Keeping NREG Clean and Corruption Free

The National Rural Employment Guarantee Program is the largest and most ambitious poverty alleviation scheme to be launched in India (perhaps the world) in recent memory which is why recent reports that the program is being implemented effectively with relatively low levels of corruption are such good news. This news is all the more remarkable considering the poor start the program got off to. Only a year ago, one of the chief architects of the program, renowned economist Jean Dreze, lamented that implementation in the state of Jharkand was doing nothing other than enrich a few bureaucrats and contractors. One observer claimed that the design of the program was so flawed that dropping money from helicopters would be more effective in delivering money to the poor.

The secret to the success of the program, where it has been successful, has been the use of social audits to keep bureaucrats and others implementing the program in line. According to the original NREG bill passed by parliament, absolutely anyone can request access to the “muster roll” (a list of who has worked when and for how long) maintained at a work site. NGOs have used this clause to keep a close eye on the functioning of the program and to pinpoint and expose cases of corruption.

While the success of social audits at the local level appears remarkable, I would also like to see more aggregated data for the block, district, and state level. Local social audits are great for shaming corrupt local bureaucrats involved in skimming from the program, but it is also important to instigate action at higher levels. A great example of a report which seeks to compare performance of different states in service delivery is Pratham’s Annual State of Education Report (ASER) The report has succeeded in shaming numerous state governments into providing better education to their citizens. I think that if an organization could step up and do the same thing for NREG implementation that Pratham has done for education, the results would be enormous.