Barriers to industrial upgrading | Eric Verhoogen keynote and discussion

[music] >> We are honored to have with us Professor Martha Chen uh as the session chair. Professor Martha Chen is lecturer in public policy, Harvard University, and the chair UNU-WIDER advisory board. May I kindly invite Professor Chen to take the chair and guide the proceedings. Please join us on the stage. Good evening, everyone. Namaste. It's such a pleasure to be in Delhi. Um and it's my pleasure to welcome you now to this keynote session. We've had uh inaugural lecture. Now we have a keynote lecture of the 2026 WIDER development conference. As all of us know far too well, the topic of this conference has unfortunately become even more timely than we could have anticipated. It is taking place at a moment when development debates are being re-shaped not only by fragmentation in the global economy, rising uncertainty in trade and finance, and the growing urgency of climate and industrial transformation, but also by the renewed intensified conflict in the Middle East and its wider repercussions. That conflict is a stark reminder that a fractured world is not, as Kunal said, an abstract concept. It has immediate consequences for people's lives, stability, mobility, markets, jobs, energy, and the prospects for international cooperation. Across these pressures runs a central question. How can countries in the global south build productive capacity, upgrade industries, and generate inclusive growth under today's constraints? And that makes the evening's lecture especially timely. So, it's my pleasure to say that our keynote speaker is Professor Eric Verhoogen of Columbia University, Professor of Economics and International and Public Affairs, and co-director of the Center for Development Economics and Policy. Professor Verhoogen is widely recognized for his research on firms, trade, technology adoption, uh quality upgrading, and industrial development in low- and middle-income countries. This evening, he speaks on a question at the core of the conference agenda, the barriers to industrial upgrading. We are also fortunate to be joined by two distinguished discussants. Mr. Ajay Shankar is distinguished visiting fellow at the ISID, and formerly Industry Secretary to the government of India. He brings deep experience at the intersection of industrial policy, government, and structural transformation. And Dr. Zainab Usman is a senior research scholar at the center on global energy policy at Columbia University School of International and Public Affairs and managing director for international programs at the energy opportunity lab. Her work brings together insights on energy, development, trade, and the policy challenges facing emerging markets. The format this evening is straightforward. Professor Verhoogen will deliver his keynote, after which our discussants will offer reflections, and then we can then open the conversation to the audience. So, please welcome Professor Verhoogen. Yes. No damage done, I don't think. Please. Yes. Yes. Yes. Okay. Um Uh it's it's a great pleasure to be here. I'm honored uh to be speaking in this in this venue. Somewhat daunted, as Suman Berry was was also saying, to be following this distinguished uh group. Uh my topic, as uh as Marty said, is is upgrading. Uh upgrading my firms in developing countries and and the factors that that shape it. Uh what do I mean by upgrading? I mean innovative behavior by firms, but I I mean that in a in a broader sense than the word innovation is usually used in. Um so innovation is often understood to mean uh things that are new to the world. Uh and in fact, the whole field of innovation economics is essentially about um uh it's it's essentially about uh firms moving the world frontier forward. Uh it's relying heavily on R&D and patents as as empirical measures. Um for developing country firms that typically, although there are some exceptions, um but typically the the goal is to catch up to the world frontier, not to move the world frontier forward. Um measures of patent patents and R&D are are less meaningful and less informative in such contexts. Um and so I think we need to broaden the notion um and think about uh think about catch-up. Think about upgrading. I like the word upgrading very much because it it incorporates this form of this this form of catching up, which would be uh which typically would not be found in the in the patents and R&D measures and other measures of innovation that are that that that are typically used. Um now upgrading in the way I've defined it, you would think in terms of catching up, uh should be easier or adopting an existing technology should be easier than inventing a new one. Um in a not very politically correct uh or now not very politically correct turn of phrase, Alexander Gerschenkron, a famous historian, referred to this as an advantage of of backwardness. It's an advantage of being a late developer is that there are a lot of technologies and products that are out there uh on the shelf, as as Human Berry was mentioning, that should be it should be possible to to to to absorb. Um but for many developing country firms, um these advantages have remained elusive. I think I it's not controversial to say that. I I describe that as um as evident. Um and so my my motivating question or guiding question for this talk is, you know, what's what's getting in the way? What are what are the barriers? Um anytime you identify a barrier, you're implicitly identifying a causal factor that could be considered a driver as well. Just if anything if nothing else but then by removing the removing the barrier. So, I'm going to use barriers and drivers as in in alternative senses, two sides of the of the of the same coin. Um now, early development economists, these questions of industrial development and I would say upgrading um were were central uh to the to development of the field. Um when Gunnar Myrdal was cited was cited earlier, when you go back to the early development economists, this is certainly something you were thinking about front of mind. But I think for for many years, at least in my sort of research career, um uh let's say the last, let's say, 20-30 years, with the rise of empirical development economics, um these issues have been a little bit on the on the periphery. Uh there we can interesting question about why why that was, but um there hasn't been so much focus or there hadn't been so much focus until quite recently on on on firms. Uh partly that's because data on firms is is hard to hard to access. Um the experimental revolution, for which a randomista would be the turn of phrase that's sort of used internally, uh for which Banerjee, Duflo, and and Kremer won the Nobel Prize, um mainly focused not on firms but on individuals, in part because it was hard to run experiments. And it's it continues to be hard to run experiments on firms. Um and so, uh as people explored this new methodology of randomized control trials and RCTs, uh these questions of industrial development got got got pushed a bit a bit to the side. Um so, happily, I'd say uh the situation's been changing. There's there's more and more research on on firms, and I would call you know, well identified in an econometric sense. So, worrying about causality and and trying to carefully measure, you know, progress uh towards upgrading, um there's that that's been a an emerging area. I now say think uh it's recognized subfield within within development, but I sometimes called refer firms in development or I like the phrase industrial development. So I'm very excited that that you know ISID has that has that in his name and is is the is the host. Um in any case, so so what I'm going to be doing part here is giving you a a bit of a progress report on on this on on this field on micro empirical research on on how firms actually behave. Typically with with some attention to causality. Would not necessarily only experiments, but but with some experiments. Okay. So I discussed there there are lots of important methodological issues about conceptually how do we think of upgrading, also how do we measure progress in upgrading. I think they're important issues. They're not very accessible to a general audience. And so what I'm going to going to do instead of that, I'm going to focus on what I consider some concrete substantive insights that's emerged that are emerging from this literature in four areas. In organizations, around customers, and it'll become clear what I what I mean by that, social learning, and competition. You can follow the bubbles along the top as as as we go. Okay. One insight that I think has emerged from this literature is that management practices are very important. This has been a very a very influential highly cited literature. The probably the leading paper in this in this literature is this paper by Bloom, Eifert, Mahajan, McKenzie, and