Shoumitro Chatterjee on Agricultural Policy

In this episode, Shruti talks with Shoumitro Chatterjee about farms and farm policy in India, particularly in terms of government subsidies and other state intervention in agricultural markets. Chatterjee is an assistant professor of economics at Pennsylvania State University. His research interests include agriculture and international trade and development.

SHRUTI RAJAGOPALAN: Welcome to Ideas of India, a podcast where we examine academic ideas that can propel India forward. My name is Shruti Rajagopalan, and today my guest is Shoumitro Chatterjee, who is an assistant professor, School of Economics at Pennsylvania State University. We discussed his recent Study of the Agricultural Markets of Bihar, Odisha and Punjab (coauthored with Mekhala Krishnamurthy, Devesh Kapur and Marshall M. Bouton) to understand agricultural markets and agricultural policy in India.

We spoke about mandis or agricultural produce markets, government regulation of agriculture, minimum support price in Punjab, his research on agriculture, trade and structural transformation in India, his intellectual influences and much more.

For a full transcript of this conversation, including helpful links of all the references mentioned, click the link in the show notes or visit mercatus.org/podcasts.

Shoumitro, welcome to the show. It is so great to finally have a chance to speak with you.

SHOUMITRO CHATTERJEE: Thank you so much for having me. The pleasure is all mine.

RAJAGOPALAN: I want to get right to this really important policy study that you’ve written with your co-authors for CASI, and the way I think about agricultural policy in India is that it’s across three major policy areas. One is that the government, both at the union and the state level, subsidizes agricultural inputs.

The second, of course, is regulating agricultural technology. This is typically in the form of GMO seeds or regulating what kind of yields or seeds that Indian farmers can import or export from other countries and so on.

The third part is agricultural markets. This has come under a lot of scrutiny recently because of, of course, the major farm reform bills in India and the protests by farmers across the country.

The agricultural market system in India has a lot of variation. One part of it is, of course, the agricultural produce markets, which are regulated—the APMC mandis—but there are various other ways of actually selling agricultural produce. Some of it is just directly to the government; some of it is self-consumption—it’s not even for sale; some of it is sold literally at the farm gate; some of it is sold through an entire chain of intermediaries.

Can you just walk us through this really complex and highly variable system of agricultural produce markets in India?

Selling Agricultural Produce in India

CHATTERJEE: In terms of agricultural markets, if there is one main finding of our study, it is that things are very complex, not only across states but also within states. Sometimes also for the same commodity within the same state. In places like Bihar and Odisha, which were two of the main states that we studied, you don’t have your “APMC markets”, in the sense that Bihar had done away with the APMC regulations in the mid-2000s. Odisha has something called a Regulated Marketing Committee, but it’s not the APMC that people understand in the popular imagination. Regulated Marketing Committees are mostly sites for procurement of paddy, so other commodities in Odisha don’t go, usually would not go, through these RMCs.

For example, chilies in Sambalpur are an important commodity. For a very long time, it has been controlled by a syndicate, a group of traders who have connections with other traders in the Azadpur Mandi, and they contract with other local intermediaries who pick up chilies at the farm gate from farmers and then arrange for transportation.

A completely different commodity in Odisha is prawns: the prawn revolution in India is a fantastic story of transformation in the agricultural sector. There the supply chains are very localized and immediately become international. You have the processing plants who contract with farmers; they send the refrigerated trucks to pick up when the harvest is ready. They process it—these are world-class facilities—and then they are immediately shipped to the Visakhapatnam Port in refrigerated trucks and exported.

Betel wine is another commodity, and again has its own chain. For all the commodities that we studied in the report, both in the main section and in the appendix, the supply chains of all these commodities have been mapped out.

What we focused on in the report was understanding the first sale to a great degree of agricultural products in Bihar and Odisha. Then, to some extent, the second layer or the third layer of the supply chain. Let me start there and not in Punjab because our understanding of Punjab models our understanding of agriculture in general, because Punjab and Haryana are very special cases, and that’s not how the rest of the country works.

What we found in Bihar and Odisha is that almost all sales were at the farm gate. So, typically, traders would show up in villages at the farm gate, and farmers would sell to them. The first thing that occurs is weighment, and so I always like saying that farm incomes are revenues times quantity minus cost. The policy focus tends to be on price, and we forget the rest of the profit equation. In order to understand my realized income, I need to know what the Q is, what the quantity is, even after production.

Surprisingly, maybe, in Bihar and Odisha, a large fraction of weighment is actually done by electronic weighing scales, even when sales occur at the farm gate level. Now you compare this to Punjab, where you would expect that marketing systems are better, weighment almost everywhere is using manual weighing scales. And weighment is one way in which intermediaries would make some cut. These are not huge margins, but this is when you start making something.

Now the commodity is weighed, the second thing is about prices. You’re promised a price on the spot, but you’re never paid on the spot. In fact, nowhere in the country or in the developing world are you paid on the spot. It’s essentially because the supply chains function on credit. Once the intermediary buys from you, they have to sell it to somebody down the supply chain, and once they get paid, they will pay you.

Now, that’s where the second layer occurs. It’s “I promise you X rupees a ton today.” Tomorrow, I say, “Okay, I bought this much from you, but 2% or 3% of that was dust, so here is the deduction, and here is what you get.” These were some of the issues that mandis were supposed to solve, and when we talk about mandis, I can come back to that.

Typically, in Bihar and Odisha, you get paid within two weeks.

RAJAGOPALAN: You actually get paid faster in the market through intermediaries than through the government. The government takes a really long time to clear payment.

CHATTERJEE: Let’s take a specific example in Bihar. Let’s situate ourselves either close to the Gulabbagh Mandi or Bihar Sharif Mandi, which are large mandis in Bihar. These are not APMC mandis, in the sense that they are not regulated, but they are markets. They’re large market sites where there is trader-to-trader interaction, or maybe large-farmer-to-trader interaction.

RAJAGOPALAN: The difference being that these are sites where farmers and traders can choose to come if they are not—it’s not a mandate to only sell there because of some kind of regulated APMC monopoly. Is that the correct distinction?

CHATTERJEE: That is one distinction, but that distinction was also valid some years ago. Those reforms had already started happening in the sense that for many commodities in many states, nobody forced you to go to the mandi any longer, especially if you were like a small farm. Even if the regulations existed in the letter of the law, if you were a small farmer, you were too small for anybody to care about you.

Even if you sold to the large farmer in the village, who then took your produce to the mandi, nobody would trouble you that why did you sell to this large farmer, even if it was not by the letter of the law.

There are intermediaries who can buy at the farm gate, and then as a farmer, if you had enough of an output such that it made economic sense for you to rent a vehicle or own a vehicle to transport it, you could go to the mandi. When you are close to these market sites, the prices in these mandis do serve as an anchor of the price that you would get at the farm gate, because we don’t have any good market intelligence system like that which exists in the US. So that’s my benchmark price.

