To decentralise or not to decentralise? The dilemma of MNREGA in Andhra Pradesh

  • Blog Post Date 14 October, 2014
  • Articles
  • Print Page

Employment generated under MNREGA has been on the decline in recent years across India. This column analyses whether political dynamics influence the implementation of MNREGA by contrasting the implementation models of two of the top performing states – Andhra Pradesh and Rajasthan. It concludes that a combination of the top-down, supply-driven approach of AP and the demand-driven Rajasthan model is required.

The Mahatma Gandhi National Rural Employment Guarantee Act (MNREGA) is the world’s largest workfare programme. Each year about a third of India’s rural population finds employment through MNREGA. A review of the MNREGA literature by Dilip Mookherjee shows that its contribution in reducing poverty has been substantial.

MNREGA confers on every rural household the right to demand and obtain employment for up to 100 days per year and to be paid a statutory minimum wage1. In other words, MNREGA is an example of “post-clientelistic” policy (Manor 2010), which implies that obtaining employment should not depend on political connections. Obviously, this does not mean that the programme is isolated from politics.

Official data shows that the average number of days of employment obtained by households under MNREGA declined from 53.99 in 2008-09 to 45.86 in 2013-14. The decline has been even sharper in the five top performing states with regard to the programme (Andhra Pradesh (AP)2, Chhattisgarh, Himachal Pradesh, Rajasthan and Tamil Nadu)3 - from 59.60 in 2008-09 to 48.24 in 2013-14, on average. However, despite the employment contraction, these states continue to be top performers in the implementation of MNREGA.

Political dynamics in MNREGA implementation

In a recent paper, I analyse whether political dynamics influence the implementation of MNREGA in AP - one of the top performing states (Maiorano 2014). In the paper, on the basis of extensive fieldwork in rural AP and in-depth interviews with MNREGA implementers and beneficiaries conducted between October 2012 and October 2013, I argue that the good performance of AP is the result, on the one hand, of the strong political commitment of the state government and, on the other hand, of a set of convergences, compromises, and clashes between the administration and the state politicians.

From a policy perspective, it is particularly relevant to investigate further one of the convergences that mark the relationship between the Rural Development Department of AP and the state politicians. This is the decision to exclude gram panchayats (local councils) from playing any significant role in the implementation of the scheme, despite the fact that the Act mandates that at least 50% of the funds should be spent through them. In other words the scheme is run in a rigid top-down manner in AP.

This makes AP’s implementation model rather different from that of all other Indian states where MNREGA is implemented through gram panchayats. An important implication of this difference is that in AP, MNREGA is supply-driven, whereas in the rest of India it is (or, as we shall see below, it is supposed to be) demand-driven, since employment is allocated on the basis of the demands of the beneficiaries to their local councils. In the wake of the contraction of MNREGA employment generation witnessed in recent years across India, it is may be instructive to compare AP’s model with that adopted in all other states.

It is not clear what has caused the decline in MNREGA employment generation across India. My interviews in Karnataka, Rajasthan, Uttar Pradesh and Delhi reveal how a common narrative is emerging among bureaucrats. According to this narrative, given that the scheme is demand-driven, the decline in employment generation reflects a declining interest of the beneficiaries in participating in the scheme. As far as the supply side is concerned, there is little that can be done. This narrative is absent in AP. On the contrary, the Rural Development Department of AP has obsessively tried to improve the supply side. This is a reflection of the fact that the scheme in AP is completely supply driven; demands for employment were not even registered until very recently. Both approaches proved to be wrong. A comparison between AP and Rajasthan is instructive to understand what went wrong and what is possible to do.

Demand vs. supply side of MNREGA employment

Indeed, there is some truth in the argument that the problem lies in the demand side. During my fieldwork in Rajasthan I came across numerous beneficiaries who explained to me how working under MNREGA is just not worthy: actual wages are much below the statutory minimum wage (in some cases wage seekers receive as low as Rs. 40 per day) and they are paid weeks or even months after the prescribed 15 days.

However, this is just part of the story. A survey that I conducted in late 2013-early 2014 among 1,200 MNREGA beneficiaries in AP and Rajasthan reveals that an overwhelming majority of the beneficiaries would like to work more under MNREGA. This applies also to those who completed the full 100 days of work. In other words, the demand for MNREGA employment is far from declining. If wage seekers do not demand work, it is because the supply side is faulty.

This is confirmed by the survey results. According to a large majority of the 1,200 wage seekers interviewed, the reason why they did not get as much employment as desired is that “work was not provided”. This (and a lot of additional qualitative evidence) suggests that MNREGA is not run as a demand-driven programme; rather, the way it is implemented in most states of India makes it a supply-driven scheme. Work may be demanded, but wage seekers get employment only when (and if) it is provided. Therefore, it seems that it is in the supply-side that the problem lies.

In Rajasthan (as in most other states) the work-supplier is the sarpanch. In practice, he/ she enjoys almost complete discretion in the allocation of work. Therefore, an important question is “why should sarpanches provide employment?” Or, in other words, what incentives do local implementers have to provide work?