Roberts. An experiment on on consulting and management practices in among 17 Mumbai area textile firms where they randomly allocated consulting from an internationally known consultant consultancy. I won't say it say the name, but it is one that you would you would all recognize. Where they did for all firms, so it's 17 is not a large number. There were 11 treated, six six non-treated. We can discuss that in the question and answer if you like about how they deal with the with the the small numbers. But the what they did is for for 17 firms got one month of a diagnostic phase of consultancy for for and then the treated firms got four months of intensive consulting. Um, the market value of which was around $250,000. That's not what the researchers paid, but that would what the firms would have paid. Um, and they tracked a long list of management practices, which are they often referred to as as modern management practices. They're management practices that this consultancy often recommends to uh to firms, including regular maintenance of machinery, tracking of inventory, uh displays, daily displays of the of the productivity of weavers, um and of each individual loom. Uh, and then uh another practice that's tracked is whether they have performance-based um pay systems. Um, the paper in part became famous for the for for photos that that appear in the paper about uh so so many there are some firms that appear not have have have adopted fairly basic, you know, management practice. So, this is this is a picture of a of the of the inventory warehouse or for holding holding uh uh holding thread and yarn uh that in principle which should be used for for textiles, but was clearly a mess. All right, so that's uh you know, it seems like they had not optimized their inventory management system in this uh in this in this in this firm. Um, what they found is that the consulting uh had had positive effects. When they track I don't know if I have a if I if I have a No, it's not going to be I don't have a pointer. Okay, but here on the on the Y axis is the share of these 38 management practices that the firms had adopted and then just time, months after the diagnostic phase. Um, so that you could see the the first four months are where the where the consultants are in in um are are there in the in the in the factories. The treatment plants and then the control plants and then they also have the tracked adoption in plants that were uh in the treated firms, but not the treated plants, right? Not where the consultants actually were, but were in other plants within the same firm. And you can see that indeed the share of these modern management practices uh went up, and including the for the control plants especially uh they they adopted some after after this diagnostic phase. They also had consulting for a month. But the treatment there's a clear separation and those treatment plants that got the four months of intensive consulting uh adopted adopted more of these practices. Um They're they're they track a number of different outcomes. One key one is quality defects. And so you can see there that the the quality defects index um declined, which is a good thing for the treatment plants relative to the control plants. And so part of what the consultants were were explaining was that rather than uh just trying to produce as much uh you know cloth as possible and then going back and reworking the quality defects, the quality defects had to be identified early on and corrected. And so in in the end that's ends up saving saving a lot of time because the quality control is integrated throughout the the manufacturing process. Okay. So I think um this has been a rightfully influential uh study showing that management that consulting can can be helpful and also that management practices are important. I I I do want to share a couple of quick words of caution. It's been interpreted as as I as identifying that uh for uh showing both that these management practices are a good thing and that firms are making mistakes by by not adopting them. And those two conclusions I think are a little bit need need some qualification. Um so so one is that the best practices may depend on the setting. So some of these practices that they identify like labeling inventory uh do seem like I would what I would call a no-brainer. They do seem like uh you know, clearly everyone should be ident- to be doing inventory better than that factory we saw we saw the photo of. Uh on the other hand, performance pay for instance, it's less clear that in every setting that a performance pay is a good thing. And I'm going to give you an example next coming next of a of an of a setting where performance pay is it's has a ambiguous uh or it's not clear whether uh it's a good thing or not. Um a second word of caution is that the consulting may have may have had an effect on outcomes not through uh the effect of outcomes not through the adoption of the 38 practices, but a direct effect. So having those consultants in the factory for four months had a direct effect on productivity or on on quality uh that may not like this is I guess an econometric point that we shouldn't necessarily identify that effect with the practices themselves. And so we need it's a word of caution caution there. And then also just that learning how to adopt new practices can be costly in itself if firms lack that know-how. And so they're not necessarily making mistakes by not adopting them if they you know for a firm to spend $250,000 in order to get the calculation in the paper was profits of $300,000 per year noisily measured. It's not clear that they're making a mistake by not making that investment you know early on. Okay. So anyway, but one thing key point here is just that management practices seem to matter and that consulting or intensive information sharing with firms can be can be helpful. Um I would Uh more broadly I would think of management practices also as a technology choice or the choice of management practices as a technology choice. Um I'm going to talk now about a a separate experiment which was one of my own um with with co-authors also thinking about a a technology where here it's actual physical technology which I'll show you about. It's about for for cutting um of artificial leather for for soccer balls or I I guess I should say footballs um depending on the on the on the I'm trying to be more international here uh in in in Sialkot Pakistan with with a surprising result. So let me tell you a little bit about this. Uh the key input into soccer balls is Rexene. It's artificial leather. It's about half of the cost is this the which is the exterior of the ball. Um the standard design I'm going to show it here has hexagons and pentagons. Um so this guy and I if we can let's see we're going to test the IT capability here. We're going to show Yes, thank you. That was very seamless so congratulations to the IT team. There's you can see they're going quite quickly. Um and this these guys are the most um experienced workers in the in the factory. You know often have been there for 15 or or 20 years. Let's show this one more more the the the video so they can do that pause. Okay. Back to the back to the slides. Yes, so that's the key step. Now, uh he was cutting hexagons there. You can see the hexagon die. Um and then the the ball the the standard design of the ball also has pentagons. And you can see that the pentagon die there is uh they imitate the hexagon die in the sense that the the sides are flush. There's a full full flush side. And here's I'm going to show you now a picture of the That's a a sheet of Rexine after hexagons have been cut out of it. Um hexagons tessellate. You can tile a floor with with with hexagons and that those cutters are pretty good at minimizing the waste, which is the most expensive uh by far the most expensive input. Uh pentagons are a harder problem. You can't tile a floor with with with pentagons or not regular pentagons. Um and you can see that there were not only the die was imitating the hexagon die, but the cutting pattern was imitating the hexagon cutting pattern. And about 20 or 24% of the of the Rexine was being wasted, right? As opposed to just 8% of the of the hexagons. Okay. So, it turns out that that cutting pattern is not the optimal cutting pattern. Um I was There's a longer story, but basically I was watching a video of a Chinese factory that was producing the 2010 World Cup ball, the Jabulani ball, and I noticed that pentagons were being cut in a different pattern than the pattern that was being used in Sialkot. Um Uh turns out mathematicians were aware of this. There's a whole branch of mathematics of which I was ignorant uh on packing problems. So, they were aware of a packing of a a superior packing, which you can see and and also Wikipedia. I could have I could have Googled just Wikipedia