Caste Structures and E-Governance in a Developing Country Rahul De’ Indian Institute of Management Bangalore, Bannerghatta Road, Bangalore 560076. India. email: [email protected]

September 2009 Abstract Castes, or endogamous kinship relationships, are prevalent in communities across the world and particularly in developing countries. Caste plays a strong role in determining community structures and political power. However, the role of caste as a factor in shaping e-governance design and outcomes is absent in the literature. This paper addresses this particular gap by examining some cases from India. The paper specifically considers whether the priorities of dominant caste groups determine e-governance design and implementations, to the exclusion of marginal and non-dominant castes. Further, it examines if e-governance introductions change or affect the relations of caste groups. The research relies on Structuration theory to provide a framework through which to study these issues. Data from three case studies from India are used to conduct the analysis, and these include the Bhoomi project from Karnataka, the Gyandoot project from Madhya Pradesh, and the VKC project from Puducherry. All three are information kiosk-based projects for providing e-governance services for citizens living in villages and rural areas.1

Keywords: e-governance, development, caste, Structuration, technology, India 1

[The exact citation for this paper will be as follows: De’, R. (2009) “Caste Structures and E-Governance in a Developing Country” in the Proceedings of the 8th International Conference on Electronic Government, EGOV 2009, (eds.) Wimmer, M.A., Scholl, H.J., Janssen, M. and Traunmuller, R., SpringerVerlag:Heidelberg, LNCS No. 5693, pp. 40-53. (The paper is available at http://www.springerlink.com/content/c5x2394545122505/). This paper was awarded the Outstanding Paper Award for the most Interdisciplinary and Innovative Research Contribution at the EGOV 2009 International Conference. ( http://www.egov-conference.org/egov2009/outstanding-paper-awards)]

1

1

Introduction

Many e-governance programmes in developing countries reach into the furthest regions of the rural countryside. These programmes intend to bring governance services, via digital means, to citizens who have little access to modern governance mechanisms. This technology ‘contact’ brings with it new assumptions and new relations of governance; it emerges in a field that is already dense with social relations that are both historically defined and are changing and re-forming in response to the onslaught of modernity. Prior research has examined the impact of e-governance in rural areas, the manner in which the systems are designed to best serve governance needs, and the manner in which participation of rural citizens can be improved, amongst other issues. This research has examined the artifact of e-government systems in the given situation, though not much effort has gone into considering the manner in which the artifact affects social structures and networks of relations. In particular, in the case of India, research ignores the manner in which e-governance affects and is affected by the structures of caste in India. This paper attempts to address this gap in the literature. Caste relations in India are a complex subject to study and understand. There is a notional understanding of caste, and an in-practice understanding based on jati and brotherhoods, which are visible through social practices such as rites, ceremonies, marriage relations, and through political formations. Castes and caste groups are changing and dynamic in nature, as they respond to the priorities and challenges of the social-political environment. The paper begins with a basic framework of caste and caste dynamics in India. This framework is used as a backdrop to understand the design, implementation and impact of a few e-governance programmes that have been implemented in rural regions. To conduct this analysis the paper relies on Structuration theory, a theory that provides a substantial analytical framework to understand the minute details of everyday practices and relations within society. The structure of the paper is as follows: a review of the literature on caste in India; a review of e-governance research in developing countries, mainly India; an outline of Structuration theory; a brief outline of three e-governance projects; analysis of the influence of caste in the projects; and conclusions.

2 2.1

Background Theory and Research Questions Caste in India

Caste in India is best captured by the phrase “idioms of association far more powerful than any other single category of community” ((Dirks, 2001); page 7). Caste is a term applied by the British colonial rulers to understand the myriad complexities of the systems of kinship, endogamy, Vedic varna categories and profession-based communities that existed across India. The most widely held view of caste is that it consists of four broad varna categories arranged in a rough hierarchy - the brahmins, the kshatriyas, the vaishyas, the shudras 2

