43069 Paper No. 109 /December 2007 Systems of Capture: Reassessing the Threat of Local Elites Benjamin Powis Summary Findings Elite capture is widely cited as one of the main threats to successful community based development, yet it remains the least understood. This paper draws on a qualitative panel study of village politics in ten rural localities in Andhra Pradesh in order to examine the changes in configurations of local elites across two local elections. These case studies highlight the complex and contextual nature of elite capture as it evolves in rural India as well as the importance of recognizing the crucial role that elites play in driving and distorting participatory development initiatives. The analysis presented here builds on recent evidence of the erosion of traditional forms of authority vested with the high caste and major landowners in Indian villages. Within this fluid social context it is, however, found that informal systems of capture can evolve that regulate and constrain the potential for divisive forms of capture. These informal systems do have a cost which is manifested in rent seeking or corruption. It is argued that ordinary villagers perceive that a portion of this cost is a reward for the vital services performed by their local leaders. Hence, at least in the short and medium term, the focus of community driven development and local governance projects should be on harnessing informal systems of capture to ensure pro-poor outcomes. SOCIAL DEVELOPMENT PAPERS South Asia Series Paper No.109 / December 2007 Systems of Capture: Reassessing the Threat of Local Elites Benjamin Powis This Working Papers Series disseminates the findings of work in progress to encourage discussion and exchange of ideas on social development issues. Papers in this series are not formal publications of the World Bank. The papers carry the names of the authors and should be cited accordingly. The series is edited by the Participation and Civic Engagement Team in the Social Development Department of the Sustainable Development Network of the World Bank. This paper has not undergone the review accorded to official World Bank publications. The findings, interpretations and conclusions herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/ World Bank and its affiliated organizations, or its Executive Directors, or the governments they represent. For additional copies of this paper, please contact: Social Development The World Bank 1818 H Street, NW Washington, DC 20433 Fax: 202-522-3247 E-mail: socialdevelopment@worldbank.org Printed on Recycled Paper Table of Contents Acknowledgements...................................................................................................................................ii Acronyms and Glossary of Terms ..........................................................................................................iii Introduction ................................................................................................................................................1 1. Reassessing the Village Elite.................................................................................................................2 1.2 Study Context ...................................................................................................................................3 2. Systems of Capture ................................................................................................................................6 2.2 Village Leaders in Andhra Pradesh...............................................................................................6 2.3 Political Resources ­ parties, positions and patronage.............................................................8 2.4 Elite Politics in Practice..................................................................................................................9 2.5 Loss to Development? ­ A model of political corruption.......................................................11 2.6 Implications for CDD and Local Governance Projects...........................................................13 3. Case Studies of Political Change in Four Villages...........................................................................16 3.1 Gradual Rise of New Leadership.................................................................................................16 3.2 First Time Victory for the TDP ....................................................................................................18 3.3 The `good' politician ­ building patronage ..............................................................................20 3.4 Re-emergence of a bi-polar polity..............................................................................................21 3.5 Outline of Study Methodology.....................................................................................................23 References .................................................................................................................................................25 i Acknowledgements This research was commissioned by the South Asia Social Development Unit of the World Bank as part of the SIDA supported `Enhancing Local Level Accountability' project. Thanks are due to Lant Pritchett, Maitreyi Das (co-TTLs) and Junaid Ahmad (Sector Manager) for their support. The paper draws significantly on the author's DPhil fieldwork, supervised by Prof. James Manor at the Institute of Development Studies in Brighton (UK), with funding support from the Economic and Social Research Council. Special thanks are due to Divakaram Rao for his dedication in conducting field-work and to Julien Bernard Labonne, Asger Christensen, Varalakshmi Vemuru, Gillette Hall and David Post for comments on various drafts of this paper. Responsibility for errors rests with the author. The views expressed here do not necessarily reflect those of the World Bank or its member countries. ii Acronyms and Glossary of Terms BC Backward Caste CDD Community Driven Development CPI(M) Communist Party of India - Marxist Mandal Intermediary tier of the panchayat system and unit of local administration (equivalent of block in other states) MLA Member of Legislative Assembly MP Member of Parliament MPTC Member of the Mandal Panchayat Committee Lakh (L) 100,000 rupees - approx. US$2500 Sarpanch Village Panchayat President SC Scheduled Caste SHG Self-Help Group ST Scheduled Tribes TDP Telugu Desam Party Up-Sarpanch Vice Sarpanch ZPTC Member of the District Panchayat Committee iii Introduction Elite capture is widely cited as one of the main threats to successful community based development. Just as the benefits of involving communities in poverty alleviation programs are well acknowledged, so too is the fact that `endogenous community imperfections' (Platteau 2002) undermine the pro-poor outcomes of these programs. In response, policy actors and analysts have sought to overcome the problems of capture through two main routes. The first has been to improve project design so as to insulate projects from elite domination. The second has been to focus on empowerment of the poor as a means to develop their capacities vis a vis the elite. Evidence suggests that it is unrealistic to expect (external) development initiatives to equalize power relations or by-pass the so called elite in the short or medium term. In a recent review Mansuri and Rao surmise that `most (community development projects) are dominated by elites, and both targeting and project quality tend to be markedly worse in more unequal communities.' They go on to argue that a `distinction between potentially "benevolent" forms of elite domination and more pernicious types of capture is likely to be important for understanding project dynamics and outcomes' (2003). This paper explores the notion that the quality of elite capture is a major determinant of participatory outcomes in India. The analysis presented here draws on research on village politics in ten rural localities in Andhra Pradesh across a five year period. The research set out to build on emerging literature on the changing nature of rural leadership in the contemporary political and developmental context in India, with the following key findings: · A new breed of leaders are emerging in rural India, who are not necessarily from the dominant caste, or part of the formal local governance system · These new local leaders perform a number of useful services, on which the poor and other socially and politically marginalized groups rely · Corruption ­ the most visible manifestation of capture ­ is widely perceived as a reward for this service · There are informal systems of capture ­ or informal codes of conduct ­ that govern the actions of local leaders, and political change at the village level more generally The first phase of research on which this paper is based was carried out in 2002 and consisted of an intensive study of participation and perceptions of local governance in four villages from two districts. A study that focused exclusively on the local elites in eighteen rural localities was also carried out in the same year and districts using semi-structured interviews with 181 reputationally sampled local leaders (see section 3.5). In 2006/7 a follow up study was carried out when ten selected localities out of the original eighteen were revisited. This was one of the first qualitative panel studies of the process of elite politics across electoral terms. This study had two main objectives. The first was to investigate the outcomes of the 2006 village election and the changes in the configurations of the elites. The second was to analyze the impact of the change in the state government on local affairs, following the landslide victory of the Congress in 2004. The first part of this paper highlights some of the key issues relating to the changing nature of rural elites and outlines the social and political context of Andhra Pradesh where the study was located. The second draws on findings from this research to develop an analytical model of the process of capture and examines its implications for community driven development and local governance programs. The third section consists of four detailed accounts of selected case studies of village politics, along with an overview of the methodology used during fieldwork. 