at the time. Pentagon is still there. I believe this design. You can see that the the vertex of the midpoint of the base, vertex of the midpoint of the base, and then flip it around 180°, and so you can get more pentagons per sheet. Um Uh many years ago I had the foresight to marry an architect. Uh so, my my wife and I sat down after having seen this video, and we just knew the size of the pentagons, we knew the size of the sheets of Rexine. We figured out that essentially using that optimal design of pentagons, you can get about 8% more pentagons per sheet and that became the basis for the new technology. There's a picture there of the of a pentagon die. You can see it's offset. It's not sharing a full side, it's sharing just half of a side. And with that die, you can get, you know, 8% more pentagons per sheet, which works out to be if you can see it there I don't know in the corner, it's about 1% reduction in cost. That's That's not an enormous reduction in cost. On the other hand, margins are very tight in this industry and profit margins are about 8%, so you would expect them to to to pay attention to a 1% reduction in cost. And in fact, we thought, you know, when we're getting ready to do the experiment, we were being very secretive about our pentagon technology and we thought, okay, it's going to spread so fast, we don't want anybody to know. And um Okay, so we gave it out in the short you know, quick summary. Gave it out, almost nobody adopted. Uh We gave it out to of the 135 firms we in town that we that we found, we gave it out to 35. 15 months later, a grand total of six had adopted this technology, which is there's a theorem, right, saying that there's more you can get more pentagons per sheet and actually you don't have to change anything else about the production or almost nothing else about the production process in order to use this technology. So that was quite surprising. Um we there was there was one big firm, again, this is for Q&A, we can talk more, one big firm with 2,000 or more employees who had adopted fully this technology. And so we were quite confident that the technology was actually working, but there was we were puzzled about why why it hadn't spread. We asked owners why not and what they told us was employee resistance. Their their workers don't like it. Okay? And in retrospect, it turned out it was clear because the those guys who've been doing this job for 20 years are paid piece rates and they just want to go quickly. They had no incentive to reduce waste, they had no incentive to be green. We said this is a green technology reducing waste, but they had no incentive to to reduce waste. And so there, if they were to adopt, this would slow them down. Right? We acknowledge that the it would slow them down at least Can I say? Uh they were worried they'd be slowed down forever. It turns out that they they got up to speed in about a month, but they didn't know that. So, they were worried their their their incomes, if there were no other change, would be permanently lower. And so, quite, you know, reasonably and rationally, they told the owners it was a terrible technology. Uh and that they shouldn't they they shouldn't adopt it. Um and so, then we ran a second experiment where we gave a bonus of a one-month salary. It's not very much, $150 for most of these cutters, which is you know, sort of the elite of the soccer ball workers, but it but it's still not very much money. And the the uh if they pass the test in the presence of the owner to that to show that the the cutting technology worked and could be could get in reasonably, uh the the workers were quite excited by that uh that experiment, and they passed the test, uh and we found a statistically significant share of of uh of of uh firms adopted the technology as a result of running this running this experiment. Okay. So, what are the What are some of the takeaways? One, the way we say it is the misalignment of incentives within firms can impede information flow. So, organizational frictions, organizational conflict can get in the way. That's a barrier, getting in the way of technology adoption. Um Another one is that inertia in labor contracts may hinder technological change. So, in a technologically stable environment, which this was in Sialkot for 20 years, you know, maybe piece rates were the optimal thing, but then once there's a new idea in town, you may be It's hard to change away from the from the contracts that you had. Hard to change away from piece rates. So, there's there's inertia in that in that way. Um I think an important point, back to the observation that sometimes is coming out of the management literature, people are just making mistakes or the firms are just making mistakes, is that uh you know, we here we have a story about a firm failing to adopt a surplus-enhancing technology even if all individuals are optimizing, right? So, there's no individual in this firm that's making a mistake. It's just that there's a contractual friction that's hard for them to resolve. That's we can talk more about that, but that's the basic That's the basic intuition. And And think the op-ed version, but which I believe is that uh workers need to expect a share in the gains for adoption for adoption to be successful. You need buy-in at all levels of an organization. Okay. So, that's that's the takeaways from the Pentagon papers. That's about what I was going to say about organizations. Okay, so that's one sort of category of of barriers or drivers of of upgrading. So, now let me let me uh talk about uh some some factors that are uh external to firms. So far, I've been talking about things that are internal to firms. Uh external factors also mat- also matter. And in particular, what I want to emphasize is that uh richer consumers are are very demanding and tend to demand higher quality, and that can have salutary effects on firms. Okay, so that when you when you increase sales to richer consumers, you essentially have to uh uh improve quality in many in many in many in many sectors. Um and then there's increasing evidence also that that in itself can stimulate learning. So, it's not just I'm producing, you know, products that I knew how to produce already that happened to be higher quality, but I learned something in the process. So, let me tell you about another um experiment, which uh I think is the best It's not the only, you know, work on this, but it's sort of the cleanest evidence on this. It's um David Atkin, Amit Khandelwal, and Adam Osman um ran an experiment among Egyptian rug producers, where they randomly allocated initial export contracts. Okay, so they they worked with a with a buyer of rugs, big international international buyer, to randomly allocate just the first contract. It wasn't not persistent, but just the first contract the uh they were willing to to randomly allocate among among a set of Egyptian rug producers. Uh they tracked quality indicators. They kept track of conversations between the buyer uh and the and the producers. And then they actually had a laboratory where they asked producers to weave identical rugs um under under laboratory conditions there there in the lab. Um I have a I have a table of results here um that I don't expect you to to uh to follow all the numbers in the table, but this is just to say um they tracked all these These are different uh dimensions of quality of rugs. How how the corners are done, the waviness, the weight, the touch, the packingness, etc. And those things are responding positively to this this experimental intervention. So, when they were when firms were asked to uh start exporting, they started producing rugs that had had that had these had these greater you know, quality and that higher higher higher values of these characteristics. Um there's there's a a technical point which is they also started uh taking longer to weave the rugs. So, if you're just measuring, you know, output productivity by by output, then that would be misleading because they're producing less square meters of rug per per hour. Anyway, just to say that's a We have to be careful about how we're measuring productivity. Um okay, the quality quality quality went up and it went up even for these identical specification rugs. So, when they said, you know, this is the kind of rug I want, these are the specifications, those producers who had received the export export contracts produced, you know, had a higher values of these quality characteristics at at no more time. I'm just quickly I'm conscious of time, so I'm I'm just quickly summarizing what's there in the table, but as such, without more time, they produced uh higher quality rugs, which looks a lot like they learned something in the process of of producing higher quality. Okay, so this uh selling to richer consumers seems to induce firms to try to produce a different set of product, higher