that drew their provenance from ancient scriptures. However, the current composition of castes is very complex - for example, the most populous castes, currently referred to as Other Backward Classes (OBCs), constitute a wide range of castes corresponding to the middle castes in the varna order. In addition, there are the Scheduled Castes (SCs), also called Dalits, corresponding to traditional manual professions, and Scheduled Tribes (STs), originally from forested tracts, who are outside the varna order. Caste plays an important and significant role in the everyday lives of individuals. In village India, the result of being in a caste implies that beginning from childhood individuals are taught rituals, responsibilities, family and community status and differences, as understood from within their caste position. Differences are sharpened and clarified, and also notions of purity for some castes are established. Everyday practices, calendrical and life cycle rituals are determined by the caste credo. Caste also shapes the political and social contours of village life, (Nicholas, 1968) determining fault lines of contest and emerging affiliations. Central to understanding caste in villages in India is the notion of the dominant caste (Srinivas, 1968). The dominant caste is identified as one having either numerical strength or a dominant status (despite not having numerical strength), which is obtained through various means, such as land ownership, Sanskritisation (adopting Vedic daily rituals) or Westernisation. The dominance provided by these two modes of living also present opportunities for caste mobility, or the ability to move away from the inherited caste to one that is perceived to be higher in the social order. Forms of modernity introduced first in the colonial period, and later, have provided means for certain caste groups to mobilise. School education introduced by the British, and the railways, provided newer forms of access and physical mobility that also enabled caste mobility. In most of colonial India, the British introduced a system of administration of villages and districts that deliberately ignored the caste dynamics on the ground. Administrators from different regions, and mostly of the Brahmin caste, were brought in to have a ‘neutral’ administration. This practice became widespread and was adopted by feudal rulers of princely estates who served the British. Post India’s Independence, this practice continued and led to a widening gap between the local elite, who held sway over village politics, and the national elite who were responsible for implementing national policies in these same villages. This tension led to the slow movement of local dominant caste groups to demand and obtain state power through electoral politics as well as through penetration in the administrative class. National politics in India is now dominated by caste groups and caste affiliation dynamics(Harriss, 1999). This, also, results in a politics of ‘appeasement’ where certain caste groups are given privileges through policy measures, in preference to others. In the popular discourse, caste biases have a negative connotation, as something abhorrent, an evil too. They are oppressive and demeaning, as, for example, those who were called ‘untouchables’ (some Dalit castes). This aspect of caste is reinforced in the media, popular literature, films, novels, and other works of art. Caste biases are viewed as racist in origin and devoid of either an objectivist or a humanist sensibility that argues for 3

eschewing caste considerations in social interactions.

2.2

Review of e-Governance in Developing Countries

There are two broad streams of literature on e-governance in developing countries. One stream identifies, describes and then evaluates projects as to their effectiveness in delivering some governance service(Bhatnagar, 2003, 2004; Rao et al., 2004). This research makes an implicit assumption that e-governance is needed in developing countries, is almost necessary, and the objective of the research is to uncover why it has succeeded (or not) and then to learn from and replicate those results. This research often assumes that technology introductions are deterministic in nature and, further, the impact of such introductions can be measured through simple and visible parameters. A second stream of research explicitly acknowledges the complexities involved in technology introductions and takes a less definite and more reflexive approach towards examining e-governance in developing countries. In this research, technology is not assumed to be a fixed and over-determined entity but a plastic catalyst that introduces change, and is also shaped by it (Williams and Edge, 1996). When information technology is introduced to assist in the process of governance it comes into a confrontation with existing modes of working and culture (Walsham and Sahay, 1999; Madon, 1992); it modifies existing practices and processes and is actively resisted (De’, 2005; Silva and Hirschheim, 2007); it impinges upon developmental issues of access, power, the role of the market and individual liberties and freedoms (Avgerou and McGrath, 2007; Madon, 2004; Prakash and De’, 2007). Though research in this stream has considered the contextual particularities (Walsham and Sahay, 1999; Puri, 2007) of the local situation, the literature is largely lacking in an explicit analysis of caste groups and how they impact or are impacted by information technology introductions. This is a serious gap in the literature, given that caste politics and relations form some of the most important and fundamental social structures in developing countries, and particularly so in India. Any developmental or social innovation in India is introduced into a field that has a pre-existing and evolving social dynamic that will engage with the innovation. This engagement has to be understood theoretically and also in its practical instantiation.