It could be possible that the intermediary who is buying from me at the farm gate will eventually sell it in the mandis.

RAJAGOPALAN: And takes a cut.

CHATTERJEE: No, but the cut—you have to remember why the cut is for. It could be an inefficient cut in the sense that it is a markup, or it could be substituting for a service.

Those are of course transportation, storage, et cetera, but also if there is a lag in payment at the mandi. Suppose, at the mandi, the large intermediary pays me after a month. As a farmer I can choose to sell to this guy, get a better price, but then I have to get paid after a month. Or I can sell to somebody in between, get paid immediately, but that difference in the two weeks of payments shows up then as a difference in the price.

How do I get paid? Almost all transactions are in cash in Bihar and Odisha. In Punjab, you do see some payment either by check or by bank transfers. What was also interesting to note is—and this is consistent with what others have found—is that Bihar and Odisha markets do not usually reward post-harvest processes, things like cleaning, grading and drying.

I don’t know if it’s a consequence or if the causality runs the other way, but because the market is not rewarding post-harvest processes, most farmers are also not engaging in post-harvest processes. The only one they try to do is drying, and that shows up as a mild—those who dry then get a 1% or 1.5% better price for their output.

The last thing is price discovery. Because you don’t have regulated markets in Bihar and Odisha, almost nowhere do you see auctions or anything of that sort. It’s mostly bilateral bargaining. Whereas, in Punjab, for non-MSP [minimum support price] commodities, where it’s not going to the government, there is almost always an auction.

Now, there is the other question of whether auctions themselves are improving prices, i.e., if we were to do auctions in Bihar and Odisha, would that improve prices? I don’t know the answer to that question. I think that depends on a whole host of things, including can intermediaries collude and beat the auction because, of course, it is possible to beat the auction.

Remember, though, that technically the way auctions are defined in the regulated acts, they are not auctions in the sense that we study them. When we study a first-price, second-price auction, it means that the price that is discovered is then binding. In agricultural markets, the price that is discovered then can be renegotiated. Both the farmer and the intermediary can ex post say that I don’t agree to this, and so let’s not go ahead with the trade.

Very often it happens that the buyers say that, “I looked at the top of the grain, which looked good, and therefore this is the price that I bid. Now that I’m looking at the full crop or the full lot, I realize that this is a bad quality, so I’ll offer you a low price,” but by that time, the farmer has already paid the fixed cost, right?

RAJAGOPALAN: Yes.

CHATTERJEE: So your stuff is already unloaded.

RAJAGOPALAN: And all the other potential buyers have walked away.

CHATTERJEE: When you have good regulated markets and strong regulated markets, these are the situations that can get resolved. Which is why, for me, the real example, the benchmark should be Madhya Pradesh because these things get enforced quite well, in the sense that if you try to cheat farmers the Marketing Committee, the secretary or somebody will say, “No, no, no, this is not what you can do. If you try to do this, I’m going to liquidate your asset which you kept for the license.”

Why Are Indian Agricultural Markets Underdeveloped?

RAJAGOPALAN: I just want to take one step back and think about agricultural markets in a way that we would think about, say, idealized markets. I’m thinking about something like Aalsmeer in the Netherlands, where the tulips come batch after batch, and you have everyone doing a Dutch auction, and they press the button, and the tulips are sold.

You have pretty much every single mechanism. You have the transportation, you have the cold storage, you have very, very thick markets, large numbers of buyers and sellers. You have flowers coming in from Kenya which are going to the United States but are being sold, or at least the deal is being struck, midway in the Netherlands.

Now, what are the kinds of structural bottlenecks, and what are the kinds of legal bottlenecks in creating these massive, idealized market conditions where you have very thick markets for agricultural produce in India?

CHATTERJEE: I guess what will help is if you take one step back and talk about when mandis exist, what do they do? Do they help in any way? Then I can come and answer this question more directly. The original idea behind why mandis were established, they wanted to solve three or four economic problems. We spoke about counterparty risk. Then there is dispute resolution. If my trader does not honor his promise, whom do I go to?

Then there’s the question about price discovery. How should we discover the price? Should it be bilateral negotiation, should it be auctions or should it be something else? For whatever reason, we thought that auctions, the way it is defined, would be the best way.

The fourth thing that mandis implicitly do is information dissemination, which is that, even for off-mandi transactions, how do I know what the market price is? In remote areas in Bihar—there’s a chapter on remoteness where we talk about this—this becomes a huge concern.

The fifth thing, which is very, very important and critical for agricultural products, is quality assaying. I bring in greens, of course, and—I like when I give seminars and people say that agricultural commodities are homogenous. My reaction is, “No. When you go and buy tomatoes at Whole Foods, you don’t just buy a random tomato.” Even basmati rice, the full grain rice is different from broken rice. When farmers bring in commodities, it’s a mix of stuff. You have good grains, and you have some broken grains, and there is some washer content and some dust and some pests. This is not like milk.

The big revolution in the milk was the lactometer. You could put an independent thing inside, and that would tell you irrespective of—independent of the buyer and the seller, what was the quality of the milk? Which I could rely on, and you could rely on. The way it works in agricultural markets in India today is that the intermediaries are performing that economic function.

Bihar is one of the largest producers of fresh vegetables. Vegetables from there get transported to Delhi. The buyers in Delhi want a certain quality of tomato or cabbage or something to be shipped. How do they know that they will get that quality? That is ultimately insured or backed by our intermediaries.

Even in MP, for the soyabean market, the private mills, they want a certain quality of soyabeans delivered to them. How do they know that they will get it? That quality-assaying aspect is done by the intermediaries. Once you think through those channels, there are many ways in which our markets are incomplete.

Our markets are spatially incomplete. Our markets are temporally incomplete because we don’t have well-functioning insurance markets. We can talk about that in more detail. We don’t have independent quality assaying going on.

Then how do you credibly convey information? I think all these are complete markets, and all actors being price takers are your basic ingredients for the first welfare theorem to hold. Those would be the basic ingredients for any market to function. There is variation across states, but we are far away from those aspects.

RAJAGOPALAN: Is the problem that the original design, that is the legal and policy design of the mandis, was flawed? Or is it that the legal design was correct, but they simply didn’t have the enforcement mechanism or state capacity for things like dispute resolution, quality control and so on?

Have they been captured by interest groups? Is there just a lack of trust in the sellers to go to the mandi? Is it remoteness and transportation because the farmers are so small that they may lack the ability to actually physically be present at these places to get the information? Can you give us a ground picture of what is it that is the roadblock?

