Social benefit not same as private gain: Abhijit Banerjee

Abhijit Banerjee, a professor of economics at MIT has been awarded this year's Nobel Prize in economics along with Esther Duflo and Michael Kremer. The Nobel committee has recognised "their experimental approach to alleviating global poverty...which has transformed development economics" while giving the award. In an interview with Hindustan Times on Saturday, Banerjee spoke about his academic work, anti-poverty programmes in India, the state of the economy and society, among other issues. Edited excerpts:

You did not begin your academic career with Randomised Control Trials (RCTs). Can you share your experiences and philosophical transformation which made you shift to this approach?

When I was a graduate student, I actually took a course in development economics and I thought it was the most boring thing in the world. I was trained as an economic theorist; my job at MIT was as an economic theorist. At some level that's still part of my identity. What's nice about experiments is that they are much more closely tied to what theorists think about the world than normal empirical research. You can design your experiment to exactly ask the question you want to ask. This is not true about normal empirical research.

Can you explain, perhaps with an example, how you use RCTs in your academic work?

There are a couple of different pieces to that. The idea of an RCT is like a medical trial. In order to know the effect of something, the best way to know it is to not have to worry about whether the people who got that 'treatment' are different in some systematic way from the people who did not get it; the control group. The way you achieve this is by choosing them at random, and then comparing the two groups. The value of that is first, it gives you relatively hard to argue against evidence on the impact of a programme. When you are doing something controversial, like what is the impact of micro credit, it is very useful to do an RCT. It is hard to argue that in the location, where the experiment has been done, I've got the answer wrong. Also, if you want to build a story about what drives poverty, it is useful to be able to test theories. What experiments often do is they give you a nice way to test your theories and hypothesis about these things.

Having received the highest recognition in the discipline, how would you like to see the future of the work you and your colleagues at J-PAL (Abdul Latif Jameel Poverty Action Lab) have been doing?

Hopefully, what it establishes is that we are competent professionals and that what we do is not silly and we know how to implement a certain kind of method. Hopefully it will open more doors. People will be more trusting of the idea that they could get their programme evaluated. We don't start from some ideological premise, which says that I'm going to always find that certain kind of programme or certain government's programme is always good or bad. We want to achieve a certain reputation for being competent and neutral. We don't care about the outcomes than doing a good job of testing it.

Your methods have also been criticized by your peers. A statement signed by three previous Nobel laureates said the following about RCTs, "But truly random sampling with blinded subjects is almost impossible in human communities without creating scenarios so abstract as to tell us little about the real world." Concerns have also been expressed that such an approach tends to ignore the "broader macroeconomic, political and institutional drivers of impoverishment and underdevelopment". How would you like to respond to such criticism?

They are two separate criticisms. About ensuring blinded or not, we try a lot to take care of it. For example, in Nigeria we were trying to find out the effect of an MTV show on HIV. We worried that it is not blind, as people knew they were watching a show. To make it more comparable, we gave the control group another TV show. It is not perfect, but it is not that we are unaware of the problem.

The fact that there are other things which are important in the world; who's arguing against it? I fully acknowledge that lots of things happen which I couldn't test in an RCT, which either work or don't work. Some disasters happen because we are unlucky and some very fortunate things happen because we are in the right place. That's the nature of life. I don't feel that we have ever claimed that we are going to explain every question in development economics through RCTs. A lot of things happen where we don't have control. My point is that we could at least do a better job in areas where we do have control.

In India, important anti-poverty programmes such as NREGA (National Rural Employment Guarantee Act) and the Forest Rights Act, have come out of political struggles. You've described the Nobel Prize as "not for us, but also for the entire movement". How do you see your movement for RCTs interacting with the political movements which have been demanding welfare programmes such as NREGA?

I've never really thought about this particular question, I'll think about it more. There's an interesting tension here. One of the downsides of many social movements is that they get frozen into a particular demand: this or nothing. This lends a certain kind of power to the movement, as you don't have to keep re-examining. The danger is that you might end up gaining something, which you really don't want when you get it. One of the ways in which we have tried to educate ourselves is to think about let's see where we are and let's think of where can we go from here. Those are different points of view. The flexibility we suggest clearly has certain gains and some potential losses.

Let us come to Indian economy now. In your speech at MIT after receiving the award, you said adhering to fiscal deficits and inflation targets will not help the Indian economy in getting out of the current slowdown. Can you elaborate a bit more?

I think this is not rocket science. It depends very much on the diagnosis of what's exactly happening in the Indian economy. I think there's a Keynesian or a demand driven downturn which feeds on itself. I don't have money, so I don't buy biscuits, so the biscuit company shuts down and they don't buy. I think the best short-term action is to pump demand into the hands of those who spend. This is what the Obama administration did in the US. This is beyond ideology. What's controversial is how much of a demand downturn are we facing. Not everybody agrees that we are facing a demand-driven slowdown. Given that the data isn't wonderful, I'm willing to accept that those who don't agree with this are not necessarily mala fide or something.

"One very real danger is that in trying to hold on to fast growth, India will veer towards policies that hurt the poor now in the name of future growth", your book says. Can you elaborate?