quality set of products, and and learn something in the process. Uh That same idea has been found by another paper. This is by uh I don't know Alfonso, I think Alfaro Ureña, Isabel Mailleux, and Jose Vasquez. Uh this is a picture who are studying of selling to a multinational corporation. Okay, so this is a picture of the Intel campus in Costa Rica where they're doing work. They're using um VAT data, which is a very exciting new data source, which India also has through the GST system where you can see firm-to-firm sales. Okay, so I think that's an important On research terms, that's an important new dimension. And so, they compare the um you know, the outcomes of firms that start supplying to multinational corporations versus suppliers to other types of firms. In particular, that my favorite comparison is is this they have a many comparisons in the paper, but my favorite one is this one where they're comparing firms that start supplying to multinational corporations to firms that start supplying to domestic exporting firms, which in other ways are similar but are not owned by by multinationals. And here and this is what here here this is just uh uh sales to others sales to firms besides the one the one multinational or the one domestic exporter that they're tracking. Okay? And so what you can see is that in year in year zero when they start selling to the multinational, they start selling to the the multi the domestic exporter, their sales to other firms goes down. There's a capacity constraint, so that's natural. Some of those they they they don't not selling as much to those other firms, but then over time it quickly recovers. And in fact for the for the for the ones that are selling to multinationals, those are the green circles, it they they they expand sales to others. Okay? Even relative to the to the firms that are selling to domestic exporters. They also look at here you can see their their their product measure productivity goes up. Those other buyers of the multinational corporations sorry of the suppliers to multinationals the other buyers tend to be larger, have higher export import shares, and have longer suppliers relationship with suppliers. So it looks like they're able to sell to a more attractive set of firms. And so the interpretation of the paper, which which I think is right, is that selling to multinationals that firms are learning something by selling to multinationals, right? Not they're they're they're gaining reputation. It's part of a branding thing. They can say now I sell to Intel, and so therefore others should buy from me. But also it there's there's lots of detail in the paper about how um you know, they they've the the demanding specifications of Intel or other multinationals, how meeting those helps them to sell to others. Okay? So what's the what's the what's what's the Um, the the takeaway here so far is that uh um being able to satisfy demanding customers requires learning and then has positive effects subsequently. All right, so that's this external orientation is quite important. And I this may be I think it's relevant also in the Indian setting. Um, when you when I think about the production linked incentives programs uh often seem to be quite domestic oriented is what my reading. You know, that they're interested in in increasing domestic output but not necessarily there's exports are mentioned but are not necessarily incentivized and so that's I I think is is relevant, right? So that I think one message we're coming from this is that those external customers are helpful have these salutary effects on firms. Okay. Uh uh someone's going to have to Mario you're going to have to help me with time. I'm not sure how how It's getting to be about 5 more minutes? 5 more minutes. I see okay, all right. So let me I'll talk faster. Maybe okay. I'll squeeze it I'll squeeze it a bit. Okay. Um, also on the uh I think a very interesting uh you know, sort of paper results um in the spirit of how selling to to um having to produce higher quality in order to satisfy international buyers can have positive effects. Um, this is there's a paper by Chris Hansman, Jonas Hjort, Jean-Marc Chuliá and Theo Tishler on Peruvian fishmeal firms. What's very interesting about fishmeal is that the output quality is actually observable. It's basically the protein content of the fishmeal and so that they can see but also at the same time the input quality, which is the freshness of the fish that are used in order to produce fishmeal is not observable. Even on the dock you can't really tell how fresh those fish are. And so what happened what they use they say they they use um quotas. So the main competitors of producing high quality fishmeal are Denmark, Iceland and Chile. So when those places hit their quotas that increases the the for Peruvian high quality fishmeal. Okay. So in those cases that they sort of have it have it that's a source of exogenous variation we would say which is trying to identify the causal effect of of of you know of increased demand for for quality. When when there's increased demand for quality, these Peruvian fish meal factories go out and buy boats. They integrate vertically. Right? Why do they do that? In order to ensure that the fish is fresh when arrives at the dock. Okay. So there that's I think a strong signal that input quality matters for output quality. The other work has shown that as well and that organizational forms can respond to these demands for quality. Okay. Is that I don't know if that's clear. Right? So there so the so the and then they also show interestingly that the the Peruvian government tracks where the boats are. There's GPS trackers on all these boats. And so those boats that are owned by the fish meal plants go out further. They they don't go out as far and they return more more quickly to to to to the port. Okay. Um firms learning from other firms. I'm again conscious of time. Let me just say I'll say very quickly there's a um How how do firms learn? One thing one way that firms learn is from from other firms in their own in their in their own sectors. Okay. How do I know that? There are a couple I think well this is probably the leading paper is by Tsai and Saidal if I pronounce the the Chinese right where they randomly put firms together in meeting groups or the treatment group were put in meeting groups and then they followed the control firms as well that met monthly for a for a year and they found large effects on on on profits and on managerial practice of those firms just from being put together in a in a in meeting groups with other with other entrepreneurs. And then one way that they one piece of evidence that they they showed is that they also injected information into into some groups either about a government grant or about a savings opportunity. The government grant was more rival in the sense there's more competition among firms. In the savings opportunity it didn't matter how many other firms had that information, and they found that um the information spread more if it was non non-rival. So, that that the savings opportunity information spread, the grant spread spread less. Okay? So, that's also interesting I think consistent with their basic story that there is information flowing and that how much competition there is among firms matters. Okay. Uh uh I wanted to tell you also now about a new brand new paper. This is in Dhaka, Bangladesh, also of mine with with co-authors about the social learning effects. Um it's relevant also for the in the case what I've been doing here besides um being hosted generously here at this for this talk is I've been doing field work in Indian Indian clusters where I think this technology might actually be be useful as well. But, it's a servo motor, which is an an energy-efficient motor. I I can Let me show I'm going to show you a picture here. So, this is for stitching machines for leather goods in our case, although similar so similar technology applicable also in in in garment firms. Um so, the servo motor the short version is it doesn't have a so, the it uses about 70 per or 75% less electricity. The clutch motor, traditional motor, it has a warm-up period. You have to turn it on for a couple of seconds, and so people tend to leave it on even when the needle's not moving. The servo motor doesn't require warm-up period. You You only uses energy when the when the needle moves, and so it uses end up using a lot a lot less energy. We randomly gave out um some servo motors in this cluster in in in Dhaka. Um and to some we some firms we just gave information, and some firms we we did nothing. Um I have a couple of tables here, which I'm just going to summarize very briefly to say oh, okay, and uh Dhaka, the firms are very clustered. This is a map of essentially uh central Dhaka. Each black dot is a is a firm. You can see they're they're very very clustered. Um and