Ritam Chaurey on Placing the Firm at the Center of India’s Structural Transformation

SHRUTI RAJAGOPALAN: Welcome to Ideas of India, where we examine the academic ideas that can propel India forward. My name is Shruti Rajagopalan, and I am a senior research fellow at the Mercatus Center at George Mason University.

Today my guest is Ritam Chaurey, who is an Assistant Professor at Johns Hopkins University’s School of Advanced International Studies. 

We talked about structural transformation in India over the last three decades, how firm level behavior responds to regulatory and fiscal changes, how firms choose between capital and labor, or permanent versus contractual labor, land use and factory location, 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.

Hi, Ritam. Welcome to the show. It’s a pleasure to have you here.

RITAM CHAUREY: Thanks, Shruti. Thank you for inviting me. I’ve been avoiding this for a while. 

RAJAGOPLAN: Yes. [chuckles]

CHAUREY: But thank you. Thank you for having me here. 

RAJAGOPALAN: Yes. So, to the viewers and listeners, Ritam is just literally across the river, maybe 20 minutes from where we’re recording. He’s been trying to avoid me, placing the river in between, and generally not giving us dates, but it’s really nice to have you here. 

The reason I wanted to speak with you is because I feel like whatever is happening in India right now, a lot of your research speaks very well to it. I’m talking specifically about the structural transformation that Indian policymakers are still trying to midwife. We’ve been in the process of doing this for about 20 years now, but also quite specifically trying to get manufacturing to become a larger part of the GDP or GVA—whichever way one thinks about it—and also, increase the employment aspect in manufacturing relative to agriculture. I feel like a lot of your research speaks specifically to that.

In particular, you look at firm behavior and how firms, on the margin, respond to particular regulatory changes, shocks or policy changes, sometimes differences between states, and so on. It’ll be a pleasure to get into all that with you, and then maybe we can broaden the conversation out to how that can help us guide policy.

The Industrial Disputes Act, Permanent Workers, and Contractors

CHAUREY: No, thank you. I think I’ve been interested in, broadly, firm behavior. I would even broaden it out to say, industrial development, since I was a grad student. Partially, the reason was that, at the end of the day, a vibrant private sector is how you generate jobs. At the end of the day, a lot of what we think of as development is getting people to work and then looking at changes over time. I’ve always been interested in private-sector development.

When I was starting out, I think a couple of things interested me as a grad student. I think this was still that phase when there were still ongoing discussions on whether labor regulations constrain Indian manufacturing. It’s not exactly the same time, but the Besley–Burgess paper was still being discussed, critiqued, whether the measures were correct, whether what they called proworker states or proemployer states, did it really line up? Once I was looking at those critiques, the overarching finding was that no matter how you looked at these things, there was something about labor regulations constraining manufacturing. In general, that was interesting. I think around that same time when I was a grad student, one thing that I want to give credit to MoSPI— the Ministry of Statistics and Programme Implementation, was that it was around the same time that they came out with the Annual Survey of Industries panel data.

RAJAGOPALAN: Yes.

CHAUREY: I think that’s the genesis—that there was a debate happening, along with, at the same time, data became available. It was an interesting time to explore questions on how labor regulations, et cetera, affect firm behavior. A lot of the work before that was at the broad, aggregate industry level, state level. I think at some level, I was lucky to be at the right place at the right time. The debate was there. The data was there. In that paper—if I were to just tell you briefly—the debate is that this Industrial Disputes Act, they have these onerous rules, especially if you’re large firms.

There are multiple rules that show up if you’re a 50-worker firm. If you’re above 100, I think you need to take government permission to fire workers—

RAJAGOPALAN: Or reassign workers and so on.

CHAUREY: Reassign workers, exactly. But I think the interesting quirk there was that it doesn’t affect contract workers. If you recall, I was working on these papers at the same time as there were all these discussions and controversies regarding that Maruti Udyog plant in Manesar. It had always been this issue with contract versus permanent workers. It was interesting to see at that time as well that Industrial Disputes Act is only related to permanent workers who are on the books.

RAJAGOPALAN: And there’s an additional quirk with the Industrial Disputes Act, right?

CHAUREY: Yes.

RAJAGOPALAN: Labor is a concurrent subject. You have the union law, which is the Industrial Disputes Act, and then states can have their own amendments.

CHAUREY: Right. 

RAJAGOPALAN: Which is basically the variation that Besley and Burgess use to either say that a particular state is “pro-worker,”, “pro-employer,” or “probusiness.”It’s not a very nuanced distinction, but we can go with it for the moment.

CHAUREY: Yes.

RAJAGOPALAN: As you said, most of the people who ran with Besley and Burgess, and this interesting variation between states that had amended Industrial Disputes Act, they were really looking at either aggregate employment or aggregate manufacturing activity, aggregate retrenchment, closures, and things like that. But what you managed to tease out in this 2015 JDE paper is, on the margin, it’s not that firms are not going to hire because they need to do what they need to do. So, what they do is they are more likely to hire contract workers if it’s a more pro-employee state, versus more permanent workers if it’s actually favoring sensible labor regulation, which is a little bit more profirm or probusiness. 

But why was that the choice on the margin? Is this because this is postliberalization, firms are booming, India seems to be growing at a pretty fast clip, so the question of aggregate employment is no longer that interesting, or some other reason?

CHAUREY: It’s a good question. I would add one thing that if we were to just look at aggregate employment, I often wonder whether contract workers get completely picked in those. I think that was one of the reasons. I think the other reason is that, which has also been explored in Chang-Tai Hsieh, Marianne Bertrand, and Nick Tsivanidis’s paper, that there was this 2001 Supreme Court law which completely changes the landscape in terms of this expansion of contract work. Just these contractors, there are such large firms came in who were supplying workers to these firms, including TeamLease.

I think a lot of these things were happening at that time. It’s a good question why—my sense around that time, at least looking at aggregate labor statistics, was that it’s possible that contract workers were not being really covered in that. 

RAJAGOPALAN: The other reason I find this entire literature interesting is, during license permit charge, there are so many constraints, right? From foreign exchange to capital to permissions to import license regime, foreign exchange license regime. But you don’t know which one is the binding constraint. Eventually, the Industrial Disputes Act is, in most cases, not the binding constraint because there are so many things before it. 

Now, once 1991 clears a bunch of that, it’s only in the first decade and decade and a half after that, you’re like, “Oh, those things are no longer the binding constraint.” Now, the Industrial Disputes Act starts acting as the binding constraint. It’s really, when you are looking at it, is really the first time we’ve had evidence for maybe seven to 10 years that contract workers are now the choice on the margin. Because before that, we didn’t even get to the labor stage. 

CHAUREY: Honestly, when I was first working on that paper, this was the interesting margin, even to the audience or the readers. I think it was also interesting at that point, apart from what was happening in India there, was that such data in general was not available also in other countries. This distinction, whether the firm tells you how many permanent workers they have, how many contract workers they have. There was really something there. A lot of my work after that has basically been an offshoot of the fact that firms always find a way. 

RAJAGOPALAN: Yes.

CHAUREY: Whether or not it’s optimal or not, but when firms are working in this second-best world, they will always find a way to adjust on the cheapest margin.

Impacts of Reliance on Contract Labor

RAJAGOPALAN: Yes. Now, the firm choice is very clear, right? You measure this in quite a sophisticated way. It’s not just quite as simple as choosing between permanent versus contractual labor. You actually look at if there’s a shock that leads to an increase in demand for the actual goods that the firm is selling, which in turn will lead to an increase in the demand for the inputs, one of which is labor, and in places—you measure this using agricultural shocks, rainfall, and things like that. Basically, on the margin, when there is an increase in demand, firms are going to produce more, so they hire more, and they hire more contract labor. 

Now, the standard version of this—so let me pick the most cartoonish version, right? The most cartoonish version is, “Firms are big, bad capitalists. Even these tiny firms that you’re talking about are big, bad capitalists. They just want to screw over labor, which means if they can, they’re going to hire them day to day on contract basis.” 

But there’s another version of this, which is, “Why would any firm that wants some continuity in its manufacturing actually rely on daily wage labor or these contract laborers?” Because it’s actually quite disruptive. You can’t invest in skilling them, you can’t invest in training them, and so on. 

So, what is the consequence of this choice on the margin that the firm is making? Because the firm is being more profitable, presumably, or at least it can survive in this terrible regulatory regime. But what’s the consequence of that, you think?

CHAUREY: That’s a good question. My journey with this paper has been fixated on this idea that you get these workers, it’s cheaper for the firm. It might, just because they are often not regular workers, they haven’t been there for that long, that it may reduce productivity. 