CHATTERJEE: I think it could be a little bit of everything. Just to go back to the policy aspect, one thing that we have to realize is that policy is not time invariant. I think when these policies were made initially—I’m going back to the 1928 Royal Commission Report and, subsequently, the initial APMC Acts. At that time it made a lot of sense. In fact, if you look at historic data, that did increase farm income substantially.

Over time, of course, what happened was, within mandis, you saw cartels amongst intermediaries. Licensing, obtaining new licenses—there should have been no barriers to obtaining new licenses. Therefore, the states should have invested more in mandis.

The first part of my dissertation exactly shows this, that in parts of the country where you have more well-functioning mandis, it leads to a better price realization for farmers. In places where you have fewer mandis, it leads to depressed prices. Mandis, for India at least, I view them as a public good. It’s nothing different than a railway station.

What this is saying is that states should have invested more in mandis. If you think about the chili syndicate that we were talking about in Odisha, those farmers whom we interviewed, they have been asking for a mandi for a very long time. They understand that if there is a mandi, that syndicate would face then competition from other buyers who would know that this is a place where you can find chili. The absence of a well-functioning mandi hurts them.

RAJAGOPALAN: Yes. Basically, what you’re describing is your classic thick versus thin market problem, right? For any well-functioning thick market, you do need some legal and physical infrastructure.

CHATTERJEE: Exactly.

RAJAGOPALAN: That is exactly what seems to be missing in the Indian mix. Sometimes for things like prawns, the physical and the legal—or contractual, rather—infrastructure has been provided by private parties, mostly people who are buyers from abroad who can afford to pay a huge premium to make this infrastructure function in their favor. For other things, like chilies and onions and tomatoes, it just simply doesn’t exist in some states.

CHATTERJEE: There’s one more point that I want to highlight, but let me just say that the other aspects that you raise are also right. India, having many, many small farmers—so the farm size distribution just being skewed to the left: that is a problem. Transportation is a problem.

Initially, if you have smaller farmers with less output, you are also not harnessing the scale economies in transportation. The initial leg of the transportation is being done in small vans and not trucks, and that is also contributing. Small things are getting added up into a big margin.

On the investment side, I think a very good example is the maize revolution that has happened in Bihar, actually. North of Ganga, in Bihar, farm sizes are bigger. I think mid-2000s, Cargill realized that there is scope for Indian maize in the East Asian market because of a unique thing that the Latin American maize does not come into the international market until July or so.

In Bihar, you could cultivate maize at yields which are comparable to yields in the U.S., for example. That maize can be shipped to Indonesia, Malaysia and so on by April or so for the poultry and feed sector. They then started investing. They got the seeds, et cetera and so on. The farmer converted from whatever they were growing into doing maize; they realized better prices. Of course, this is all linked to the international markets, so you are now also more subject to more volatility.

What is interesting to note, though, is that Cargill is not directly buying from farmers. Although they are involved in the ecosystem, the trade is through intermediaries. It is intermediaries who buy from farmers and then deliver it to Cargill, and then Cargill exports. Even if you look at investments in storage, for example, which is being done by other private players, there is a difference. If you look at the capacity of the processing mills in Bihar, it’s about 120, 130 tons a day, as compared to if you look at Andhra Pradesh, the mills are like 1,000 tons a day.

This aspect of the thing has nothing to do with laws; part of it is just investment. It could be electricity; it could just be the farm size distribution resulting in a certain structure of market.

On Landholding Size

RAJAGOPALAN: It just seems to be a whole bunch of things, which eventually all come down to small farms and low productivity. I want to touch upon that for a minute. So one of the things I have argued multiple times, because my work is mostly on the legal end and talking about property rights, is that in India, we have really, really small agricultural landholding. There is, of course, some variance across states, because different states have different land reform regulation, but also different states have different urban and rural land ceiling legislation that is there.

The average across India, I believe, the average size of the holding in the last agricultural census, was something like 1.08 hectares. This is about two and a half acres.

Now, do you think the root cause of the problem is solving the land problem, which will then, in a sense, fix everything else? Because there will be greater productivity, there will be fewer farmers, there will be greater investment and so on—or is this just one more thing, and it will not solve a host of other things?

CHATTERJEE: It’s a very, very good question. I think it goes back to two things. One is obviously the seminal 1954 paper by Sir Arthur Lewis, but also back to my profit equation, which is my income is P times Q minus the cost. And part of my income realization will—in the West, the increase in farm incomes has mostly come from increasing the Q and not the P, because when you tend to focus on increasing the P, we forget that increasing the P also hurts consumers, and a lot of those consumers are farmers themselves.

I might be a paddy farmer, but if I’m buying a cotton shirt, and if you raise the price of cotton, then that reduces my real income. Yes, solving or thinking about how we can deal with the landholding—the small landholding size issue in India is one of the key elements of understanding how we can improve incomes in the agricultural sector.

However, the flip side is that we also have a very large population. Then it partly relates to my work with Arvind Subramanian, where we talk about exports and so on. We really don’t have jobs anywhere else. The point is, if you think about people moving from agriculture to somewhere else, we have to also understand what that somewhere else is. That somewhere else cannot be unemployment.

Another friend and colleague of mine—Rohit Lamba, who’s at Penn State, and a PhD student of mine, Abhishek—we tried to look at the labor force data. Of course, a lot of people are just unemployed. The maximum employment, nonagricultural employment, that you see, about 10% is in manufacturing, productive manufacturing, but a bulk of it is construction, wholesale-retail trade, things like your local rediwala, a little bit teachers and so on. But both construction and the odd jobs that you’re doing in the service sector, these are not reliable—

RAJAGOPALAN: They’re seasonal.

CHATTERJEE: It’s not even seasonal, in the sense that it’s not the case that I will get it every season. I was a construction worker last year; this year COVID hit; I don’t have that job anymore. If you have those kinds of jobs, even if that gives you economic growth, that is not going to create structural transformation because, remember, the two things that are important for structural transformation are, one, in most models of structural transformation, you need agricultural productivity to be increasing. That’s one of the first ingredients for people to then move because then you have surplus labor. I can move outside agriculture without production falling.

Two, I have some reliable source of income in whichever sector I move to, and we are lacking on both aspects. I just don’t know how this—of course, I think we will never become the U.S., just in terms of the landholding distribution, just because there are just too many people, and where will they go, both in terms of jobs and in terms of locations? I mean, Delhi and Bombay are already overcrowded. So unless new cities come up with new centers for jobs, where will they go? Definitely, we have to think about how we can get out of this trap. Diego Restuccia, who is at Toronto, has been doing some work on this.

In Odisha, actually, one other way of trying to address this problem is, you may not want to buy or sell, but you may want to write rental contracts. A lot of sharecropping contracts and rental contracts tend to be on the gray margin between legal and being illegal in Odisha. We were surprised when we were doing these surveys, and initially these numbers were coming in, we got alarmingly low rates of sharecropping in Odisha.