2.3

Review of Structuration Theory

Structuration is a theory of sociology that explains, or draws cause-and-effect relations for, human actions and events in society. This theory builds on the idea of ‘structures’ that are abstract rules or ways of doing, or the meanings associated with seeing and doing, and operate through implicit and explicit codes. Structures have three dimensions, those of signification, domination and legitimation. Structures of signification provide meaning and associations for everyday objects and acts. These structures reside in objects and artifacts of daily use that justify the structures themselves and reinforce their signification as humans continue to use them. Structures of domination are relations of power evident in 4

conversations, writing and other forms of discourse and in human action. These structures are reinforced by humans acknowledging and acquiescing to power relations, such as to those in authority. Structures of legitimation are those rules of conduct and action that are considered appropriate and are sanctioned. These structures determine what is acceptable and habitual in the everyday practices of humans. Structures are not fixed or unchangeable (Giddens, 1984). They are created through practice, by humans, and thus reside in the everyday and ordinary activities of communities of people. The theory of structuration deliberately avoids prior theories that relied on fixed structures in society, thus making human action passive, and also theories that relied inherently on human action, while ignoring the issues of power, social relations and constraints (Jones and Karsten, 2008). Structuration argues that structures existing in society and human action are mutually constitutive and form a duality. So observed social phenomena are not a product of one or the other but of both. Structuration theory enables a richer understanding of the issues that pertain to the use and management of information technology in organisations. For instance, important interactional influences of IT within organisations (Orlikowski and Robey, 1991) are: shaping of IT by humans, facilitation of human action by IT, embedded norms in technology, and the shaping of structures with IT. These interactions arise in an ever-changing and evolving context of the organisation. Structuration theory enables interpretive flexibility (Orlikowski and Robey, 1991), which assumes that technology is never objectively fixed and its nature is changing based on its use. Further, structuration achieved with the introduction of technology is emergent and recursive by nature. More specifically, the “...structures of technology use are constituted recursively as humans regularly interact with certain properties of a technology and thus shape the set of rules and resources that serve to shape their interaction”(Orlikowski, 2000).

2.4

Research Questions

Caste practices are a form of structuration on their own. Caste affiliations and jati practices create structures of signification, domination and legitimation in village life: use of caste symbols (by choice of clothes, as marks on the forehead) on an everyday basis signal caste position and power, as also reinforce the position; access to village resources and affiliation with those in power mark caste domination through everyday living; and living in castesegregated villages provides legitimation of caste-based practices. Introduction of e-governance in villages in India creates a set of interactions between technology and human relations. It is hypothesised that relations between those in different caste groups are affected by the technology, may be mediated by it, and caste groups and affiliations impact the manner in which the technology is designed, implemented and used. In this context the two important research questions that arise are: 1. Do the design and implementation of e-government systems reflect existing caste priorities and practices? Are the systems designed, implicitly or explicitly, to address the priorities of the dominant caste? 5

2. Do e-government systems change or affect in any way the existing caste order and practices? Do they impact the existing relations between different caste groups?

2.5

Methodology

The research questions are addressed by considering three e-governance projects from India. The three projects are the Bhoomi project from Karnataka, the Gyandoot project from Madhya Pradesh, and the Village Knowledge Centres (VKC) project from Puducherry. These projects were chosen as they are all mature projects, they have sustained for more than five years, are widely used, and all have a focus on rural populations. Data for two projects, Bhoomi and VKC, were obtained from primary sources - multiple site visits, several hundred field interviews, and structured surveys - as also from published secondary sources. (Details of the extensive primary data collection are omitted for reasons of space.) Data for the Gyandoot project was obtained only from secondary sources. All secondary sources are cited.

3 3.1

Data and Analysis The Bhoomi Project

The Bhoomi (meaning ‘land’) project of Karnataka was launched in 2001 in a state-wide effort to computerise land records. Bhoomi kiosks enable two services, of providing copies of land records to farmers and of providing a queue system for logging land mutation (sale or transfer) requests. The kiosks are located in all 177 sub-districts of the state and cover the entire revenue land (agricultural land). Farmers go to the kiosks to obtain land records, which are essentially records of their land holdings (listing the pattern of farming, a record of loans taken, etc) that they can obtain for a nominal fee and also quite quickly. Bhoomi kiosks are widely used, particularly for obtaining land records that are later used to obtain loans from banks. The kiosks were designed with the specific intent of giving easy access to land records for farmers. Traditionally, the records were maintained and issued by village accountants (VAs), officials of the revenue department located in villages, and they were known to take a long time to process the records and also demanded bribes (Chawla and Bhatnagar, 2001). Now, the records are maintained at a kiosk located at a sub-district office, and are updated three times a year by the village accountants with data on crops grown by the farmers. The kiosks are managed by fresh, young hires who were specifically trained for the job.