It may lead to—I don’t know. There’s a paper [Krueger and Mas] from earlier where they talk about these—I think it’s about a tire company. And there’s this time where there is a shock or a strike by the permanent workers, and then they bring in these contract workers or these part-time workers, and it leads to lots of defects in the tires. So, that was the model through which I was looking at this.

But if you look at some of [the] more recent work on contract workers, especially the paper that I was talking about with Chang Tai-Hsieh, Marianne Bertrand, and Nick’s paper, there they find that, at least for larger firms after this expansion in contract work, or the ability to hire contract workers, productivity has actually gone up. In fact—I hope I’m not misremembering—I also think that they find that these firms introduce new products. So, it’s interesting. 

Just also going back to this Maruti case that we were talking about. I remember from that case that the basic point there was the contract workers were saying that they’ve been working at the factory for a long time. They’ve been doing basically the same things. They just don’t get paid as much, and then they don’t get the benefits. 

So, I’m not sure. I have to say that it’s been hard in general to—we’ve tried it, but I think it’s been hard in general to figure out the productivity implications for firms that hire different amounts of contract workers.

RAJAGOPALAN: Yes,I can see it can go different ways, right? So, if you have a very large pool of permanent workers who are highly unionized, then they have no incentive to work. The contract workers may presumably want to come and prove themselves every day, which could give you a bump in productivity, but the lack of skilling and training or churn and retrenchment could actually drop productivity. On the other hand, it’s also flexibility, so you can hire more complementary labor. So, you’re right. It’s an empirical question. We don’t know which way this is going to land.

CHAUREY: But one interesting point on that is that I don’t know how much of that is specific to manufacturing in general. Because, for example, Prakhar Misra— he’s working on this project where he’s saying that the public sector is increasingly hiring more contract workers. 

RAJAGOPALAN: Yes.

CHAUREY: And one of the reasons that many bureaucrats give for this increased hiring—of course, outside of all the other reasons—is that they are more productive. It’s just that in the case of manufacturing; I’ve not wrapped my head around whether that’s definitely the case. But I can see for the public sector or for a service sector, that may be the case.

RAJAGOPALAN: Yes.The firm behavior in this case is crystal-clear. If I were teaching this in the classroom, the way I would think about it is, if a firm makes final goods which are in a highly elastic market, it’s going to be a lot more sensitive to the way it hires its inputs, which means it’s much more likely on the margin to switch from permanent workers to contract workers. 

Do you see patterns like that? Are firms that are exposed to export markets and therefore global prices much more elastic, or firms which work on very tiny margins, do you see more of that? Or is this just overall, the regulatory overload is just so costly and onerous, it doesn’t matter which market you’re in? You just have to make the switch if you can.

CHAUREY: My sense is that although I don’t exactly look at this margin in the paper, my sense is that contract workers are in—it seems like labor-intensive industries which are also the more highly elastic industries, are adjusting more on this margin compared to others. I feel like I’ve seen work on this by others, but that’s—

RAJAGOPALAN: Yes, but not so specific to India.

CHAUREY: Yes.

Labor Protections and Their Impact on Firm Behavior

RAJAGOPALAN: So now, this brings me to another paper of yours, which looks at employment protection and Social Security benefits and those kinds of benefits. As you know, the new labor codes have been passed. The Industrial Dispute Act, the infamous one that you studied, has actually been rolled into one of these four codes. The 50 and 100 has been expanded to—the threshold is now 300, I believefor Industrial Dispute [Act]–related protections. Now your paper may have to be redone in a decade or so, and we’ll see what happens.

Chaurey: Yes.

RAJAGOPALAN: There’s another interesting debate going on in India about introducing social protections for contract workers, especially gig workers. 

CHAUREY: Right.

RAJAGOPALAN: They’re not in manufacturing. They’re much more in services. The big outcry is, “Hey, gig workers are people too, or contract workers are people too, and we need to treat them well.” Now the problem is that the expense may not justify hiring them at the same levels. It’s a highly price-sensitive market, so actually the amount of hiring will go down, and this is literally being debated in the newspapers every day as we go. 

Now, you actually look at this very specifically in Andhra Pradesh in your paper with Gaurav Chiplunkar and Vidhya Soundararajan, and what you find is exactly that in Andhra Pradesh, right?

So, on the margin, the moment you start saying contract workers also need protection, basically firms, instead of hiring fewer workers or even fewer contract workers, just decide to become informal altogether so that they don’t fall in the gaze of the regulatory state at all to completely escape all of the costs. That’s an astounding result.

CHAUREY: Yes. I really liked the paper also, thanks to both my coauthors. I think what was interesting there—first, two things were interesting about that law in particular. At least before we started working on that paper, I hadn’t realized the kinds of pushback that were coming from labor unions about the Contract Labour Act. This idea that contract workers were working on core activities, which is exactly opposed to what the Contract Labour Act says.

There was an active debate on this. Of course, I was not—before this paper, I hadn’t really invested much of time on that. But I did realize that there was a lot of support for banning contract workers from unions. It was interesting, in that paper…The one thing that was interesting there is that there is, in general, how we think of informality. There is this great Reviewarticle by Gabriel Ulyssea on this. I guess the idea is that there are two margins of informality.

One is that you’re hiring these contract workers who are in the firm, but they are off the books. Or as a firm, you just decide that, “Oh, actually, I just will not register with anything. I’ll just become informal, totally.” I think the interesting part of this is this paper and this policy, and how it turned out was that at the same time we were able to look at both these margins, which had not previously been possible. Because, if you just look at firm-level data, as I had done for my JDE paper, then you can only look at that one margin which is whether you’re hiring permanent workers or contract workers.

RAJAGOPALAN: Yes. And the other people who studied this typically think of it as a capital–labor substitution, right? So, they think that there is going to be more fixed-capital investment or more machinery and plant and things like that as opposed to workers.

CHAUREY: Right.

RAJAGOPALAN: So, that, people have studied outside of India, but I haven’t seen this. You’re exactly right.

CHAUREY: AndI think, again, going back to how I was talking about the first paper, a lot of it has to do with India did have good data then on a variety of these margins. You could look at the Annual Survey of Industries, look at formal firm behavior on these two margins. What kind of workers they are hiring. But you also had the economic census and the NSS. Economic census is where you could just look at the universe of all kinds of firms in India, of nonagricultural firms. And there, it was interesting because one thing that the economic census used to ask is, “Have you registered with any government entity?” It helped us figure out this extensive margin of informality.  

Then NSS, of course, asks each worker, “What kind of contract are you working on?” It could be an informal contract, or you get paid weekly wages, so on and so forth. But I think one of the most interesting results there was that after this law change that you were not allowed to use contract workers at all. There was a ban on the use of contract workers in 2004. We see that very large firms, in fact larger than 50 workers… 

RAJAGOPALAN: Yes, that’s a very large firm in India, just FYI. [laughter]

CHAUREY: —are claiming that they are not registered anymore. 

RAJAGOPALAN: Yes.

CHAUREY: If you had asked me before I started the project if that’s something I would have expected, that’s clearly not. That was a surprise finding for us, especially firms of that size saying that they are no longer registered.

RAJAGOPALAN: So,I would have expected that new firms entering the market would reach the 49 threshold or something like that and then stop. I wouldn’t have expected previously registered firms to switch. You’re exactly right. That just means It’s costly at a very different margin because now, you’re losing all the benefits of formality.

CHAUREY: Yes. I actually want to add one thing on that, this 49-versus-50, 99-versus-100.

RAJAGOPALAN: Nine-versus-10.

CHAUREY: I’ll go back to the nine-versus-10 in a bit. [laughter] People have tried for a while thinking about whether if Industrial Disputes Act really is a binding constraint then we should find lots of firms saying they have 99 workers because at 100, there are all these regulations. Or there’ll be a lot of firms at 49 because at 50…And, years and years, many, many people have tried. They’ve not been able to find this.

RAJAGOPALAN: Yes, and I have a theory for that.

CHAUREY: OK.I’ll go back to your theory. I’ll first tell you that there is, of course, a paper that finds that there is bunching at nine.

RAJAGOPALAN: Yes, at nine. This is Amirapu and Gechter’s paper, right?

CHAUREY: So, that’s the only paper that finds some evidence of bunching— 

RAJAGOPALAN: Yes.

CHAUREY: —but I want to hear your thoughts on that.

Binding Constraints and the Rise of 'Invisible Workers'

RAJAGOPALAN: So, we have—labor regulation is not one or two things, right? We had about 44 union-level laws.

CHAUREY: Yes.

RAJAGOPALAN: If you start counting state-level laws, you’re talking about more than 200. Now they’ve compiled them into four codes or something like that. Now, the thing is, the Factories Act used to tell you everything from the color of the uniform to the paint on the walls to the font on the wage slip to my favorite regulation, which is urinal height and the number of spittoons. Spittoons are something you spit paan into, and things like that. The Factories Act was actually far more onerous when it came to the labor inspection regime for most small firms relative to the Industrial Disputes Act, which really kicked in much later. I feel like the Factories Act is the binding constraint, and in most states, Factories Act kicked in at 10 workers, and some states, I think like Rajasthan, had amended it to 20 workers.

