ISSN (Online) - 2349-8846
-A A +A

Bureaucratic Rationality, Political Will, and State Capacity

MGNREGS in Undivided Andhra Pradesh

Rahul Mukherji ( and Himanshu Jha ( are with the Department of Political Science, South Asia Institute, Heidelberg University.

The successful implementation of the Mahatma Gandhi National Rural Employment Guarantee Scheme in undivided Andhra Pradesh underlines the triumph of citizen formation over patron–client politics, aided by a democratic politics. This article argues that its success in Andhra Pradesh depended heavily on how the ideas within the rural development bureaucracy interacted with the political executive. This synergy engendered a state’s capacity to insulate a committed bureaucracy from powerful farmers and construction companies who had a clear interest in thwarting the programme. Elections can in a democracy elevate citizen concern over particularistic populism, when political will is matched by ideational conviction in the bureaucracy.

This study was undertaken prior to the formation of the state of Telangana, carved out of Andhra Pradesh on 2 June 2014. Hence, this article discusses MGNREGS in the context of undivided Andhra Pradesh. It may be noted that Nizamabad and Mahbubnagar districts, where we carried out field visits in 2013, are now located in Telangana. Wherever the article mentions Andhra Pradesh, it is the undivided Andhra Pradesh that is referred to.

The Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS), also referred to as NREGA, being implemented all over India, is one of the world’s largest employment guarantee schemes. Though its implementation is hobbled by rampant corruption in many Indian states, undivided Andhra Pradesh was an exception. The poor were demanding and obtaining work through the scheme, and the targeting of resources to the poor is being effectively implemented in the state. Women, Scheduled Caste groups and Scheduled Tribes, the most oppressed sections of society, freely participate in a programme that aims to alleviate the misery of citizens. Successful programmes such as this highlight the triumph of citizen formation over patron–client politics, aided by a democratic politics.

This article argues that the success of MGNREGS in undivided Andhra Pradesh depended heavily on ideas within the bureaucracy and the capacity of the subnational state, especially in the ability of Chief Minister Y S Rajasekhara Reddy, to insulate a committed rural development bureaucracy from powerful farmers who had a clear interest in thwarting the programme. It highlights the role of state capacity in installing an architecture that checked corruption, and even strategically exploited actors in society to achieve its ends. It argues that elections in a democracy have the potential to elevate citizen concern over particularistic populism driven by ethnic considerations that have characterised large parts of India. This becomes possible when ideational conviction within a bureaucracy is matched with political will. Democratic politics is not all about the capture of the state by vested interests when politicians and technocrats find it rational to help the poor.

The bane of poverty alleviation programmes in India is targeting. Caste and patron–client relations, it has been argued, engender corruption in India.1 But MGNREGS in Andhra Pradesh has, to a considerable degree, been an exception; its performance there surpassed other states in targeting resources to the poor, which is what prompted us to study the MGNREGS in Andhra Pradesh.

Dutta et al (2012) found that the demand for work of rural households in the programme as a proportion of rural households was related to the proportion of the population below the poverty line (BPL) in a particular state. However, the participation rate of households was only weakly correlated with the incidence of poverty across states. Low participation rate in a state with a substantial BPL population is bad news for MGNREGS as an effective instrument of poverty alleviation.

Jha et al (2008) studied Rajasthan and Andhra Pradesh more intensively, by randomly selecting 942 households from three villages in every district in the two states. This study revealed that the poor were being targeted much better in Andhra Pradesh than in Rajasthan.

Likewise, the Comptroller and Auditor General of India credited Andhra Pradesh with having implemented MGNREGS in a manner that has improved the well-being of the poorest, despite challenges to its implementation.

Our choice of Andhra Pradesh was therefore necessitated by studies suggesting that MGNREGS needed to break the patron–client logjam in India’s poverty-stricken rural areas. The figures proved that a large state like undivided Andhra Pradesh, with substantial poverty and a history of feudal land tenurial relations, was enabling marginal farmers and the landless to earn a decent livelihood at the cost of large landholder farmers who now needed to pay higher wages than before as MGNREGS was indirectly boosting agricultural wages. If such successful experiments persisted, India could make significant headway in improving levels of well-being among its poor.

This article is based on our study conducted between June and July 2013. We spent a week in Delhi interacting with members of the National Advisory Council, a week in Hyderabad working with the Department of Rural Development and the social audit office, Society for Social Audit, Accountability and Transparency (SSAAT), and a week in rural Mahbubnagar, Nizamabad and Vizianagaram districts. This followed up on three weeks spent in Hyderabad and parts of rural Mahbubnagar, Anantapur and Kurnool in 2011.

The second section of this article details where Andhra Pradesh’s success in delivering work to the poor fits among theories of the state. We first review that literature and then situate the state in Andhra Pradesh in that context. An autonomous bureaucracy was central to the achievement of delivering services to the poor. But that required the presence of a supportive chief minister who insulated the pro-MGNREGS bureaucracy from its powerful opponents—farmers with large landholdings, and contractors. There is a clear relationship between state autonomy from powerful vested interests and its capacity to deliver wages to the poor.