3.2

Caste Interaction with Bhoomi

In Karnataka (Manor, 1990, 1977) the dominant caste groups are divided principally into Lingayats and Vokkaligas. Lingayats are a coalition of castes that formed from a particular 6

religious movement, whereas Vokkaligas are landed peasants who also belong to different endogamous castes. These two groups have dominated village-level politics in Karnataka, and in the post-Independence era also at the state level. An important aspect of this dominance is that almost all chief ministers, who have come from these caste groups, have always played to the interests of these communities, with few exceptions. Government policies, handouts, jobs, schemes, scholarships etc were all targeted at these two communities with the neglect of others, particularly the lower caste groups. As these caste groups began to gain political power, they started replacing Brahmins and non-Hindus in the bureaucracy. This further strengthened their political power in the state. At the village level the bureaucratic functions of revenue collection and land settlement was also dominated by the two caste groups, led by the village headman, and their allies. This combined with the fact that most farmers in Karnataka were owner-cultivators presented a strong class interest of farmers that was dominated by the caste-group interests. Bhoomi’s introduction in Karnataka followed the priorities of the dominant landed castes. It was designed to provide easy access to land records and to make efficient the land sale or transfer process. These benefited the land-owning castes the most, as they were in the best position to use the easy availability of land records to obtain loans and also participate in land transactions. Dalit and lower castes work mainly as landless labour and as tenant farmers in the state and they had marginal use of the Bhoomi system. In a study (Prakash, 2008) of two different land-tenure regions, one in which the majority of farmers were owner-cultivators (the district of Mandya in Karnataka) and one in which the majority were tenant farmers (the district of Koppal), it was found that Bhoomi clearly benefited larger farmers over small farmers, and also land-owning farmers over tenant or landless farmers. In Koppal, elimination of VAs from the land records issue process helped the landowning and absentee land-owners to ensure that their titles did not get diluted (Koppal is dominated by landlords who allow tenants to farm on their lands; if this fact is recorded on the land record by the VA, it could mean loss of rights for the owner). Any entry in the record now requires approval at the sub-district offices, which the land-owning castes have better access to and they can control adverse recordings (eg. name of the legitimate tenant cultivator). The fact that since the time of the introduction of Bhoomi no fresh tenancy applications have been filed in the office of the Koppal district administrator (as also the use of Bhoomi data for disposing off any pending tenancy applications) points to the insignificance of Bhoomi for the non-dominant tenant castes of Koppal. A strong motive for Bhoomi’s design and implementation was to reduce/curtail the powers of the village accountant (De’, 2008). The VAs invariably belonged to the dominant castes and traditionally served the dominant groups well, however, they were also the most helpful officials for Dalit and lower castes in all matters related to access to government services - such as scholarships, subsidies, certificates, loans, etc. With a reduction of the powers of the VA, the non-dominant populations are affected as their everyday access to governance and services is also reduced. In the everyday practices of Dalit and landless farmers, Bhoomi has not acquired significance. It remains remote and alien, also physically removed from the site of living and 7

work. Bhoomi remains disconnected from and irrelevant for acquiring seeds, negotiating with moneylenders, sowing, etc. It is not a carrier of information about prices, weather, subsidies, government schemes, etc. In its distance from the village, it privileges those who can travel to the sub-district headquarters, can afford to lose a day’s wages, can apply for and get bank loans, and have property to transact.