They really find this. They do a regression discontinuity design, if I remember correctly, and it just falls off the cliff, right? At the 10th worker, you can see this huge difference, and in India, most firms are microfirms. I mean, even generally, anecdotally, you have pretty good reason to believe that starting with the 10th or 11th worker, most firms will—it’s not that firms aren’t getting bigger, it’s that they’ll keep them off the books.

So, it’s not even contract workers. They’re just invisible workers or someone who really doesn’t exist in any meaningful way in the firm’s books, and you have two sets of books for all accounting purposes. So, that’s my theory, that the Industrial Disputes Act is not the binding constraint for most firms; the Factories Act is.

CHAUREY: I see. No, there’s another story—although I completely believe their paper. There’s another story going on right now, which is about, there’s something to do with enumerators. The idea is that each time a new economic—last economic census was 2013. Until 2005, if there were more than 10 firms, then you— 

RAJAGOPALAN: Ten workers.

CHAUREY: Ten workers, sorry. More than 10 workers in a firm, then the enumerator who had been sent by the Ministry of Statistics and Programme Implementation had to fill up an additional set of questions. So, it’s also burdensome on the enumerator. [laughter] There’s a paper on this, interesting paper on the burden on the enumerator to fill out all these details and whether this can explain why there are more nine-person firms as opposed to 10-. If you look at 2013, that cutoff has now changed to, “You need to fill up an additional form if there are 12 workers.”

RAJAGOPALAN: Oh, so we can test for that now.

CHAUREY: We can test for that. I think they do that in the paper. In any case, I do believe that this Factories Act is a big constraint whether or not we see the bunching exactly happening there or not.

The Case for Larger Firms

RAJAGOPALAN: And we haven’t raised that limit when we collapsed it into the codes. All the attention was on Industrial Disputes Act. I still think the codes aren’t very good. In fact, what I think we should do is, we should increase the threshold at which any labor regulation kicks in to 1,000 workers or 5,000 workers.

The reason is that—OK, so, there is an ostensible reason and the real reason, right? So, the ostensible reason why all these laws were brought in is there is a big, bad capitalist. Imagine very much Karl Marx, Charles Dickens, capital is extremely powerful. Now your MSMEs and your 10-employee factory and your 50-employee factory is really not that capitalist. Right? [chuckles] This is actually a pretty broke guy or a broke family who are just trying to do something fairly basic, who are extremely vulnerable to stress and can go under anytime. This is not the capital–labor asymmetry that we’re seeing. If you really want to address the original reason for labor protection against capital, then it needs to be 1,000 workers or more. That’s one reason.

The other reason is exactly what you find in your work, which is, the reason this is happening is there’s a lack of skilling, and you’re working in very elastic markets for final consumer goods or even intermediate goods. So, there’s no way to square this unless you can give a break to the firm to make hiring labor cheaper. Just again, Econ 101, right?

CHAUREY: Yes.

RAJAGOPALAN: If you want more of an activity, make it cheaper.

CHAUREY: Make it cheaper. [laughter]

RAJAGOPALAN: So, if you want people to hire more and you want more employment, just make it cheaper. So, for these two reasons, I think just increase the threshold to 1,000 and then see what happens.

CHAUREY: Only one point that I will push back on about the 1,000 thing is Indian firms are just so small. [chuckles]

RAJAGOPALAN: Exactly.

CHAUREY: So, you’ll probably find a handful of firms above 1,000.

RAJAGOPALAN: But that depends on what the binding constraint is.

CHAUREY: Yes.

RAJAGOPALAN: If labor law was the binding constraint you’ll find more. But if something else is the binding constraint, I agree with you.

CHAUREY: I think Arvind Subramanian has a great paper on this, that, “How come just these large firms don’t exist in India?”

Yes, just going back to the 1,000 point, at that point, it’s basically that the government is cherry-picking the firms [chuckles] it wants to have those laws on.

Labor Supply Preferences in Structural Transformation

CHAUREY: In general, this is—again, I don’t know how much of it is my personal view, but one strand of literature that’s interesting in this regard, also in general about structural transformation, is this issue about labor supply. Most of the times when we talk about why structural transformation hasn’t happened in India, it’s been that, “Oh, firms are not hiring. We should remove regulations on firms.” 

There’s also now this growing strand of literature, a couple of papers just off the top of my mind. Most of them are by Supreet Kaur. There’s a paper by Suanna Oh— 

RAJAGOPALAN: The caste paper.

CHAUREY: The caste paper. It’s not clear people want to exactly work in factories. 

RAJAGOPALAN: Yes.

CHAUREY: There are these other kinds of norms. There’s another interesting paper in Ethiopia on similar lines. I think it’s a Chris Blattman paper where they randomly give people these vouchers to go work in a manufacturing plant. I think people just don’t want to do it. [chuckles] I don’t know how much of just the labor supply side is playing a role in structural transformation or the lack of it, where we want them to move from agriculture to manufacturing and then to services. People are like, “I don’t want to work there.” [chuckles]

RAJAGOPALAN: You know, yes and no. If this was a labor supply preference issue, it wouldn’t explain the variation across states in India.

CHAUREY: For sure, for sure.

RAJAGOPALAN: Why is Tamil Nadu able to become this highly industrialized state, and then Kerala, which is next to it, is like the labor can’t wait to become old enough to leave.

CHAUREY: Yes.

RAJAGOPALAN: Right? I agree with you that over a period of time, maybe the way Western countries developed, which was your post-world war, massive influx into manufacturing. That’s how the first generation broke out of poverty or stress, and then each successive generation moves into better jobs, may not be the story for India. But still, there’s too much variation within India. And I hate explanations of, “Biharis don’t want to work in manufacturing, whereas Tamilians do.” You know?

CHAUREY: Yes

RAJAGOPALAN: Those things always make me a little bit uncomfortable. And there is also that narrative going on in India. That would be my basic response. But of course, this is not empirically driven in any way. [chuckles]

CHAUREY: No, no, no, I completely agree. I was just saying that a lot of the times, when we are focused on structural transformation, it’s about the demand side. “There’s no labor demand. Factories are not generating enough jobs.” I just wonder if this is the entire story.

RAJAGOPALAN: No, fair enough. It may not be the entire story. But this is a single case of one firm like Apple.There’s so much discussion on female labor force participation in India and so on and so forth. But then magically, you give them dormitories, and they are safe and they’re right next to the factory, and guess what? You can find 20,000 women to work there from all over India. So, this can’t be, “Oh, Indian women can’t leave the house, and it’s about seclusion, and it’s about supply side.” There is something there. Now, it may be that there is certain stickiness for certain custom or certain preferences. But as economists, we know that just means the compensating differential in wagehas to be higher, which in India it isn’t.

CHAUREY: Yes.

RAJAGOPALAN: Right? That could be the source of stickiness. But I don’t know if I’m being too price theory about this.

CHAUREY: No, eighty percent of me thinks more like that. 

RAJAGOPALAN: Yes.

CHAUREY: There’s just this 20 percent where it’s like I find this literature examining labor supply stuff interesting.

RAJAGOPALAN: Yes.You have epistemic humility which is outside of your price-theory imperialism. [laughs]

CHAUREY: But, going back to your point about labor codes, two things that are interesting, I would say. One is this new thing about fixed-term employment. I want to say they are substantially different than contract workers, but I’m not sure.

At least how it looks on paper, they look different than contract workers being hired through an external contractor. I don’t know how much of a difference that’s going to make. The other thing I also found interesting, because you mentioned that 10 years later, we’ll have to revisit the paper. I think there are going to be versions of Bessley–Burgess in 10 years because different states haven’t yet ratified what they want to do with the labor codes. That’ll be interesting to see.

RAJAGOPALAN: Yes. The Besley–Burgess always—with labor codes, it will be more interesting because there are four labor codes. I think it’ll be interesting to look not just at the effect of Industrial Disputes [Act], because I do think it was not the only constraint, other than for much larger firms. In fact, the constant argument I have with some of my more communist-leaning friends in India is I’m like, “If you want a genuine labor movement, increase the threshold to 1,000.” It’s only when you have 400, 500 workers that they will even attempt to unionize.

Right now, we’ve killed the labor movement because we’ve killed organized labor of any kind, right? If you really want a strong, left-leaning labor movement in India, increase the factory size. Yes, I think that would be interesting, but I think they would test a lot more beyond Besley and Burgess.

CHAUREY: I hope so.

RAJAGOPALAN: I hope so, yes, the way it is set up.