And we were just surprised. In Bihar, you get 30%, 40% sharecroppers. In Odisha, how can there be just 2% sharecroppers? Then we did more research, and it turned out that it’s illegal. When you show up with a survey instrument, farmers were not willing to tell you that they were doing bhaga chasi, which is what they would call sharecropping.

RAJAGOPALAN: Yes, because it puts the property titles in jeopardy. That creates a different problem with the state.

CHATTERJEE: Right, even those kinds of—the first step is to even fix the rental market before you do the buying and selling.

Big Corporates in Agriculture and Current Farm Reforms

RAJAGOPALAN: I want to go back to one discussion we were having earlier about Cargill making huge investments in Bihar. Now, one of the really big fears—and here I think India seems to be quite polarized on the policy front. One huge fear is, there are going to be these huge players like Cargill and Ambanis who come in, and they are going to replace the APMC monopoly—if it were to exist in the state in the first place—and strong-arm farmers, almost like the East India Company, into giving away their produce for free. There’s this entire narrative around a fear of big players.

On the other extreme, there’s this huge narrative which says, “Hey, we need big players to come in. We need really large sellers. They are the ones who are going to make investments in cold storage. They are the ones who are going to make investments in warehousing. They are the ones who are going to move us from a tempo to a refrigerated truck,” and so on and so forth.

A small player, a small retail establishment intermediary, simply cannot make those investments. Where do you, yourself, lie in this completely polarized narrative? What has been the contribution of really large players? Would it actually be better for farmers, especially those who are in states like Bihar and Odisha, relatively poor states where most of the population is in farming?

CHATTERJEE: I want to say two conceptual things before we get into the details. The two conceptual points are, you’re exactly right that this whole debate with the new farm laws has become so polarized that the discussion that we should have had was to go clause by clause and understand what makes sense and what does not make sense for us and what is the ultimate objective of the laws. As far as where I stand, and that’s how I look at it, for me, there are parts of it that make sense. There are serious design flaws which we should think about just in terms of what the end objective is.

RAJAGOPALAN: Can you give an example of that?

CHATTERJEE: Think about dispute resolution. Right now, both in the contract farming law and in the free trade, dispute resolution rests with district magistrates. Imagine a small farmer who has been cheated shows up at the DM’s office, says “Find this random guy who cheated me.” I’m not saying that the current system is good. These are the things that you should have thought about and improved upon. Instead, you have gone and done something that just seems not enforceable to begin with.

This is how the new laws have been sold, that they are providing more options to the farmers. But are they really? In the sense that you think about market integration—let’s suppose that market integration was indeed the objective. Market integration is about homogenizing things in space. What we have done is actually market segmentation. You had the APMCs. You did not do anything with them. They continue to exist, and now you have created another trade area with a different set of rules. Earlier there was variation across states; now you have variation within states in how agricultural produce will get traded.

RAJAGOPALAN: Unless one of them ends up dominating and becomes like a Schelling point market, you’re basically going to get five thin markets instead of one thick one.

CHATTERJEE: Yes, and then we have to decide which is the way that we want to go. The third thing is, even if you and I agree on what the end is, the means to the end is as important. How do we get from where we are to what our end objective is? That seems to be the problem of what is happening in Punjab. To come back to the question about corporates, I think the first thing we have to ask is, where are we in that landscape?

The big corporates in some spaces do exist. So ITC is producing atta [wheat flour]. The Reliance Fresh stores exist in cities, even in my hometown now, in Lucknow. There are other cities where these stores do not exist, and the question is why. Is it because of overregulation or bad regulation, or is it because of . . . One other explanation, for example, would be that there isn’t enough demand for Reliance Fresh–type stores. I know friends whose parents just don’t want to go and buy tomatoes there. They prefer buying it from the local rediwala [street cart].

You ask them why, sometimes they will say that in the Reliance Fresh store we don’t get the quality of the tomato that we want. Sometimes they would say that I get a better price here. It’s more convenient. If those are reasons, then what we are trying to do is not the solution. First, there could be other reasons. So in Bihar, although there has been entry of the private sector in maize, they are not still buying directly from farmers, and there hasn’t been entry of the private sector in other commodities.

One very simple explanation is that there are contracting frictions in the sense that, in the U.S., if I had demand for X tons of maize, I would at most look at 10 farmers and that’s it, I am done. In India, I would have to do the same with 1,000 farmers, and that scales up my cost of just having to interact with so many small farmers. If I look at it from that perspective, then this vertical supply chain, where there is an aggregator who picks up stuff from farmers, and there are bigger aggregators who pick up some from—so that for the private farm, then they are dealing with one or two persons at the most. It is just an efficient response to the underlying realities of India, which is small farm sizes.

RAJAGOPALAN: Will that result in a better price, is the question. Will it actually improve farm incomes? I have two ways of thinking about it.

One is, of course, the farmers are still selling to the same 10 intermediaries, and the intermediaries are getting aggregated, and then Cargill deals with only one or two people at most. So that’s the feeding system. On the other hand, Cargill opens up the produce to a global market, right? Now there are people not just in Delhi and Lucknow bidding for the produce. Now it’s people in Indonesia, people in the United States—literally, the world is your oyster.

Now, if Cargill gets a better price globally than they would, does that get passed down to the farmer in Bihar or would it not?

CHATTERJEE: I don’t think there is an obvious answer to your question. There are two forces that act in opposite directions. Of course, this is assuming that the government of India keeps its trade policy not constant. When onion prices and cotton prices rise, if you put export bans, then that’s the end of it; you’re not getting any better prices.

Having said that, so suppose that that wasn’t a concern, then the case is whether the corporates have market power. In the event that they have market power, you will see less passed through, and how much gets passed on will depend on how large the market power is.

The second is, can you get rid of the intermediaries at all in the sense that are they not performing any economically valuable function? Let me give you a couple of examples to illustrate this. There’s a very nice paper by Dilip Mookherjee about the edible oils market in Bangladesh. It so happened that, I think early 2000s, I’m forgetting it, but the Supreme Court in Bangladesh banned intermediaries from the edible oil supply chains. They said, despite legislation, “You cannot have intermediaries. This is illegal.”

What happened—the thought process was exactly this—that they are making margins, and so if they go away, farmers will make more and things would improve. But the supply chain just collapsed.

The reason supply chains there collapsed was because everything, as I said, in India was working on credit and trust. The farmers were giving it to intermediaries who were then selling it to the processors. The processors trusted intermediaries, the farmers trusted intermediaries, and so they were solving an economic function. Now, doing it overnight meant that there was no substitute for this. You could not create alternative systems, and if you did not get alternative systems, things broke down.