3.3

The Gyandoot Project

The Gyandoot Telecentres were launched in the Dhar district of the state of Madhya Pradesh, in the year 2000. The telecentres or kiosks were small rooms consisting of a computer connected to a network via modem, a printer, an un-interruptible power supply (UPS) box, some furniture, and some additional peripheral equipment (like scanners). 34 kiosks were set up initially, in various towns and villages in the district, with one kiosk in each town, and each was connected to the rest via a local area network and to the district headquarters, in Dhar town. The kiosks were initiated and funded on a public-private partnership basis, with local youth being selected as the kiosk operators or soochaks. The project champion was the District Collector (the senior-most official in the district and also a member of the elite Indian Administrative Service (IAS)). Gyandoot kiosks were designed to provide e-governance services such as complaint registration, access to government forms, purchase of land records, applications for government schemes, along with a host of commercial services such as commodity price reporting, sale of goods, and information exchange. Gyandoot was launched under a formal committee called the Gyandoot Samiti that selected the soochaks and the location for the kiosks. The objective of the Samiti was to have local control, through panchayats (village councils) over the kiosks. The soochaks had to pay an annual fee of Rs 5000 to the Samiti to retain their access to the networks. They earned revenues by levying nominal charges for all services rendered through the kiosks. Gyandoot did not have the impact that was initially promised. The project was awarded the Stockholm Challenge Award in 2000, soon after its initiation, based more on the concept and its bold objectives. Gyandoot had been located in the Dhar district particularly because this was a ‘backward’ district, with about 60% of the population below the poverty line. After the initial phase, and particularly after the champion was transferred out of the district (which is routine for IAS officers), the governance services declined dramatically as the departments responsible for addressing the needs arising from the kiosks lost interest in servicing them. Studies show (Kothari, 2002; Cecchini and Raina, 2004) that there was not much impact on uplifting the poor residents of the district; and the few interesting cases of redress of complaints, better price availability for sale of products, faster response for crop diseases, remained isolated cases with no substantial impact. The kiosks also suffered from problems of connectivity and a steady supply of electricity. Gyandoot kiosks do survive and are in use. Recent studies show that they are thriving on providing a host of services in addition to those they were designed for; the most sought after services continue to be for accessing land records (Tiwari and Sharmistha, 2008), followed by information on market and prices, and the access to exam results from school 8

boards.

3.4

Caste Interaction with Gyandoot

The state of Madhya Pradesh is located in the what is known as the ‘Hindi belt’ that is characterised by a majority of the population speaking the Hindi language. It is also a state that has historically been dominated by the upper castes, as it was amalgamated from princely estates after India’s Independence. Political power was consolidated within the three varna castes, also based on their access to control over most agricultural land in the state. Post Independence, the Congress party dominated politics and held power continuously, with all the chief ministers drawn from the dominant castes. The Congress party retained a long history of factionalism within the state and no individual chief minister could sustain complete control for too long. However, the later leaders made sure that their base was consolidated by incorporating some of the backward castes, SCs, OBCs, and STs, which constitute a fifth of the population in the state. The Gyandoot project was initiated with the explicit developmental intention of addressing the needs of the poorest people in Dhar, those below the poverty line, and this meant largely the SCs, the STs and a fraction of the OBCs. Gyandoot’s three-part objectives of information and service enhancement, entitlement enablement, and capability enhancement (Tiwari and Sharmistha, 2008), were implemented in a manner that enabled the landed middle-class and literate population to have a far better use for its services than the marginal populations (Jafri et al., 2002; Tiwari and Sharmistha, 2008). STs are typically marginal/subsistence farmers who do not produce surplus that they have to market, typically do not have title access to land and so have no need for land records (one of the most popular applications of Gyandoot), and do not have the literacy or wherewithal to use the facilities for training in computing technology. Gyandoot kiosks were of two kinds, those controlled by the village councils, and funded by them, and those controlled privately. The soochaks were appointed by open advertisement and competitive selection by the village councils; they were trained to use the equipment and were provided assistance with setting up the kiosks. There is evidence that the private kiosks were assigned more on the basis of connections and caste affiliation a particular case in point (Sreekumar, 2007) is in Badnaver town, where an upper caste soochak was able to set up a kiosk despite there being one in existence already, on the grounds that he was able to entertain the district officials at his home. Surveys of users and non-users (Jafri et al., 2002; Tiwari and Sharmistha, 2008) show two prominent tendencies: when a kiosk is located within the village or town, the community is more likely to use its facilities; and there is high awareness about its facilities amongst most non-users. The first point is apparently ‘obvious,’ that proximity enables use, but it also underscores the fact that for an alien technology to be integrated into the daily practices of a community, it is legitimated by proximity and a familiarity with the kiosk operators. The second finding shows that awareness does not translate into use; e-governance is brought to the local community by a central power, its mandate is more to centralise services rather than to decentralise (De’, 2008), and as such it represents 9

a prohibitive symbol of state power that is captured and manipulated by the local elite (Sreekumar, 2007). Since the awareness of and knowledge about the kiosks carries mainly through word-of-mouth (Jafri et al., 2002), it substantiates the power relations already existing within the caste structures of the communities.