The third section describes the findings of our village studies. In conclusion, we relate Andhra Pradesh’s success in reaching wages to the poor to its ability to insulate the bureaucracy from powerful vested interests opposed to it. This does not mean that all is well with the state. We reflect on the social and political conditions under which corruption varies in different locales in the state of Andhra Pradesh.

Bringing the State Back In

Comparative political economy suggests that the impact of the state on institutions and policies can be understood in two distinct ways. The first set of arguments suggests that the state is so profoundly influenced by powerful social actors that if we can discern who these actors and their interests are, we will understand why the state behaves in a certain way. The second set of arguments suggests that the state is an autonomous actor with a mind of its own and acts in a manner that is not influenced by the interests of powerful social actors alone. For these sets of arguments to become viable, we need to understand why the state thinks in the way it does. Policy ideas within government create interests within the state that differ from the interests of corporate groups and society. Hence, one cannot comprehend the behaviour of the state without paying close attention to the dominant ideas within the state.

Scholars typically make two kinds of social structural arguments. The first suggests that the state is the handmaiden of the capitalist class, an argument that is as old as Marx and Lenin.2 Subsequently, scholars such as Moore (1966) and, more recently, Acemoglu and Robinson (2012) argued that the bourgeoisie produced democracy. Where the bourgeoisie was weak, the state would not follow a liberal democratic constitution. Przeworski (1991), for example, demonstrated that democracy arrives at a certain high level of per capita income. Capital accumulation must occur before the state adopts a democratic path and the interests of the state are tied to those of the bourgeoisie. The welfare state, according to these scholars, arose because of the need for capital to find social peace. This is the reason why, from a few notables, voting rights were given to more and more citizens, including women and workers. These scholars, therefore, cannot understand why India took a liberal democratic path in 1947, when it was even more poverty-stricken than today.

There is another set of social structural arguments that locate the behaviour of the state within social structure, though not necessarily with the interests of the capitalist class. Hamza Alavi’s (1974) classic conception of the bureaucratic–military oligarchy in Pakistan challenged the view that the postcolonial state was a handmaiden of the capitalist class. Alavi argued that the bureaucratic–military oligarchy in Pakistan enjoyed relative autonomy from the metropolitan and the indigenous bourgeoisie of the country. The state in Pakistan, argued Alavi, was a bureaucratic–military oligarchy that mediated conflict and cooperation between three powerful groups—the landed, domestic capitalists, and the metropolitan bourgeoisie, without being captured by capital. The relative autonomy of the state in this argument does not need to dwell on the dominant ideas within the state.

Bardhan’s (1984) account of the dominant coalition of industrialists, farmers, and professionals explained rather well why import substitution persisted in India during the 1980s. The state in India, like the one described by Alavi, mediated conflicts of interest between these powerful classes, which had an interest in preserving the status quo. Likewise, Kohli’s State and Poverty in India (1987) related India’s land reforms to the social basis of the party in power in three Indian states. In West Bengal, where the landed and the capitalists had been excluded by the Communist Party of India (Marxist) after it won elections in 1977, success in land reforms was the most spectacular. In Karnataka, where the Congress party’s social base included the landed, success was less spectacular, despite the ideological commitment of the Congress party. In Uttar Pradesh, under Janata Party rule—which was largely supported by small traders and the landed—there was almost no land reform. Kohli thus relates the power of the landed within the ruling party with the varied success of land reforms in the three states under consideration.

We argue that the successful implementation of MGNREGS in Andhra Pradesh has more to do with the way the bureaucracy within the state of Andhra Pradesh thought. This falls within the second category of state-centred explanations. A powerful state-centred explanation for policy and institutional change needs to view the state as being even more autonomous from social pressures than mentioned above. Such arguments hold in cases where the state, especially its bureaucracy and technocrats who power and puzzle over the failure of past policies, pursues ideas to improve upon them unimpeded by powerful pressure groups within society. The interests of the bureaucrats cannot be reduced to those of the dominant classes or groups within society.3

This research agenda would benefit from an exploration of the dominant ideas within the state. For example, Hall (1993) argued that ideational changes within the technocracy—given the policy puzzles faced by Keynesian policy frameworks—led to first- and second-order changes in the direction of monetarism in Great Britain on the eve of Margaret Thatcher’s ascendance to the prime ministerial position. Kaviraj (1997) made the case for an ideational passive revolution within the Nehruvian state after the death of his powerful political competitor—Sardar Patel. Development within the Indian state was to be premised on capital-intensive import substitution that stressed economic self-reliance over global economic interdependence. This ideational approach to development, discussed above, cannot be reduced to the interests of the industrial class. Mukherji (2013, 2014) argued that the economic reforms of 1991 took place primarily due to ideational changes that occurred within the state since 1975, and which reached a tipping point during the balance of payments crisis of 1991. The change in the policy paradigm from state-led import substitution to globalisation and deregulation was thus not led by the interests of the industrial class or the interests of the International Monetary Fund and the World Bank at the time of a foreign exchange crisis.