What this also meant, by the way, was not only did farmers and processors lose out, consumers also lost out because, in general equilibrium, oil prices went up, and that hurt consumers. There’s a recent paper now by my friend at the LSE. She has looked at the markets in Kenya and the entry of agricultural companies, agricultural processing companies in Kenya, and how that affects farmer prices.

RAJAGOPALAN: Swati Dhingra, right?

CHATTERJEE: Right. What she has found is that in Kenya, what these companies did was, they kept the consumer prices where they were, and they pushed the margins down towards the farmers, creating a greater surplus for themselves. And so, net, they ended up hurting farmers.

Now, this is not to say again that this is exactly what will happen in India. It will depend where you are. It has been the case for the prawn farmers in Baleswar [in Odisha] or for the maize farmers in Bihar, you have seen increases in their incomes. Unfortunately, I don’t think there’s a clear answer. But in terms of regulation, therefore, what you need is a better Competition Commission.

If you see instances where companies are exerting market power, you have ways to curtail it. But where we are in our development path, we are far. If you know the amount of regulation on competition that exists in the United States, forget agriculture, just in manufacturing, we are far, far from it.

RAJAGOPALAN: A lot of the competition actually ends up getting hurt by just government policy. Simple things like bans on exports, bans on importing low-price inputs, really high-quality seeds, and other things like that. In some sense, one huge barrier to competition is, of course, people, those who have market power. But it’s also this really ad hoc policy of the state when it comes to agriculture, which is completely depending on the political climate, has nothing to do with farm incomes or some systematic policy towards farm productivity.

Insurance and Futures Markets

RAJAGOPALAN: Now, I want to ask you two more questions, which are ubiquitous when you see markets in the developed world, and they’re completely missing in India. One is insurance markets, and the other is futures markets. Now, both of them actually help smooth out a particular kind of volatility. Insurance, of course, can more directly reduce counterparty risk and other kinds of risk, weather-related and so on.

Futures markets actually are a great signaling device, in some sense. When they function well, then farmers can actually decide well in advance how much—the P times Q part of the equation. They get some signal from the market before they actually show up at the mandi.

And as you already pointed out, once you show up at the mandi with your produce, that’s what you have. You don’t have too many options beyond that point. What would it take to develop these two markets further in India? Is it legal, is it structural?

Now, this is early Feb., so we’ve just had a policy announced, an increase in FDI in insurance up to 74%, so there might be some movement there. But what about futures markets and insurance markets more generally?

CHATTERJEE: Insurance markets in agriculture is something that has puzzled me for a while. I don’t have answers, I have conjectures. But I would like to point you to a few facts, which is that insurance markets in agriculture, by the way, don’t work anywhere in the world. I was surprised. When I was thinking about crop insurance in India, I thought there are probably market failures, and you’re not being able to get the symmetry in information, people are reneging, I don’t know what is going on. Let’s look at the West and let’s look at how insurance markets there are doing.

It turns out, in the United States and in Europe, the government actually subsidizes insurance. It’s not like auto insurance markets, where the private players are just doing it by themselves. In fact, most of the bill of insurance is footed by governments in the West. When I started to think more about why that might be the case, that’s when I realized that maybe—and I don’t know whether this is the answer—it’s because food is one market where the government or all governments actively intervene to keep prices low.

If we think about insurance as just hedging between the good state and the bad state, if I am actively intervening to keep the good state low, then I will automatically not get emergence of insurance markets.

Having said that, there are various kinds of insurance that governments in the U.S. and in the West provide to their farmers. The yield insurance is the one that we understand, and even that is not fully successful in India. Then they also provide insurance against price risks of two types. One is just market fluctuation. So, your prices fell by three standard deviations, it’s an unexpected shock, you need to be insured for it.

The important thing is that this is not crop specific. In any crop, if you face a price shock, then you get insurance against policy risks. If the government follows countercyclical trade policy that has an impact on your prices, then they compensate you.

In India, what is going on is that all these policy objectives are being sorted by one instrument, which is the minimum support price when it works. In fact, the minimum support price is trying to do more, which is food security, price stabilization and increasing farmer incomes on top of insuring risks. If the government will actively intervene to keep prices low, insurance is something that they will have to provide, but it may not work for India.

I wrote an op-ed in the Hindustan Times where I argue that maybe the most sensible thing to do for us is to start with a base of income transfers, where you say that you get a base, sort of income support, and then you do pieces above it.

In the states of the world when there are extreme events, then you compensate for that specific loss, because otherwise developing insurance markets for India might just be too complicated given that we don’t have records of most things. I don’t know how large a farmer is. What about sharecroppers? What about renters? What about landless farmers? It might just be best to do something that works for India.

On futures markets, I’d say they are trying to develop it. It’s at like a nascent stage. For futures markets to work, again, one big constraint would be how do you solve the quality-assaying problem, which is why the whole eNAM thing backfired. I am a remote buyer, I want to buy something that is 1,000 kilometers away, I can’t see it. How do I know what I am bidding for? Right now somebody goes and looks at it physically—either I go myself, or my agent goes in the mandi and does it for me.

This is in space. Now we’re talking in time. How do I know that—of course, the futures market will tell you a price for a particular type of commodity. Again, the thing about small farmers will overlap with this and interact with it because it’s much easier for larger farmers to maintain a certain quality. For smaller farmers who are just producing for self-consumption and are only selling a little bit extra in the market, I don’t know how they’re going to participate in the futures market.

RAJAGOPALAN: Yes, the transaction costs are too high because futures markets don’t operate on such a small volume. They operate on very large volume, so it’s only the intermediaries who can potentially participate in the futures markets in some sense.

CHATTERJEE: I think that’s exactly right.

Minimum Support Prices

RAJAGOPALAN: India has relied, it seems, too much on minimum support prices in one sense. As you said, it tries to solve multiple problems, not just one of food security or farm incomes. It has also led to a lot of other problems. As you’ve pointed out, it has led to, among others, a host of environmental problems. Minimum support prices are propping up certain crops in certain geographical areas that ideally should not be grown in those geographical areas. They just require too much water. This is the example of paddy in Punjab and so on. What is the minimum support price policy in India, and what can be done to reform it if not the current farm bills?

CHATTERJEE: Again, I think it would be instructive to go back to the history on how this started to understand how we have come here. You will see the common theme that runs through is that policy should not be time invariant, but policy tends to be time invariant. What might have made sense 40 years ago does not make sense anymore. The first part I think is sort of well understood, but let me, for the sake of completeness, let me just say it. Pre–Green Revolution, when we were a food-scarce country, you really wanted a Green Revolution to happen in India. A Green Revolution, by the way, was investment in agriculture and was not a legal change. It was hard investment.

RAJAGOPALAN: It was a technological shock, in one sense.

