nationalism and popular democracy. Cambridge, U.K.: Polity Press. Drize, J., and Gazdar, H. 1996. Uttar Pradesh: The burden of inertia. In Indian Development: Selected regional perspectives, ed. J. Drize and A. Sen, 33-128. Delhi: Oxford UniversityPress. Dube, S. 1998. In the land of poverty: Memoirs of an Indianfamily, 1947-1997. London: Zed Books. Duncan, I. 1999. Dalits and politics in rural north India: The Bahujan Samaj Party in Uttar Pradesh. Journal of Peasant Studies 27:35-60. Dutta, N. 1997. Arya Samaj and the making of Jat identity. Studies in History 13:97-119. Godbole, M. 2000. Co-operative sugar factories in Maharashtra-Case for a fresh look. Economic and Political Weekly 35(6):420-24. Gupta, A. 1995. Blurred boundaries: The discourse of corruption, the culture of politics and the imagined state. American Ethnologist 22(2):375-402. 1998. Postcolonial developments: Agriculture in the making of modern India. Durham, N.C., and London: Duke University Press. Gupta, D. 1997. Brotherhood and rivalry. New Delhi: Oxford University Press. Harrison, G. 1999. Corruption as "boundary politics": The state, democratisation, and Mozambique's unstable liberalisation. Third World Quarterly 20:537-50. Harriss-White, B. 1997. The state and informal economic order in south Asia. Paper presented at the Moscow School of Social and on Sciences Economic Colloquium Exploratory/InformalEconomies: Substance and methods of study, January. 1999. How India works: The character of the local economy. Cambridge, U.K.: Cambridge Commonwealth Lectures. Hasan, Z. 1989. Self-serving guardians: Formation and strategy of the Bharatiya

Kisan Union. Economic and Political Weekly 24:2663-70. . 1995. Shifting ground: Hindutva politics and the farmers' movements in Uttar Pradesh. In New farmers movements in India, ed. T. Brass, 165-94. Ilford, Essex, U.K.: Frank Cass. . 1998. Quest for power: Oppositional movements and post-Congress politics in Uttar Pradesh. New Delhi: OxfordUniversity Press. Herring, R. J. 1983. Land to the tiller: The political economy of agrarian reform in south Asia. New Haven, Conn.: Yale University Press. Hirsch, L. V. 1961. Marketing in an underdeveloped economy: The north Indian sugar industry. Englewood Cliffs, N.J.: PrenticeHall. Jeffrey,C. 2000 Democratisationwithout representation? The power and political strategies of a rural elite in north India. Political Geography 19:1013-36. . 2001 A fist is stronger than five fingers: -Caste and dominance in rural north India. Transactions of the Institute of British Geographers 25(2):1-30. Jeffrey, C., and Lerche, J. 2000. Stating the difference: State, discourse and class reproduction in Uttar Pradesh, India. Development and Change 31:857-78. Jeffery, R., and Jeffery, P. 1997. Population, gender and politics: Demographic change in rural north India. Cambridge, U.K.: Cambridge University Press. Kalecki, M. 1972. Social and economic aspects of "intermediateregimes." In Selected essays on the economic growth of socialist and mixed regimes, ed. M. Kalecki, 162-70. Cambridge, U.K.: Cambridge University Press. Kaviraj,S. 1991. On state, society and discourse in India. In Rethinking third world politics, ed. J. Manor, 72-99. Harlow, U.K.: Longman. Khan, M. H. 1996 The efficiency implicationsof corruption, Journal of International Development 8:683-96 2000 Rent-seeking as process. In Rents, .. rent-seeking and economic development: Theory and evidence in Asia, ed. M. H. Khan and K. S. Jomo, 70-144. Cambridge, U.K.: Cambridge University Press. Lerche, J. 1995. Is bonded labour a bound category? Reconceptualising agrarianconflict in India.Journal of Peasant Studies 22:484-515.

CASTE, CLASS, AND CLIENTELISM IN NORTH INDIA

41

. 1999. Politics of the poor: Agricultural labourers and political transformations in Uttar Pradesh. In Rural labour relations in India, ed. T. J. Byres, K. Kapadia, and J. Lerche, 182-243. London: Frank Cass. Lieten, G. K. 1996. Panchayatsin western Uttar Pradesh; Namesake members. Economic and Political Weekly 31(39):2700-5. Lieten, G. K., and Srivastava,R. 1999. Unequal partners: Power relations, devolution and development in Uttar Pradesh. New Delhi: Sage Publications. Madsen, S. T. 1998. The decline of the BKU. at the European Paper presented Conference of Modern Asian Studies, Charles University, Prague, September. Malik,A. 1993. Role offemale population in the Meerut. Unpublished Ph.D. thesis, Meerut University, Meerut, India. Office of the RegistrarGeneral, 1991. Census of India, Meerut District, Uttar Pradesh, Village/Town PrimaryCensus Abstract. Pai, S., and Singh, J. 1997. Politicisation of Dalits and Most Backward Castes: Study of social conflict and political preferences in four villages of Meerut District. Economic and Political Weekly 32(23):1356-61. Perry, P. J. 1997. Political corruption and political geography. Aldershot, U.K.: Ashgate. Robbins, P. 2000. The rotten institution: Corruption in natural resource management. Political Geography 19:423-43.

development agrarianeconomyof district of

Rs. 24 crore cane scam unearthed. 1998. Economic Times 18 May. Scheper-Hughes, N. 1992. Death without weeping: The violence of everyday life in Brazil. Berkeley: University of California Press. Sen, A. 1997. Structural adjustment and rural poverty: Variables that really matter. In Growth, employment and poverty: Change and continuity in rural India, ed. G. K. Chadha and A. N. Sharma, 78-109. New Delhi: Vikas. Sharma, R., and Poleman, T. T. 1994. The new economics of India's Green Revolution: Income and employment diffusion in Uttar Pradesh. New York: Cornell University Press. Singh, J. 1992. Capitalism and dependence: Agrarian politics in western Uttar Pradesh 1951-1991. New Delhi: Manohar. Trouble at Simbhavli sugar mill. 1997. Amar Ujala 28 January. Wade, R. 1985. The market for public office: Why the Indian state is not better at development. World Development 13:467-97. . 1988. Politics and graft: Recruitment, appointment and promotions to public office in India. In Corruption, development and inequality: Soft touch or hard graft? ed. P. Ward, 73-110. London and New York: Routledge