these are not These are These These firms are are extremely close together. So, this is uh this the the bottom right there, that's Old Dhaka or Bangshal. Um this is a picture in the bottom left is a picture of the That's one street right in Bangshal, of which for which we have a picture. That's that street. And you can see there are probably 150 along that couple blocks 150 firms a couple block radius there. Um and so uh they're extremely clustered. Um essentially the the basic finding is that having a neighbor who receives the motor has a significant positive effect on whether a firm adopts the motor itself. Okay, so uh that that that essentially being exposed to a firm that was in our T2 group if you were in Let me go back to my sorry. I'm going to here. Uh if you're in T2 group which you got the motor yourself, but sorry, if you're in T1 group you just got a video or control. If you're in T1 or control and you have a neighbor who got T2 who got the the motor you're much more likely to adopt. You're 19% more likely to adopt the the motor um than if you if you didn't have such a neighbor. Okay? And that effect is actually larger than being shown of being shown a video of that of that neighbor. Okay? We could talk I could talk more in details about how we control for the fact that some firms are centrally located and that's not maybe not random who's centrally located. We have ways of controlling for that for that fact. Okay. So I think it's pretty strong evidence that um distance matter that spill that there are spillovers. Okay? And we also find that distance matters. So if you This is the coefficient if you instead you you ran it with a regression on the number of neighbors that are treated and then you look by distance you can see that within 250 m or 500 m walking distance that matters a lot and then it dies out. Okay? And then actually so the the that effect we find is much stronger. We also looked at the effect of having someone in your communication network, the people instead of people that you say you talk to on the on the or you know, on the phone or in person about about technology uh or or having um you know, shared suppliers or shared repair technicians. That would be another form of networking or having other sorts of shared links. We asked people where they go to pray. All those things, we didn't find much of it evidence of adoption of of spillovers there. We did find evidence of spillovers on very narrow very narrow geographic scale. Okay. Um just to say if anyone's interested in this project, we're you know, we've been in clusters in Agra and Mathura and Delhi in the last in the last 3 days and uh this technology is right for for diffusion um and so uh it's something to uh that maybe could be explored in India as well. Um Okay, I'm going to I'm going to skip because conscious of time. I was going to say something about uh the effects of competition on on upgrading. I think there's some effect but but the effects are somewhat unclear. It's sort of harder to hard to hard to um interpret just the effects of competition. So, I think we shouldn't just say, you know, I mean I'll say about that is just that increase in competition is not a panacea. Right? For it's not Upgrading is not necessarily going to follow from greater competition. Okay. Let me try and and sum up and then we can get to the to the discussion comments. Some drivers that seem to be important, so practices that I encourage information sharing within firms um seem to be important for upgrading. Pressure from demanding customers is important. Learning spillovers from nearby firms. All those things seem to be important drivers of the upgrading process. Um some other uh topics that I haven't had a chance to talk very much about but are in this review paper that I the that I mentioned, the role of family ownership and control and the role of input quality availability. I'll Maybe we can if people want to talk about that, we can do it in the in the Q&A. Let me just say about input quality. Often uh imported inputs tend to be on average higher quality often than domestic inputs. And so, having access to those high quality imported inputs can be important for for for for the upgrading process. So, another way that sort of international engagement may be important for upgrading. Uh and the big challenge, I think, is how to promote uh learning learning in firms. Um So, what are the three takeaways, I think? Now, trying to step back from this big literature on or this this emerging literature on firms development, one is that I think there's increasing evidence that spillovers are important. That provides a a natural rationale for policy interventions to speed up grading. Often firms won't fully internalize, you know, the positive effects that they're having on other firms by by upgrading. And so I think, you know, maybe this is this is a friendly audience for this idea that, you know, industrial there's a rationale for industrial policy interventions. Um But I I, you know, I I do think as I was saying before, international engagement is also quite important. Right? So, selling to demanding consumers is quite important and access to high-quality inputs is is important. And then also I would say back to the organizations uh uh you know, part of the of the lecture, learning is not automatic. Just having a sector, you know, subsidizing a sector so the sector exists, you're not necessarily going to going to have a learning. Right? It It You have to be aware, be attuned to the organizational barriers that might get in the way as well. I have not said much about specific policy interventions. Uh in part, I think that's because the the literature that's sort of careful empirical micro literature evaluating industrial policy intervention is still pretty thin. Partly for the reasons that I discussed before, that it's, you know, for for many years there wasn't much you know, there wasn't much literature on these on uh Where? Sorry. This wave of micro empirical literature when it started out was not mainly focused on firms. And so we don't have I don't I think we need a whole, you know, generations of researchers to go out and evaluate industrial policy interventions and to and and to and give us the results. That's going to take time, obviously, and some policy decisions have to be taken before before that happens, but I do think that in parallel that would be a good a good um a good thing to happen. And let me just say, if I had to identify one uh sort of policy direction to to explore and talk about, it would be export promotion. Um and I haven't said very much and again, I'll maybe we can talk in the Q&A. I've recently done an experiment in Tunisia on export promotion where they ran The Tunisian government randomly allocated matching grants for exports where the government would basically sub provide a 50% subsidy for fixed costs for exporting. That had big effects on on exporting. They find they they found large effects. I think exporting has all these salutary effects in part because consumers tend to be richer, it tends to induce quality upgrading, that also tends to induce learning. And so I think if I had to identify put my finger on one area where uh you know, with a promising on on policy terms, it would be policy upgrading. Okay? Um the non-experimental evidence is somewhat mixed, but we have this we have this RCT that that that had positive effects. Let me stop there. Thank you very much for your attention. NOW I'LL TAKE QUESTIONS. >> [applause] >> THANK YOU VERY MUCH. IT'S MY FIRST exposure to this new wave of literature, so thank you for that. And I'll turn now to the two discussants and maybe we'll start with Ajay Sharma. And um I'm told we should not turn the mics on or off. They're all set to go. So um they each have 10 minutes. Yeah? Okay, thank you. Uh thank you. Let me first of all begin by expressing admiration for whatever I've heard. And in this age one doesn't hear a lecture and feel that one has learned a lot. So I can add that. And I must compliment you for choosing this topic. Because I think firms don't get the attention in policy analysis in economic literature as they do. So they're all about macroeconomic issues or other issues, but firms are the real players in the market. The economies are basically populated by firms and and firm behavior is what is very critical. And I'm so glad that you brought it center stage and it's the inaugural keynote address. Uh few observations based on my experience of India. So the first point you made I think is terribly important for developing countries is that at some stage of development you don't need R&D and patents. You need just to catch up with what is there in the world. So in India we mourn that we are not spending enough on R&D and we are not getting enough patents. Well, all the things you've talked about are low hanging fruits. The