CHATTERJEE: Right. That’s something to keep in mind. What came for the Green Revolution was that it was a risky technology. It was new. Farmers did not know whether it would pay off, but it required investments. You had to put pesticides, fertilizers, and you needed a stable supply of water. At that time, it just made sense that places that have deep aquifers, places where if you had two years of drought, you wouldn’t run out of groundwater, that’s where it made the most economic sense to introduce them.

Initially, to incentivize farmers to take up the Green Revolution technology, the seeds, the government brought in price support. They said that if you invest in this, in the event that prices fall too much, we are going to guarantee a minimum price.

That’s how the minimum support price came into being, but initially, the objective of minimum support price was only to keep the price from falling below a certain level. In other words, you were still not conflating it with food security. PDS and MSP were two different policies. Over time, of course, now there is inertia in this. You started with MSP, at some point then you started providing free power in Punjab, and then free or close to free power in other states. Then you get politically locked into these systems.

Remember, for farmers who are used to getting an MSP, this is like a salary. They are used to it. I have also argued, in a macro sense it might not make sense for the country, but for them it makes sense. Unless you provide them with an alternative, you don’t know what to switch to. What is well understood is that the MSP is provided mostly in paddy and wheat, and maybe some other crops.

Now, Odisha is also starting procurement a little bit in millet. How does the procurement activity itself happen, and who gets paid? That is at the heart of understanding also some of the protests. In Punjab and Haryana, the site of procurement is the mandi itself. What happens is, after I harvest my paddy or wheat, I take it to the mandi. Whom do I sell to? I don’t interact with a government official. I sell it to the Arthia or the commission agent. The commission agent then takes it to the miller or the processor.

The payment from the government either goes to the miller or to the Arthia, and then the Arthia or the miller pays me, the farmer, in Punjab or Haryana. Which is why, when there is any threat to this system, the system that I as a farmer have known for over 40 years, my immediate reaction is, what am I going to do? If these guys are going to disappear, how will I get this thing that I have got for 40 years? In other parts of the country it’s different. I think, first let’s take MP and then go to Bihar and Odisha.

MP as you know has emerged as the other big procurer of wheat in the country. Now, in terms of volume, it’s the largest, but in terms of coverage, it is not, because MP is just a much larger state than Punjab. Not every farmer in MP is getting it, but the coverage is good. Again, the distinction is, how is the procurement done, and who gets paid? In MP, the government procures at the village. The site of procurement is not the mandi, although there are strong mandis in MP. They set up procurement centers in the village, and the payment is made directly to farmers’ bank accounts. Even if there’s a threat to the mandi, they would not see it as a threat to the minimum support price system, just because of the way it is operating.

Then you look at places like Bihar and Odisha, where procurement is very weak. Less than 15% of the farmers are actually getting access. Less than 15% of the farmers who actually have marketable surplus—because a large fraction don’t have marketable surplus to begin with—they get access to sell to the government in Bihar principally through PACs, the Primary Cooperative Societies; in Odisha, through the regulated marketing committees. But then local politics comes into play. Because there is limited procurement, because not everything will be procured, there is a little bit of corruption.

The bigger farmers or the sarpanch or whosoever is in charge, first, they would delay procurement. In MP, post-harvest, you have somebody sitting to procure; in Punjab, procurement is almost up to you. Anytime you sell, you will be able to sell. Immediately after post-harvest in both Bihar and Odisha, you will have almost no procurement, or very limited procurement. That forces the smallest farmers, who are liquidity constrained anyway, to either sell to larger farmers or sell to intermediaries who might then eventually sell to the procurement agencies.

Sometimes they also put these lists. Outside procurement centers, they would say, “On day one, this person should come and sell. On day two, this person.” There are proper lists assigned. You are assigned a date as a farmer on which you can come and sell. This does not happen in Punjab. Punjab, you can come and sell any day. That list we have found in our research often is regressive in the sense that the larger farmers get first dibs. The moment I fulfill my quota of how much I wanted to procure, the smaller farmers are again getting left out of the net.

RAJAGOPALAN: Now, when it comes to the procurement in, say, Punjab and Haryana, as you said, it started with a particular historical legacy. It made sense at the time. Now it has reached a point where two, three things are happening. It is not particularly progressive in the sense that a lot of resources from taxpayers are being spent on farmers who tend to be relatively wealthy. Given the income and wealth levels of farmers across the country, Punjab and Haryana farmers are some of the richest farmers. That is one element of subsidization which I think one mustn’t lose track of.

The second aspect of MSP is it doesn’t induce or incentivize anyone to change course, and this means just huge amounts of environmental degradation. It takes many forms, of course: there’s a groundwater depletion by growing paddy in Punjab, there’s air pollution because of crop burning, which is not directly related to MSP but indirectly related to MSP. There is also just an unwillingness to look for alternatives. The way I think of the farmers’ protest from Punjab and Haryana is, of course, the government has put them in this situation in the first place, but we have a bit of a transitional gains trap.

CHATTERJEE: There are two main reasons that at least I have heard that the Punjab and Haryana farmers have complained about. One is that they think, although it is not mentioned in the bills themselves, that they think that either the MSP will go away or the government will start procuring at the farm gate, in which case again their local mandi is threatened. I gave you some reasons why that might cause anxieties.

The second is that they fear about this corporate power. They feel that once corporates come in, they are going to push down prices. If there are no mandis, where’s the anchor price going to come from? All those are, again, sound arguments. Empirically, what will play out, we don’t know.

RAJAGOPALAN: Inefficient as the system is, their anxiety comes from a real place.

CHATTERJEE: One hundred percent. Having said that, we agree that we have to move away from this for the sake of just the national welfare, environment, but more importantly, for the welfare of the farmers in Punjab themselves, because with deep aquifers, once you run out of groundwater, replenishment will take centuries. We are looking at desertification of Punjab on the close horizon. It is no longer a 50-year thing. It might happen sooner. There have been proposals, by the way. People have put forward proposals where you said that you start replacing the MSP.

Essentially, if you hear them argue, from whatever ground reporting I’ve seen—independent journalists, not media houses—what they want is some reassurance in income. The argument goes something like this: “I have X hectares of land. Can you tell me with whatever crop I grow, will I receive a stable income six months from now? Or is it a completely risky investment?” They would point to things like, look at farmers who cultivate fruits in UP. These are highly tradable commodities, but they come and tend to work as laborers on our farm because their incomes are highly volatile. How can you smooth that here? For that, alternatives have been suggested.

One is just replacing the minimum support price with an income support, but that replacement requires establishment of trust. What I had argued is that you start doing this in phases. You provide some income support to first assure that that works. Because we have promised farmers MSP, but the state does not procure grains from other parts of the country, the inference of that is that the state does not fulfill all promises that it makes. Now when I’m promising that I will go from MSP to an alternative system, how do I trust that that transition is going to go?

