Note: This interview, originally published on December 20, 2017, is being republished on the occasion of Abhijit Banerjee winning the 2019 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, also known as the Economics Nobel.
In conversation with the economist on his work with randomised control trials, demonetisation and the GST, the ‘Gujarat model’ and more.
Jahnavi Sen: Hello and welcome. I’m Jahnavi Sen from The Wire and with me today is Professor Abhijit Banerjee. He is a professor of economics at MIT and also founder and director of J-PAL, the Abdul Latif Jameel Poverty Action Lab. Thank you so much for joining us.
Abhijit Banerjee: Thank you for having me.
JS: Sir, now you’re celebrating ten years of J-PAL South Asia and on that I wanted to ask you – in 2012 when you wrote Poor Economics, it was something of a phenomenon. It had a lot of success and since then your work with J-PAL has also been in the limelight. How would you say that work has affected both policy makers and aid donors, in India and otherwise?
AB: Let me separate me from the story. J-PAL has 180 affiliates, these are professors at different universities, and the work we do, the reason why it has some impact, is because it’s a sum total of a huge body of talent. And the part that I have contributed to it is tiny. So mostly I would say is that one thing J-PAL has done successfully is brought together lot of talent, and lots of those people in fact work in South Asia, so it’s a specifically South Asian success story, in the sense that we’ve managed to attract really the top talent of the world to come and work here. And that, I think, is first and foremost what J-PAL South Asia has been able to do.
Then if you continue: what have we achieved? Well I think we now have a place at most tables. In the sense that people know what we do – they may decide that they don’t want to talk to us, either because of the work we do, or they don’t like what we say or we’re just too much trouble for them, it takes too long to do RCTs [randomised control trials] or they don’t know what RCTs are – but they know our name. So we exist as a kind of a player in the policy space in a sense in which we didn’t exist ten years ago. Right now, I don’t think we need … It’s not that everybody knows about us, but you can find champions inside any government who’ll know what we do and why if at all we’re valuable.
JS: In India, you have a lot of projects running. Would you have any examples of your specific RCT projects that have had a direct impact on some sort of policy change?
AB: Sure. One place where I feel that we’ve been chipping away at a problem for many, many years, I think starting in the late 1990s, is this problem of why kids who are in school don’t learn. This is now well established, we’ve all seen the ASER report which comes out and everybody shakes their head, then they forget about it. We’ve been sort of pushing that particular question for a long time and we’re pretty confident that we know a significant part of the answer.
In other words, one very big part of the answer is simple the fact that when teachers teach children, they teach the syllabus and not what the child needs to learn. And at least in lower grades, that’s a huge problem because if you don’t have the basic skills, then it doesn’t make any sense to be taught anything else. If you can’t read, then teaching you history is pointless. Many children in Class IV, where they do teach social studies, can’t read. So then it’s sort of a pointless exercise to write things on the board, which they try to copy down, but they can’t write, so they can’t copy it down. So there’s a certain amount of meaningless ritualisation of the syllabus which has been accepted as the standard thing to do. And I think when you don’t do it, we have now probably seven or eight of RCTs showing, you immediately get large amounts of progress. And you can do that through various different models, it’s not that there’s one way of doing it, but all of them share this idea that you better teach a child what s/he needs to learn, rather than what the syllabus dictates what s/he needs to be taught.
I think that idea has had a fair amount of eventual leverage, because I think you see now many governments prepared, like Karnataka is the latest one – I was in Karnataka about a month and a half ago, talking to the government – and you feel there’s a lot of interest in how to make the education system work better. And one idea their people are increasingly receptive to is that you somehow have to organise your classroom such that what children need to learn is given primacy. You might think that’s obvious, but it’s actually not, for reasons I don’t entirely understand I must say, something to do with the sociology of the eduction sector. But it’s something that I think people are beginning to realise that, it’s so obvious, it better be true.
JS: A lot of your work is based on what can be seen as a relatively small interventions, or like you said it may seem obvious, and how they could have big impacts that people have been for some reason ignoring. One criticism has been that may be that ignores political processes or the larger context. What would you say to that?
AB: I’ll say two things. One is, sure. I think that if I had a lever which made our democracy work perfectly tomorrow, if there were a magic wand which makes it happen, I’m going to wave that wand. I just feel like it’s an easy thing to say because you don’t feel the obligation to actually deliver on it. It is not clear to me that other than solving the problems we can solve, there’s much else we can do. So its great to talk about how we need to make our society more democratic or more egalitarian, all those things I completely agree with. But I don’t feel that I have an immediate lever, given right now the politics of the country – I don’t see that I have one lever which says that, you know, tomorrow the country becomes much more democratic and everybody will participate equally and we will get the service delivery that the founders of the country intended. This would be great, but I don’t see that and that’s one answer. One answer is that, we do what we can do.