The State in Andhra Pradesh and MGNREGS

Let us now turn to the state in Andhra Pradesh. Prior to the enactment of MGNREGS in 2005, the state had a committed rural development bureaucracy, which seized the opportunity to promote a rural development agenda when the Congress party came to power with Y S Rajasekhara Reddy as chief minister in 2004. The previous government of N Chandrababu Naidu had focused on building up governance systems, industrialisation, and information technology. The bureaucracy pointed out to the new chief minister that growth had been promoted, but rural areas had been neglected. This was also somewhat evident to Rajasekhara Reddy, he having walked 1,500 kilometres of the state on foot before making his electoral bid. Electoral victory signalled to Reddy in 2004 that Andhra Pradesh’s urban governance, industrialisation, and the information technology boom did not necessarily help garner votes. Reddy’s first important decision as chief minister was to provide free electricity to farmers, a step that would hurt power sector governance in the state. The political climate had turned pro-poor at the centre and the state when the Congress party came to power both at the centre and the state in 2004. This was an opportune moment to grant citizen-centric rights rather than patron–client favours based on caste and religion, to garner votes.

Democratic politics was demanding welfare. Why was it that Andhra Pradesh did better than most states in implementing MGNREGS? The rural development bureaucracy in the state, which had been endowed with civil servants such as S R Sankaran and B N Yugandhar, had fought for social justice in a feudal state for many years. This had engendered a team of committed civil servants led by a Dalit, K Raju, the principal secretary charged with rural development. Raju had seen poverty first hand, and had spearheaded the formation of women’s self-help groups in the state. The empowerment of women through self-help groups had many positive consequences for employment in the state. The Department of Rural Development in Andhra Pradesh had nurtured civil servants of remarkable capability and integrity. This was evident, for example, from the manner in which the government selected project development officers.

Reddy’s ascent to the position of chief minister played a vital role in insulating this committed bureaucracy from two powerful lobbies—the farmers’ lobby and the construction lobby.4 Farmers would oppose the right to employment as this would indirectly boost the wages of workers toiling on their farms. The MGNREGS programme engendered class conflict between farmers with substantial farmland who needed labourers, and marginal farmers and landless labourers whose extreme deprivation required the state to guarantee employment and wages. The larger the land owned by the farmer, the greater would be the opposition to MGNREGS. We interviewed one former rural development minister who argued quite forcefully that MGNREGS would stifle the agricultural potential of Andhra Pradesh. An earlier food-for-work programme also had suffered from problems of “capture” by local groups. MGNREGS had, therefore, to be safeguarded both from patron–client caste politics as well as from farmers for its benefits to reach the poor in the state.

The construction lobby was also powerful. Politicians depended on contractors for rents. Some ministers were keen that MGNREGS be associated with road construction, an activity that would inevitably attract contractors.

For MGNREGS to achieve any success, the bureaucracy needed to be insulated from both these powerful opponents of the programme. The rural development bureaucracy, led by K Raju, convinced Chief Minister Reddy that MGNREGS would be electorally advantageous if it were insulated from its opponents. Reddy was thus convinced that the programme needed protection from farmers and contractors. He decided to take a position of staunch support for the programme, despite the political opposition it would face. The state worked hard to devise a way to make the work enticing for marginal farmers and labourers but unremunerative for construction companies. He also called this programme his Ayyappa programme after the Southern Indian deity of Sabarimala in Kerala—it would not be polluted by corruption. Reddy had enough grass-roots experience to foresee the electoral benefits of successfully implementing MGNREGS. The lure of electoral gain was a powerful incentive. Further research reveals the fascinating process of how a political class entrenched in patron–client relations comes to unwillingly accept a break from that process.5

The insulation of the bureaucracy from vested interests helped K Raju and his team to work expeditiously on the design and implementation of the programme. They moved with such speed that much was achieved even before the opponents could realise their intent. There was a core team of six government officers who worked tirelessly on this project. The strategy of implementing the programme began in Andhra Pradesh in April 2005, even before the act was passed in Parliament on 5 September 2005. Implementation began in 2006. Considering the synergies between the governments led by the Congress party both at the centre and in Andhra Pradesh, the bureaucracy was able to administer the act in a manner that would lead to speedy and successful implementation. The implementation of MGNREGS hence began immediately in the state after its enactment.

A number of measures were taken to create an enabling architecture for MGNREGS implementation. First, provisions were made to keep contractors out by defining work in a manner that inhibited their intervention; work was made labour intensive and unremunerative for contractors. Studies were carried out to see how much digging was possible within eight hours under diverse soil and climatic conditions. The daily amount of work was matched with the amount of payment to be made to the worker. Second, Tata Consultancy Services (TCS) provided complimentary transaction software that would help track funds through the state.

Third, while gram panchayats would decide the nature of work, the work itself would be brought by a field assistant at the level of the village. The head of the gram panchayat has excellent incentives to seek rents from clients. The job is not well-paid and the sarpanch exercises authority over everything from issuing birth and death certificates to ration cards needed for accessing subsidised foodgrains. So a field assistant was assigned as the lowest level government functionary who was to be the regulator of the excesses of gram panchayats. It was hoped that the field assistant would help check the problem of upper caste dominance in elected village bodies. In fact, the field assistant had wider powers apart from bringing in the work. The field assistant’s responsibilities included assisting the panchayat secretary in maintaining records, muster rolls, and the materials register; marking out the worksites; providing technical support to labour groups; measuring works; and submitting the muster rolls to the mandal programme development officer (Department of Rural Development 2006).

Fourth, payment was to be made through the post office. This led to a surge in postal savings accounts from 17,000 to eight million in five months. Over time, Axis Bank was given charge of making payments. No financial resources flowed into the hands of panchayats. Fifth, it was decided that every village keep a muster roll that noted the name of the villager, work done by him/her, and related compensation. This would aid the process of social audit, whereby audit officers would go from village to village to check whether the work corresponded with what was on the rolls. These rolls had to be made public to any Indian citizen who wished to access them. We will discuss the architecture of the social audit office as a regulator of MGNREGS corruption in the next section. Much of this governance architecture appears to have been prepared even before the principal secretary approached the chief minister. Since the bureaucracy needed the protection of the chief minister from the powerful political opponents of the programme, it was imperative that they had a sound plan before they approached him.

The bureaucracy in Andhra Pradesh was committed to providing the right to employment. The political conditions were appropriate at the national as well as state level. The bureaucracy was able to fulfil this commitment because the executive protected it from the political opponents of reform. This was not a state driven either by capital or farmers; it was a state that understood welfare as a means of garnering votes with a civil service that could exploit political opportunities for redistribution.

Bringing Society within the State: Creation of SSAAT

How could MGNREGS in Andhra Pradesh be insulated from patron–client relations? Chief Minister Reddy would not support the programme if employment guarantees leaked to contractors and farmers. It was well known that the patron–client driven National Food for Work Programme (NFFWP) had suffered from problems. In fact, the programme did not even maintain a proper register of participants and their work. The Department of Rural Development, and especially the senior civil servants within its fold, had to convince the chief minister that MGNREGS would not suffer from the same flaws as the NFFWP, that it was possible to insulate and regulate funds in a manner that would enable most of it to go to marginal and landless labourers. It was the challenge of insulating the programme from powerful vested interests that led to the creation and consolidation of the SSAAT.

The department invited activists from the Mazdoor Kisan Shakti Sangathan (MKSS) to audit the NFFWP in 2006. MKSS is a non-governmental organisation (NGO) led by former Gandhian civil servant Aruna Roy, who resigned from service to work among the poor in the village of Devdungri in Rajasthan. Roy, along with Nikhil Dey and Shankar Singh, pioneered the idea of transparency and accountability by pointing out that if funds allocated to the poor actually reached them, they would be considerably better off than they were. In the initial stages of their activism, MKSS would obtain information about funds allocated for rural development. They would then go to the concerned village and mobilise people to find out how much was really spent on what was claimed by government. They also became well known for exposing fraud in drought-relief work in Rajasthan, in this manner. It was this mobilisation that led to the right to information (RTI) legislation in Rajasthan and subsequently, the national level legislation in 2005. MKSS activists had also worked closely with government. They were not about to bring about a revolution; they were out to shock government and work with sympathetic civil servants and politicians to bring in legislative change.

Principal Secretary K Raju knew about MKSS’s work having attended their public hearings in Rajasthan. He was therefore aware of social audit as an innovation of MKSS. The aim of getting MKSS to audit the Food for Work Programme in Andhra Pradesh was to trace the roots of patron–client relations and train a cadre of state-level officers. These audits in Andhra Pradesh revealed rampant corruption in NFFWP and catalysed the department to institutionalise the process. A Strategy and Performance Innovation Unit was set up within the Department of Rural Development in 2005, with a small grant of ₹7,00,000.6 A civil servant from the department headed the unit. Its first director, Karuna Vakati, noted the total insulation of the unit from political pressures, thanks to Chief Minister Rajasekhara Reddy.7 An audit of 12 villages in Nalgonda district revealed that 90% of these funds had been consumed by corruption. There existed well-entrenched connections between gram panchayats, local officials, and politicians.

The period March–December 2006 was spent in devising a training methodology for auditors. Initially, NGOs such as Action Aid and MKSS were given this task. Over time, reliance on NGOs for conducting social audits was considered a bane for the following reasons: first, it would be difficult for NGOs to achieve scale. Second, there had to be standard procedures and not differing methodologies. Third, these organisations were sometimes too confrontationist for the government.

Andhra Pradesh had also pioneered the idea of women’s self-help groups, a successful project funded by the World Bank. Could these women become auditors? It was found that these women generally did not need MGNREGS work and therefore did not have a stake in the programme. Finally instead, village-level social auditors were chosen from among the youth in families that benefited from MGNREGS. Personnel were hired for the Strategy and Performance Innovation Unit and were trained on different aspects of social audits. By December 2006, the unit had a 25-member state resource team and 260 district-level resource persons. These district-level resource persons were tasked with training personnel at the village level.