This is the experience of so many countries. What happened in the UK or US? After growth slowed down in these countries in the 1970s, it never recovered. They had no idea why it slowed down. At that point, the first reaction was that the high tax and high redistribution policy frame was to blame for it, and therefore cutting them was the solution to it. That was Regan and Thatcher style economics.

Are you saying that we are heading towards that kind of economics in India?

No, I'm just saying that it's the natural reaction of governments when growth slows down. We are saying that we should be warned that such policies did nothing for the US and UK other than blowing up inequality. In some ways these policies created the way for economics which leads to Trump and Brexit.

What would you do to revive the economy?

I'm not a macro economist. But if I were a policymaker then I would first collect a lot of data, and if I were convinced that it is a real demand-driven slowdown, then I would put lots of money into the hands of the poor. That is something I believe in absolutely. That is what the Obama administration did in the name of quantitative easing. That's not unheard of.

Your book also talks about how fighting corruption is not a costless endeavour. In the last few years, fighting corruption has been an important driver of politics and policy in India. Cancellation of resource allocations to demonetisation are some examples. Do you think it has contributed to the current economic situation?

Whether it is anti-corruption, or the fear of being seen as corrupt, or maybe corruption was important in greasing the wheels of the economy and it has been cut down, lots of my business friends tell me that decision making has slowed down. They feel that there is a set of institutions, but they seem to be reluctant. I think certain amount of forbearance is crucial at the heart of market economy. You have to tell people that you are innocent until proven guilty. You have to assure people that mistakes are not malfeasance.

The book terms bulk of R&D efforts being directed towards machine learning and other big data methods as a distraction from truly path-breaking innovations. Are you making a case that private dominance in R&D might not lead to socially optimal innovation?

There is a pretty good case to be made that private incentives are not social incentives. We don't like people being unemployed, but private companies don't care. If it is cheaper to replace an employee with a machine, a private company will do it. The Korean government is the first to declare that if you replace people with machines you have to pay a tax. It's a tax on robots. They make private companies internalise the social cost of unemployment. Social benefit is not the same as private benefit. We have to realise this.

What do you think about the importance of international cooperation to deal with the current economic situation globally? We live in the times of trade wars. History tells us the breakdown of such cooperation preceded the Great Depression.

History suggests that people should be very scared of this moment. I despair that the Chinese will not be able to persuade Mr Trump. If the US makes China shrink, then we will all pay for that. China is also a very large consumer of commodities, which are supplied by a lot of poor countries.

Your book argues that expression of contempt for those who express racist sentiments or vote for such leaders serves to only reinforce those sentiments. Would you like to explain this in the context of India?

I have always argued that we need to be able to understand people's reasons for voting for even polarising leaders. If you take the view that because somebody has done it, therefore they are a bad person, you are undermining all possible ways for democratic discussion and resolution. It is my belief that lots of good people get pushed down certain political alleys by forces they do not fully understand or have control over. Therefore, both morally and practically, judging them for doing so is not correct.

Are you saying that democracy has not let down liberal values, but liberals have let down democracy?

Correct. That is a beautiful way of saying it. I have always taken that view that we, as liberals in India, have not created a deep enough liberalism. We should stick to our liberalism and not lose the tolerance which is built into it.

"Many economists have a philosophical objection to manipulating preferences", you've written in your book. We have just celebrated the 150th birth anniversary of Mahatma Gandhi, whose economic outlook, often bordering on changing preferences, is criticized by most economists. Do you think Gandhi is relevant in the current fight against global poverty?

I think most economists don't know much about Gandhi, and a particular generation of economists might have been anti-Gandhi. But in today's world, it is hard to not appreciate his extraordinary commitment to not label people and try to be inclusive of people who were despising of him. I think that is a very powerful commitment which I deeply admire. Even though he got angry temporarily, he would try to reach out to the same people once his anger went away. Gandhi's philosophy was remarkably deeply liberal in a way. It came from saying that in the end, everybody has good in them. That liberalism is deep liberalism.

The Nobel Prize will significantly increase your following beyond the discipline of economics, especially in India. What would you like to say to India's youth, especially those pursuing their academic careers?

You should love what you do. I talked to a bunch of people in finance who say I've made a lot of money but I'm bored. I think they would have been better off doing what they enjoy. Passion is an extraordinarily handy thing to have in life. I'm 58, but I'm still extremely passionate about what I do. I want to just keep doing the research I do. That's a huge reward.

Do you think we in India have been able to provide such an eco-system to our young generation?

I think less than when I was growing up. When I went to college and to JNU, there were people who had very different political views, but they were all extremely talented and passionate people. That's what made me want to be an academic. If we don't get really talented teachers, we are not going to get young people interested in doing exciting academic things. Partly because industry has become so much more lucrative etc. We don't attract the most talented people in academics. If we don't have good teachers, we can't have good students.

Can you list two factors each which makes you optimistic and pessimistic about India?

There is still a deep optimism in India that the future will be better and it will be Indian. Then there's the effortless sense of immediate warmth in India, which you do not find in other places.

I worry about the fact that there is a whole generation of people who've been short changed by the education they have got. At some point they have a right to be angry. They invested in education, were first generation learners. So their expectations have gone up, but they might crash. Second, we need to be more tolerant of dissent. It is a part of our tradition. In my family there were all kinds of Hindus; Vaishnavas, Shaivas, Shaktas etc. There was no particular homogeneity. We need to embrace that.