I don’t mind people taking the opposite view, which I think is also useful, to say ‘All we do is protest’. And protesting is valuable, I don’t at all want to imply that just the act of taking the time out to protest is not valuable. It is valuable because it kinds of creates the pressure to do something – but that’s not me. I would rather spend my time trying to solve a problem that can be solved. I don’t at all think that it’s the only thing to do, but I do think that to say that this is useless until we transform the society is kind of not helpful. Second point which I wanted to make, which is, we actually do work a lot on governance and a lot of our work is on trying to, for example, if you inform voters, do you get better outcomes? And we do see some evidence of it, so I don’t want to say that democracy is somehow beyond RCTs – it’s not at all. Its a very natural place to try out different things and in some ways it’s a place where there’s lots of possibilities, people are uncertain of what the right way to move is, so it’s a natural place to experiment. I think of it is being very much within the domain of what we do and indeed we do a lot of work on it. That’s not to say that I have a recipe for fixing democracy totally tomorrow.
JS: On the topic of RCTs, if a researcher or a policy maker is looking at a certain results, that are from a particular context, how do you think s/he will be able to judge whether or not that could extrapolated to a larger context or to a different setting, what do you think could help make that judgement?
AB: So you put it perfectly. Eventually there is no way to avoid a judgement. I don’t think there is a sense in which we could say ‘We did it here and here, and now we know it’s true everywhere’. That kind of claim at some level would be great to be able to make, but I don’t think we can ever make. I think that’s mostly true for everything and not particularly true only for RCTs, any form of social science has that of problem, in the sense that, any evidence we see comes from some place and may not apply to some other place. Therefore, there is always some judgement. I mean we try to do the best we can, which is try out things in multiple locations before we go to town with it, we try also vary the implementer, we go from implementation by an NGO to implementation by a government department. We try all the different forms of variation. Hopefully that’s enough. But you’re right, there is a point at which we say, ‘Well, we’ve tried enough and we can’t wait forever’. Fair enough, I think it’s the nature if social science knowledge.
Also read: Enrolment Rates Are Climbing. So What Explains the Sorry State of India’s Education Sector?
JS: Just to move topics a little bit. Recently you’ve been writing about economic policies currently in place in India. So what would you say about the current government and their economic strategy, successes, failures, have there been any major missions?
AB: So what I would say is, for most policies what’s striking is how little…maybe they are doing a lot but certainly what shows up on the front page looks like UPA-III. There isn’t really a deep policy thought that’s different. I think at the beginning there was a lot of talk about freeing the people from the yoke of the government. I was not sure what that meant, but I now see that mostly that’s not a big part of what’s going on. Having said that, there’s a couple of weird things that the government did, one is the demonetisation – I don’t think there was any serious economics in it and there was no particular reason why it would do much good. My sense is that, even though, people in government won’t say it, I don’t think they are totally totally convinced that it was great success either. I don think anybody is saying that… I think the government can’t say it, for some reason, but I don’t think it was a…
JS: But maybe the fact that they kept changing the stated aim in indicative of that.
AB: Yes, that’s something they did. And then I think the GST implementation was not great. I think the GST idea in the long run is a good idea, but I think the couple of political battles that needs to fought to get it implemented well, and also I think it was worth missing the deadline to get all the processes along the way sortedout. And I don’t think they made the right judgement call. But there I don’t, I feel it a little bit less, may be it was kind of a judgement call. Maybe they got it wrong, but I don’t think it was perverse. I think the GST idea, in the long-run most countries have gone to a GST. We have to think about how to deal with the informal sector. We haven’t really fully got it. But I think it’s not going to be…if we waited for the informal sector to be all formalised, it’ll be a long, long time. So I think there had to be some imperfect movement, so I don’t want to be too harsh on the implementation. I think it annoyed a lot of people, clearly, there were lot of problems, but there’s no perfect way to get from where we are to the GST. So in that sense I feel that it was something that could have been done better, but it wasn’t an enormous disaster by any means.
JS: In 2015, when the 14th finance recommendations were accepted by the Centre and the state’s share in the Centre’s taxes went up by 10%, there was also a lot of talk, including from the prime minister, about how this would help the states shape their own development schemes and implement them better. Have you seen any impact of that on the ground?