A social audit of Anantapur district, carried out in September 2006, proved to be an important learning exercise. This was a confrontational, albeit successful audit that received substantial media attention. It also generated such local political opposition that no audit could be conducted in that economically backward district for the next one year. The audit alienated both the local bureaucracy and the politicians. The Strategy Planning and Innovation Unit realised that it was more important to work closely and patiently with persons at the local level to make this audit a success. The unit tried to make connections with panchayats to build a long-term support base for the process of social audit. A number of government orders were passed to institutionalise the process. Key officials were to be present and all records were to be made available to the auditors. Steps were also taken to curb any violence that could arise from taking the guilty to task.

In 2008, MGNREGS was scaled up from 13 of the poorest to all 22 districts of Andhra Pradesh. There was an inherent contradiction in the government auditing itself. Could the government be objective about itself? The model of an independent regulator was arrived at inductively by setting up a society within the Department of Rural Development. The society was christened the SSAAT in May 2009. At its helm was a new director, Sowmya Kidambi, a former MKSS activist hired on a contractual basis. The significance of this move lay in the fact that a government institution felt the need to bring in a civil society activist in order to introduce energy, competence, and capacity in its social audit programme.8 The SSAAT was set up like an independent regulator with a portion of MGNREGS funds going directly to it. It was financially independent from the Department of Rural Development.9 The interactions between the principal secretary, Subrahmanyam Reddy, and its director, Sowmya Kidambi, were important for institutionalising the process. These interactions not only reflected the view—that a regulator of this type could be a governance innovation in India—of the Department of Rural Development but also of the leading lights of MKSS like Roy and Dey, who were made part of the SSAAT’s governing board.

Mobilising through Social Audit

Social audits produced revelations but grievance redressal was weak. The sample chosen from SSAAT’s database was quite reliable. We detail briefly in the next section the reasons behind the varying degrees of corruption in MGNREGS in villages we visited in July 2013. In order to correct this situation, Subrahmanyam Reddy set up a vigilance cell within government that could act upon the problems highlighted by SSAAT.

It became evident from our village surveys that SSAAT did excellent work in uncovering levels of corruption. One can independently seek these village locations, as we did, and check the validity of its data. However, we also learnt from SSAAT that though they knew where the problems existed, they could not act against them, especially in the more feudal areas like Mahbubnagar district in Telangana. Nor were people mobilised through the audit process.

There exist two views about the audit process. One is that the process has deviated from the MKSS’s strategy of mobilising people from below. Auditors do an excellent job of fact-finding by going from door to door and matching the muster rolls with the availability of work. But this is a top-down, state effort rather than a public hearing where mobilised people come out in the open and shame local officials. This makes the process appear like one arm of the government—an independent regulator—checking the other arm involved with policy. Those who believe in the process of mobilisation for checking corruption therefore believe this is not the optimal path. They have begun effectively mobilising unions of MGNREGS workers in Lingampet mandal of Nizamabad district.

There is also a powerful defence of the current methodology of a top-down audit process. When society has been trapped in exploitation and oppression for centuries, it is unlikely that the people will rise against the perpetrators of injustice. Under such circumstances, the state has to defend its poor and marginalised citizens from the social power of the oppressors. It is precisely this role that SSAAT is playing.

Village Studies of MGNREGS

To what extent did the architecture of MGNREGS implementation check corruption at the village and mandal levels?10 This investigation throws open variations in the “why” and “how” of success, failure, or that which lies between the two. In-depth examination of the villages unravels the local dynamics of social and political processes, which could throw up important variables contributing to the state of MGNREGS implementation. We decided to visit the villages on the basis of the following parameters:

(i) Mandals with high MGNREGS expenditure and low deviation (HE&LD): High expenditure implies that the mandal has been able to substantially utilise the resources available under MGNREGS, and low deviation means that the money has been well spent, with less leakage and corruption. This implies that these mandals were characterised by low levels of corruption.

(ii) Mandals with low expenditure and high deviation (LE&HD): Low expenditure suggests that the concerned mandal has not been able to utilise the money available under MGNREGS; high deviation denotes high leakage and corruption.11

Following this methodology, we visited two villages each in Peddamandadi and Amrabad mandals in Mahbubnagar district. Peddamandadi was known for its high expenditures and limited corruption (HE&LD) while Amrabad recorded low expenditures and high levels of corruption (LE&HD). Similarly, we visited one HE&LD village in Gantyada mandal of Vizianagaram district. We also visited two villages in Lingampet mandal in Nizamabad district because it presented itself as a “model mandal” in the implementation of MGNREGS (HE&LD). This was an opportunity to examine the variations in social and political processes that may have resulted in successful implementation.

These field visits brought to light various facets, which either aid or hinder implementation of MGNREGS at the village level. Two significant points emerged from these visits.

Local Politics: Cases of Exploitative and Virtuous Nexus

First, we discovered the positive role of bureaucratic autonomy at the village level in insulating the programme from vested interests. There was a link between state autonomy at the macro level and the insulated autonomous bureaucracy at the village level. Where vested interests, such as farmers and contractors, have been able to capture the programme, it is because the local bureaucracy was often not insulated from these forces.