3.5

The Village Knowledge Centres Project

The Village Knowledge Centres (VKCs) project of the M.S. Swaminathan Research Foundation (MSSRF), an NGO based in Chennai, is a developmental initiative targeted at using information and communication technologies for development. These knowledge centres follow a kiosk model of providing computing technology to rural communities, a part of which includes providing governance information and services. (As opposed to Gyandoot, the VKCs were not designed for providing mainly governance services.) Currently located in the Union Territory of Puducherry, itself a small land mass surrounded on the east by the Bay of Bengal and on the other sides by the state of Tamil Nadu (two districts of Puducherry are not in this land mass), the VKCs provide kiosks that connect the village community through a local network to MSSRF’s resource centres from which information is transferred and exchanged. 15 VKC centres are located in Puducherry, around the central facility at Puliyarkuppam. The centres provide information on market activities, data on government schemes, computing facilities for local industry (such as milk and dairy production), training for computing skills, and Internet surfing (this last is charged). The centres are financially supported by MSSRF through international funding. MSSRF is a prominent non-governmental organisation (NGO) in India that was set up in the late 80s. When MSSRF conceived of the VKCs in the late 90s, they actively engaged the local communities in villages to determine what services to offer, where to locate the kiosks, whom to employ for manning the kiosks, and how to manage them. There was an explicit emphasis on inclusion of women and Dalits in running and usage of the kiosks. The VKCs currently run quite successfully in that they are actively used in most of the villages they are available in. Their financial viability was never an issue as they are funded by MSSRF. Most of the usage is for training in computer skills, which is sought by young persons seeking employment in nearby towns and cities, for information on market prices and for maintaining accounts for local milk producers. Many of the kiosk operators are women and there is a large participation of women in the kiosk usage, though less than that of men.

3.6

Caste Interaction and VKCs

The region of the Kaveri river delta, where the VKCs are located, is a site of intense caste-based politics. Many villages in that region are strongly divided, physically, along caste lines. For example, in a village of Sripuram, in the district of Tanjore in Tamil Nadu, Beteille (1996) noted the existence of three distinct neighbourhoods, those for the

10

Brahmins, called the agraharam, those for the Non-Brahmins (a collection of many middlecastes, including OBCs), and those for the Dalits (or lower-castes), called cheri (Beteille, 1996). It was common practice for members of either community not to visit the neighbourhoods of the others, and if by chance they had to, they performed rituals to ‘cleanse’ themselves. All living practices and rituals were conducted in the region bound by caste lines, and within these bounds too there existed divisions along sub-castes. In the hierarchical order of belief, the Brahmins considered themselves to be the highest and the Dalits to be the lowest. Part of this dominant order also arose from land rights, where Brahmins historically had access to land whereas the Dalits worked as landless labour. The middle-caste Non-Brahmins were land owners but did not have the political power of the Brahmins. A political movement of the Dalits, referred to as the Dravida movement, began in the 1920s, which brought a sense of assertiveness and political organisation to the Dalit groups (Gorringe, 2007). After independence the Congress party dominated politics in the state of Tamil Nadu, but by the 1970s the Dravida movement asserted itself, based on a call for Tamil assertion and also a rejection of North Indian language imposition. State politics was taken over by the Dravida party, which in the course of time developed its own internecine differences and split. By the late 90s and early 2000s, the Dravida parties had fractured into many, with two dominating the political process, however several parties relying on the middle-castes (Non-Brahmin castes) had also gained political power through sheer numbers. Over the years, affirmative action policies by successive state governments had replaced the mostly Brahmin bureaucracy by persons from the Dalit and middle castes. These actions have also led to claims for and contest for power, leading to violence amongst caste groups (Vincentnathan, 1996). When the VKCs were introduced in the late 90s, the promoters explicitly engaged in a dialogue with people in the villages to understand the manner in which caste would interact with the usage of the kiosks. One of the issues that had to be resolved was the location of the kiosks. They had to be in a place considered to be neutral for access. Certain temples and panchayat offices were considered to be suitable for the kiosk offices as they were accessible by many different castes (Beteille, 1996). After the kiosks were set up, some were actively resisted by certain caste groups as it was not in their part of the village and they used political pressure to have them relocated (Parthasarathy, 2004). The VKC kiosks remain largely unused by Dalits (Parthasarathy, 2004; Kenny, 2005). The services and facilities provided are of no use for them. The kiosks have no meaning for their everyday practices as landless farmers. The kiosks are modern inventions that are used by the upper and dominant castes, for accessing various government services and for learning new skills. Even in a village where the kiosk is located in the Dalit areas, the usage is low as the main residents do not have much use for the kiosk’s services, and also because upper caste residents will not venture into the facility, it being outside their caste boundary.