AB: No, to be honest, I haven’t. But there was a little bit of a smoke and mirrors there, because to some extent what happened was, there were state entitlements under specific schemes which was also buried under that extra money they were getting. So it was not easy for the states to sort of say, ‘Oh well, now we’re going to default on some of our obligations’. Obligations are kind of set out, people kind of know that the Indira Awas Yojana will give me a house or whatever, and now they’re telling them, ‘No, we changed our mind’. It’s politically difficult. Eventually it is possible that this will have consequences because, maybe, eventually some of the new programmes states will not agree to take up and will develop their own. There are some state initiative that are interesting. I mean, there is this Bhamashah programme in Rajasthan which is kind of novel. States are doing some new things, it’s not huge yet…
JS: And do you think it’s an outcome of this change?
AB: I don’t know. The only point I’m making is that we do want the states to become laboratories of different strategies and that would be a good thing. Whether the funding has much to do with it, I don’t know.
Also read: The Truth Behind the Gujarat Growth Model
JS: Talking about state strategies, recent there’s been all this talk about the Gujarat model, thanks to the elections. What is your take on the Gujarat model of development?
AB: I never understood what that exactly meant, other than that I feel that Gujarat does have very good infrastructure, it’s explicitly business friendly. If I take that to define the Gujarat model – the results are actually, given that, not spectacular. You might think that Gujarat may be much faster growing than the rest of the country, you don’t see that. There’s two issues there. One, the big issue that scares me a bit is something that economists call convergence, which is that places that grow fast for a while, inevitably slow down. But if that’s whats happening in Gujarat, Gujarat isn’t that rich. So then we should be very worried. I don’t know why Gujarat is not doing better, it’s not one of the absolutely fastest growing states in the country, its among on the top but it hasn’t shown a dramatic change, for example, since the BJP has taken over, the growth rate is pretty stable, going downside.
That part could either be because the policies aren’t doing anything, or because the policy are doing something, it’s business friendly or whatever, but on the other hand there’s a general tendency to a slowing down, specially in the richer states. If that’s the case, then that bodes very ill for the whole country. Because Gujarat at its level, this is what is called the middle-income trap and I worry if Gujarat represents our step toward the middle-income trap, where you get stuck at a relatively decent level of per capita GDP but don’t go towards the really well-off countries. And that’s happened in a bunch of countries and I worry that maybe we are slowing for that reason, so maybe the policies are doing something, at least in terms of attracting investment, but on the other side, just the extra capital doesn’t drive development.
Now, why do I say that, I feel that one reason that might be the case, and that’s sort of one very intuitive idea, is that Gujarat has not done a great job in investing in human capital. It’s one of the states which is a laggard in terms of investment in human capital. Now, that might eventually become a constraint on growth, because you are investing more and more on infrastructure, capital etc., but you don’t have the corresponding improvement in human capital. And that’s often the reason why countries get into the middle-income trap. I hope it’s not true actually, but it could be that’s what’s going on. I mean, as a country we have done very badly in investing in human capital. We have been growing pretty fast despite that, but we kind of don’t know why we are growing given how badly we have done on human capital. So it could be, basically, yes, you get to some point – $5,000 PPP GDP per capita, $6,000 – but then you stop growing. That’s a concern we should be taking more seriously than we are now.
JS: Just one last question. You’ve advocated before for shifting to cash transfers in India. What do you think will be the gain from that and what kind of schemes do you think it should replace?
AB: Well, replace…there are two parts of the question. One is, I think that, despite all the claims made for the PDS, it’s an extremely costly scheme to run. It’s getting much better. I still don’t believe it’s one of the most – I think the particular assumption of it which is that if you give people food, they eat more – I don’t think there is any evidence for it. And I just don’t see it being…there is no particular reason to take all the cost of maintaining the FCI, the procurement, every single step of the government infrastructure. I don’t see a reason why to have it. So, is it the case that … now you may say that there are places where food is really scarce and supply is bad, I’m willing to be persuaded that there places where you want to keep the PDS. But to take grains from Punjab and redistribute it in Punjab seems senseless to me. You might as well give people cash, it’s not very different. I think there is plenty of evidence suggesting that in places where supplies are adequate, it makes no difference to nutrition. There is a bunch of RCTs looking into that. Then there is the question of what do you do about something like MNREGA and there I think it’s a more interesting question because there, you may say that what MNREGA was intended to do, it doesn’t do. So the one thing it is not, it is not cash on demand. It’s something that, your village has to put up a set of proposals, then they get approved, then you get money…something like six months after you need money, you get money. One other problem we need to think about as a nation is how do we provide adequate emergency relief. We talk about suicides and things like that, these are usually symptoms of a crisis, also of a long-term problem but especially of crisis and we don’t have a crisis response system. The best thing we have is MNREGA, but MNREGA is not crisis response system. It’s a very slow responding system. So we may want to think about a way to allow people, possible based on self-selection like MNREGA, which is where people get access to cash quickly when they really feel desperate, rather… we haven’t really talked through that entire structure.
JS: Thank you so much for your time.
AB: Thank you.