CHAUREY: I must say, off of—everyone loves to criticize that—

RAJAGOPALAN: It’s very elegant.

CHAUREY: There’s some elegance to it. It’s like, no matter what you do, you find things are not working exactly in those states.

RAJAGOPALAN: And, you know, it was also the first time in India we started looking at state-level variation for a lot of things and Besley–Burgess was part of that movement. Right? If I look at Indian structural transformation papers, you start with everything is macro-aggregates and then India just looks like a disaster there.Then you start looking at state variation, and then you, of course, get the whole RCT revolution, which is looking at very granular, specific things which have no generalizability. And now, we are zooming out again, but I think the state level and the firm level is where I usually feel most at home and I find the most valuable insights. But that could just be a bias.

CHAUREY: No, I think I would be on the same side of the bias.

RAJAGOPALAN: Yes.

CHAUREY: Yes.

A ‘Bad Law,’ Musclemen, and the Evolution of Debt Law in India

RAJAGOPALAN: Yes, for structural transformation, I think that’s a useful place to look. 

There’s another. This is probably my favorite paper that you’ve written. This is in Journal of Law and Economics, and it’s looking at SARFAESI, which might be my favorite bad law. OK? 

Do you want to set up SARFAESI first for our listeners, because this is not something a normal person knows other than just check-bouncing?

CHAUREY: OK. [laughter] I think, let me say that before SARFAESI—let me go back even before, a lot before SARFAESI.

RAJAGOPALAN: Yes.

CHAUREY: If you took a loan, you’re a firm, you took a loan from a bank, you didn’t pay back, they had to take you to civil courts— 

RAJAGOPALAN: Which we know is very clogged.

CHAUREY: It’s clogged, and it’s very, very clogged because it has all kinds of cases, including check-bounce cases. There is an incredibly large number of check-bounce cases in India. Basically, nothing ever happened. You could basically keep delaying the case, and you didn’t pay back. The value of your assets would keep going down. The bank could not repossess the assets. Then comes ’91. Liberalization is happening.

RAJAGOPALAN: Bank credit explodes.

CHAUREY: Bank credit explodes. The government comes up with debt recovery tribunals, which, for a long time, were working very well because you could bypass the civil courts. You could go directly to DRTs. Basically, they were working for a while.

RAJAGOPALAN: Or at least working better than civil courts.

CHAUREY: Better than civil courts. As usual, like all good things, there was always this option of appealing, and that could just delay forever. That was one killer blow on DRTs. The fact that you could appeal and you could take it to the appellate tribunal, so on and so forth. This is going on in the late ’90s. Around late ’90s, early 2000s, I think there’s also the second Narasimhan report on this.

RAJAGOPALAN: Sorry, if I may interject, this creates a particular problem for banks because the moment a loan becomes nonrecoverable, they have to mark it differently in their balance sheet. But they simultaneously have to keep spending on legal fees to recover the asset, which has already been marked off their books, which creates all sorts of interesting problems for bank balance sheets, specifically. More generally, this is also nonbank financial firms and things like that.

CHAUREY: Totally. I think the listeners would benefit more from your explanation given your legal background on this. Then comes SARFAESI where it’s like, as a bank or as a creditor, I need to give the person that has taken a loan from me a 60-day demand notice. After that, I can bypass the courts. I can just show up and recover my assets. And it’s a Supreme Court law, so it’s going to overrule anything that the high court or any other court does. 

RAJAGOPALAN: What is the precise effect, I guess, would be useful? Is it just as simple as they change the default? So, before SARFAESI, if someone defaulted on a loan, the assets remained in the custody of the debtor or the defaulter until it was proven beyond any doubt in a court of law.  And if someone came to repossess, it was basically like they were sending goons, like the heavy musclemen to repossess something.

But from what I understand, SARFAESI basically switched the default rule in favor of the creditor, where now you could just repossess. You could send in the heavy musclemen, and you could repossess, and that was no longer a crime, essentially.

CHAUREY: The additional part was what I was talking about in the case of DRTs, you could appeal. Here under SARFAESI, if you wanted to appeal, you still had to deposit 75 percent of the loan. [chuckles] I guess the argument was that if I had that much money— 

RAJAGOPALAN: I would have just paid the loan back.

CHAUREY: I would just pay the loan back. In some sense, it was—

RAJAGOPALAN: Again, the debtor had to deposit the money. That’s right.

CHAUREY: Exactly, exactly. The debtor had to deposit that money. I guess that was the discussion around that time, basically that if they had the money, they would have repaid the loan.

If you just look at the law, it sounds great. It’s like, OK, banks can now recover their loans, NPAs will go down, now they’re going to make more loans. Firms are going to take on all this debt, India’s going to grow, so on and so forth.

RAJAGOPALAN: Yes, so, just to pause there for a second. You say it’s a great law. On paper, it always sounded a little bit draconian to me.  But I guess it’s great in terms of second best. We don’t have good loan recovery in India, but we were also a country which was really starved of formal credit. Right?

It’s great on a couple of points, which is one, we expect that banks will be more likely to give loans. Second, that there will be a selection among those who are taking on loans because they will not go to these formal institutions for very risky projects where they think they’re going to lose their shirt or something like that. So, there is automatically a better selection mechanism at play, so you would expect fewer defaults to come out of it because now they could just repossess with no other legal rules required. Then the third part of it is the expectation is once you can formalize a firm through credit, it will grow, and it’ll formalize in other margins. There will be all these other benefits that come from getting access to formal credit. Is there something else I’ve missed?

CHAUREY: No, I think this was, from a credit-supply perspective, that’s basically it.I think the discussions that I read about this; they were very optimistic about the credit-supply response.

RAJAGOPALAN: Can you tell us when you started working on this, what is it you thought you were going to test for? Then we can talk about what you find, which is really fun.

SARFAESI’s Unintended Consequences

CHAUREY: So, long story short, even before we started the paper, there was a wonderful paper already on SARFAESI. It was in the Journal of Finance. It’s by Vikrant Vig.

The paper, it makes the stark point that although the expectation was that credit supply would explode, banks would want to give out a lot of these loans. But Vikrant actually finds in his paper that overall secured credit actually goes down. That was very interesting because it came up with this idea that firms actually didn’t want these loans. [chuckles] Firms actually didn’t want to put up a collateral, knowing that if they defaulted, it could be repossessed by the bank just with a 60-day notice. There were also these discussions around that time that banks would have this—their powers would just go out of the roof in the sense that they could just declare that you have a fake NPA. Right?

I could just show up at your place and repossess it. If you read Vikrant’s paper, there are these footnotes where he describes these—there are these descriptions of firms actually telling the banker that, “Please don’t have my personal assets attached to it as a collateral.” I think, to me, it was a fascinating paper.

It was, to us, I think, with my coauthors, Shashwat and Vasudha, it was interesting because we wanted to see—what we wanted to explore is how does an increase in creditor rights affect input choices of firms?

RAJAGOPALAN: Yes. Just before we get to the input choices, there’s another interesting thing happening in the early 2000s, just if we zoom out to the macro.

In 2001, you had that famous WTO dispute resolution, which now says that it’s not just going to be intermediate and capital goods. It also includes consumer goods competition. So, you see this huge influx of global competition when it comes to consumer goods in India, right? 

Also, the landscape in which Indian firms are operating has suddenly changed quite a lot. Their regulatory environment hasn’t made them more competitive. But the global competition that they face, or even the local competition they face, suddenly is a lot higher. Which means this is not the sleepy socialist economy where nobody went out of business, which is useful to caveat in the decade of 2001 to 2011, I guess.

CHAUREY: No, no. I was actually going to come back to this point. I think between 2000 and 2008, maybe. Now, I forget how India did after 2008, but that seven, eight years, the manufacturing sector was doing quite well. They were quite competitive. All kinds of good things that you associate with a strong manufacturing country were present in India. This is actually in that backdrop that this— 

RAJAGOPALAN: Yes. No, but that tells you—when manufacturing is usually doing well, there are also more bankruptcies, and there are also more firms coming in and going out. There are also new products being launched, and more workers being hired and fired. There’s a lot of—under the headline of, “Manufacturing is picking up and growing well,” is all this turbulence going on, which will affect creditor rights and debtor rights, right?

CHAUREY: Absolutely. I think in that sense, I want to say that the timing was also off with SARFAESI.

RAJAGOPALAN: Yes. 

Chaurey: It was actually one of the most dynamic periods in Indian manufacturing. You come up with this law. I’ll come back to it. Let me first go through the inputs. The interesting part here was that when we looked at—we had Vikrant’s results that firms did not want to take secured debt, and their debt maturity had gone down. In fact, they were hoarding a lot of cash. It was interesting because, if you look at anywhere else in the world, as soon as there are stronger creditor rights, you invariably find that capital investment goes up. Then, whether or not labor is going to go up is going to be dependent on whether it’s complementary or it’s substitutes.

