work, but he seemed to be arguing more on principle than based on any experience.
Indeed, it is impossible not to be taken in by NREGA’s sheer scope — measured by the number of potential participants, it has to be largest transfer programme in the world — or its vision of an anti-poverty programme that avoids making the poor rely on the munificence of some functionary — health worker, teacher, ration-shop owner — to get what’s owed to them.
The challenges are, of course, enormous as well, and much has been written about them; the implementation is patchy at best, with many of the poorest districts generating fewer jobs (by a factor of ten or more), than their richer counterparts. The processes, carefully designed by the framers to check abuse, are often subverted or ignored: most families who want them have job cards, but not necessarily in their possession; those who apply for work are rarely issued receipts; and social audits often turn into audits by some committee appointed by the sarpanch, even in Rajasthan, where things are supposed to be going well.
The worry, of course, is that a lot of the money is lining the pockets of the local elites. A survey of about 1,400 households who were reported to have participated in NREGA in Gajapati, Koraput, and Rayagada districts in Orissa in 2007-2008 by Paul Niehaus and Sandip Sukhtankar, who were then Ph.D. students at Harvard University, found that households claimed to have worked a third or less of the workdays that were reported on their behalf, and only 60 per cent say that they did any work.
Some of the sheer egregiousness of this no doubt reflects the fact that these were the early days of the programme. Orissa is also one of the states where implementation is known to be weak. My sense is that in some respects things have gotten better over time and will continue to improve. For one, in many states the money now goes directly into the beneficiary’s accounts, which makes it harder to pay ghost workers. People are also increasingly aware of their rights under NREGA and this might make it easier for them to fight back. I would also expect that the next round of panchayat polls will see a lot more emphasis on delivering the NREGA goodies — after all, for many villagers, this is the biggest thing that the panchayat controls.
My guess is that over time we will see a shift in emphasis towards a rather different set of problems: problems less of elite capture than of popular capture. Take the work requirement, for example — why would a village leader, who wants to be popular, want to insist that people need to work the whole day in order to be paid? Would it not be better for all concerned for the village to extract all it possibly can from the government and then share it out, so that everyone is happy? Maybe those who work more will get paid more, but why not make sure that everyone gets something? Neither social audits nor political competition will stop this, because no villager has a reason to complain.
There are two reasons to worry about this: one, the programme will then start costing serious money — Rs. 1,00,000 crore or more — and it is not clear that the government has budgeted for it. Equally importantly, the fact that people can get paid minimum wages for doing, say, half a day’s work, effectively doubles the minimum wage facing the private sector, and this might start to have negative productivity effects.
The answer to both these problems is better enforcement of the work requirement. What is not clear is who is going to do it? Niehaus and Sukhtankar found even greater over-reporting of work when work was paid by piece rates, suggesting that the government officials in charge of the programme (who must have noticed that the claimed amount of work was many times what was actually done), were not too bothered.
A possible solution would be to pay more attention to the demand side of the labour market. One idea, which in India goes by the improbable name of convergence, is that the NREGA work days would be assigned to, say, the irrigation or surface transport or agriculture ministries, so that in order to complete their projects, they would need to use NREGA labour. This would create a stake in getting work out of people. There are, of course, problems with this: what do you do if people don’t put much effort into their work, since they cannot be fired? How do you estimate how much NREGA labour will show up on a given day? How do you deal with contractors? And so on. On the other hand, this makes it more likely that some really high quality rural infrastructure will get built, which has to be worth something.
Abhijit Banerjee is Director, Poverty Action Lab, Massachusetts Institute of Technology
The views expressed by the author are personal