MGNREGS social audit lessons from AP
At the heart of the Congress party’s narrative on the rights-based welfare state is the idea that rights laws, to quote Sonia Gandhi, “put pressure on the executive to be more responsive and accountable” and in doing so bring about an “empowerment revolution”. To enable this revolution, rights laws have had built into them procedural requirements for greater transparency and accountability. Prominent among them is the mandatory provision to conduct social audits on the implementation of the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS).
Despite the mandatory provision, Andhra Pradesh (AP) is the only state government that has conducted social audits regularly and at scale. Since 2006, the state has conducted over six rounds of audits in 21,000 gram panchayats (village councils). How effective have these audits been in unleashing the empowerment revolution and ensuring a responsive and accountable government?
To answer this question, my colleague Soumya Kapoor Mehta and I studied citizen experiences with participating in social audits through in-depth interviews with more than 200 MGNREGS beneficiaries in eight villages of Medak district in 2012. Here is what we found:
In these seven years, social audits have emerged as a credible mechanism through which citizens interact with government and voice complaints. Most of our interviewees had heard of and participated in the audit. Crucially, 89% of those who had complaints voiced them in different stages of the audit.
The audits perform two other critical functions relevant to the rhetoric of the empowerment revolution. First, they help raise awareness about entitlements. One of the earliest studies on the social audit in AP surveyed 800 MGNREGS beneficiaries to find a significant change in awareness levels after the audit. Thirty-one percent respondents were aware that the law guaranteed 100 days of work per household before the social audit. This increased to 88% after the audit.
Second, the audits serve to shift, if only temporarily, the power dynamic between beneficiaries and officials. In the normal course, beneficiaries interact with officials as petitioners requesting and receiving work and wages. During the social audit, beneficiaries can demand action and, in part because of the public nature of the event, the government has to admit to its mistakes and justify its actions. A field assistant’s (frontline work site manager’s) description of the audit best articulates this shift: “The villagers complained that…wages (were) not paid completely. The MPDO (Mandal Parishad Development Officer, a senior administrator)…ordered an enquiry. After verification…complete wages were paid.” And this temporary shift seems to have influenced beneficiary perception of MGNREGS officials: 85% of our interviewees said that the audit had increased their confidence to seek information from officials about MGNREGS.
But has increased awareness and greater confidence put pressure on the government to be more responsive and accountable? Early findings present a complex picture. For one, the audits are yet to result in effective redressal. In our study, of the 125 beneficiaries who had complained during the audit, 112 said their complaints had not been resolved. That redressal is slow is also evidenced by data from the AP social audit society. In six rounds of audits, Rs.141 crore worth of fraud had been identified, of which Rs.23 crore had been recovered. In addition, punishments have been imposed on 27% officials. While this kind of recovery is unprecedented, the scale of the redressal is a cause for pause, especially so because the government has put in place an institutional system to ensure redressal. Our analysis suggests that the problem stems from administrative failure. Departments don’t talk to each other, making it difficult for the rural development department to hold key officers responsible for MGNREGS, like the postmaster, who makes payments but reports to the postal department, accountable.
In the absence of sanctions, the number of complaints has not decreased. In our study villages, six years later, the primary concern voiced by interviewees was still related to wage payments. An empirical study of three rounds of social audit reports from 300 panchayats by economists Farzana Afridi and Vegard Iversen (forthcoming, India Policy Forum) found a similar trend.
Finally, has the social audit influenced the corruption market? Afridi and Iversen’s work highlights an intriguing shift in the corruption dynamic away from wage theft towards more sophisticated forms of fraud. Their analysis finds that non-wage related irregularities reported in social audits increased by a whopping 173% between 2007 and 2010. Audit reports in our study villages reflect a similar trend. It is likely, although there is no empirical proof, that the social audit by increasing the transaction costs of wage-related corruption has pushed the corruption market to adapt to newer ways of making money.
But making money from non-wage related fraud is complicated. To fake a worksite for instance, the entire local machinery from the mandal officer to the field assistant need to work together through well-oiled networks that are necessarily political. Tackling this nexus is a long-term and difficult proposition. And while audits may be critical in making this fraud transparent, significant reforms will be needed to strengthen the sanctioning capacity of the government.
AP’s experience with social audits holds important lessons for the Congress’s empowerment agenda. Clearly, mechanisms like social audits are a powerful tool. But for social audits to result in a responsive and accountable government, they need to be supported by a political commitment to reform the administrative system and challenging political arrangements that support corruption at the local level. In the absence of this, social audits will remain a powerful idea that simply hasn’t gone far enough.
Yamini Aiyar is a senior research fellow and director of the Accountability Initiative of the Centre for Policy Research.