11

4

Conclusions

We revisit the research questions and examine answers to them in the light of the three case studies presented above. The first question was whether the design and implementation of e-government systems reflected the priorities of the dominant castes. For the Bhoomi and Gyandoot cases there is clear evidence to conclude that the services of the systems meet the needs of primarily the dominant castes. Inscribed into the implementations of the kiosks was a rationality that clearly favoured the practices and priorities of the dominant castes in Karnataka and Madhya Pradesh. For the case of the VKCs, there was a deliberate attempt to include the needs of non-dominant and marginal castes, however, through use and emergent configurations, the priorities of the dominant castes prevailed. Caste, an “idiom of association,” manifested itself in everyday practices of access to, mobility around, sharing of knowledge about, and use of the kiosks. Caste affiliation and privilege played directly into the equation for extracting the new resources made available by a powerful technology. Everyday practices of the dominant castes facilitated, and were facilitated by, the easy appropriation of the technology. The second question was whether the technology introduction changed or affected the existing caste order. The evidence from all three cases is conclusive: although the nondominant castes may not be worse off from the presence of the e-government systems, the dominant castes are certainly better off. With easy and better access to markets, information and governance services, the dominant castes have improved their own relative economic advantage. There is strong evidence for this in the Bhoomi and Gyandoot cases. Historically, caste mobility was possible by effective appropriation of new technologies by caste groups, and this appears to be the case with e-governance also. With some caste groups being able to leverage the new technologies better, they are in a position to mobilise their jati with respect to others. In the fiercely competitive space of political and economic dominance, e-governance plays its, possibly, intended role. Caste and kinship groups exist in all communities of the world, and, in particular, in the developing world. So far, research in e-governance has ignored the interactions that such technological innovations have with caste groups (and this paper makes a first attempt to fill this gap in the literature). The findings of this paper underscore the fact that caste concerns not only impact the design and implementation of e-government systems, but also are themselves impacted by the technology. Structuration theory provided the lens by which to examine the everyday practices of individuals in communities as also to examine and interpret the recursive influences of technology and social structures. It helped constitute an a priori theoretical framework with which to read the particular context of e-governance introductions in a rural community. Another strong contribution of this paper is in demonstrating that Structuration theory is valuable in understanding the sociology of caste and e-governance.

12

References Avgerou, C. and McGrath, K. (2007). Power, rationality, and the art of living through socio-technical change. MIS Quarterly, 31(2):295–315. Beteille, A. (1996). Caste, Class and Power: Changing Patterns of Stratification in a Tanjore Village. Oxford University Press. Bhatnagar, S. (2003). Development and telecommunications access: Cases from South Asia. In Avgerou, C. and Rovere, R. L. L., editors, Information Systems and the Economics of Innovation, pages 33–52. Edward Elgar, Cheltenham UK. Bhatnagar, S. (2004). E-Government: From Vision to Implementation. Sage Publications, New Delhi. Cecchini, S. and Raina, M. (2004). Electronic government and the rural poor: The case of Gyandoot. Information Technologies and International Development, 2(2):65–75. Chawla, R. and Bhatnagar, S. (2001). Bhoomi: Online delivery of land titles in Karnataka, India. Web Document. Accessed in April 2005. De’, R. (2005). E-Government systems in developing countries: Stakeholders and conflict. In Wimmer, M., Traunmuller, R., Gronlund, A., and Andersen, K., editors, Proceedings of Electronic Government, 4th International Conference, Copenhagen, Aug 2005, LNCS 3591, pages 26–37. Springer-Verlag. De’, R. (2008). Control, de-politicization and the eState. In Wimmer, M., Scholl, H., and Ferro, E., editors, Proceedings of Electronic Government, 7th International Conference, EGOV 2008, volume 5184 of LNCS, pages 61–72. Springer-Verlag. Dirks, N. (2001). Castes of Mind. Princeton University Press. Giddens, A. (1984). The Constitution of Society. Polity Press, Cambridge, UK. Gorringe, H. (2007). Taming the Dalit Panthers: Dalit politics in Tamil Nadu. Journal of South Asian Development, 2(1):51–73. Harriss, J. (1999). Comparing political regimes across Indian states: A preliminary essay. Economic and Political Weekly. Jafri, A., Dongre, A., Tripathi, V., Aggrawal, A., and Shrivastava, S. (2002). Information communication technologies and governance: The Gyandoot experiment in Dhar district of Madhya Pradesh, India. Working Paper 160, Overseas Development Institute. Jones, M. R. and Karsten, H. (2008). Gidden’s structuration theory and information systems research. MIS Quarterly, 32(1):127–157. Kenny, C. (2005). Behind the digital divide. The Economist, 374(8417):22–25. 13