second observation which is India-centric but I think it's worth making. That till we opened up our economy to the rest of the world we had a closed economy. And within that closed economy for historical reasons we had family-owned enterprises. And family-owned enterprises were really people who understood money rather than the production processes. So as long as they could make money through what you'd call rent seeking they didn't pay much attention to the production process. And your photographs were very revealing because I've seen many such factories in the course of my career. And a management which is completely indifferent. We have had an added problem in India in terms of our social hierarchies. So so that again is a great barrier to any attempt at increasing productivity or understanding how productivity can increase. But in the Indian context we have now today huge contrasts and the contrasts are worth I think [snorts] exploring in greater detail. So our first real success stories came from the point you made about the consumer. So the first real success came when we got the Japanese to help us set up our car industry and the auto component industry. So these were not traditional entrepreneurs. Many of them were just you know selected on a what you call a beauty parade kind of thing. I I like you. Somebody sent you. You'll set up the factory. You'll produce. But, because they have started from scratch and didn't inherit a culture which you demonstrated in the photographs. So, they became extraordinarily good. And so many of them won quality prizes of the Deming Awards. Then, I saw similar transformation in another sector when some Indian government producers started exporting garments. So, the moment they were to sell for an exporter, then they knew that whatever they were doing till then wouldn't work. So, they modernized and modernized very well. And now I'm seeing that amazing transformation with what Apple has brought about in India. So, so they've grown at a phenomenal pace and they have for once and all decisively shattered the myth that Indians cannot manufacture well. And that large-scale manufacturing is not possible or managing a large workforce is not possible. Employing about 130,000 people by now. I think that total volume production is about 20% and with the whatever government incentives are there to make good the cost disadvantage. In the international system, they're as good as production in China. So, so this issue of the consumer or the being the MNC is quite important and I think India has benefited enormously by opening up its economy. But, where we still have a problem and that's the the thing you highlighted is that the bulk of our firms are still not experiencing the transition at the pace they should. And I like To last point. That's the space for government policy. But government policy thinking is all about some subsidies, interest subsidy, capital grant, or some quota, or even the PLI is is a heavy-handed instrument. So, the the subtle tailor-made policy interventions is an area which is weak for all countries and certainly weak in our country. But I think your work certainly drives home the point that that's a low-hanging fruit with very little cost, you can get very big results. And India, it's also crisis because a lot of these old enterprises are actually being marginalized or driven to decline or disappearance. So, I think this is a very important area for sector-specific, cluster-specific studies. And government money spent there and you know, whatever the consultant did in some firm, it's very little money and the results would be enormous. There is one other positive factor that that I think gives me great hope in the future is that the uh family-owned enterprises, which understood money, now have a second generation which is better educated, understands engineering and technology. And what they're doing is amazing. And and that is yet to be seen in terms of its effect. It'll take a few more years, but I'll just narrate two unusual success stories of Indian industrialization. For reasons of political correctness, they're not celebrated. But these are technology-driven. So, one is Sula, the wine industry. Born out of nothing. When I was young, we didn't know how to drink wine. We couldn't distinguish between good wine and bad wine. When I was secretary industry, I went to Nashik and I was being offered a wine tasting tour. And Financial Times had a front-page story on Sula. And the other success story is more recent. Young Indian entrepreneurs have created a factory where they make single malt whiskeys and they claim they win global prizes and they price themselves sky high. So, with those comments I would end, but I would again say that this is very very useful because industrial policy has to have a granular understanding of the grassroot reality and then prescriptions. With macro level prescriptions we've done a lot in India. We spent fortunes over the decades. And outcomes are I think quite modest when you look at the amount of money we've spent. Thank you. Thank you for your very grounded informed insights. That was very helpful. So, over to Zainab. Thank you very much. That was a very informative presentation. I also read the paper and a couple of publications of yours on that front. Definitely very refreshing to see empirical work on this front particularly around as you mentioned firms, but also firms that export. Um What I'm going to do is uh I guess I would like to pose three key questions around the feasibility and the viability of um the pursuit of policy. So, basically picking up where you left off on the policy discussion. So, the feasibility and the viability of uh uh industrial policies basically uh to promote industrial upgrading or to facilitate industrial upgrading at this particular moment in time. And uh hoping that those questions are going to touch on maybe two elements of uh your presentation on rich customers and on social learning. Um so the questions are on policy autonomy, export market access, and um coordination. I suppose the first question is whether low-income countries can seize this particular moment to pursue effective policies that facilitate and promote industrial upgrading. And the reason for posing this question is obviously we've discussed uh the fracturing of the global economic order and its various drivers, and we know a lot of those drivers, but there's one in particular that I'd like to pick up on. And this is the debate that has been happening for a couple of years, uh you know, maybe since the 2010s and maybe accelerated by the COVID pandemic, and perhaps that is still happening right now in uh many countries, but particularly in high-income countries, around uh the relationship between the state and the market, right? Which is why we're seeing um the prioritization of national security uh in economic policy, uh the privileging of uh national security also in foreign policy. Um but as part of this debate is and in a sense it's almost as if that debate has almost been settled, but it's really overturning an orthodoxy that existed for decades that uh the role of economic policy should be, among other things, to um uh reduce consumer prices. Uh what we're seeing right now is that there's a prioritization of reshoring, um and in fact privileging industrial production and employment. You know, this is happening in the US, it's happening elsewhere. Um and this debate has, of course, kind of um is it ended or it has dissolved that orthodoxy and the kind of Washington Consensus that underpinned it. But this is no longer just a domestic phenomenon. We're already seeing the impact on the multilateral trading system, the multilateral financial system, and also on technological diffusion, right? We are seeing all of all of the barriers that are being erected. Um, so what what then uh, this means for low and middle-income countries is that there are a lot of uh, risks involved, risks associated with them. We don't have to go into the details here. Uh, the flow of development finance, right? Uh, that multilateral systems that upheld certain kinds of principles uh, also maybe under assault or they are weakened. But there are also opportunities. Opportunities around having policy autonomy because the truth is development finance that I know a lot of us are kind of we, you know, we are in despair about the future of development finance, but that development finance did come with conditionalities. Some of the conditionalities were you know, helpful, some were not so helpful. Um, conditionalities specifically around preferred policy reforms that the providers of development finance uh, uh, um, uh, you know, uh, kind of encouraged in low- and middle-income countries. So now that the flow of development finance is very uncertain in some respects and depending on the numbers you looked at you