RAJAGOPALAN: Yes. What’s the underlying logic or mechanism for that?

CHAUREY: My understanding from a lot of the other literature is that you just take out more loans and buy machines. A lot of the work has happened from regulations in the ’90s or 2000s, so I guess capital was not always labor-substituting. [chuckles]

RAJAGOPALAN: OK, so basically, the assumption there was firms always wanted to buy more capital investment or heavy machinery. It was really the credit that was the constraint. Before the credit became the constraint and license permit charge, it was the permit and the foreign exchange which was the constraint. Now, all that went away, so now credit became the new constraint. 

And the moment you make creditors more comfortable about the fact that their loan is recoverable, that constraint is going to relax, and they’re going to give more. Is that the correct mechanism?

CHAUREY: That’s the correct mechanism. There’s also an interesting paper by Abhijit Banerjee, Duflo, and Shawn Cole, I think, in the Restat, where they find that even large Indian firms were extremely credit-constrained which is not how we think of large firms in general.

So, I guess that was the initial motivation that, for all practical purposes, Indian firms are credit-constrained. You have these creditor rights. Everywhere else in the world, when there is an improvement in creditor rights, firms take on more loans, they get in more fixed capital. Most of the times, you find that because workers are complementary to the machines that they buy, that there’s basically an expansion in firm size. We basically came to this research question with the same intuition or prior. Prior is a better word. We came in with this prior that let’s figure out—this seems a bit odd, given Vikrant’s results.

RAJAGOPALAN: Yes.

The Drop in Fixed-Capital Assets 

CHAUREY: Just to preface the results, we find that actually when there was an increase in creditor rights in India after SARFAESI, firms that had a lot of tangible assets before SARFAESI, in the sense that these assets are their existing machinery, their plant, their buildings.

RAJAGOPALAN: It could also be like the desk in their office televisions, transistors, all the things that can just be picked up by heavy musclemen and walked out.

CHAUREY: All the things that could be used as collateral to get a bank loan. Firms that used to have a lot of these collateralizable assets, there is a sharp decline in their fixed-capital investment after SARFAESI comes in. Interestingly, they start hiring a lot more workers, in some sense becoming more labor-intensive. It was very interesting, and we thought a lot about it. Then we explored that in the paper, that maybe what SARFAESI did was increase this threat of liquidation. It’s that the banks can just send some goons, [chuckles] and everything that could be collateralized is just taken away.

RAJAGOPALAN: When I was reading your paper, I didn’t get there directly. I got there eventually. I initially thought that the reason they do this is bank loans and bank recovery have now become a little bit more stringent. Labor allows you more flexibility, which means you can adjust your wage bill week to week, month to month. When there is turbulence, and when there is a chance of default, the reason they switch to labor is actually because they want the flexibility. 

But then you do a number of other tests to show that that’s not exactly the case. It’s actually the recoverability. So, how did you figure out that distinction?

CHAUREY: I want to add that I think that’s there in the background. I think that is happening. That’s the reason why that’s not the only story. Let me just add that it’s a story we cannot completely rule out, and it’s a story we don’t want to completely rule out. The one thing that we additionally do here is that we basically divide states into whether you have a strong judiciary or a weak judiciary.

RAJAGOPALAN: Yes.Basically, a judiciary full of delays and clogs or not.

The Impact of Judicial Delays on Labor-Capital Substitution

CHAUREY: Exactly. If you have a lot of delays per capita, then you are a state with weak judicial capacity. Then we find that, especially in states where the judicial capacity is weak, is where you find this substitution even more. It’s that firms are not intending to invest in capital and they are hiring more workers. And it seemed to us that that’s totally in line with this threat-of-liquidation story in the sense that banks, especially in states where they know that the judicial system will not help them recover, are going to additionally use these SARFAESI requirements and go after your capital.

RAJAGOPALAN: I thought you also ruled out the labor flexibility issue because you look at whether it’s a prolabor or a proemployer distinction, which is the same distinction you use in your other papers, right? Which is why I thought you had ruled it out. 

CHAUREY: We do that. I agree. I was going to come back to that point by saying that that’s the beauty of Besley and Burgess. Going back to the point, we do find there that in proworker states, after SARFAESI, firms, when they hire workers, are more likely to hire contract workers. And the opposite is true for proemployer states where firms are willing to hire permanent workers on their permanent payrolls. I do want to add that I still don’t think we were able to rule out completely the interesting and important channel that you raised, that firms are expecting these shocks. They know that they could default. They know that the wage bill of workers can be through just the revenues and just working capital.

RAJAGOPALAN: Also, you can default on wages. You can say, “We’ll give you wages two months later.” People don’t switch employers that quickly just because there’s a delay in one or two months. We see this even with very large firms. When the airline companies go bankrupt, pilots haven’t been paid for five, six months sometimes. You do see that you have some greater flexibility when it comes to defaulting payment to workers or delaying payment to workers.

Chaurey: Just going back, I think the Besley–Burgess [paper] sort of points to that. I don’t think we can completely—yes. Completely negate it.

RAJAGOPALAN: Do you say we can’t completely get away from it becausestates which are proworker or proemployer and also states which have slow courts?

CHAUREY: I think, yes. I think there would be a high correlation. I think there would be a high correlation. It’s something we don’t test in the paper, but I do think that there would be a very strong correlation between those two questions.

RAJAGOPALAN: Yes, because that would help us tease out this flexibility issue.

CHAUREY: One thing that we could have done was to test for both of those things at the same time. Whether states have a weak judiciary and whether it has this proemployer or proworker labor regulation. We don’t do that. That would have been the ideal test.

RAJAGOPALAN: No, you can do it 10 years later. Now, we have new labor codes.

CHAUREY: Yes. I agree.

RAJAGOPALAN: I think all your papers, [laughter] you can replicate in a 10-year period. 

No, I’m laughing. The reason I found this paper so interesting is, one, because I didn’t expect the result at all, and that’s and that’s usually the loveliest result. It’s one of those things where the second you say it, it becomes obvious, but until then, no one intuits that result.

But the second reason is the obvious policy implication of this is not, “Let’s have crazy stuff like SARFAESI on the books to increase employment,” because that’s not the actual mechanism. We actually needed to move toward a more sensible liquidation and bankruptcy system, which actually India eventually did. So, what do you make of the journey out of SARFAESI and into the Indian bankruptcy code and so on?

The Bankruptcy Code and Marginal Improvements

CHAUREY: I think the bankruptcy code was very much needed. There were multiple things in the SARFAESI that were questionable.  I think one very important thing that IBC did—I’m not entirely sure if SARFAESI had that—was this Section 29 that the promoters themselves can’t go and, [chuckles] again, buy the firms that they’ve been promoting now that the value of the firm has gone down. I am not entirely sure if this provision was there in SARFAESI.

RAJAGOPALAN: I don’t remember.

CHAUREY: Yes. I have a feeling that there was this blatant abuse where once the firm value went down, the original promoters would then go and— 

RAJAGOPALAN: Would swoop in. They’re basically shorting their own company by defaulting on their own debts, right?

CHAUREY: IBC has all these— again, it’s one of those things which, again, on paper, it is great. I am not sure. I hear that, again, the delays and all have been creeping up.

RAJAGOPALAN: No, but in India, it’s always marginal, right? It’s still faster than the courts and it’s faster than pretty much anything else we have going there. So, it may still be an overall improvement, even though it’s not what was promised.

CHAUREY: Yes. And I think there’s some very good work now looking at the IBC and the fact that I think it reduced misallocation of capital between firms. Which is that the right firm, the firm that was going to grow, got the loans. Not the one that had connections. For all practical purposes, it’s been going in the right direction whether it’s the perfect law or not. The thing with a lot of these second-best environments— and the reason I keep saying second-best is that there are so many distortions frictions, market failures in India that even a law that looks perfect how it’s going to impact, given all these other distortions in the market, it’s never clear.

RAJAGOPALAN: Yes.For me, the other learning from your paper is also that no matter how wonderful formal credit is and how much we want to pursue that as a policy goal if we do it without fixing other regulatory barriers and courts and contract enforcement and things like that, you’re going to end up in a very distorted firm space. In this particular instance, the distortion worked out well for us, but it may have meant a reduction in total-factor productivity or something else which was totally avoidable. We don’t know if the firms got more productive, whereas—

CHAUREY: We don’t know at all. In fact, it may be the case—

RAJAGOPALAN: It may be the opposite.

CHAUREY: Opposite, yes.

RAJAGOPALAN: And they did employ more labor, which is ideal and good. For me, that’s the bigger lesson coming out of the paper, which is the legal and regulatory regime is deeply entangled, and you can’t just pick at one or two things like that.