So it has to be done in phases. There it was like, how do you do the design thing? Then maybe you set the benchmark at the 75th percentile or the 50th percentile, where you would cover most of the farmers in terms of land sizes. Then you do an income transfer equaling to the income of that farmer so that farmers on the lower end of the distribution get more than what they are getting. Then the top few—and they would be very, very few farmers—they would lose out a little bit, but in the aggregate, it would be politically implementable because you would have a buy-in from most farmers. Again, the job is, how do you convince that this is actually going to work? In the current system, this was not proposed.

RAJAGOPALAN: And now a complete breakdown in trust anyway after the protest. One, it wasn’t proposed, and now it seems very difficult for buy-in on any side of the table for a proposal like this.

CHATTERJEE: Arvind Subramanian, when he was the CEA, there was a committee to propose a minimum support price for pulses. This was the year when the toor dal was selling at 100, 150 rupees a kg. So I contributed to that report, and we had computed a kind of MSP for pulses that incorporated the external benefits of growing pulses, i.e., the environmental benefits that you get. You not only help the environment, you use less water, but you also reduce dependence on urea, because it’s a legume.

Then to make it politically implemented, I think the suggestion was that you keep the MSP of paddy and wheat fixed in nominal terms so that inflation weeds it out. Then you keep increasing the MSP for the pulses, but then you actively go in and procure. You slowly start weeding out procurement of paddy and wheat, and you intensify procurement of pulses. That will automatically cause a shift. It is still a little bit micromanagement. You have to decide which policy you want to go with. There have been some policies that have been put forward. All of them are sensible.

Input Subsidies

RAJAGOPALAN: Now, I want to just move on to the last of the three big questions. This is on inputs. Of course, the input subsidies given to agriculture are just enormous. Leaving aside price support, there is the huge fertilizer subsidy. There is, of course, farm credit, which is giving a very small amount of crop insurance. One of the biggest subsidies probably is electricity, free or subsidized electricity. Irrigation. Now, another subsidy which has been coming up over the years is credit, loan subsidy. This is not a credit subsidy which is given in advance ex ante, but ex post loan waivers.

All these subsidies put together—of course, there are different sets of calculations, but a few years ago, again, from the economic survey, it’s about 2% to 2.5% of the GDP. It’s a fairly large input subsidy. One problem, of course, is, again, it tends to be regressive; the larger farmers get most of the subsidy. You know all these arguments. The other part of the input subsidy is, once again, almost all of them lead to environmental degradation.

They have also other unintended consequences. The entire power industry is in jeopardy in a sense. All your electricity, state utility units and vehicles are in jeopardy because of this free electricity being given and diverted towards agriculture. Then, of course, there’s the environmental question. If you get free electricity or subsidized diesel, you’re much more likely to run huge pumps and indiscriminately get groundwater out, and depleting groundwater tables and aquifers and things like that.

What is the way out of this subsidy entanglement that we’ve created? They are so large, they’re almost impossible to get rid of. They also affect Q in your equation, as well as cost. If your subsidies are both improving your quantities and they are reducing cost because they are subsidized, if we take those away, suddenly all farm incomes might collapse, especially for large farmers. What is a good way to think about reforming the subsidy culture?

CHATTERJEE: Let me just say two subtle points and then answer the bigger question. One is that the depletion of groundwater, using free power, there are two aspects to it. One is, of course, the power is free, but the other also is the power is irregular. If you made power regular, if you said, “We’re going to give it for two hours, but these are the two hours for which you will get power,” that itself has some adjustments. But over and above that, I think power should be priced.

You cannot have free power because then you’re not optimally adjusting for it. Then there are—even within fertilizer you are subsidizing, and too much, and you’re not doing anything with P and K, and that leads to overuse of nitrogen. If you’re overusing urea, then you’re also overusing pesticides, and there’s that entire—and water—there’s that entire cycle.

The other point that I wanted to make, but in the grand scheme of things, if we even combine all the subsidies, Indian agricultural subsidies are actually lower than agricultural subsidies in the West. Agriculture, I think, tends to just be a sector where there is a lot of subsidy. Part of it may have to do with politics; part of it may also have to do with poverty, in the sense that most of the poor are also in this sector.

To make things more efficient, I think the only sensible way is to slowly move towards a system where you don’t have too many of these small subsidies, but move towards something that is an income transfer. There’s one small glitch, and the small glitch is that supply chains everywhere are not complete. That’s something to be mindful of. That should not stop you from reform, but that’s something—this is the whole thing even with PDS: should I just give you income transfers and let you buy off the market? If the supply chains are not complete, especially if I am in a remote place, I might not get access to food, not because I don’t have money, but just because the . . . So, similarly with agricultural inputs, if there’s a remote region with few farmers, private companies might not find it profitable to supply stuff there. Then I’m not getting inputs, not because I don’t have money, just because the supply chains are not complete.

You might argue that in the larger scheme of things, maybe those people should not be farming in the first place, but those have welfare consequences. I think the only economically sensible thing is to slowly move towards just blanket income transfers irrespective of what you’re doing. Something on the lines of UBI [universal basic income], which are all good ideas, but it will be a phased transition where you do it slowly. And then you realize there are glitches, and then there won’t be a perfect solution, to begin with; you will have to move towards it. We will make small baby steps, we’ll realize there are certain imperfections, you fix them and then you do another, and gradually you move in that direction.

Intellectual Influence

RAJAGOPALAN: Now I want to move on to a little bit about you. Can you tell me a little bit about how you became an economist and how you became interested in agricultural economics in India and so on?

CHATTERJEE: Everything happened by accident. I grew up in Lucknow, in an average upper-middle-class family. When I was growing up, in high school, the buzz was engineering or medicine. For me it was, I’ve always liked math, so it was always like, you wanted to be an engineer, but I was terrible in chemistry.

It so happened that in school, in class one day, our teacher was asking everybody, like, “Who wants to do what?” Everybody basically said engineering, medicine. There was this one friend of mine who said, “I want to go to St. Stephen’s College.” I hadn’t heard about St. Stephen’s until that day, but that raised my curiosity, so I asked him what it is.

Then I found out a little bit more. Of course, I was in Lucknow. This was still when you did not have internet at home. You had to go to a cybercafe and Google and see. St. Stephen’s had a pretty bad website, but slowly one found out there was something interesting happening here.

Then I found that there was a senior of mine who was actually in math in Stephen’s, and I asked my schoolteachers to put me in contact with him. My parents had no idea about this. Then I asked him, he gave me the usual spiel. He said, “Well, you go do engineering, and then you do an MBA. In engineering, you do four years, and then you do an MBA. Here, you do the same thing in three years and you can do an MBA. Why would you want to spend four years?” And I said, “Okay, that sort of makes sense.”