look at uh uh, that flow is uh, less of a stream and more of a trickle in some respects. Um there might be a window of opportunity there for countries to be able to decisively define their development uh priorities. But the question is, will those countries seize those opportunities? Do they even understand that those opportunities exist? Specifically around um industrial upgrading, right? So, export promotion. Uh you know, when we strip export promotion down to some of its fundamentals in a low-income country that has limited infrastructure, that would require building roads, roads that may not necessarily make a lot of immediate financial sense, but they need to be built to connect rural areas to urban centers, urban centers to ports. Uh there's the need for obviously electricity. Uh there's a need for rail lines. And in all of the infrastructure that is being built, there needs to be redundancies. These are not infrastructure projects that typically were financed by development uh lenders or policy banks. So, this is the policy space that is being created by some of the uncertainties that we're seeing right now. So, that's really the first question. Whether countries will seize this opportunity to um uh shape uh their development policy priorities. The second question is around export markets. And it's around whether low- and middle-income countries can't can secure in some cases or maintain their access to export markets in high-income countries. And this goes directly to your point around rich customers. I mean, the evidence as you presented, but you know, there's also a large literature here that um you know, firms just perform better across different metrics when they are able to export. Uh because export uh so so some of the literature has to do with export just makes exporting makes them more competitive, right? And then some of the literature has to do with as you mentioned, um you know, improving quality. Um but clearly with the uh um upheaval in the global trading system right now, and with the trade barriers that are being erected across the board, there's a lot of uncertainty around the continued access to export markets in high-income countries. So, I give one example here, the Generalized System of Preferences, which is a preferential program that's meant to give um low-income, I think even uh lower-middle-income countries access to um uh rich country markets. At least the US version expired in 2020 and has not been reauthorized, right? So, that's just like one specific example. And then even having a free trade agreement clearly is not a guarantee of having uh access to an export market. And maybe the final point, and I'll just pose it without going into an elaborate explanation, has to do with coordination. I think this is quite important, and it goes to the uh social learning point that you mentioned. Whether if a country decides to pursue industrial policies, and maybe can reasonably pursue those policies effectively, whether that country can step back and think about coordinating with neighboring countries. And this particularly applies to countries that have small economies, countries that are low income, countries that have populations that are dispersed, they have they have perhaps a limited uh pool of labor, and for a whole range of reasons that industrial policy perhaps might need to move beyond being national to being more regional. Um I'll stop there. Sorry, I took more time. >> [applause] >> Um very important points and bringing in a comparative perspective to um India and Bangladesh. And what were the other country? Pakistan, South Asian countries. Um so, I have to ask the organizers whether in fact we have time for Q&A. 10 10 minutes? All right. Um So, oh wow. Oh my gosh. Okay. Uh can we I'll I'll start back by the cameras. There was a hand up back there. And please identify yourself. >> It's uh I am Nityananda from Council for Social Development. Thank you for the wonderful presentation. Earlier I saw only um hexagon and T-shaped patches football. Now I see lot many pentagons from Adidas. I'm happy to know that you had a role there. Uh but my question is uh you know, I was also involved with uh what we call resource efficiency initiative in some Indian firms some 7 8 years ago, Professor uh Mr. Ajay Shankar might know him. >> Resource And there resource efficiency initiative. And there I found that the greatest resistance actually came from the entrepreneurs, not the not the employees. Uh because uh in most cases it involved some upfront cost, though uh ultimately there would be cost savings. You know, when you reduce wastage, when you improve your material use efficiency. But, this was not understood. They thought there are risks involved. So, that was my experience. So, my question to you is that when you say that 8% savings in material, but only 1% um savings in cost, does it also include that specially in the short run, there can be increase in labor cost because the labor's efficiency will go down? So, that could be a kind of disincentive for the employers also. Secondly, my experiment, you know, not my, but our experiment was with firms which were supplying only um domestic consumers, final consumers, or even B2B consumers. Uh and we did not have this option of, you know, approaching Adidas. Okay. Kind of, you know, Did you have any advantage that you could approach it in the initial >> lots of people who want to ask questions. So, I would say we take three questions and then um So, the third row back there. So, Yeah? At the Yeah, on the aisle, the woman in black. Third row back. Thank you so much. My name is Amelia Santos from UNCTAD. Um I just wanted to ask about the learning effect that you mentioned. I know you didn't have much time because we did some work about demonstration Hold the Hold the mic up. Now, it's okay? Yeah, okay, perfect. No, it's just about the learning effect that you mentioned because we did some work on demonstration effect. How local firms learn from multinationals operating in the in the local environment. And one question and one area where we saw a lot of impact was through gender. How women moving from multinationals to domestic firms brought with them all the social norms and etc. from the MNE. So, I don't know how much in your context you can see those type of demonstration effects through through labor. I would be interesting to know. Thank you. Uh please Rakesh. Please. Thank you. I had a question on one of your takeaways. Uh in uh I had a question on one of your last takeaways, which is role of government policy on some of the micro interventions that you were talking about. My question there is why shouldn't it be industry associations who do that rather than the government? Because you're talking about clusters and so on. The government can never in some sense, at least what I know of the Indian government, can be effective in doing that kind of work, which is very important what you demonstrated. So, why not talk about industry associations or cluster associations, etc. Also, the kind of experience in Germany of the Steinbeis Institute uh and then Baden-Württemberg and also the Fraunhofer institutions, which are more public-private in the sense they're not really totally government. Thank you. One more from this side. Raquel. Uh thanks so much for an excellent presentation. I'm sorry you didn't have time to >> Introduce yourself. Raquel Fernandez, uh professor of economics at New York University, a member of the board. Um you didn't have time to talk about the competition. And I was wondering why you think why it was that competition did not have the same type of results as being an exporter. Because really, in the end, the more demanding customers, it must be that if you're not fulfilling the demand, you're going to your profits are going to be lower or you're going to be wiped out of the market, which seems to be a competitive effect. So, what do you think was giving you the difference between domestic competition and exporting and therefore international competition? Okay. Uh Justin Lee. Okay, and then we'll have answers. Well, Justin from China. Thank you very much for the very informative presentation. And then my question is that the term you use as industrial upgrading seem to be too narrow. You know, from all the studies study you mentioned seem to be the improvement of management or practice or technology used by a firm in a given industries. But for us in the developing country when we talk about industrial upgrading in general we said move from agriculture to manufacturing and then in manufacturing climbing up industrial ladders from more labor intensive gradually to more capital capital intensive industries like from the textile to machine building and so on. And uh your study did not touch upon that you know, at all. And then my question is that what kind of barrier for those kind