CHAUREY: I just want to mention here that when I was talking about IBC there’s great work by Manaswini Rao. Also on judicial capacity in general.

RAJAGOPALAN: Yes, yes. She has multiple papers on this, which are really excellent.

Spillover Benefits of Relaxing Land Use Regulation

RAJAGOPALAN: Now we’ve discussed labor. We’ve discussed capital. I guess the next obvious thing for me is to go to land. For the listeners, if you think labor and capital are badly regulated, I think the crown goes to land because it’s some of the worstregulation. Land has another peculiar quality, which I’ve been writing about. It’s not peculiar; it’s obvious. Labor can move. Capital can move. Land cannot move. The moment you have bad regulation in land, basically, you trap it in a way that labor migrates and capital flies. We talk about capital flight. When it comes to land, having even one really bad regulation which is binding on land use makes it very, very problematic. Then it traps all the other things around it.

You actually manage to measure that quite well in your paper on Karnataka. This is basically relaxing what is rezoning or what is land use. India has crazy land-use conversion laws. Here, I will refer to one of my own papers at Mercatus which is also written in coauthorship with one of your students, Ankit Bhatia, and Shreyas Narla, Kadambari Shah, and Ankita Dinkar. We look at 20 states which have the worst land-conversion laws.

But you look at something very specific in Karnataka where we’re not talking about relaxing all of the regulatory regime. You’re just talking about relaxing it in one space and saying what was previously agricultural land can now be zoned as industrial land. You find that that one change gives you surprising spillover benefits. One, can you walk us through that paper? Then I’ll have a bunch of other questions on what we can extrapolate from that paper.

CHAUREY: That paper has a history. My coauthors, Ram Fishman and David Blakeslee, they were working on a separate project in Karnataka on groundwater. It was part of their other project that they came up with. They had all these conversations with KIADB. Which is the Industrial Areas Development Board of Karnataka. They were surprised how the KIADB—they were having these conversations. It was basically that, “We think about SEZs and we think about all of these other place-based policies where you have all of these other incentives. You have to give infrastructure, you have to give tax benefits.” But they said, “We basically did this real estate promotion,” which was that they took contiguous parcels of agricultural land, rezoned it into industrial land, and then sold it at market rates. We basically look from, I think, ’97 until 2005 or something. Then we basically have the list of all of these industrial areas that are rezoned. 

We find two very interesting results. One is that just the fact that you converted this from agricultural land into industrial land without any of these other benefits, without giving infrastructure. You’re not even sure if there is a skilled workforce there. These were agricultural parcels of land, so most of the population was rural. Without any of these fiscal benefits, without any infrastructure benefit, just the fact that you rezoned, there was lots of large-firm entry into these zones.

RAJAGOPALAN: That’s the binding constraint, which is the way an economist thinks about that, right? [chuckles]

CHAUREY: Exactly. So, it was a very interesting finding that just this land—land has always been a binding constraint, but just the conversion of land would generate those effects. There, we also find, very interestingly, that the surrounding areas, those that are…

RAJAGOPALAN: Sorry, just to interrupt, the reason for this partly is if you don’t have the land conversion certificate, nothing else that’s supposed to come after that can be easily gotten. So, if you want to set up an industry, you need a bunch of permits, and you need a water connection, and you need to abide by the building code. You need to get an electricity line. You need to get all these other things in place.

But you can’t even begin that journey [laughter] without a land conversion certificate which is one potential reason why, for any firm which is a decent size and wants to actually set up [a] plant—and it’s not like a tenement, this becomes the binding constraint. Sorry, please continue.

CHAUREY: No, no, exactly. So, we find a lot of large-firm entry. Now I don’t remember the number, but there were lots of large firms coming in, creating jobs. But interestingly, we also find that the nearby areas which are still agricultural, because those have not been rezoned, there is a lot of economic activity in those regions, in those areas. Especially, lots of small firms show up. Lots of people move out of—or at least they claim they’ve moved out of agriculture. Now they have these small service-sector firms. These small agricultural firms. Most of them are very small like one- or two-person firms. I thought that that was one of the more interesting results. Of course, the fact that land conversion itself would lead to these huge gains was interesting.

RAJAGOPALAN: But it is also interesting because it tells you something about the location-based and the location-specific benefits. Even if it’s not a land-conversion story, you have various states that are trying to attract a data center or they’re trying to attract a mining company, or whatever it may be, to set up in a particular location. It’s actually useful to know that even if it’s in a cronyist way, if you can manage to relax constraints for one or two large firms, what could be the potential spillover to the surrounding area?

CHAUREY: For sure. We definitely saw evidence of a lot of spillovers in the nearby areas. We found it quite exciting. Whether or not spillovers always show up, I’m not sure the evidence is very clear on that.

RAJAGOPALAN: Yes. I think it depends on the kind of firm to begin with.

CHAUREY: Yes. It definitely depends on the kind of firm. It depends a lot on what there is in the nearby areas.

RAJAGOPALAN: Also, the kind of labor, right? 

CHAUREY: Yes.

RAJAGOPALAN: One of the big debates is if it’s in a tribal or a scheduled area, then you don’t have the complementary labor in the immediate area to actually even do small firms and so on. Or they may not receive the relevant capital that is required or something like that. There are a bunch of other things which are worth looking into.

CHAUREY: In fact, Ankit has a great paper on this. 

RAJAGOPALAN: Yes. 

CHAUREY: I love the scheduled area paper. I wanted to point that out. It’s true that those complementary investments matter a lot. I think often there are other constraints. I have a bunch of other papers looking at place-based policies. Even in my work, the broad conclusion that I’ve come to is that sometimes they work, sometimes they don’t work.  I don’t think ex ante you could predict whether they would work or not. [chuckles] There are so many of these different constraints, different settings. Sometimes people want to move to those places and take up those jobs. In other contexts, no one wants to move to hilly areas. And then it works very well for people who live there.

RAJAGOPALAN: And sometimes people don’t switch out of agriculture because they’re getting free electricity. And the moment they become a small industry; it is going to be metered. It could be one of the million reasons that India is as complicated as it is. But your first result still holds. You have two results in there.  One is basically firms swooping in the moment you rezone.

And the second is all the spillover benefits. But the fact that firms, large firms, formal firms immediately want to come in the moment you make the land bottleneck go away is a pretty good result in itself.

Location-Specific Subsidies and Barrier Relaxation

RAJAGOPALAN: So, now, if I look at your location-specific regulatory barriers, this is not regulatory barriers. It’s also fiscal barriers, right? You look at a paper where you’re talking about tax benefits or subsidies and so on. What is a good way to think about relaxing barriers or giving subsidies that are location-specific?

CHAUREY: This is a very, very interesting question, very important question. I think if I were to just do Economics 101, just look at—

RAJAGOPALAN: That’s what I do 99 percent of the time, by the way, and it holds us in good stead. [chuckles]

CHAUREY: If efficiency is the only thing and I just want to increase aggregate growth, then giving place-based benefits doesn’t make sense.

RAJAGOPALAN: Yes.

Chaurey: Right? What is the market failure that we are addressing? One could argue that if you bring in more firms to these areas that don’t have any firms, there’ll be these agglomeration effects, right?

RAJAGOPALAN: That’s about it.

CHAUREY: Lots of these existing firms will have these productivity spillovers. They will get these knowledge spillovers. They will learn about technology. Is that the only way one should think about it? There’s also this equity perspective, which is that if—

RAJAGOPALAN: So, if I may pause. Even before we get to the equity perspective, there’s also a question of how mobile inputs are, right? Because the 101 assumes that capital is completely mobile. It flies as if it has wings in bank accounts. Labor is extremely mobile. Labor is not just mobile. Labor is highly flexible in terms of the skills that it brings in, right? There’s no highly specific human capital. There’s no highly specific land-related land use and so on. I guess that’s useful to caveat when we step away.

CHAUREY: You are exactly right. A lot of those things are playing a huge role in how we think about the efficiency perspective. That if workers are mobile, if firms are mobile, then there is no point of having these place-based policies. Now, whether that’s always the case or not, whether that’s the case in the specific context in India versus somewhere else, I think that’s why empirical work in this helps.

A Case Study: Uttarakhand and Himachal

CHAUREY: At least, even in my research, even though I work on India about place-based policies—I find that in Uttarakhand and Himachal, for example, this was this 2003—it was called the hilly area scheme or something. I think it’s also around the same time that Uttarakhand had become a state. And there were all these fiscal benefits. If you were a firm, you set up a plant in Himachal or Uttarakhand, you get excise tax and corporate income tax exemptions for 10 years, I think.

RAJAGOPALAN: Which are nontrivial.