Kothari, B. (2002). Comment (Gyandoot: When you look closer, the moon has craters). Regional Development Dialogue, 23(2):60–64. Madon, S. (1992). Computer-based information systems for decentralized rural development administration: A case study in India. Journal of Information Technology, 7:20–29. Madon, S. (2004). Evaluating the developmental impact of e-governance initiatives: An exploratory framework. The Electronic Journal of Information Systems in Developing Countries, 20(5):1–13. Manor, J. (1977). Structural changes in Karnataka politics. Economic and Political Weekly. Manor, J. (1990). Karnataka: Caste, class, dominance and politics in a cohesive society. In Frankel, F. and Rao, M., editors, Dominance and State Power in Modern India, volume 1, pages 322–361. Oxford University Press. Nicholas, R. W. (1968). Structures of politics in the villages of Southern Asia. In Singer, M. and Cohn, B. S., editors, Structure and Change in Indian Society, pages 243–284. Aldine Transaction. Orlikowski, W. J. (2000). Using technology and constituting structures: A practice lens for studying technology in organizations. Organization Science, 11(4):404–428. Orlikowski, W. J. and Robey, D. (1991). Information technology and the structuring of organizations. Information Systems Research, 2(2):143–169. Parthasarathy, B. (2004). Information and communication technologies for developement: A comparative analysis of impacts and costs from India. Technical report, International Institute of Information Technology Bangalore. Prakash, A. (2008). Development Paradigms, Social Contexts and Patterns of Technology Use : Study of Land Record Computerization in India. PhD thesis, Indian Institute of Management Bangalore. Prakash, A. and De’, R. (2007). Importance of developement context in ICT4D projects: A study of computerization of land records in India. Information Technology & People, 20(3):262–281. Puri, S. K. (2007). Integrating scientific with indigenous knowledge: Constructing knowledge alliances for land management in India. MIS Quarterly, 31(2):355–379. Rao, T., Rao, V., Bhatnagar, S., and Satyanarayana, J. (2004). E-governance assessment frameworks. Technical report, National Institute for Smart Government, Hyderabad. Silva, L. and Hirschheim, R. (2007). Fighting against windmills: Strategic information systems and organizational deep structures. MIS Quarterly, 31(2):327–354.

14

Sreekumar, T. (2007). Decrypting e-governance: Narratives, power play and participation in the Gyandoot intranet. The Electronic Journal on Information Systems in Developing Countries, 32(4):1–24. Srinivas, M. (1968). Mobility in the caste system. In Singer, M. and Cohn, B. S., editors, Structure and Change in Indian Society, pages 189–200. Aldine Transaction. Tiwari, M. and Sharmistha, U. (2008). ICTs in rural India: User perspective study of two different models in Madhya Pradesh and Bihar. Science, Technology & Society, 13(2):233–258. Vincentnathan, S. G. (1996). Caste politics, violence, and the panchayat in a South Indian community. Comparative Studies in Society and History, 38(3):484–509. Walsham, G. and Sahay, S. (1999). GIS for district-level administration in India: Problems and opportunities. MIS Quarterly, 23(1):39–66. Williams, R. and Edge, D. (1996). The social shaping of technology. Research Policy, 25:856–899.

15