Take, for example, Shetapally and Mengaram, HE&LD villages in Lingampet mandal. These two villages are served by an excellent local bureaucracy, social audit, and MGNREGS unions. A proactive field assistant (FA) and responsive administration at the helm of affairs led to the creation of a virtuous nexus in these villages. An earlier attempt to siphon off funds and influence the field assistant to be a signatory to the approval of work amounting to ₹3.5 million, which never took place, was thwarted collectively by an honest and committed team of district collector, programme officer (PO), assistant programme officer (APO), and field assistant.12 This prevented a corrupt nexus from taking root at an early stage.

In contrast, the collusion of vested interests between the local politician, village head, mandal officer, PO, APO, and the FA led to the emergence of an exploitative nexus in the case of Chitlomkunta village (Amrabad mandal) and Ghotlakhanapur village (Peddamandadi mandal). For example, in Chitlomkunta,13 we discovered that the anomalies in the implementation of MGNREGS in the village were due to the nexus between the local politician, mandal development officer (MDO), village head, and field assistant. The field assistant and the village head were related to each other and colluded for each other’s benefit.14 There were improprieties in the muster rolls, carried out in collusion with the field assistant, village head, and the MDO.15

A similar picture emerges from Gotlakhanapur,16 a poor village in Peddamandadi mandal. Even though the local population participates regularly in MGNREGS work in this village, the people are dissatisfied with irregular payments, unavailability of work, and the irregular distribution of work. This was due to the collusion between the village sarpanch and the field assistant. Interviews revealed that the sarpanch influences the nature of work to be undertaken and the technical measurement for the work. There is a lack of collective decision-making in such matters, which makes MGNREGS implementation lopsided towards the favoured few. Villagers pointed out that the field assistant and the village head were Congress party workers, which also skewed the distribution of work.17

Our village cases show a significant role of an autonomous bureaucracy operating independently of an exploitative local politics. The field visits established the role of the state and an effective bureaucracy, which sought to insulate itself from the vested interests operating locally, in creating what we term a virtuous nexus. On the other hand, the case studies discussed above amply illustrate that wherever the programme is enmeshed in local politics, it has resulted in the creation of an exploitative nexus. The capacity of the bureaucracy to insulate itself from vested interests is critical for the proper implementation of MGNREGS.

Role of Social Audits

The field visits show the importance of an effective implementation architecture, drawing lessons from rent-seeking in a similar programme, Food for Work. As is evident from the cases, despite an exploitative nexus taking root, the effectiveness of the state administration and the implementation architecture of the programme have been able to detect and expose the nexus of corruption, resulting in prompt action by the state administrative machinery to rectify such anomalies.

Four patterns emerge from our field visits regarding the social audit process in Andhra Pradesh. First, the social audit in Andhra Pradesh follows a top-down approach, where the role of the auditors—rather than citizens being mobilised to complain—is more significant. Such a social audit produces reliable statistics on corruption and prods the state to act, even though the process itself is less capable of dealing with corruption. In this regard, we argue that the SSAAT needs to be viewed as an autonomous regulator, which monitors and ensures the accountability of MGNREGS at the village level. This institution serves as an important link in the chain of governance of the programme.

Second, the process of social audit in Andhra Pradesh is different from the conventional methodology of social audits, which are carried out through public hearings elsewhere. This is reflected in the way the social audit was experienced in the villages we visited, where the audit officials gathered information door-to-door. We found that the social audit process had contributed to the success of the programme. Social audit teams inspected individual job cards and muster rolls, which revealed anomalies in the implementation (as we found in the cases of villages discussed above).

Third, the top-down autonomous social audit model enables the bureaucracy to be in touch with the existing micro-realities. The fact that the social audit office in Hyderabad provided credible expenditure deviation figures right up to the mandal level suggests that the state’s administrative machinery has its ear to the ground.

Fourth, the penetration of social audit teams right up to the doorstep of the beneficiary has heightened the sense and understanding of legal entitlement in rural areas. This is not the case in other states where the level of understanding and awareness of entitlements is still very poor.


Andhra Pradesh is a front-runner in the implementation of MGNREGS, despite the problems of patron–client relations discussed above. The state has a history of feudal relations as was evident from politics described in villages where MGNREGS implementation has been far from perfect. Moreover, the literature in Indian and comparative politics generally codes the Congress party as being more pro-upper caste than the Dravidian parties and other lower caste-based parties in India. How then did a Congress government in power implement a programme that clearly benefited marginal farmers and the landless at the cost of richer farmers and construction companies who exploited joblessness and low wages?

The answer to this puzzle lies in the way a proactive rural development bureaucracy convinced the chief minister about the electoral benefits of a citizen-enhancing right to employment, which successfully challenges a social order steeped in clientelism. It was not easy for the politician to give up assured rents from a politically powerful support base. But ideas within the bureaucracy and the relationship between the political executive and the bureaucracy proved critical for the programme’s success. The latter needed the programme to be insulated from vested interests. This insulation could only be provided by the political executive.