In the engineering entrances, I did not get into IITs [Indian Institute of Technology]. I was admitted to slightly lower-ranked places. Because I was interested in Stephen’s—and you had to still then get a form, you could not apply for this online—so I had sent my uncle to actually go and get the form for admission. They were against universities because their imagination was, “Lucknow University, where you’re not studying all the time and other things are happening.” And when he came back and he brought the form, he said, “This is a fantastic place, it’s spick and span, there is no dirt, there is no paan lying around, it’s very clean, everybody’s properly dressed, and there are morning assemblies.”

They were immediately on board. I applied, then there was the interview. I still didn’t know what economics was. Accidentally, there were two forms: the first form I filled was for math, and the second form, just because you got it for free and it was extra, I filled it for economics. Then I got into both, and then was the decision, should I go into math or econ? I then went to this friend of mine because he was really my only source of information.

He said, “Well, in math, there are 15 courses and you get a pay package of 10 lakhs. In economics, there are 12 courses and you get a pay package of 15 lakhs.” This was such bad decision-making.

RAJAGOPALAN: So you did economics?

CHATTERJEE: That’s why I started with a bachelor’s in economics, but the whole—I remember, the imagination still is, I would be working for McKinsey or Bain and doing an MBA. But then when I graduated, that was the year after the financial crisis, so the jobs had completely dried up. I sat in a couple of interviews, and those interviews were all these consulting interviews, and I didn’t particularly like them, because they were saying, “How will you increase McDonald’s business by making them not give ketchup in packets, but in smaller sachets?” I was like, “Okay.”

At the same time, what happened—and this is where the luck part was very important—at St. Stephen’s they were notoriously known for not following the rules of the university in the sense that, if the university said you need to have written exams, they would make you write research papers. So in my third year, I wrote two very basic research papers, one on game theory, another on development. That’s the first time I got a taste of what original research is—that is, embark on a puzzle and you don’t have an answer. And I found the process fascinating, plus I hated these interviews.

That’s when I decided I want to do a masters, and Delhi School was across the street. I wrote the entrance and then I went to Delhi School. Delhi School is where I seriously started enjoying economics. I could see synergies with policy and stuff. Although, at that point I wanted to be a theorist, because economic theory is really the strong point of Delhi School. A game theory mechanism design, topology, those are the things that were stronger than . . .

RAJAGOPALAN: Did Sudhir Shah teach you topology?

CHATTERJEE: Sudhir did teach me topology. In fact, I did not do a single course in econometrics.

RAJAGOPALAN: Yes, he’s the gateway drug, right, for everybody who goes to D-School.

CHATTERJEE: That’s right. I hadn’t done econometrics until that point, and then I worked with two of my mentors, Mausumi Das and Tridip Ray, her husband, who used to be at ISI for a year as an RA. Actually, my now wife, both of us worked together, and then we got into graduate school. When I came to graduate school, that’s when I realized that everything I’ve learnt until now is probably like 10 years old. The frontier is just too far out. Then I met Angus Deaton, and that changed my life in many, many ways.

RAJAGOPALAN: This is at Princeton?

CHATTERJEE: This is at Princeton. The agriculture thing again happened by accident, because first, second year I was completely lost. In fact, I remember in my second year, I went to Angus’s office and I asked him, “What courses should I do?” He asked me, “What do you want to do?” I said, “Development economics.” He said, “But development is not a field.” Like, “What do you mean? You teach a course in development economics.” He said, “No, your core training should be in some core fields of economics, and then you can choose to answer questions in development with those tools.” He said, “You should definitely do some econometrics. You can take the development sequence if you want to.” And at that point, I had started liking international trade, so I did the trade sequence, but I wanted to write a classic international trade firms dissertation.

I had some idea which was not super impressive, I was working with it, and at that time I was again lucky and got the opportunity to work with Arvind [Subramanian], and this was his second year as the CEA. He was actually looking for somebody who would help him with trade, and there are just fewer Indians who take international trade—I think there are four or five of us in total. He hired me to do something in trade, and then, as it happened, he had to write the economic survey. And in his office there were these two whiteboards, and all the chapters were listed with names of people written against it as to who would help with what chapter. Chapter four was agriculture, and it was blank; nobody’s name was written beside it.

I asked him, “Why is this blank?” He said, “No one will write something about agriculture. There’s nobody whose primary interest is agriculture,” so we had to figure out something. I said, “Okay, why don’t I do it? I’m doing the trade thing, I’ll also do agriculture.” That’s the first time I read about agriculture in India and realized what the issues are. I think what worked in my favor was that I had already trained in international trade, and I could see how you could use many of the tools that I’d learned to think about stuff in agriculture.

That is also the time when I met Devesh [Kapur], and he started mentoring me a little bit, and in fact, at CASI, Mekhala [Krishnamurthy] had come to give a talk. And I went to CASI, I requested, “Devesh, can you set a meeting so that . . .” I wanted to talk to her about the potential idea of my dissertation, to ask whatever I am doing, will it even make sense? Because at that point, she was, amongst the young people, she was literally the only one working on agriculture.

I had these things, but I hadn’t done fieldwork and I wanted to learn more, so that’s when I met her, and then sort of that’s how our association began. Then, in the 2015–16 economic survey, I helped write the chapter, along with Devesh, on agriculture, and then, from that, we wrote our Six Puzzles paper. Then, Mekhala, Devesh and I, we were constantly talking, and we realized that this thing is so complex that not one discipline will be able to even—and there are so many open questions still that there has to be like a systematic thing.

RAJAGOPALAN: What is your writing process?

CHATTERJEE: I learned this when I was in high school. I had an English teacher who was, again, very kind and generous and has been one of my mentors. His name is Stephen Daniel Subba, and he is the first one who actually taught me writing. I keep jotting down my thoughts. Once I have a paper or an op-ed or a narrative, I first begin by just writing in short sentences what is the one big idea that I want to convey, and what should be my sequence of argument. As both my advisers stressed, the sequence of argument should be linear.

It helps to just write down 10 sentences, and then those 10 sentences become paragraphs. Other than that, I just sort of stick to short sentences, not using complicated words, the usual thing about not using adjectives and adverbs.

I try to write something every day. I did take the course by your friend, Amit Varma, to improve. But that is what I do.

RAJAGOPALAN: So did I. The final question, which is, what are you binge-watching during COVID?

CHATTERJEE: Binge-watching is usually during dinner. Both my wife and I, we are big Bollywood fans. Anything that comes out, and nowadays all the good content is coming on the OTT platforms, we don’t know for how long—I think they’re going to regulate it. The last thing I think we saw was Tandav, which we thought was very good. Initially, I did not realize what the controversy was about and only later figured out, “Oh, this is what was going on.”

RAJAGOPALAN: Thank you so much for doing this, Shoumitro. This was really, really such a pleasure.

CHATTERJEE: Thank you so much for having me.