CHAUREY: They were nontrivial. You also get this capital investment subsidy.  The fiscal benefits were very, very large. In that particular setting, I do remember that I think one of the—I find huge effects there. I find—there are two reasons for these huge effects. 

RAJAGOPALAN: Tell us what the effects are.

CHAUREY: I find huge effects in—there were lots of firms coming in lots of employment generation, fixed-capital investment had gone up, and real wages of workers had gone up. In fact, during that project—I remember this was one of the papers when I was a grad student. I actually went to Baddi in Himachal because my adviser at that time, he said, “You got to go there and figure out whether it’s really working or these are just artifacts of data.” It really had a lot of factories. I remember Udham Singh Nagar had Tata Nano factory and there were lots of pharma factories in Baddi, in Himachal and a lot of pharma factories in Uttarakhand as well. I think Nestle, they were making ketchup there. It definitely felt dynamic.

RAJAGOPALAN: I remember this from the other side because I remember environmentalists going bananas. There were cases being filed, and there were so many issues in both Uttarakhand and Himachal. Which is how I even learned. I didn’t learn it through the tax and the subsidy side. I actually learned that there was industrial activity increasing because protesters were going nuts.

CHAUREY: In fact, that was when I was in Baddi, I think that was some people were talking about that. “Oh, these tax benefits, but it has completely destroyed the local economy. The environment is gone. The mountains were so clean.”

RAJAGOPALAN: Yes.

CHAUREY: I do remember those conversations. But I want to add one thing going back to our previous conversations that, again, for here, the way I learned about this policy was—basically there was this article in EPW then by Professor Bishwanath Goldar. He was just looking at state-level GDP, and the question was—he was just documenting that India was growing really fast between 2000 and 2008. India was very dynamic in manufacturing, and all the states were doing well. He was just basically questioning what explains this. [chuckles] I remember in that it was a commentary article in EPW, and he talks about—if you look at all the states there, Uttarakhand and Himachal were growing really fast.

RAJAGOPALAN: Well, the base was also very low.

CHAUREY: Very low. [laughter] I think he makes this point that, “Oh, it’s possible that some of these benefits may have something to do with it.” Then he goes on to discuss other things. But, that’s the genesis that India was, in general, doing very well in that case.

RAJAGOPALAN: And now we know that some of it was just cheap-capital-fueled. There were other things. There were global headwinds, which were the reason. This was all pre–Financial Crisis, which was when the world ended for most of us. There were a lot of things that were going on, right? [chuckles]

CHAUREY: Oh, for sure. Coming back to this Uttarakhand thing. There, interestingly—next, of course, the question is that you’d give these text benefits to—you’d have these place-based policies. You give tax incentives to Uttarakhand and Himachal. You want new firms to come in. Where are they going to come from? Are they closing down somewhere else and reopening here? Is someone just moving their capital from some other state here? In that instance, I looked a lot, or I tried to study spillovers. There, I don’t find any [of] these cross-state spillovers in that paper. Even my visit there broadly confirmed that these were new firms. These were not—

RAJAGOPALAN: They were real.

CHAUREY: They were real. I think that visit really helped, in case others are working on such policies.

RAJAGOPALAN: In general, I strongly recommend students, when you work on something, try to actually go and touch and feel it. You have better instincts than you imagine.

CHAUREY: I think that the fact that there were no spillovers—the one interesting—again, going back to agglomeration the way we think of agglomeration is that you bring in a new, big firm, and the older firms around that area are going to just become more productive. I don’t find any evidence of that.  I think the existing firms did not have any increase in productivity. All the productivity benefits that we saw there were because these new firms were coming in.

RAJAGOPALAN: Basically, the sense I got from that was these were already firms that knew what they were doing somewhere else. They were pulled in just by the fiscal benefit. They brought in all that technology, know-how. They even brought in their line manager or factory head or whatever it may be.

CHAUREY: I think a lot of that was going on because you saw a lot of these—Tata Nano was there. Nestle was there. Lots of these pharma companies were there. I think that’s definitely true. I did a cost-benefit analysis where I was looking at whether it made sense for the subsidy, for the government to subsidize—give all these corporate income tax exemptions. There, I think what I found was if I just added up the profits that were being made by these firms, and I looked at all these wage bill gains of workers, it was larger than the subsidy amount that the government had given. I think a large, big reason there was I don’t find in-migration into these states. Generally, if lots of people come in, then housing prices go up. You may get some nominal wage gains, but prices have gone up.

RAJAGOPALAN: It was mainly the locals working.

CHAUREY: Mainly, the locals were working. I think the firms, although they brought some managers, but the workers were local. I think that really helped Uttarakhand and Himachal, at least the residents there, in that sense. I think this fact that there was low labor mobility.

RAJAGOPALAN: They’re also small states where you don’t see too much labor mobility from within the state to these districts also, right?

CHAUREY: Exactly, exactly. Because there’s always this, “Oh, I am from the mountains. I’m from the valley.” There’s always this divide as well.

RAJAGOPALAN: It’s a huge divide.

CHAUREY: Yes.

The Impact of New Bank Branches in Underbanked Communities

RAJAGOPALAN:James C. Scott has written in enormous detail about hill people and why they are hill people and so on. What you may find in Kerala, Karnataka, Andhra Pradesh, and this quite easy movement in the plains south of India, you may not really find in the hills.

Now we’ve discussed spillovers. The other part of it is not exactly credit constraints, but also the broader monetary system, which is the last set of your papers, which I wanted to discuss if we still have a little bit of time. There are two big ones. One is on the banks and informality. How do I think about it? Which is when bank branches enter into areas where you previously didn’t have bank branches.

The reason I call it a monetary policy issue is this is very much coming from the Ministry of Finance and the RBI on how we do banking as a whole, which largely was state-owned banks then. A little less so today, but still largely a state-owned bank business. Then how that impacts firms’ productivity, firms’ formality, and so on. The second is, of course, demonetization, which we’ll get to in a minute. This is the last bit that I want to discuss before I let you go on the—I have a captive structural transformation expert, I feel, so I feel like I should make the most of it. [chuckles]

Can we start with the banks and informality? Let’s just start with, what is the prior, or a good way to think about what happens when a bank branch opens up in a particular place and then everything that follows.

CHAUREY: Right.I think the big prior for us was that you read a lot of “How to formalize,” “How should we increase formalization?” Large literature says neither the carrot nor the stick has worked. Despite saying that, one thing that is always hanging around there is that if you increase access to formal finance. So many firms are credit-constrained. 

RAJAGOPALAN: Things start happening.

CHAUREY: If they are informal, they are going to get registered because they want to access that finance. I think that was the starting point.

You think of a lot of papers by David McKenzie on this, on how to try and get informal firms to formalize. Lots of RCTs, and they failed. It’s really hard to get informal firms to formalize. We’re like, “OK, maybe we can look at this.” And this is a huge policy shock. The bank branches are coming in with this explicitgoal and mandate.

RAJAGOPALAN: Of priority-sector lending and things like that.

CHAUREY: Exactly, of priority-sector lending. This should motivate informal firms to formalize. We look at it from two different angles.

RAJAGOPALAN: Just to pause for a second, when you say it should incentivize informal firms to formalize, there are two aspects to it, right? One is a firm is informal because the cost of formality is just too high. They’ll go under if they actually file all that paperwork and comply, so they prefer to stay informal. The second way in which you can say that informal firms formalize is that now credit becomes that much cheaper. Let’s say they were going to their usurious moneylender or to their friends and family network. The interest rates were just so high that it never made sense, and actually the constraint of formality was more the interest-rate constraints or the usurious rates than it was this. What was your prior when you went into that? Was it that this was very much a credit-based story, or was it a regulation-based story of informality?

CHAUREY: Let’s see. I think our story here was similar to both the aspects that you’ve covered. I think, for us, it was that the ongoing benefits are going to exceed the ongoing costs. The ongoing benefits are that I don’t have to go to the moneylender. I’m going to get this cheap credit. It’s going to help me grow. Even if I have to comply with these onerous regulations the benefits from the cheaper credit are going to be—

RAJAGOPALAN: Yes.Are going to outweigh the regulatory costs. Perfect. That would be my intuition, too.

CHAUREY: If you read a lot of these financial inclusion documents from the World Bank, you’ll see that this is one of the big-picture ideas they have. You have financial inclusion. It’s going to help formalize. With my coauthors Prabhat and Rithu Devi, we look at this. We look at it from two different angles. First, we look at only small firms. Small firms are most likely to be informal. We look at household firms. We look at NSS data on household firms. 

Let me also give a little bit on the policy. The policy was that if your existing bank-branch-to-population ratio was less than the national average. Then you would get more of these bank branches. [laughter] Also, on this…

RAJAGOPALAN: No, I’m smiling because this is how we make policy, but it’s not the worst thing.