The Department of Rural Development in Andhra Pradesh is a case of unusual state autonomy from powerful social groups threatening the MGNREGS. Obviously, Chief Minister Reddy insulated it from other sections of the state that would favour richer farmers and contractors. This kind of state autonomy is unusual considering the votes and financial power that emanate from these groups.

This narrative is therefore very different from narratives that link the state either with capital or with other powerful interest groups. More research is required regarding how bureaucrats powered and puzzled over earlier policy failures and attempted to reach goals despite political constraints. It lends support to the literature on the state that argues that ideas within the state—the bureaucracy and technocrats, rather—are important for explaining a state’s capacity to transform social relations, between marginalised groups in rural Andhra Pradesh and richer farmers and contractors. It was an autonomous rural development bureaucracy firmly committed to the poor and insulated from powerful political pressures that designed the key elements of the MGNREGS architecture in Andhra Pradesh.

So what were the important elements of the successful architecture? First, MGNREGS works were brought to a village government by a field assistant. This created the possibility of dealing effectively with an upper-caste dominated panchayat. We found that in many failed cases, such as in Chitlomkunta and Ghotlakhanapur villages, the field assistant was influenced by local politics that did not favour the poor.

A committed local bureaucracy was essential for the effective implementation of the programme. In Shetapally and Mengaram, our best villages, we found the entire local bureaucracy—field assistant and above—very committed to the poor. These were also villages where MGNREGS unions were most active and biometric measurement of work had been implemented most effectively. And these villages in Lingampet mandal had made the best use of biometric measurement to register both work and payment.

Could it be that a combination of a sound local bureaucracy and MGNREGS unions spurred good governance in Shetapally and Mengaram? Field assistants, after all, can also collude with the gram panchayat and the local elite. Therefore, the integrity of the local bureaucracy and checks on it in the form of social audit and MGNREGS worker unions could lead to significant governance innovation.

Second, funds flowed straight to workers through a post office or a bank. Gram panchayats did not see any cash in hand. This was clearly a governance innovation that helped check corruption with the aid of transaction software provided by TCS.

Third, the social audit office in Andhra Pradesh was a significant regulator of corruption. Social audits helped to bring society within the state. Rather than depend on activist NGOs like Action Aid and MKSS, their contribution in the form of orchestrating public hearings in villages to curb corruption was incorporated within the SSAAT social audit office by inviting a former MKSS activist to direct the office. It also had a governing board comprising eminent social activists. This state–society interaction was significant in unearthing corruption in Andhra Pradesh. We found that social audit is effective in pointing to corruption even though it cannot always solve the problem. But we were impressed that we found the levels of corruption in our villages consistent with the data provided by the social audit office. Moreover, workers were aware and vocal about corruption and at times action was taken in the case of villages like Chitlomkunta. Since social audits are fairly universal in Andhra Pradesh, good governance has now to focus more on how to insulate the local bureaucracy from corruption.

Andhra Pradesh presents a unique case where macro-level processes have been able to fine-tune themselves to the existing diversities at the micro level. If it can be argued that an effective public policy programme should be able to strike a judicious balance between the centralising elements at the macro level (unity) and the existing micro diversities (diversity), Andhra Pradesh presents a success story (Jha 2008). This could be seen in the way the autonomy of the bureaucracy goes all the way down to the village level, insulating it from vested interests. Even in cases where we saw the formation of a corrupt nexus of vested interests, the implementation architecture of the programme enabled the system to detect and rectify the problem. The administration drew lessons from previous welfare initiatives to devise an effective architecture that addressed the gaps and problems from the previous programmes. Social auditing is part of this well-worked-out architecture and presents yet more evidence to establish the argument about successfully balancing the macro and the micro. This has not been the experience of many other states where the operationalisation of MGNREGS in relation to the reconciliation of grass-roots realities has not been effective (Jha 2008).

Regulatory capture still dominates poverty alleviation programmes in many states. It is therefore important to understand how democratic politics can deal with regulatory capture. Our paper suggests that innovative ideational evolution within the technocracy and its synergies with political will can help the state deal with the enemies of the developmental projects favouring growth and industrialisation. Mukherji (2014, 2017) has pointed to such synergies in explaining the successful reforms of 1991 as well.


1 On caste and patron–client relations, see Mukherji (2014: 101–56).

2 See for example, V I Lenin, Imperialism, the Highest Stage of Capitalism, New York: International Publishers, 1939.

3 See for example, Skocpol (1985) for an excellent review of this literature.

4 Interview with Subrahmanyam Reddy, Principal Secretary, Rural Development, Government of Andhra Pradesh. Personal interview. Hyderabad, 3 June 2013. An important government order that corroborates this point is: Government of Andhra Pradesh, Administrative Reforms: Selection of Suitable Government Officers for Posting as Project Officers (ITDA), Project Directors, DRDA/DWMA/MEPMA, Vidya Mission (Sarva Shiksha Abhiyan) by the Screening Committee, Government Order Ms No 143, Hyderabad. This point has also been corroborated in a number of interviews with civil servants in Andhra Pradesh.