1 1. Reassessing the Village Elite The changing nature of elites in rural India is a recurrent theme in recent research on local politics and participatory development in rural India. With few exceptions (Kripa 2007) there is a broad consensus about the decline of the traditional axis of rural authority along the lines of social status. As Krishna argues, the traditional elite have been replaced by a new set of political entrepreneurs or naya netas who are younger and more educated than their counterparts (2002). Manor (2000) refers to these people as fixers who have particular skills in mediating between villagers and higher level bureaucrats and politicians. Importantly, the status of these elites is not conferred upon them exclusively by their caste or landownership, but rather on their ability to adapt to the new compulsions of democracy and development in rural India. The erosion of pre-existing structures of social1 and political organization has not been a uniform process across Indian states2. Yet there are clear indications of this trend from a range of different contexts. For example, in his study of Madhya Pradesh and Rajasthan, Krishna finds no evidence of a permanent grassroots-level political party structure. `Rather than aligning themselves with any particular political party, ..., young development-orientated leaders in villages usually drive hard bargains by playing off one political party against another' (2002: 152-156). Even where well organized party structures are found at the local level, namely in the left dominated West Bengal, recent ethnographies have shown that the controlling influence of the parties is far more limited than is often assumed (Veron et al 2003, Williams 2001)3. The implicit conclusion of these recent studies is that the quality of local leadership remains crucial in determining developmental outcomes at the local level. While the terminology differs, there is a broad recognition that this new breed of rural elites genuinely serve the people who they claim to represent and play a crucial role in driving and distorting the democracy and development in rural India (Manor 2001). `Most fixers tend to be more concerned with their own interests and careers than with the welfare of the people whom they serve, but they are not strangers to altruism. Many take real pride in having helped to arrange benefits for grassroots groups. More crucially, nearly all fixers understand that villagers in most parts of India are now so politically aware and assertive that it is insufficient for them merely to avoid large-scale profiteering. Fixers must also deliver something tangible. Only then will they be able to generate the esteem at the local level on which their careers depend. So the distinction between their self-interest and their ability to assist rural dwellers is something of a false dichotomy' (Manor 2000:822) The resilience of these codes of conduct that act to regulate or constrain the business of brokerage is likely to be an important determinant of the pro-poor outcomes of emerging forms of `elite capture' in rural India. Far from being homogenous groups who are aloof from ordinary villagers; the new elite perform services on which the poor rely and, as such, are assessed according to their performance and the fixing fees that they demand. Furthermore, a more fluid local society can potentially expand the pool of aspiring elites and increase competition between these local actors. Where this occurs, one can think in terms of informal 1The declining importance of caste in rural India is highly debatable, especially in light of the macroscopic picture of caste based exclusion and differential development outcomes. However, there is emerging evidence that challenges this on the basis of micro level ethnographies (Krishna 2003) and a broader literature on the decline of traditional models of the village social order, or the dominant caste model proposed by pioneers like Srinivas. Clearly this is an important issue that demands further research. 2Theses studies highlight the extent to which the transformation of the rural socio-political power structure varies across a state and even districts. In the latter half of his paper, Manor provides an interpretive account of inter-state variations in the stock of fixers and the use of this resource by party leaders (2000). There is a need for more detailed empirical research on these differences so that the implications for the design of CDD and decentralization projects become better understood. 3The myth of the CPM as an enabling influence on decentralization has rarely been questioned. The two accounts referred to here are some of the only empirical studies to have tested (and in the process, significantly challenged) the assumption of the party's pro-poor organizational capacity. 2 systems of capture4 which are influenced both by the broader process of social change in the rural context as well as the change in the formal systems of democracy and development at the grassroots level. As Lauth (1999) argues, there is a dynamic relationship between formal and informal processes of governance and democratic participation. Whereas the former is openly codified through Acts or project level guidelines, the latter draw their legitimacy on the basis of their effectiveness. Accordingly, initiatives that seek to strengthen local governance and to promote community driven development play a crucial role in shaping the context of capture. For example, a recent study of four south Indian states shows that men from the traditionally marginalized sections of rural society are more likely to attend gram sabha (village meetings held by locally elected councils) than others and that there appears to be more targeting of development activities towards these groups where these meetings are held (Besely, Pande and Rao 2005). These findings underline the pro-poor potential of local governance institutions in India, and in particular the possibility that institutional design can overcome the monopoly of the elite over information and resources. Though in most states the process towards democratic decentralization has been limited, rapid advances have been made in terms of involving communities in local development activities through single-sector initiatives5 that take place outside the framework of democratic decentralization. The effect of this new trajectory in participatory development is a contested issue. While these parallel institutions can serve to develop the capacity of the poor and voiceless to participate in local affairs, there are concerns that such processes tend to provide over-representation on user committees for elites and pliable individuals within local arenas (Manor 2004:199). Moreover, donor and governmental support for new approaches to participatory development have significantly expanded the stock of resources that has infused the process of capture with a new dynamism. The following analysis seeks to emphasize the fact that the new elite in India are distinguished from their predecessors not only by their social status but, more importantly, by the nature of the linkages between the elite ­ which I refer to as informal systems of capture. The common theme in the case studies in part three of this paper is that the process of political change at the village level is shaped by the gradual realignments between groups of leaders that are loosely defined along party lines. The dynamism of these groups is based largely on the financial and symbolic rewards that emerge from new policy initiatives to involve `communities' in the development process. In many ways, these groups are remarkably similar to the traditional forms of factionalism and patronage that pre-date the deepening of democracy in rural India that have been described in detail by political anthropologists writing during the 1960s6. While there is little doubt that the social basis of rural elitism is undergoing a process of transformation across the country, in Andhra Pradesh the emergence of new informal systems of capture has advance more rapidly than in most states. 1.2 STUDY CONTEXT During the late 1990s the Government of Andhra Pradesh initiated a radical and controversial experiment to promote new forms of participatory development in rural areas. As in other parts of the country, these initiatives effectively by-passed the formal system of democratically elected local governments and focused on directly empowering communities to take up rural development activities through newly created committees and self-help groups. The case of Andhra Pradesh stood out both in terms of the scale and the political 4A distinction needs to be made between the informal systems that are discussed here and the more commonly used notions of informal institutions. In the Indian context the latter tend to refer to structures of caste, class and kinship. 5The rapid proliferation of self-help groups is one important aspect of this new trajectory in participatory development. The other is the vast array of user-groups for local resource management that have been established under projects managed by specific ministries, notably education, health, irrigation and rural development. 6There is a great deal of literature on the Indian faction leader, most of which predates the democratic upsurge that began in the 1970s. In particular, I am referring to the work of Myron Weiner and Paul Brass and theories of the early Congress Party System. A more detailed discussion of the process of village politics would highlight the relevance of this model in the current context. 3 support for alternative institutions following the launch of the ruling party's flagship development initiative called Janmabhoomi (Box 1). The importance of these initiatives needs to be understood against the social and political context in the state. Janmabhoomi was launched just eighteen months after a change in leadership in the ruling party. Unlike his predecessor, Chandrababu Naidu lacked much in the way of charismatic or popular appeal and hence sought to use this high profile initiative as a means to both consolidate his control over the party cadre and the bureaucracy as well as undermine the local power base of the opposition Congress party (Rao 2004, Manor 2003). Less acknowledged is the fact that Naidu had a far more ambitious objective, one of building a local party cadre, something few regional parties have attempted in the past three decades in India (Powis 2003). Box 1: The New Policy Framework for Participatory Development in Andhra Pradesh Janmabhoomi (literally, land of one's birth) was the Telugu Desam Party's flagship program. Based on a concept from Korea, it was launched in January 1997 and was formulated on three primary concepts: a) bringing the government to the people, b) mobilizing voluntary contributions to development, and c) micro-planning at grassroots level. It was implemented in rounds; initially there were four rounds every year; later this was brought down to two. Each round had a particular theme, such as health, women, water conservation, etc. which was decided by the Chief Minister in response to current issues. For every round, which would have previously required them to travel to the mandal or district headquarters, officials went to the villages and conducted Janmabhoomi meetings where routine office work was conducted. The meetings were also designed as a forum where complaints and demands could be lodged, and where community works could be identified. Where possible, issues were resolved on the spot. Where this was not possible, officials were required to present an `action taken report' at the next village meeting. No specific funds were allocated for Janmabhoomi. The initiative was run by the Planning Department at the state and district level rather than the Ministry of Rural Development and Panchahyati Raj. It drew on a range of state and centrally sponsored schemes including Employment Assurance, rural road maintenance, rural water supply, education and others. In effect, this meant that Janmabhoomi became the main mechanism for the delivery of all government development programs. Distribution of funds or sanctioning of groups (including grants to self-help groups) happened at the time of the Janmabhoomi meetings. For instance, pensions were distributed, revolving self-help group funds were sanctioned, and gas connections and house sites allocated at the time of Janmabhoomi. No less important was the role that Janmabhoomi played in publicizing new schemes and the broader development achievement of the government, through the Chief Minister's address to the people which was read at the start of each village meeting. Janmabhoomi was an alternative system of decentralization that by-passed the panchayats both directly and indirectly. According to a World Bank report, the panchayat institutions were ignored, marginalized and starved of funds The quarterly Janmabhoomi meetings effectively usurped the institution of gram sabha and many of the user-committees received considerable funds. As a result there was considerable confusion over the role of the elected gram panchayat. Of crucial importance in the organization of Janmabhoomi were stakeholder associations, user groups (related to, for instance, watershed development, forest, education, etc.) and in particular women's self-help groups. Five years after Janmabhoomi was launched it was estimated that there were over four hundred thousand women's SHGs in the state (about half the national total). In addition to these, 37,885 youth employment groups were established, 6616 joint forest management groups, 10,292 water user associations, 36,436 mother's committees, 99,618 school education committees and 5499 watershed committees. One of the most important changes that took place was that beneficiary committees were required to contribute up to 30% of the cost of work either in cash or labor through a system of voluntary contributions in the process of executing public works at the village level. In the majority of cases this amount was deposited by local contractors who took up the work independently and `adjusted' this amount on final payment in collusion with the local government officials. Most observers agree that the contribution system had a significant impact across the state, both in terms of institutionalizing corruption and politicizing development at the village level. References: World Bank 2000, Powis 2003, Mooij 2002 The second important feature of the Andhra Pradesh social and political context that needs to be highlighted is the high degree of political awareness and the complexity of the caste based alliances that have prevailed since the emergence of the Telugu Desam Party in 1983. Among the major Indian states, Andhra Pradesh ranks roughly third in terms of voter turnout in state elections, exceeded only by the left dominated states of 4 Kerala and West Bengal. Though competition between the two main upper castes ­ the Reddys and the Kammas ­ tends to dominate the macroscopic view of state politics, electoral success has increasingly hinged upon mobilizing political support among the numerous backward castes and the two main scheduled caste groups, namely the Malas and Madigals. The adoption of development as a political platform and the attempt to mobilize political support through appeals to non-caste groups ­ especially women and the youth ­ can be seen as a reflection of the highly competitive and complex political equations in the state (Ayyangar 2005). 