CHAUREY: It’s not the worst thing for writing papers.

RAJAGOPALAN: Exactly. It gives you a clean separation.

CHAUREY: First is that we see lots of entry of private banks in these really underbanked districts, which is interesting because RBI had this rule that if you’re ICICI, you want to open one branch in Bombay, then you have to also open a branch in…I don’t know if it’s one-to-one or if it was one to some other ratio. We see a lot of private-bank entry.

RAJAGOPALAN: The RBI has draconian powers to enforce this. Banks can’t easily duck this.

CHAUREY: You have to give an entire list of where you’re going to open your branches. Otherwise, you’re not going to get that approval for that Mumbai branch. In any case, we find a lot of private-bank entry. When we look at firm-level data, which is informal firms, we find that informal firms in these districts have higher fixed assets. They have more sales revenue compared to firms in districts that did not get these bank branches.

Surprisingly, they say no. “Do you have credit from the bank?” It’s basically you find no differences between firms in districts that have more of these bank branches versus those that don’t. 

That was surprising. It’s like, “OK, how come they’re not taking more loans?” OK, that’s one. Then if you ask them their formalization status, in fact, we find a very strange result that informal firms in districts that got these bank branches are less likely to formalize than firms in places that already had bank branches. We were surprised by this. We were puzzled by this. Then we look at household balance sheets. It’s a great thing about NSS of the past. 

CHAUREY: They used to have great datasets. There’s also this dataset called All-India Debt and Investment Survey. It’s a household survey where they ask you about your financial debt and assets.

RAJAGOPALAN: Also who you borrowed from and things like that.

CHAUREY: Not only who you borrowed from, but why for what purpose. There, we find a very interesting result that, first, you see that if I just look at household balance sheets, more households say that they have bank accounts. More households say that they have finances. More households say that they have loans from the bank. When you ask them the purpose of the loan, they say it’s for their nonfarm business.

It was interesting because it completed the story. It’s almost like it’s onerous for the firm to get a loan because you have to comply with all these regulations. But the household and the firm are the same thing. [chuckles]

RAJAGOPALAN: Yes, and the individual can get a loan a little bit more easily because the individual is more legible. And now with SARFAESI, potentially more compliant.

CHAUREY: Exactly. That’s what we find. We find that households are taking loans as households not as firms. The purpose of the loan is nonfarm business. They are most likely putting it in the business, and the businesses are growing.

RAJAGOPALAN: This gives you the very clean separation between the credit constraint and the regulatory constraint.

CHAUREY: Exactly.

RAJAGOPALAN: It actually does relax the credit constraint, but not up to a point to make the regulatory barrier worthwhile or overcoming it worthwhile, which tells you how onerous the regulatory barrier is. [chuckles]

CHAUREY: Totally. At least I find this paper very interesting. Yes.

RAJAGOPALAN: This paper is fantastic because it confirms—I like your papers, even the ones that don’t confirm my priors, but this really confirms my prior. You don’t measure this exactly, but you could measure the difference between what the moneylender is charging and what the bank is charging is effectively the cost of regulation and compliance for a small microenterprise in that district effectively. That number is nontrivial.

We know that is a few percentage points compounded over a period of time, which can be quite enormous.

CHAUREY: Yes. One hundred percent. Here, actually, I think we looked at loans from moneylenders. I remember that that didn’t seem to be different between districts that got banks versus not. Either you can say that no one is taking loans from moneylenders or that both kinds of households are taking loans at the same rate.

RAJAGOPALAN: Or they don’t say moneylenders. It could be friends and family networks where your money lender is effectively your third uncle or something or third cousin, right?

CHAUREY: That’s true.

RAJAGOPALAN: Then you don’t call them the moneylender effectively, which is how these things usually go in India.

CHAUREY: Yes.

The Demonetization Shock

RAJAGOPALAN: Now, the other interesting banking-sector shock is the demonetization shock. Demonetization is basically a liquidity shock in most cases. The expectation is there was this 50-, 60-, or at most 90-day period when there was this huge shock when it came to banknotes. For firms that are able to manage that liquidity shock, they’re able to survive. The firms that are vulnerable to that shock may either struggle or not survive and so on.

And the expectation is that firms that are mostly export-oriented firms are going to have not as bad a problem because their receipts are not going to be paid in these banknotes. And their receipts are going to be paid by people who are not impacted by the banknote shortage, effectively. What you find is actually it’s a pretty systemic impact. It’s basically everyone around an exporting firm, whether it’s people providing inputs or anything else, that actually impacts even an export firm’s ability to function because of this huge liquidity and banknote crunch. Is that a good way of thinking about it?

CHAUREY: That’s a good way of thinking. I think another way of thinking, which is you already pointed it out, I think that it’s hard to think of exporting firms as just pure exporting firms. They work and they are embedded in domestic supply chains where they are buying imports or, in fact, even selling a part of what they make to domestic firms, domestic customers, who themselves may have been reliant on cash.

RAJAGOPALAN: But it’s also their employees who may not be able to get a bus or their car gets repossessed or they can’t fill petrol that day because they don’t have cash.

CHAUREY: 100 percent.

RAJAGOPALAN: It could be something so trivial-sounding effectively, right?

CHAUREY: But it was interesting there to find that there was such an immediate drop in exports following demonetization, especially for firms that are more embedded in domestic supply chains. I think one other interesting result there, which goes back to the workers point, is that all this export decline is export-revenue decline. It could have been through prices or quantities. But the fact that the quantity of exports has gone down suggests that  were also worker issues.

RAJAGOPALAN: Yes.People may have gone back home because they can’t deal with the stress of what’s going on and so on.

CHAUREY: It was actually like production quantity had gone down. It was interesting because the prior, again, there was that exporters should be completely insulated from such a shock. First, they don’t deal in cash on the international side, but even on the domestic side, it’s very unlikely that they would be affected.

Still, there was lots of discussions on demonetization, but I think there was always a discussion that there would be problems in the domestic supply chain, whether or not permanent or temporary. Both the Gita Gopinath paper and even in our paper, we see that it comes back up, but the fact that it affected exports was unexpected.

RAJAGOPALAN: The way I think about this is when we think of the export market, we only think of global headwinds and we think about global prices or elasticities being the major constraint in how export firms operate, and tariffs and nontariff barriers, of course, being core to that. Now, the interesting thing is when we think of tariffs, tariffs are a tax on the foreign producer and the domestic consumer. Nontariff barriers are a tax on the foreign producer and the domestic consumer and the domestic producer but everything else we have, in terms of India’s regulatory crap, [laughter] is a tax only on the domestic producer and the domestic consumer. And demonetization is the mother of all crazy taxes on the domestic producer and the domestic consumer. It’s that crazy shock. That’s the way I interpret your paper that it’s obviously going to be a shock. It’s going to go down, it’s going to go back up, but there is a reason they’re not able to survive it even for a 90-day period without seeing those drops.

CHAUREY: Yes, yes, yes, 100 percent. But I think the other—

RAJAGOPALAN: Is that the correct mechanism?

CHAUREY: I think the mechanism, of course, is correct. The other thing is that it was also eye-opening to me how embedded exporting firms are on the domestic things that are happening in the domestic economy.

RAJAGOPALAN: Yes. COVID, look at the migration shock. You can rewrite this paper for COVID lockdown.

CHAUREY: I think some people have done that.

RAJAGOPALAN: OK. Because there, you have the demand-side shock and the supply-side shock, right?

CHAUREY: Yes. Here, in some sense, it was cleaner because there was no shock on the international side, so we could really only claim—

RAJAGOPALAN: No, and money is supposed to be neutral. [chuckles]

CHAUREY: Exactly. This money neutrality has played a big role in our thinking in this paper, that, “Is it neutral?” Also working on the project with Pravin and Ryan, the dataset was exciting. We got customs transactions data.

RAJAGOPALAN: Yes, that’s fantastic. Actually, you have very, very good customs transactions data, which I imagine will be popping up in other papers because it’s very hard to get that.

CHAUREY: Yes, yes. I’ve seen more and more people now beginning to use it. I had a call with a grad student who just recently got it. We’ll see more exciting papers now from India on customs transactions. We have lots of exciting papers on India, but, yes, I think a lot of the times we had to deal with papers from Belgium and Portugal to study those questions. Hopefully, we’ll have developing-country evidence.

RAJAGOPALAN: This was fascinating. Thank you for doing this. You have to come back and promise that you’ll come back because we’ve covered only about half your papers at this point. There’s another whole bunch of papers on nontradable services and H-1B visas and a bunch of things that we didn’t get to because of time, but hopefully you’ll come back and we’ll have another chat.

CHAUREY: I will, I will. I had a great time. I will definitely come back.

RAJAGOPALAN: I’m happy to know that. Thank you so much.