While international development practitioners debate and discuss the best tool for people’s monitoring, the Indian government takes a page out of the book of the Right to Livelihood and Right to Food movements and of the Mazdoor Kisan Shakti Sangathan (MKSS) and institutionalizes social audits by mandating them in the National Rural Employment Guarantee Scheme (NREGS). The onus is now on the state to ensure that its own performance is monitored and evaluated by the people. The implementation of these bi-annual audits has been patchy; states differ in their strategy of “rolling them out”. While Andhra Pradesh has established a large state machinery that facilitates social audits, Orissa has decided to appoint the National Institute of Rural Development (NIRD) as the coordinator. In Rajasthan, the seminal work of MKSS has made people’s monitoring and social audits into something of a movement. Other states have left it to NGOs. Still other states have taken minimalist steps. At some level, the relationship between social audits and the state reflects the relationship between grassroots NGOs and the government. State governments have thus variously circumvented NGOs, replaced them, co-opted them or ignored them in the conduct of social audits. Another reason for the patchy take-off is that social audits mean different things to different people. There are still states that have not been exposed to the techniques. Concomitantly, there is a dearth of NGOs that can conduct effective social audits and bring pressure on the state. Notwithstanding this, thousands of grassroots NGOs have seen the value of social audits as a tool to demand accountability – not just in NREGS, but in other programs as well. The fact that the state mandates these audits strengthens the hands of grassroots organizations. In Orissa, the state that has come under so much recent fire for poor implementation of NREGS, NGOs have set up what is perhaps the first electronic listserve of grassroots organizations working on NREGS and of those monitoring the program. This Google group serves not only as a networking and solidarity forum but also as a virtual update from the remotest corners of the state on the status of NREGS. How can social audits and other mechanisms of people’s monitoring be truly imbibed by the state? In Andhra Pradesh an independent evaluation of the impact of social audits (undertaken at the behest of the government) shows that social audits enormously increase awareness about and information on the program. This result holds across caste groups and gender. But participation in audits has no impact on participation in the program. Here lies the real crux of people’s monitoring. How can results of such monitoring or of evaluations be integrated into better design and implementation? India has taken the first step of mandating social audits within a governmental system. Through different ways, social audits are actually conducted and ordinary citizens put their lives on the line to challenge power structures, stand up - often in the face of intimidation - and speak up. That is step one - and a huge step forward in empowerment and confidence building. Then what? What happens once these audits are conducted and the people and government find that there are irregularities and malfunctioning? How do the results of the social audit feed into the design and implementation of NREGS in this case? That is step two. Both Andhra Pradesh and Orissa (and probably other states) are keen to move from step one to step two. And the answer does not lie just in “training” or “sensitizing” local level functionaries, in the hope that they will take social audits seriously. The answer lies in the strong message from above that social audits are as, if not more, important feedback mechanisms than administrative reviews. And that administrative reviews will be conducted on what happened to the results of the last social audit. Then comes step three. Perhaps the reason why biannual audits are mandated in NREGS is that the state needs to “report back” to the people who participated in the first audit on action that was taken on their grievances. So step one empowers and increases awareness by providing information and a forum for airing grievances. Step two is to take the results of these audits and act on them at the local level, with strong monitoring by higher levels. Step three is to report back to the communities. If the state does not take all these three steps, social audits run the risk of generating frustration, anger and hopelessness and in the process doing more harm. And if the state uses social audits as just another item on the review agenda, it runs the risk of making this a perfunctory exercise. People often talk of the “demand” and “supply” side of accountability – people’s movements are considered “demand side” and state response is considered “supply side”. If we think of social audits or indeed any people’s monitoring systems as a seamless loop between people and the government, then the distinction between demand and supply sides will show up to be artificial. The issue of what happens to the results of independent monitoring and evaluation is not peculiar to social audits. The same holds true of studies that empirically evaluate impact through surveys and randomized experiments. What really happens to them? Does it matter who commissions them? If the agency (whether government or NGO) that implements also commissions the study, can it be truly independent? Are statistically non-significant results socially significant? The importance of these exercises lies certainly in the learning they can bring to other groups doing similar work--and good publicity for the organization. But to what extent does it change the way the program functions? Is that not where we should be putting greater attention?