5 2. Systems of Capture During the first few months of field work in Andhra Pradesh in 2002 the importance of the sub-set of villagers who were known as the political leaders was immediately evident. While the term elite carries with it connotations of social status, villagers in Andhra tended to use the term `political leader' (often in the English form) which is suitably differentiated from references to primordial identity or the traditional social order. The leaders are distinguished not only in terms of their knowledge of local politics and the various development schemes, but also in the fact that it was these men (because this group is predominantly male) who were most often encountered at the local mandal office7, in the nearby town, or traveling between villages in the area. As the research progressed, the importance of the horizontal stratification of leaders with different levels of status and experience, as well as the vertical divisions between the elite that gives rise to the formation of groups of leaders became apparent. This two dimensional framework of elite formation is crucial for understanding the process of political change and the articulation of interests at the village level. Competition between these groups - or networks of leaders - is most evident during village elections but in Andhra Pradesh it also played out on an on-going basis, albeit less visibly. These systems of capture in which the elite operate were found to be more important than the individuals who inhabit them. While the form that these systems take varied significantly between localities, the research identified certain generalizable patterns or properties of these systems. 2.2 VILLAGE LEADERS IN ANDHRA PRADESH Interviews with leaders from eighteen rural localities confirm what many other recent studies of village politics have shown, that there is the emergence of a relatively young, educated group of elites, who are playing an active role in village affairs. As table 1 shows, there is also a high representation of backward and scheduled castes in this group, though forward castes accounted for over a third of this sample ­ and therefore are over-represented. Once villagers were questioned directly about who their local leaders were it became evident that this was a broad term that included everyone from senior party leaders to newer political aspirants. In an attempt to understand this stratum of political leaders better a rapid assessment method was devised that involved asking prominent persons to list the main leaders for each of the political parties that were active in the village. To ensure that this list was accurate, opinions were taken of key informants from various social groups as well as supporters of the various parties or factions within the village8. In this way it was possible to reduce what was often a long list down to a core set of leaders that averaged around ten in each village9. These leaders were asked what it is that they do as part of their leadership role. Unsurprisingly, they responded that their most important responsibility is towards party work and election duties, including mobilizing people for meetings, developing the party base and contributing to election campaigns. However, these respondents also considered social and developmental activities an important part of their role as leaders, and in particular attending gramsabhas and resolving conflicts between villagers. 7Mandals are the local unit of administration. These are similar to the block offices in other states, but following reforms in the panchayat and administrative structure in Andhra Pradesh during the 1980s, these units were made smaller than most blocks in other states, containing approximately twenty gram panchayats. 8Along with most other studies and analytical work, there is an assumption that there are a set of elites for a village. In reality, this is often not the case, as some leaders are very important to one group (caste or clan) but irrelevant to another. This is an important complication, though one that lies beyond the scope of this paper. 9The term village or locality is used interchangeably. An important part of the research process involved defining the effective unit of local social and political organization. Briefly, while village politics are formally organized around the unit of the gram panchayat, it was found that in most cases politics tended to be centered within the main village and thus outlying hamlets and colonies were, in some cases, not directly involved. 6 Table 1: Social composition of leaders across 18 villages N=181 Men 164 Gender Women 17 Age <30 22 30<40 53 40<50 72 50<60 20 60+ 14 Caste Category SC/ST 28 BC 85 OC 68 Education No formal schooling 19 Primary 24 Secondary 79 Higher 59 The importance of these social activities was a common theme that emerged in discussions with aspiring leaders. As one youth said: `I started as cadre, worked hard, made relations to higher people ­ this made me a leader. You cannot become a leader just by being wealthy or spending money for village development. Leaders need to be available to the public and need to have positions, good image, political party affiliations as well as relations to higher leaders.' (From author's field notes 2002) Two important insights can be drawn from this comment. Firstly, the leaders in Andhra Pradesh are, with very few exceptions, linked to political parties. While this does not preclude the possibility that caste leaders and other non- political actors play a role in local development, key issues and events tended to be dominated by party politics. As mentioned above, this is an aspect of village politics that is often overlooked, in part due to the fact that the degree to which parties have effectively penetrated to the local level tends to be far less than is the case in Andhra Pradesh. Furthermore, the debate on the interface between party politics and local development has focused on the unique cases of the cadre based parties at the far left of the political spectrum. One key message that arises from the case of Andhra is that the less formalized overlaps between party and local politics demand more attention than has been given to date. Secondly, there are clear distinctions between different types of leaders within the village ­ from apex party leaders at the one extreme to followers or party workers at the other (Box 1). By disaggregating the village elite in this way it is found that the extent of elite capture is dependent both on the presence of a powerful apex leader for each party or group, and even more importantly, on the capacity of these leaders to build an effective base of followers by creating opportunities for advancement or rewards for good service. 7 Box 1: A Typology of Village Leaders · Persistent lead role in events (i.e. over time/election terms) Apex Leadership · Possess independent political relations with leaders outside the village · Key source of patronage locally · High (personal) investment in politics · Dependence on higher leaders Secondary Leaders/ · Own sources of patronage, used to build personal support base Aspirants · Importance in one or few events · Prepared to invest in party development · Politically very knowledgeable · Working for the leaders and the party Party workers/ · Reasonably politically aware, but rely on information from local Followers party leaders · Dependent on local leaders for positions, opportunities and promotion 2.3 POLITICAL RESOURCES ­ PARTIES, POSITIONS AND PATRONAGE There has been a tendency within the literature on elite capture to focus on the financial rewards that the elite earn for performing the service of fixing or brokerage. This reward can be gained in two ways. The first is by assisting individuals to access government schemes like housing grants or crop loans. The second are profits that accrue from the implementation of small scale civil works undertaken on behalf of local governments or other community based organizations. This latter set of activities dominated the development process following the launch of Janmabhoomi in the late 1990s in Andhra Pradesh. As many observers have documented, this gave rise to a new culture of contracting at the local level, whereby local political-contractors took up these works, often in the name of a user committee or self help group. Through political connections these contractors were able to negotiate the payment of the deposit against the cost of the work. As a result, it was widely acknowledged that this new breed of political-contractors often spent less than half the allocated funds on actual construction and the rest was channeled back into bribes and profits (see below). Contrary to what is often assumed, very few of the leaders interviewed were actually involved directly in contracting work on any significant scale10. Roughly half of the leaders interviewed had executed works themselves or in conjunction with other leaders. Among these, approximately sixty percent had undertaken works valued at 2 lakhs or less. This implies that just 28 leaders had executed works of approximately 162 lakhs, over half the total (approximately 300L) value of contracts undertaken in 18 village in the last three years. The dominance of one or a few political-contractors over the largest sums of works as well as the distribution of minor works across a number of cadres is a pattern that is common across all the villages studied (Graph 1). 10This information was collected during the first phase of research, in 2002. As mentioned, it was common for leaders to take work on other's names and hence official records in village level works are unreliable. This information is drawn from interviews with the leaders themselves ­ and is best understood as illustrative. The focus here is on the twelve villages where this information was considered reliable. 8 Graph 1: Distribution of value of contracts undertaken by leaders over the last two years, (in lakhs ­ Rps 100,000) 20 18 16 14 Value of 12 contract undertaken 10 Number of Leaders (L) 8 5 6 2<5 4 <2 2 no works 0 1 2 3 4 5 6 7 8 9 A B C Village Code The fact that a significant proportion of leaders were not directly involved in potentially profitable financial dealings challenges simplistic models of the economic basis of fixing. What, therefore, is the motivation to work for leaders who were cornering the vast majority of profits? The first answer to this question is that followers anticipate rewards in return for their service in the future, and hence their present involvement is predicated on this expectation. The second is that the majority of local leaders are in fact genuinely committed to party political and development work with no expectation of a direct financial return. It is also possible to turn this question around and ask: What are the other forms of rewards that leaders are able to offer their followers? One clear conclusion is that the status of `being a leader' is of far more significant value than has been credited in the current literature. As a leader one not only moves with the key political figures in the village, but one has access to the higher level leaders outside the village, be it the local member of the legislative assembly or mandal or district level officials. These connections are an important resource which the individual can draw on when in need of support in dealing with the local administration, for example. It was also found that effective leaders were those who were able to recognize these non- financial aspects of leadership by conferring positions within the village to their followers, as members of the village education or watershed committee or with the party committees where they existed. 2.4 ELITE POLITICS IN PRACTICE The need for this kind of formal recognition is vital for understanding the process of political change within villages. In most of the localities studied there was more than one group of leaders, most commonly representing the two major parties in the state. Unlike the vanguard model of parties that is said to exist in left dominated states, these village level parties operated with a high degree of autonomy from the high command, and selection of candidates for village elections and other every day political issues were primarily negotiated between the village elites themselves11. Nonetheless, the importance of building credibility within a 11The importance of this flexibility or local adaptation is the main subject of the author's dissertation on which this paper draws heavily. In particular, it was found that the TDP, under the leadership of Chandrababu Naidu was particularly successful in creating political resources (primarily through participatory development approaches) that enabled the party to adapt to different local contexts. This is an entirely new phenomenon that has not been documented in India in recent years, and draws strong parallels to the early era of the Congress party. 9 village level party structure was apparent, notably in the low levels, or highly strategic nature, of defections both by party sympathizers and party activists. Hierarchy and groupism form the basic grammar of the systems in which the elite operate. The village elite can neither be understood as caste or hereditary leaders, nor can they be understood as atomistic or self- serving fixers. Leaders would tend to defect across party lines only when they sensed that significant space was opening up in the opposition or when their progress within their own party was blocked either by ineffective apex leaders or rival aspirants. The understanding of the processes of democratic engagement and political change more generally is significantly enhanced by recognizing the extent to which these informal systems influence the working of village democracy. As the cases that are documented in part three of this paper show, these outcomes can differ significantly across different localities. These cases focus on the different patterns of political change that took place across elections in four localities. A number of important themes can be identified from these examples of politics in practice. As Yadav (1999) argues, one of the defining features of Indian democracy is that the intensity of electoral politics increases as one travels down the various levels of the political system. We can go one step further, and argue that village elections are fundamentally different from any other election that takes place within the democratic system, and assume a far higher degree of importance than any other electoral contest principally due to the proximity of the candidates and the issues involved12. Far from being an open contest, the choices before the voter are carefully managed by the local leadership who support the candidature of strategic aspirants. Furthermore, evidence from this study shows that sarpanch candidates spend an average of 2.3 lakh on their elections13, more than ten times the limit imposed by the Electoral Commission. The requirement to invest in expensive campaigns further limits the pool of candidates who would be willing to put their name forward for election. Reservations play an important role in bringing the lower castes into the ambit of the democratic system ­ but this process of inclusion is mediated through the systems of elite capture. Only in a few cases are these candidates mere proxies, or dummy candidates acting on behalf of high caste patrons14. Rather, the vibrancy of the party system ensures that upper castes who did not adequately reward or include lower caste followers rarely survived in the long term. Finally, the study showed that the lines between personality and party are often blurred but that fundamentally, the importance of having a `good' political leader is regarded as an asset by the broader village society. The issue of corruption in various forms ­ including high electoral expenditures and profit seeking from local development activities ­ is a recurrent theme in these case studies. The accumulation and distribution of political resources are the driving force in the dynamic process of elite formation. However, these political resources are drawn from development programs that are aimed to benefit the poor and the marginalized at the village level. In the next part of this paper a way to analyze this `loss to development' is discussed. 12To test this, as part of the initial phase of this research carried out in 2002, official voter turnout figures at the parliamentary, assembly levels as well as district, mandal and village election were compared. While a gradual increase can be seen with a move down the levels of the political system, a significant jump of nearly ten percent take place when the last two of these levels are compared. 13This average figure is based on data collected from eighteen villages during the first round of field work conducted in 2002. Candidates from the higher castes spent more on these elections than lower castes, but reservations for female candidates had no discernible impact on spending. 14The same argument rarely applied to women, as proxy candidates for women's reserved seats were common and rarely contested. 10 2.5 LOSS TO DEVELOPMENT? ­ A MODEL OF POLITICAL CORRUPTION As Rao and Mansuri argue, the success of community based development is dependent less on the ability to overcome the influence of the local elite than on the extent to which the local elite are benevolent. This paper seeks to extend this argument by focusing less on the individual and more on the character of the system in which the elites operate. In so doing it is implied that much of the losses to development that are a result of capture are used not simply for personal gain but rather feed into the system of capture itself. In an attempt to analyze this process a model of political corruption is proposed as a means to highlight some of the challenges to improving participatory outcomes in rural India. Corruption is the most visible outcome of elite capture. As mentioned above, the focus of the literature to date has tended to be on transactional corruption between the elite, the state and the beneficiary. As mentioned earlier the rules of corruption have changed over the last decade due to the promotion of democratic decentralization and participatory development approaches. One of the most significant outcomes of this new development approach is that the volume of funds vested with the community has increased considerably. The other important outcome is that the elite have been forced to work more closely with the community; resulting in high levels of dependence of the community on their elites. By constructing a model of these exchanges the extent to which this interdependence carries a specific set of costs to the development process can be highlighted. For the purpose of this discussion a hypothetical case of an infrastructure project for one lakh rupees that is being executed by an apex village leader, or political contractor, can be conceived. A similar logic could be applied to elite mediation in the distribution of an individual scheme like a housing grant to beneficiaries. In this case the total public cost of a development activity is given by X in table 2. Out of this a certain proportion A is spent on the actual work itself (or given to the beneficiary) and hence, the remainder X-A=Z can be considered as the loss to development, or corruption. This loss (Z) can be disaggregated in a number of ways. The proportion that is paid directly in bribes or speed money (B) to local bureaucrats can be considered as a distinct category. In the same way, the payments that are made to other political intermediaries (C) can also be isolated by identifying two sub-components. The first are payments made directly to higher level patrons to get the works allocated to this particular village political-contractor (C1). The second are expense incurred in carrying out the process of political fixing, including entertaining local politicians and traveling between the village and the nearby town where they reside (C2). Together the cost A+B+C are borne directly by the village level political-contractor and hence the remainder could be considered his (normal) profits. This argument can be taken further, by factoring in the additional costs incurred by the village level political- contractor to affirm his status at the local level, without which he would not be able to enter into such a contract. These (D) political investments could have two components. The first (D1) are the cost of supporting cadre or followers in the village and day to day expenditure on village party work like meetings and informal gathering. As is seen in the case in part three of this paper, contributing to election campaign funds is another important part of this type of investment. The second component (D2) includes expenditure to support other senior leaders both in the village and at the higher levels. Unlike fixing fees, these political investments are considered to be regular or recurring costs which are not directly related to a specific activity. 11 Table 2: Disaggregating Systems of Capture ­ A model (Super Normal) Profits E Normal profits D2 Supporting patrons or senior party D or leaders Loss to Political D1 Supporting Cadre Development investments C2 (indirect) Entertaining officials, Z C Political Fixing travelling between the block and the village Public Fees Cost C1 (direct) Payments to political brokers X Bribes B Financial transactions with government officials Spent on Development Actual Costs A The expenses incurred as political investments may be substantial, and the extent to which normal profits are ploughed back into political investments is clearly at the discretion of the individual political-contractor. It is not impossible to imagine that he could decide to invest to the extent that he incurs a loss in this particular activity if there is a likelihood that this could earn him further rewards in terms of larger contracts or even a political position of importance. Regardless, the actual extent of corruption, or cash in the hand for the political-contractor at the end of the work, will be determined by the size of E. Of the various types of losses to development the focus of attention to date has been on the transactional costs made between fixers and local officials. Community based development aims to overcome both the financial and allocative inefficiencies that arise when the remote state is charged with the business of village level development. However, as has been underlined in this paper, evidence suggests that there is an associated set of costs that arise from putting funds and responsibilities in the hands of unequal communities. Understood from normative perspectives, these costs could be considered as the transitional manifestation of the imperfections or systemic failing of rural social structures. However, a more radical model can be suggested wherein these costs are a reflection of social realities of inequality, leadership and the dialectics of power and competitive process which, if promoted, could balance out at some `equitable' equilibrium. 