of industrial upgrading that we normally understand in the developing country because for us if we want to fulfill aspiration to be a high-income country, we need to move up industrial ladder and that is what normally we use in the industrial upgrading. So, a great diverse set of questions. Okay, thanks. Uh maybe I'll try to respond in in reverse order. Uh Uh so, Justin, structural transformation, I agree, is crucial. Um it's more ambitious than I feel that we have the knowledge to be at this point. I You know, so the uh Yeah. So, I'm I'm I'm being trying to I I think that that our state of our knowledge about what works and what doesn't industrial industrial development is is relatively you know, undeveloped. Uh and so, I I don't have confidence in you what's going to drive structural transformation. That's a that's bigger question. I feel like we're making some progress on what's getting individual micro firms to move up the ladder a little bit, right? But the And I totally agree that, you know, eventually you want structural transformation, but I just don't know how I don't have confidence about how to how to bring that about. So, it's not it's not a I'm not disagreeing with what you're saying, but it's just I think we have to be a little bit humble at this stage of our of our knowledge. Uh Raquel, yeah, so for competition, part of the point I was going to make in those slides was sometimes it clearly matters. There's a There was an Indian state-owned rail mill that they threatened to have entry the SAIL, it's called, S A I L. Uh There There was going to be threatened entry into production of rail, and all of a sudden their productivity increased, right? So, it was clearly a positive effect of competition. Uh why am I less confident? Sometimes it's because competition kills off the the poor performers, and that looks like productivity is going up, but it's just because the low performers have been killed off. That's one That's one answer. Uh a a second answer would be because sometimes there's there the return there are scale effects that, you know, the more competition there is, the smaller the scale of producers, and so the less they can recover those. But I think the most interesting story is the one about is one about relational contracts that that upgrading often requires, and I didn't say much about this, but requires sort of mutual investments different between different uh firms along the supply chain. And sometimes when there's more competition, those there's fewer of those relational investments. So, Rocco Macchiavello and Ameet Morjaria have an interesting paper on coffee in Rwanda, which is basically about that, making that point. And so, there anyway, just to say the you know, words of caution, but but in general, I don't want to disagree that, you know, competition also matters. Um Uh Rakesh, yes, how about Oh, why can't industry associations do this? Yeah, so there are interesting So, Sialkot, where I was working, the industry association was basically the reason why there's an airport. There's an airport in in Sialkot, it's basically the industry association did. Everyone in Pakistan talks about the fact that this is rare among industry associations that they're able to do this. It's It's basically the Chamber of Commerce is extremely active. So, some things you would expect industry association not to be able to do, sometimes they can do. But But I think it's an exception. It requires a lot of coordination. There's lots of coordination problems. There's collective action problems that we can't necessarily, I think, expect firms or groups of firms to solve on their own. Maybe in extraordinary circumstances they can, but some things it's just It's again, it's just like too ambitious to say, you know, we're going to start uh I don't know what it'll be, you know, build a road, or going to build a what I'm what I'm talking about or or, you know, infrastructure projects or uh you know, uh So, one thing that often lacking is quality certification procedures and agencies. Sometimes that can be solved uh you know, um by individual agencies entering, but sometimes you need more of a a you know, a collective more than the individual firms or or associations can can can do. And so, it's just a question of, you know, can those coordination Is it feasible that firms are going to solve the coordination problems, even though sometimes they can. Amelia was asking about learning learning through workers. I I think that's a fascinating topic that we want to study about you know, how worker flows, which may be could be gendered or could be not, how how often workers, as they move, carry information from one firm to another. There's a There are a couple of papers. There's one about in Portugal, where they have employer-employee data. One firm starts exporting, and the workers move, and then other firms start exporting where where they move start start exporting. But we just don't have many many studies, and we we need more. Partly, it's hard to track worker flows. You know, in in uh Dhaka, it's just very hard to know, you know, it uh it's hard to interview workers and then it's hard to hard to track the flow. So, that's a great topic for research, I think it would be. There was a question about entrepreneurs sometimes are the problem. It's not always workers, sometimes entrepreneurs are the problem. In in the case in our case in in Sialkot with the soccer balls, the increased labor costs were about smaller than the gains from reduced waste. And so, by order of magnitude, actually. So, clearly there was a surplus there that was not being realized. Can I go back to also to say that there was very quick? Yeah. But just to say both I thank you very much for the discussion. That that was great. I totally agree that the granularity is important and and I would love to hear more, you know, India-specific examples. And then as I know I was thinking about this, you know, the complicated the demise of the international, or international governance in some ways, which maybe we maybe experience. Maybe may there could be some silver lining in it in the sense that, you know, the WTO was quite restrictive in its rules. So, things like conditioning incentives on export success were not WTO compliant. And so, now that the WTO is essentially or becoming defunct, you know, that the rules are not being enforced, may actually perversely increase some space for for doing industrial policy. That's a very interesting idea. It's a controversial thing to say, but I can I can say it. I don't have any any any political, you know, force to answer to. But but I think it's I think it's an interesting very interesting idea. Okay. Uh the 10 minutes of discussion are up. [laughter] They're gone. And I just want to thank to begin with the audience for your being here, for being attentive, and for your a really good set of questions, and I wish we had had time for more. Um I really want to thank the speakers, the speaker, and the discussants. Um and we're very grateful to you for being here at this time and also to Zainab coming a long way and >> [laughter] >> Mr. Sharma, I think you live right here, right? But all I think you're in everyone's insights were really, really important. Um, we've had a round of thanks for the earlier speakers, but I think it was very it was a really engaging first round and I won't go through because everyone's tired. Um, but I want to take time to just thank the Indian Habitat Center for um, hosting us this evening and to ISID and our colleagues Anjali and Ganesh and others for their partnership and support in this process and um, as I said, thanks to all of you and I think we've gotten the conference off to a lively start and um, dinner will be beginning shortly. Where? Out there? But I heard the rumbling of thunder. I heard so Yeah, so was there rain? Are we Oh, we're going to eat inside. Yeah, I just kept hearing thunder and I thought somebody's scrambling to reorganize dinner. So, we'll be eating in here and thanks again to everyone and we are off to a good start and we look forward to the next two days. Thank you so much. >> [applause] >> There's there's a little ceremony. Can we can can we request the audience to hold on for just another couple of minutes? Uh we have a tiny momento presentation for this session. May I uh request Professor Deepak Nayar Sir, for Yeah. For to present to present the momento to uh Professor Eric Woogon. >> [applause] >> And uh I'm going to request Professor Sir, I would uh >> [clears throat] >> May I request Professor Nayar to present the momento to Professor Martha Chen. Sir. Thank you, sir. >> [applause] >> I'm requesting Mr. Ajay Shankar to please present the momento to Dr. Zainab Usman. Thank you, sir. And uh Professor Nagesh Kumar to present the momento to Mr. Ajay Shankar. Thank you. Um Could we have a group photo of the speakers?