5 We are deeply indebted to K Raju for these insights. They were shared in a number of meetings at the National Advisory Council in New Delhi in May 2013. This was also corroborated in meetings with civil servants dealing with rural development, MGNREGS, and SSAAT.

6 It subsequently received funds from the Department for International Development (DFID), UK, in 2006.

7 Interview with Karuna Vakati, Director, MGNREGS. Personal interview. Hyderabad, 2 June 2013.

8 The joint director was a government official, but state team monitors, and the state, district, and village resource persons were all contract staff.

9 On the evolution of SSAAT, we depend on interviews with K Raju in Delhi in May 2013, and with Karuna Vakati and Sowmya Kidambi in Hyderabad in the first week of June. See also Aiyar et al (2011).

10 Andhra Pradesh is divided into multi-tiered hierarchical administrative units at the district, mandal and village level. Undivided Andhra Pradesh was divided into 23 districts. Each district was divided into approximately 3–4 subdivisions, and the state had a total of 79 revenue subdivisions. Each district had between 30 and 60 mandals. Each mandal consisted of approximately 15–20 villages that comprised about 40–50 habitations. There were a total of 1,125 mandals in the state, and around 25,000 villages.

11 Expenditure and deviation figures were obtained from the SSAAT office, which was on the basis of social audits carried out in different districts in 2012.

12 An early incident to influence the field assistant was shared with us. The programme director and the local politician used physical threats to get him to sign a document, which showed a payment of ₹3.5 million without any work being undertaken under MGNREGS. He was locked in a room and was pressurised to sign. He then called the APO, who informed the district collector. The locked-up field assistant was advised to sign the document as he was under physical threat, but the collector later revoked the signed document.

13 Visit to Amrabad mandal, Mahbubnagar, 7 July 2013.

14 The former field assistant is the older brother of the sarpanch.

15 For instance, five weeks were shown as 10 weeks and 40 groups were bumped up to 60 groups. The differential amount was then pocketed by the nexus.

16 Visit to Gotlakhanapur village, Mahbubnagar, 8 July 2013.

17 The field assistant had earlier been suspended in 2008 due to some fraudulence in the accounts and muster rolls, but was later reinstated through the influence of a local politician.


Acemoglu, Daron and James A Robinson (2012): Why Nations Fail: The Origins of Power, Prosperity and Poverty, New York: Crown Business.

Aiyar, Yamini et al (2011): “A Guide to Conducting Social Audits: Learning from the Experience of Andhra Pradesh,” Accountable Government: Policy Research Series, Centre for Policy Research, New Delhi,

Alavi, Hamza (1974): “The State in Post-Colonial Societies: Pakistan and Bangladesh,” New Left Review, Vol 74, pp 59–81.

Bardhan, Pranab (1984): The Political Economy of Development in India, New York: Basil Blackwell.

Department of Rural Development (2006): Andhra Pradesh Rural Employment Guarantee Scheme: Operation Manual 2006, Hyderabad: Government of Andhra Pradesh.

Dutta, Puja et al (2012): “Does India’s Employment Guarantee Scheme Guarantee Employment?,” Economic & Political Weekly, Vol 47, No 16, pp 55–64.

Hall, Peter A (1993): “Policy Paradigms, Social Learning, and the State: The Case of Economic Policymaking in Britain,” Comparative Politics, Vol 25, pp 275–76.

Jha, Himanshu (2008): “Policy as Instrument for Reconciling Diversity,” Policy Issues in Federalism International Perspectives, Vol 5, John Kincaid and Rupak Chattopadhyay (eds), Forum of Federations Canada/Delhi: Viva Books.

Jha, Raghbendra, Raghav Gaiha and Shylashri Shankar (2008): “National Rural Employment Guarantee Programme in Andhra Pradesh: Some Recent Evidence,” Departmental Working Papers 2008–05, Canberra: Arndt-Corden Department of Economics, The Australian National University.

Kaviraj, Sudipta (1997): “A Critique of the Passive Revolution,” State and Politics in India, Partha Chatterjee (ed), New Delhi: Oxford University Press, pp 45–88.

Kohli, Atul (1987): The State and Poverty in India: The Politics of Reform, Cambridge: Cambridge University Press.

Moore Jr, Barrington (1966): Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World, Boston: Beacon Press.

Mukherji, Rahul (2013): “Ideas, Interests, and the Tipping Point: Economic Change in India,” Review of International Political Economy, Vol 20, No 2, pp 363–89.

— (2014): Globalization of Deregulation: Ideas, Interests, and Institutional Change in India, New Delhi: Oxford University Press.

— (2017): “Governance Reform in a Weak State: Thirty Years on Indian Experience,” Governance, Vol 30, No 1, pp 53–58.

Przeworski, Adam (1991): Democracy and the Market: Political and Economic Reforms in Europe and Latin America, Cambridge: Cambridge University Press.

Skocpol, Theda (1985): “Introduction: Bringing the State Back In—Strategies of Analysis in Current Research,” Bringing the State Back In, Peter B Evans, Dietrich Rueschemeyer and Theda Skocpol (eds), Cambridge: Cambridge University Press, pp 3–38.

Updated On : 8th Dec, 2017


(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top