12 2.6 IMPLICATIONS FOR CDD AND LOCAL GOVERNANCE PROJECTS The process of elite capture has undergone a significant change over the last few decades in India. The gradual breakdown of the traditional authority of the high caste landed elite is part of a long term, and on- going process of social and political transformation that has reached different stages across the country. In addition, there is evidence that donor and governmental support for community based development and local governance has infused new resources into the increasingly fluid local area which has had the paradoxical effect of extending the scope of the business of brokerage and developing a culture of acceptance of this phenomenon. Assessing and Monitoring Elite Capture Undertaking contextual analysis of the nature of elite capture as part of social assessments during project preparation Establishing systems of monitoring and learning that can assess the dynamic interactions between project inputs and social processes This research yields three main messages for policy actors in India regarding the design of local governance and community driven development projects in India. Firstly, the process of capture cannot be generalized across rural India and greater emphasis needs to be given to assessing nature of social relations and process in operations at the state level. Additionally, variations within states can also be significant in influencing project outcomes. Incorporating an analysis of local elites into social assessments conducted as part of project preparation is one important means to improve the quality of project design and hence developmental outcomes. Owing to the complex and contextually embedded nature of local level elite capture this would require more intensive approaches than are typically employed for social assessments in order to uncover the hidden transcripts of rural power in specific state and regional contexts. It is equally important to stress that local governance and CDD projects can be expected to alter the process of capture during the project cycle by infusing the local polity with new resources and hence altering the incentives for local elites. Large scale quantitative baselines and evaluations are limited in their capacity to track these highly localized processes. Therefore, additional monitoring systems that utilize qualitative approaches to assess the social dynamics in a small sample of villages could yield important information that would enhance the learning process. Both these recommendations would have significant cost implications for project preparation and supervision. These costs would need to be assessed against the potential benefits that would accrue in terms of improving project design and formulating more effective risk management strategies. Cultivating New Leaders Developing systems to identity positive leaders from among project beneficiaries and enabling them to fulfill their potential to represent the interests of their social group Pro-active disclosure of information on local development ­ in appropriate formats ­ to undermine information asymmetries at the local level 13 The second key message of this research lies in the importance of improving the quality of local elite and hence the pro-poor outcomes of systems of capture. On the basis of the evidence presented here, it is argued that it is unrealistic to expect (external) development initiatives to equalize power relations or by-pass the so- called elite entirely, at least in the short or medium term. In short, `fixing' is not a dirty word and until the overall governance structure and processes can truly become pro-poor, local leaders or elites will continue to perform services on which the rural poor rely. Thus, one important means to improve the quality of capture is to cultivate new leaders at the local level who are inclined to acting on behalf of the poor. In the context of social heterogeneous villages, whether or not a caste or particular social group (especially women) have their own leaders who can articulate their interests is a critical determinant of their ability to access development entitlements. As Krishna (2002) argues, social capital ­ or the ability of villagers to network and act collectively ­ is an important but insufficient prerequisite for poverty alleviation. The potential of this capital is significantly enhanced by the presence of active agents, or leaders, who act on behalf of the `community' or their social group. The World Bank, over the last decade, has been supporting a range of projects that aim to support the ability of villagers to act collectively to improve their livelihoods and manage local resources. As many of these projects enter their second phase, one important area of emphasis could be on identifying potential leaders from among project beneficiaries and providing them with additional capacity building support so that they can act as benevolent fixers and overcome barriers to contesting for local leadership. Asymmetries in access to information regarding development schemes and procedures are another important aspect underpinning elite capture. As mentioned above, the villager leaders often travel between the village and local government offices and other political centers in rural areas; and the elite tend to be male and possess some level of education or status that allows them to communicate with lower level bureaucrats and politicians. However, many of the most important potential leaders in rural areas are women, lower castes and the illiterate. To enable these people to become benevolent fixers, greater emphasis needs to be given to reducing the physical and social barriers to information access through the pro-active dissemination of appropriate information materials which are accessible to the poor. Making Informal Systems More Accountable Recognizing the potential for party involvement, electoral expenditure and corruption at the village level to enable more socially rooted project processes Sensitizing local level project facilitators ­ and policy makers ­ to the process of capture and establishing realistic targets to assess project success Finally, making informal systems of elite capture more accountable to the poor is likely to be the most effective means to improving developmental outcomes in the short and medium term. This is also the most challenging area for agencies like the World Bank to engage. One of the recurrent themes of the case studies presented below is the role which political parties play in distributing development assistance and the high degree of expenditure in local elections. The ineffectiveness of policies and Acts to control these phenomena contributes to the prevailing conspiracy of silence over the role of parties in the process of voice and representation and the perceived utility of various forms of corruption. Like the process of capture more generally, these socially embedded transactions are unlikely to disappear in the short term. Moreover, while sector specific projects may be able to minimalize corruption or elite capture, more often than not the result is at best a temporary displacement across sectors, schemes or institutions. An alternative approach is suggested here; one that aims to strengthen these informal social processes through strategies that aim to enhance pro-poor outcomes. 14 There are at least two possible routes towards engaging with informal processes of participation and representation. The first relates to the policy level, and the need to open up channels for a more open debate on these issues in order to reassess the policy frameworks for community based development and local governance. The second relates to the project level, and the need for greater attention to both sensitizing project staff to social realities and setting more realistic targets for countering the negative forms of capture as part of a long term ­ or incremental ­ strategy. A broader theme arises from this analysis, one that calls for notions of `community' and collective action to be examined more closely. Enhancing the voice of the poor and marginalized groups in local development underpins both CDD and local governance approaches. As the case of Andhra Pradesh shows, at least in the short and medium term, the participation of these groups tends to be mediated by local actors ­ leaders or elites ­ whose involvement results in a distortion of developmental outcomes. When exactly the long-term objectives of equality of voice and collective action can be expected to emerge remains a question. The dynamics of the micro-political economy of local development is likely to provide the context in which voices of the poor and community development are articulated in the foreseeable future. Hence greater efforts need to be made to understand these informal systems of capture in order to enhance their pro-poor outcomes. The following case studies serve to highlight the vibrancy of the local political economy of development as it has emerged in rural Andhra Pradesh. The focus of these studies is on the process of political change across village elections. However, parallel participatory development programs that have limited formal links with the local governance system are intrinsically linked into the process of elite formation and hence, electoral outcomes. We also found that village elections are, to varying degrees, influenced by supra-local politics at the block or district level. These findings further emphasize the importance of understanding community based development not only as a political process at the local level, but also as a part of the broader political system. 15 3. Case Studies of Political Change in Four Villages This section consists of four stylized accounts of the complex process of village politics that have been compiled from research carried out in ten localities over a five-year period. These cases have been selected to reflect on contrasting processes of political change in specific rural localities. In all cases, names of people and places have been excluded or changed to preserve the anonymity of the people involved. Elected Post in the Panchayat System in Andhra Pradesh Villages vote directly for all members marked in bold District Panchayat (Zilla Parishad) President (indirectly elected) Members (one per mandal) Mandal Panchayat President (indirectly elected) Member (one member from constituency roughly approximating the gram panchayat Gram (village) panchayat Sarpanch ­ directly elected Ward members 3.1 GRADUAL RISE OF NEW LEADERSHIP This case documents the gradual emergence of a new Yadav leader from the backward caste in PM village that has long since been dominated by a strong bi-polar struggle between high caste Reddys. Shrewd strategy and caste-based support enabled a Yadav leader to establish a strong foothold at the village level gradually over two electoral terms. The importance of this case is that previously political aspirants from the lower castes had been managed and manipulated by the entrenched Reddy elite. In this case, the new Yadav leader has been careful to work with these structures and as a result is the first major non-Reddy political leader the village has had. Approximate Breakdown of Background - The village of PM is a major panchayat, located Caste Wise Votes in PM near a national highway and 5km from (***) mandal headquarters. OC Reddy 55 The village has a famous temple, and the high price of real estate Kamma 25 makes this one of the richest panchayats in the area. Politically this Brahmin 30 is a very important village, home to a number of important BC Goud 400 mandal level leaders. While most stay away from village politics, Yadav 400 these big leaders play an important role in managing village Chakali 200 politics. One example of this is that, in the recent election the Other 365 mandal Congress leader (a native of PM) gave Rps 50,000 to a SC/ST SC Madigal 650 secondary level factionalist to ensure his defection from the TDP SC Mala 150 to the Congress. The management of these subtle realignments is Total ST 45 vital in understanding village politics in PM. The presence of big leaders also ensures that both major parties ­ the TDP and the Congress ­ have their own strong independent sources of funds (or political resources) that have supported a large pool of village level leaders from both parties. As a result, despite the fact that this is regarded as a Congress village, the main opposition party (TDP) and the process of horse-trading between parties, is very vibrant. Village politics in PM has long since been dominated by the Reddy lobby, more specifically by two entrenched Reddy leaders, one from the Congress and one from the TDP. Despite being numerically dominant, BC and SC leaders have tended to be used as pawns by the Reddys. Congress has held the post of 16 sarpanch, except in 1995 when the post was reserved for an SC. In this election the TDP was successful in attracting the most prominent Madigal leader resulting in a virtually unanimous election. However, the TDP has held the post of MPTC for the last two terms, through a BC (Goud) leader in 2001 and a Kamma leader in the 2006 election. This is the village where the highest election expenses were recorded across the villages studied. In the last sarpanch election the winning candidate spent 12L and the opposition spent in excess of 5L. The previous election was a contest between the two major Reddy leaders who spent in excess of 13L between them. Background of Yadav - The Yadav leader, a thirty-two year old BA degree holder and owner of ten acres of land, made his entry into active party politics in 2000. Before this he had been a TDP sympathizer but was not very active. In the lead up to the 2001 panchayat election he argued with the TDP leadership that there was a need for a strong BC at the village level. The party did not respond as they had been following a strategy of backing the Goud community to build a support base among the BCs. Instead the Congress offered him a ticket for the District panchayat elections as they did not have a BC candidate for this reserved seat. Furthermore, they feared him as a potential leader of an anti-Reddy front and hence felt it prudent to attempt to co-opt him. With the support of the senior Reddy leaders, Yadav was able to win this election having spent only 3L. Yadav's victory was also aided by the fact that another BC factionalist (from the Padmashali community) also wanted the TDP ticket. When it was given to a Goud, he worked behind the scenes in support of the Yadav (Congress) nominee. 2006 Sarpanch Election ­ The sarpanch post was reserved for a BC woman in 2006. Yadav saw this as an opportunity to consolidate his power within the village and got support from Congress leaders including the MLA. He fielded his mother as a candidate, as his wife was too young. He had shown visible development in village during his tenure as the ZPTC. As a result, even the CPI(M) cadre in the village supported him, despite having a state level alliance agreement with the TDP. Yadav also promised mandal leaders and the MLA that he would shoulder all village level election expenses ­ giving 20,000 to each ward member as part of the total 10L that he spent on his own (mother's) election. Part of this electoral fund was used to distribute Rps10-15,000 to SHGs and Ryot Sangam (farmers groups) leaders, as well as 100rp per household. Three days before the election he brought a key factional leader (mentioned above) from TDP into the Congress by offering him the post of chairman of the temple trust committee, as well as a cash payment of Rps 50,000. He also promised that he would allow him to take contracts directly from the MLA and the MP and made this man his campaign manager for the sarpanch election. As manager, this new recruit distributed all funds; and his house was used to store the `election alcohol' and became the nerve centre of the campaign. Yadav fell short of Rps 3L and during the end of the campaign borrowed money off the cousin of the former TDP activist. The following conditions were drawn up: If Yadav's mother gets elected he would give the Overhead Tank contract to the money lender No interest would be paid for first 1L if paid within one year. The remaining 2L would be interest free if paid in year one, else one contract has to be given to the money lender This money was very important in giving Yadav the advantage in these elections, which he won by a margin of approximately 100 votes. Again, factional struggles between various groups of BCs from the TDP played an important role in splitting the TDP vote in his favor. 2006 MPTC election: The recent panchayat election saw another significant shift in village politics. This was the first time that the MPTC was unreserved, and a very senior mandal level TDP (Kamma) leader took the opportunity to ensure that a Kamma was elected, for two reasons. The first was to overcome the factional squabbling between the Reddy leaders and to utilize the Kamma leader's strong links to the BC community. 17 Secondly, as a relatively better off leader, the Kamma was made to promise that, when elected, he would take care of the financial needs of the party in the village. In response, the Congress fielded a BC Goud in the open category in an effort to show that not all power was vested in the Reddy elite. In return for this symbolic gesture, the BC Goud was promised that if he lost he would receive contracts to cover his election expenses. Following his defeat he was given contracts (from the MLA and MP) worth Rps4L in total. Summary: The recent GP election saw the young, educated Yadav leader coming into active village politics. Having been a ZPTC in the previous term, it is important to emphasize that this apparent downward move in the political system is quite the opposite in reality. Entering the village arena entailed confronting deeply entrenched personal and party based dynamics. In so doing it is clear that he was not used by the Reddy elite. Through the careful use of strategy he was successful in strengthening both the Reddy base and in consolidating BC support for the Congress. This may be a long term trend which may result in a stronger BC force against the Reddys, but this will depend on the ability of BC aspirants to work together. However, presently Yadav has to work closely with the Reddys and his success will depend on his ability to balance the need for Reddy support with the pressure to overturn their dominance. One thing is clear, through his election he has become more equal to the Reddys than any other BC leader in the past in this village to date. 3.2 FIRST TIME VICTORY FOR THE TDP The 2006 election was the first time that a TDP candidate won the post of sarpanch in IP village, an event that was the culmination of a gradual decline of the old guard Reddy elite from the Congress and CPI(M) and the emergence of the backward and lower castes. IP used to be a CPI(M) dominated village in 1980s. The CPI(M), Approximate Breakdown of headed by a Reddy, used to have a strong following among the SCs Caste Wise Votes in PM and certain active youths from the BCs. However, during the mid- OC Reddy 160 1990s there was a gradual shift away from the CPI(M) by some of the BC Mutraj 300 BCs and SC youth who were not getting opportunities within the Yadav 350 Reddy led party. The village Congress was dominated by a powerful Other BC n/k group of five Reddys who controlled the majority of contracts and SC/ST SC Madigal 600 committees at the village level. Though parties have always been SC Mala 150 important, dominance was really expressed through Reddy power, Total ST 70 and parties have colluded over time to perpetuate this dominance. This Reddy collusion was powerfully played out in the 2001 when a remarkable alliance was struck between Congress, BJP and CPI(M) (contrary to the state level party alliance) to stop the new TDP leaders from coming to power. In this election the post of sarpanch was reserved for an SC. The Reddy alliance agreed they would support a Congress candidate for sarpanch, and arrangements were made to allocate ward member seats. A financially crippling struggle between brothers (one from TDP and one from Congress) in which each spent more than Rps1.5L resulted in the Reddys getting their candidate elected. The BJP were allocated the post of up-Sarpanch and an SC woman from the CPI(M) was elected as MPTC. It was widely known that the SC sarpanch was a puppet of the Reddys - who took all contract work in his name - which further contributed to discontent in the village. While not holding power, the non-Reddy leaders were very active in village development programs launched by the TDP government, including running the night school, youth groups and in supporting self-help groups. However, one of the key factors that held the village level party back was the lack of a single powerful leader. While there was a strong support base for the TDP, this village stood out among all those studied in terms of the sheer number of parties ­ as some of the disaffected former CPI(M) followers had joined the 18 regional parties. This was the only village where the BJP had a following. Despite this apparent vibrancy, it was clear during the research in 2002 that the village lacked a single credible non-Reddy leader. The only candidate for this role was a BC, called AB, who was a former CPI(M) state level youth leader and was well connected with the local MP and other mandal leaders. AB was an insurance agent and contractor in his thirties, and was educated till the intermediate level. However, AB did not live in the village and between 2001 and 2006 was not interested in getting involved in politics in IP. A final important feature of this village was that all communities were divided geographically. Within the main village there were clearly defined areas where the BCs and the SCs lived ­ and behind the panchayat office was a street of large concrete houses where the Reddy leaders lived. There were also five small tribal `thandas' but these residents always remained very removed from the affairs of IP ­ and many felt closer to the neighboring gram panchayat. 2006 Sarpanch Election - By 2006 AB had been appointed as the mandal youth leader of the TDP and when the post of sarpanch was reserved for a BC he was keen to take the opportunity to contest. There was no dispute over this from any other TDP aspirants. One of the reasons for this was that AB was able to invest Rps5L for this election, and no other BC was prepared to match this. To do so however, AB needed to borrow money from a Vaisha man in the nearby town. It is said that he was offered financial support from the TDP mandal president, but that he refused. However, the support of this key leader played a vital role in giving him the confidence to contest in a village where the Reddy Congress had long since dominated. The innovative strategic thinking employed by AB needs, at least in part, to be attributed to the influence of this key ally. The first element of this innovation was the decision to strike a deal with the CPI(M) leaders for their support. In return the TDP did not put up a candidate for the MPTC (the CPI(M) lost all the same) and an SC ward member from the CPI(M) was made up-sarpanch. The alliance between the TDP and the CPI(M) was a village level deal ­ though since the Congress won the state election both parties formed part of a loose anti-Congress front. More importantly the mandal TDP president was also a Reddy (as well as a very powerful local politician ­ and ex-CPI(M) cadre) and therefore able to broker the deal with the village level CPI(M) (Reddy) president. This was also the first time that a candidate campaigned actively in the tribal thandas. Previously strategies to garner the tribal vote were limited to providing trucks to bring them to the polling booth. Besides spending liberally on cash inducements and on alcohol, AB brought cell phones for some of the key tribal leaders. In addition, he worked out a cooperative arrangement with the Congress leader from the neighboring village (who held great sway in the IP thandas and who was fielding an ST candidate in the MPTC elections) by allowing him to continue managing development work in that area. This move was important as it broke the long term understanding between the IP Reddys and the Reddy leader from the neighboring village. The Congress fielded another key Yadav caste leader, of the same caste as AB, who was a supporter of one of the main IP Reddy contractors. The Yadav spent only Rps 2L on the election and was promised contracts in return for contesting. The Congress candidate was from a higher sub-caste within the Yadavs than AB ­ and AB had, over the past few years, spent over Rps 1L on activities to support the Yadav community in the village in an attempt to unify them as a group. As a result, he and the Congress nominee were long standing rivals. Further, AB took the opportunity to wage a strong anti-Reddy campaign in the village, taking up the case of the treatment of the former sarpanch who was widely seen as a proxy for his Reddy puppet masters. By turning this election into a caste issue he was able to secure the support of a few SC leaders who worked with him to ensure greater support from the Madigals. One important factor in this BC/SC alliance was that AB ensured that the defeated (TDP ­ SC) candidate from the last election was given the post of village education committee chairman by asserting his influence through the higher level party leadership. 19 Summary: The political changes in IP are closely linked to caste, and in particular the stranglehold that the Reddy Congress and CPI(M) leaders had long since maintained. However, these issues were articulated through party politics ­ and the TDP became the only party that would give space for alternative leaders. However, it was not until a skilled leader emerged at the village level that this transformation took place. Importantly, translating a lingering anti-Reddy sentiment into an electoral victory involved a complex deal brokerage that drew on the support of higher-level (Reddy) party leaders. 3.3 THE `GOOD' POLITICIAN ­ BUILDING PATRONAGE This case study focuses on BBL - the native village of the then mandal TDP president, called RR, who played an important role in helping the BC leader turn the political tide in IP village. RR was a major figure in local politics ­ and came from a family of Reddy Patels who had a long history in the area ­ formerly as CPI(M) leaders. RR was the first in his family to jump to the TDP in 1999 and since then has molded a political career for himself with great success. What this case shows is that, despite being mandal level president, the mark of a successful politician is his ability to distribute patronage among village level cadres. A major gram panchayat, 10km from *** mandal headquarters, BBL was known for village factionalism. In the 1980s this gram panchayat was controlled by a powerful Reddy leader (MMR) who had a great deal of support from all communities and was the first elected mandal president (in 1988) and then became head of the mandal CPI(M) committee. At the village level, MMR was supported by the family of Reddy Patels who held the post of sarpanch for three terms. The last time was 1995 when the son of the former Reddy Patel leader, called RR, was elected sarpanch. In the lead up to the 2001 election the political equation in BBL was undergoing some important changes. From the late 1980s a factionalist called SK - a Munur Kapu (BC) ­ had started to build a support base by mobilizing non-Reddys against the Reddy leadership. Factions intensified during the early 1990s and the 1995 election was a fierce contest between RR and SK. In 1999 RR had shifted to the TDP sensing the declining opportunities to promote his cadre through the CPI(M). In 2001 when the post was reserved for a BC, SK successfully promoted his wife who became sarpanch. In this election RR and MMR fought hard to stop the emergence of SK, investing 2L in the election. They were successful in getting eight out of 12 ward member seats for the TDP. In the same election, RR won the MPTC post with little contest from the Congress who, sensing inevitable defeat, fielded an SC for an unreserved seat. SK was a known as a `rowdysheeter' ­ or troublemaker ­ and tried to dominate other BC leaders in the village. This was a particularly challenging task as there were at least 6 major BC castes in the village. He had very little standing with the mandal level Congress party but no one else was forthcoming as a village political leader. During his (wife's) tenure his main objective was to make money from contracts, and he gave no opportunities to anyone else in the village. In the meantime, RR was very active at the village level and established an NGO that was supported by his brother, a non-resident Indian in America. The main focus of the NGO was education. In addition to this, RR proved his skills as a political leader, operating at the mandal level to forge ties with the two major factions in the TDP. Probably the most important event during this time took place shortly after the 2001 sarpanch election when the chairman of the village education committee was killed in BBL. Details around this event are unclear, but many believe that this was in some way related to factional politics. As a result, the atmosphere became very tense and, during the initial round of fieldwork in 2002, few people in the village were prepared to discuss village politics. Many villagers were accused in this case, including a number of RR's key cadre who had spent two months in prison. RR was widely understood to have played a managing role in securing the eventual release of the TDP cadre, thus proving his capacity to deal with local police, as well as the level of responsibility he assumed for the protection of his followers. 20 2006 election: In 2006 the post of sarpanch was reserved for a tribal candidate. BBL had six tribal thandas (habitations) making the selection of candidates crucial. In this election RR's wife was elected as the ZPTC but he nonetheless paid careful attention in managing the sarpanch election in the village. There were two main contenders from the TDP and RR chose the one who was less well off, but more firmly rooted in village politics. This candidate spent a total of Rps 3L on his campaign, of which 1L was donated by RR. The campaign was a complex negotiation of various community bases. For the Yadavs, they brought cooking utensils and constructed a temple to a local god in their locality. In one of the thandas, a motor and water supply pipes were put in place. The Vaderra community, many of whom were working in construction sites in Hyderabad, were persuaded by a caste leader to return to the village to vote. Out of their 550 votes, the TDP claimed to have secured 530. Another key group were the Malas, among whom over Rps 35,000 was distributed and a promise was made that one member of their caste would be nominated for the next election, as it was expected that this seat would be reserved for an SC. In addition, RR supported the election expenses for the ward members and other expenses to the tune of an additional Rps2L. RR's ability to work the equations of caste and leadership in the village was based on a long standing effort to ensure that his cadre was rewarded, even during the time of the Congress sarpanch. The main instrument for this was the allocation of positions in village level committees as well as the distribution of civil work contracts that he was able to secure through his mandal and district level connections. The village controlled three of the four village education committees, the single window director post (of the agricultural cooperative) as well as the chairmanship of the water user association. Most importantly, RR ensured that these posts were filled with key party cadres who held some independent support base and, more importantly, allowed these position holders to undertake work directly, offering only support where needed. Analysis: This case has been included to highlight the extent to which successful leadership is not something that is conferred by caste or traditional status ­ but rather by the active and skillful management of followers and vote banks. While RR had developed his own independent base at the mandal level, the strength of the base was critically dependent on his ability to maintain a stable foothold at the local level. Compared to village leaders, his ability to do so was ensured by the fact that he was not personally dependent on an income from village level contracts and was therefore able to distribute patronage and win a loyal support base. 3.4 RE-EMERGENCE OF A BI-POLAR POLITY In DVV village in Prakasham District, a TDP candidate held the sarpanch post since 1995. Prior to that, the former apex leader of the Congress (HR) was sarpanch. The strength of the TDP at the time of this research was indicated by the fact that this was the only village where a building (formerly a school run by an NGO) had been designated as the party office, situated at the head of the main T-Junction in the heart of the Kamma colony. The apex leader of the TDP, HB, held the post of sarpanch through the election of his wife in 1995, and in 2001through proxy control of a scheduled caste woman. In 2006 he won the election and, for the first time, became official sarpanch of the village. However, as this case shows, his attempt to consolidate his power, together with changes in the district level political structure, actually served to develop the opposition party in the village. During the initial research period in Andhra Pradesh, DVV stood out Approximate Breakdown of among all the villages studied for two reasons. The first was the Caste Wise Votes in DVV numerical dominance of the Kamma (forward) caste; the second was the OC Kamma 800 political dominance of one semi-professional political leader, called HB, Reddy 100 from the TDP. With its wide streets and lines of concrete houses the Brahmin 10 village looked like a small town rather than a village. At the head of the Vaishya 90 main Kamma street was a building that had been designated as the TDP BC Chakali 85 office. At the insistence of the HB, all preliminary meetings were held in Yadav 80 this office, rather than the panchayat building. Kummari 30 Kamsali 30 Mangali 20 SC/ST SC Mala 650 21 SC Madigal 250 ST Yerukula 76 This case study examines the declining dominance of the HB and the TDP that occurred in 2006. However before looking more closely at the events that took place around this election, some background of the political scenario in 2002 needs to be established. The professionalism of HB as a village political factionalist was striking in comparison to other leaders met during the course of this research in 2002. In the preceding years he had developed a set of very pliant and unthreatening followers ­ or secondary leaders ­ by allocating a vast array of positions to them as leaders of village education committees, water shed committees, and the milk society. In addition, he had personally created additional institutions, including the village development committee and the village level TDP committee and allowed members of his faction to chair these committees as well. Having managed the last sarpanch election he directly controlled this post as well, for which he had supported a member of the scheduled castes. At the time, he personally held only the post of tank committee chairman, a committee which had not received any funds. However, it was widely acknowledged that he personally executed all the contracting work that was being done under all these committees headed by his faction, offering only minimal rewards to the official chairpersons. That HB was a faction (rather than party) leader became evident during the course of the research. The