Sarpanches have two main reasons to provide employment through MNREGA. The first one is that this will make them popular and therefore their chances of re-winning the elections should increase. This is certainly true, but there is a strong counter-incentive in this regard. Throughout India, big farmers have been complaining that MNREGA led to an increase in the wage seekers’ bargaining power, which in turn caused an increase in rural wages4. Therefore, there are strong pressures on the sarpanches from farmers to provide as little MNREGA employment as possible, especially during the agricultural season. Sarpanches must thus find a compromise between accommodating (powerless) wage seekers and appeasing (powerful) farmers. As a result, the political incentive to run MNREGA well is severely weakened.

The second reason why a sarpanch is incentivised to provide employment is because he/ she can extract something out of it. However, the introduction of bank/ postal accounts in 2008 for paying the beneficiaries’ wages made stealing from the programme a very hard endeavour5 (Jenkins and Manor, forthcoming). Although definitive evidence is not available, it is probably not a coincidence that the decline in employment generation coincided with the introduction of bank accounts. As a result, sarpanches (and administrators at higher levels) in many parts of India do not bother much about MNREGA as they think that it is a lot of work for little reward.

In AP, the situation is different. Here the key actor at the local level is the Field Assistant (FA), a contract employee of the state government. According to a senior official of the Rural Development Department, “if the FA doesn’t want to give jobs, he doesn’t”. Similar to the sarpanch, the FA enjoys great discretion in allocating MNREGA jobs.

FAs in AP have different incentives to generate employment. Of course, given that they are not elected officials, they have no political incentives. Moreover, AP is the only state where social audits have been institutionalised and are conducted regularly, making stealing from the programme even more difficult than in other parts of India. However, FAs have a strong administrative incentive to run the programme well: if they want to keep their jobs, they have to generate at least 15,000 person-days per year6. This obviously further diminishes the demand-driven component of the scheme. However, this incentive mechanism rightly assumes that the demand for employment is somewhat ‘structural’ to rural India’s political economy, given the severe shortage of jobs in rural areas, particularly during the agricultural lean season.

This is just one of the many attempts by the Rural Development Department of AP to improve the supply side at all levels. The results have been fairly good (AP is one of the top performers in terms of employment generation). However, this does not mean that the right to work is ensured in the state. Wage seekers obtain employment at the bequest of the FA. Also, one of the objectives of MNREGA was to foster poor people’s ability to interact with the public sphere and to obtain work on demand – a potential source of empowerment. The top-down model adopted by AP clearly does not serve this extremely important purpose. In short, improving the supply side is a necessary but not sufficient condition to ensure the right to work and to unleash MNREGA’s empowering potential.


To sum up, two key lessons can be learned from this comparison. The first one is that the work-supplier must be given strong incentives to generate as much employment as possible. Such incentives are virtually absent in most Indian states, and the sarpanches political incentive is clearly not strong enough. The solution adopted in AP is giving good results, but it is far from enough to ensure that the demand for employment is fully met.

The demand side, and this is the second lesson, must be strengthened too. The Rural Development Department of AP has recently introduced a policy measure that might be worth replicating elsewhere. This is the unionisation of MNREGA workers. All wage seekers in a village are now organised as a federation that deals with the village authorities as a group. This is strengthening the voice of the wage seekers’ enormously. Moreover, the FAs are now accountable to the federations: each month the federation’s representatives must sign a form that will release or block the FA salary. This constitutes another very strong incentive for the FA to meet the wage seekers’ demand for employment.

The results are encouraging. The villages where the federations were introduced on a pilot basis witnessed a steep increase in employment generation, in contrast with the state (and national) trend. Moreover, MNREGA in these villages is now completely demand-driven: wage seekers demand work through their federation’s representatives, obtain written, dated receipts and get employment within the prescribed 15 days. In other words, their right to work is ensured.


  1. The minimum wage ranges from Rs. 120 to 200 per day, depending on the state.
  2. This column refers to undivided AP.
  3. These are the best performers in employment generation among India’s largest states.
  4. It is not clear whether MNREGA is responsible (and to what extent) for the increase of rural wages in recent years. However, that MNREGA increased poor people’s bargaining power is certainly true (Roy, forthcoming).
  5. This does not mean that corruption is absent from MNREGA – far from it. However, many sarpanches and administrators think that it is easier to steal from some other sources (interviews in AP and Rajasthan, October 2013-February 2014). Yamini Aiyar (2013) argues that, in order to steal from MNREGA large corruption networks need to be formed, which is something not so easy to do.
  6. This is a state-level policy.

Further Reading

  • Aiyar, Y (2013), ‘Rights, Particpation and Accountability: A case study of Social Audits in Andhra Pradesh’, Paper presented at the Second Azim Premji University international Conference on Law, Governance and Development, 2nd August 2013, Bangalore.
  • Jenkins, R and J Manor (forthcoming), Politics and the Right to Work: India’s Mahahtma Gandhi National Rural Employment Guarantee Act, Hurst&Co, London.
  • Maiorano, D (2014), “The Politics of the Mahatma Gandhi National Rural Employment Guarantee Act in Andhra Pradesh”, World Development, 58, 95–105.
  • Manor, J (2010), “What do they know of India who only India know? The uses of comparative politics”, Commonwealth & Comparative Politics, 48(4), 505–516.
  • Roy, I (forthcoming), “Reserve Labour, Unreserved Politics: Dignified Encroachments Under India’s National Rural Employment Guarantee Act”, Journal of Peasant Studies.
No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter