One does not have to agree with his views to be intrigued by the possibilities opened up by Prime Minister Narendra Modi’s emergence as communicator/harangue-master in chief. Public conversations about who we are and who we want to be are key to the vitality of our democracy, and leaders can seed those conversations when they speak out their own views. When I hear people in the Delhi metro or waiting in line to buy sweets, discussing Modi’s latest Tweet, I do feel that some part of our democracy is coming alive.
That being said, I am disappointed that Modi has opted to play it so safe. Who would disagree with the fact that we need to be cleaner or that our bureaucrats need to work harder — whether we can get there is a separate question. But where we really need conversations is where there are disagreements that demand mediation.
Take the MGNREGA for example. The idea that it would be scaled back was announced with very little fanfare and essentially no explanation. As I can tell, Modi did not even Tweet about it, let alone use one of his speeches to explain the logic behind the shift. Yet there are a lot of people who are beneficiaries of this and will miss it when it is gone and no doubt many who are happy as well (mostly farmers who are having trouble finding people to do their work). Both sides (and indeed the entire nation) deserve to be told why this is the one UPA programme that has been singled out for the axe, if for no other reason than to inform our views about the design of future programmes. Modi is a genuinely compelling speaker and would have been the ideal person to do it — as it is I will try to do my (inadequate) best.
Let me put my cards on the table. I have never been a huge fan of the MGNREGA (I will come to why in a bit). But it has done something for the rural poor, which is not something one could say with any confidence about most other government programmes. In 2013-14, 2.2 billion person days of work were generated by the programme, covering around 50 million households and over 200 million people, which makes it one of the largest welfare programmes in the world. And contrary to what is sometimes suggested, these days, most of this money does not end up in the wrong pocket: Clement Imbert from Oxford University compared the number of days of MGNREGA employment reported in the National Sample Survey in 2011, with the corresponding report in the MGNREGA database, and found a gap of 20% — implying that 20% of the reported days of work did not happen. But 80% did, and this is a huge improvement from the 50% in 2007-08 that Clement had found in his previous work. The various reforms of the MGNREGA, including the fact that in many states the money goes directly to the beneficiaries’ bank account, has helped a lot.
It is also not true that the work is entirely unproductive. Both in Bihar and Rajasthan we made an effort to track down the projects that were supposed to have been constructed using MGNREGA labour; somewhat to our surprise, we found almost all of them.
One index of the overall impact of the MGNREGA is migration to the urban sector. Clement and his co-author John Papp use the fact that the MGNREGA was phased in to show that those villages that had early access to the programme have less temporary out-migration. Reassuringly this is only true after the MGNREGA begins and not before. People who had access to this did not need to go to the city to find work. Whether or not you think that this is a good thing (after all there many who would say our migration rate is too low), this is clearly evidence that the programme is doing something.
Where the programme fails is in being flexible and responsive. This was the one great potential advantage of a programme like the MGNREGA — since it relies on self-targeting rather than bureaucratic verification of entitlements, if I suddenly discover that I really need some extra cash, in principle all I would need is to show up and ask for work. In practice however there are many reasons why it does not work this way: To start with, almost all the work is in the dry season. And I can get work only where I live — so if I am away in the city, tough luck.
Moreover, the panchayat I live in might not have worksites this year because our panchayat leadership has decided it is not worth their effort. The degree of political pressure to make MGNREGA jobs available varies massively from state to state — which is why access to MGNREGA jobs is worse in a very poor state like Bihar than in a richer state like Andhra Pradesh. The fact that the system is now much more vigilant in terms of preventing fraud makes this problem worse — there is more and more paperwork and less and less money to be made, which might explain why the enthusiasm for the MGNREGA among the panchayat leadership has been waning in many states and correspondingly, NREGA spending has been going down.
Even if my panchayat wants to create jobs, it may be shackled by the fact that the state may not have access to MGNREGA money. This recently happened in Bihar because Bihar as a whole had not met the reporting requirements of the programme. Money simply stopped and so did job creation.
And even when jobs are created, the earnings from them take a while to come. In a state like AP, where the MGNREGA works exceptionally well, this is a matter of several weeks. Elsewhere it usually takes several months.
In terms of targeting the needy when they need it, where they need it, the MGNREGA therefore does not do particularly well. Many poor people get left out because they live in the wrong place or woke up to their need at the wrong time. This is why Rinku Murgai, Martin Ravallion and Dominique Van de Walle from the World Bank find that a universal cash transfer with the same budget would reach the needy more effectively than the MGNREGA. This is a (hypothetical) programme that, being universal, makes absolutely no attempt to target the poor!
This is why the possible scale back of the MGNREGA should be used to start a debate about what could come in its place. A universal cash transfer in the form of a minimum guaranteed income would mean that automatically everyone has something to fall back on without having to deal with the vagaries of their local panchayat. Being universal, it would require no one’s approval, no signature from anyone, no BPL card. It is true that some of the money would go to those who don’t need it, but we would be rid of the entire apparatus of verification — panchayat representatives I meet complain about the enormous burden that the MGNREGA imposes on their (very limited) bureaucratic capacity — projects have to be proposed, budgets have to be made, muster rolls have to be uploaded and so on. Now they will be free to focus on their main job of providing local infrastructure. Moreover, the fact that the (relatively) small amount of money that the guaranteed income represents needs to be collected every week by showing up somewhere would discourage a lot of the wealthy from applying, so a certain amount of targeting would happen automatically.
I have no doubt that the supporters of the MGNREGA have many arguments why this is too optimistic and some of those arguments will surely be worth taking seriously. But whatever we ultimately conclude about the relative merits of these approaches, we need to talk.
Abhijit Banerjee is Ford Foundation International Professor of Economics and Director, Abdul Latif Jameel Poverty Action Lab, MIT
The views expressed by the author are personal