Congress Party, the only other party in the village, was very weak and had only one leader of any standing, called HR. It was immediately clear that HR was totally powerless against HB, capable neither of building any power base nor of challenging HB on what were widely acknowledged to be substandard village works (the case of a crumbling water tank served as an ever present reminder of this). However, there were two other key leaders in the village ­ VR and RR ­ both TDP supporters and both with a track record and status as political leaders approximating that of HB's. However, since 2001 HB had worked tirelessly to sideline these adversaries and as a result, RR and VR identified themselves as `the TDP rebels'. The formation of this rebel faction dated back to the 2001 panchayat election when RR's wife won as MPTC against HB's candidate. However, despite holding this position, the TDP rebels failed to make any inroads into the village political scene, largely due to the strong equation that existed between HB and the TDP MLA. As a result, the TDP rebels maintained that they were loyal to the TDP but were waiting the higher level leaders to step in to settle the factionalism that had set in within the village party. From all accounts, this process of dispute resolution within the party failed to emerge. Instead, the decisive change occurred only in 2004 when the party command decided to shift the sitting MLA to another constituency. The new (TDP) MLA, a former MP from the area, changed the power equation in a number of important ways, largely due to the fact that HB relied on the protection of his former patron. Another important change that took place was that HR ­ the former Congress leader in DVV ­ died in 2003, thus opening up a new political space that the TDP rebels exploited. Perhaps sensing the potential for the TDP to loose power in the state election, the rebels decided to shift their allegiance to the Congress, and in so doing emerged as the overnight leaders of the village level party. Accordingly, the emergence (or return) of bi-polar politics in DVV was, at least initially, the result of changes in the supra-local political scene. 2006 Sarpanch Election: The sarpanch election appeared to be a straight contest between HB and RR's wife (ex-MPTC) and thus a real test of the village factions. HB won with a margin of only 36 votes. However, this contest needs to be understood within the broader context of the other local elections that took place in the same year. In reward for shifting loyalties to the Congress, RR had been given the party ticket for the ZPTC elections and won. For the concurrent MPTC election he nominated one of his supporters and succeeded in winning this post too. However, in retrospect, this was a tactical blunder on the part of RR. By nominating his wife for the post of sarpanch, he was perceived to be asserting a new regime of a single-man dominance (this time 22 under the Congress) which served to help HB in his bid. Secondly, having lost the MPTC by a significant margin (around 500 votes), HB was stirred into action and campaigned with vigor and diligence. With his political foothold at stake, HB took personal responsibly for the door-to-door campaign. In comparison the RR campaign was lackluster, as he was under the impression that as money was given (500 per household) this would suffice. RR spent more than HB ­ (Rps6L vs Rps5L) and hence the outcomes need to be understood as the triumph of strategy. It is rumored that HB succeeded in co-opting an SC from the RR group who leaked out the Congress' strategy of distributing money and that HB used this information to target the cash campaign effectively. Further, the RR group did not concentrate on the ward members ­ whereas HB spoke to all localities, all caste leaders and made sure that the TDP got the majority of the wards (which they did by holding seven out of the ten wards). HB's victory in this election takes on additional importance when the changes to the constituency level politics at the time are considered. Initially HB had wanted to contest MPTC and even approached the new MLA for a ticket and their support to make him mandal president (it is widely understood that, in many cases, this indirect election is dependent on such high level support to manage the result for a favored candidate). The MLA refused to help HB as he was a loyalist of the former MLA. This was part of a deliberate strategy by the Congress leader to undercut HB's fortunes. The Congress saw RR as instrumental in nullifying the TDP's prospects. Hence, winning the sarpanch election was critical for HB, and it was poor strategy on the Congress' part to allow RR to put forward his wife as a candidate. A more shrewd approach would have been to promote a new leader (someone from the BCs for example) as a signal that the party would accommodate leaders from other castes. Summary: The case of DVV shows how village politics is shaped by political changes at the district and state level. However, it also serves to underline the extent to which these changes are mediated by the issues, events and actors that are firmly embedded in the local context. This case is also important as it shows how leaders are calculating in their decision to open up new political spaces at the local level ­ as demonstrated in the time it look for RR and other TDP rebels to make their move to the opposition party. Finally, another key message of this case is that, while personal feuds and fiefdoms drive the process of village politics, parties provide a structure through which political changes are channeled. 3.5 OUTLINE OF STUDY METHODOLOGY Following an initial round of research on participation and perceptions of local governance in Andhra Pradesh, which was carried out in the first half of 2002, a methodology for a more rapid assessment of the configurations of leadership was developed and applied to eighteen villages over a three month period. The approach was to sample elites through a `reputational' analysis. Mitra (1991) uses a similar model for sampling goan ka netas, however the emphasis is on position holders, including panchayat representatives, school teachers, co-operative chairmen etc. A key finding of this piece of research, as others (Ruud 2003), is that these formal posts are not adequate in understanding the actual dynamics of village leadership. In each village, the study team met with the present and ex- sarpanch and other important persons for the first two days. Given the highly politicized context, these big leaders were relatively easy to identify. During these meetings basic information on the village was collected, including details of the last two elections, information about key parties and leaders and other important political and developmental issues in the village. The next stage focused on identifying at least five individuals ­ representing different castes and parties (including at least one woman) to list all the important political leaders in the village. By cross referencing these `long lists', `short list' of key local leaders were arrived at. Nowhere did this list exceed fifteen persons. Across eighteen villages 181 leaders were interviewed, through this selection method. 23 These interviews were semi-structured, focusing on political experience, activities as leader and engagement with development in the village. In addition to these interviews, further key informant interviews were conducted to help develop a political history of the village, including details on elections and other important political events. In 2006 ten of these village were selected for a repeat study that was carried out between September 2006 and March 2007 independently by a researcher who had had no involvement with the previous study but had a great deal of experience with interacting with village level leaders. This approach was deemed necessary to ensure an unbiased assessment of the village political scene and to avert the inevitable suspicion that would have arisen should the same foreign researcher return to investigate sensitive topics in these villages. During this research the author was in constant communication with the researcher and met with him twice for a five to ten day period to compile findings after the completion of work in each district. The objective of this second phase of research was to use informal interviews with local political leaders to understand the process around the last elections and other local political events though narrative accounts and to document the present configuration of local leaders and the positions they held through formatted tables. This required the researcher to spend at least five days in each village. Additional time was needed to establish contact with key mandal leaders to understand the broader issues of political change at the constituency level. 24 References Ayyangar (2005) `Inclusive Populism through Development Programs: Some Evidence from India', Working Paper, Campbell Public Affairs Institute Besely, Pande and Rao (2005) `Participatory Democracy in Action: Survey Evidence from South India' Journal of the European Economic Association 3(2-3):648-657 Fuller, C. and V. Benei (2000). Everyday State and Society in Modern India. New Delhi, Social Science Press Gupta, A (2005) `Narratives of Corruption: Anthropologica; and fictional accounts of the Indian state', Ethnography Vol 6(1) Kripa, AP (2007) `Rivalry or Synergy? Formal and Informal Local Governance in Rural India', Development and Change 38(3) 401-421 Krishna, A. (2002) Active Social Capital: Tracing the Roots of Development and Democracy, New York: Columbia University Press Krishna, A. (2003). `What is Happening to Caste? A View from Some North Indian Villages.' Journal of Asian Studies 62(4). Lauth, H (2000) `Informal Institutions and Democracy', Democratization Vol. 7, Issue 4 Manor, J. (2000) `Small-time political fixers in Indian States: `Towel over armpit'' Asian Survey Vol 40 No 5 Manor, J. (2001) `Centre-State Relations,' in Atul Kohli, ed., The Success of India's Democracy, Cambridge Manor, J. (2003). `Democratic Decentralisation in Two Indian States: Past and Present.' The Indian. Journal of Political Science 63(1). Manor, J. (2004). `User Groups: a potentially damaging second wave of decentralisation?' European Journal of Development Research 16(1): 192-213 Mansuri, G. and V. Rao (2004). `Community -Based and -Driven Development: A Critical Review.' The World Bank Research Observer 19(1). Mitra SK (1991) `Room to maneuver in the middle: local elites, political action and the state in India', World Politics, 43:3, pp.490-413. Mooij, J. (2002). `Welfare Policies and Politics: A study of three government interventions in Andhra Pradesh'. Working Paper 181, Overseas Development Institute, London Platteau (with A. Abraham), 2002, `Participatory Development in the Presence of Endogenous Community Imperfections', Journal of Development Studies, Vol. 39, N° 2, pp. 104-136. Powis, B (2003). `Grass Roots Politics and the "Second Wave of Decentralisation" in Andhra Pradesh.' Economic & Political Weekly (June 28th). Rao, V. (2004). `Symbolic Public Goods and the Coordination of Collective Action: A Comparison of Local Development in India and Indonesia', Development Economics Research Group, The World Bank. 25 Ruud, A. E. (2003). Poetics of Village Politics. New Delhi, Oxford Véron, R., S. Corbridge, et al. (2003). `The Everyday State and Political Society in Eastern India: Structuring Access to the Employment Assurance Scheme.' The Journal of Development Studies 39(5): 1-28. Williams, G. (2001) `Understanding "political stability": party action and political discourse in West Bengal' Third World Quarterly 22 (4) 603-622 World Bank (2000). Overview of Rural Decentralization in India - Volume 2. Washington, World Bank Yadav (2000) `Understanding the Second Democratic Upsurge: Trends of Bahujan Participation in Electoral Politics in the 1990s' in Francine R. Frankel, Zoya Hasan, Rajeev Bhargava and Balveer Arora eds., Transforming India: Social and Political Dynamics of Democracy, Delhi: Oxford University Press, 2000, pp. 120-145 26