Hello, this is Kobe Time, a podcast series on Markets and Economies from D BS Group Research. I'm Tamur Big chief economist, welcoming you to our 130th episode. Today we have a first for copy time. After over four years of running the show, we have the first instance of two people from the same household grace. This show on separate occasions. Back in 2023 we had World Bank South Asia chief economist Francisco Zorg on episode 113.
And for this episode, we have her husband Shekher is a non-resident fellow at rural, a visiting scholar at the Johns Hopkins School of Advanced International Studies and a visiting professor at the National Council of Applied Economic Research and C er until 2023 shaker held a number of senior physicians in the International Monetary Fund.
More recently as division chief in the research department where he helped coordinate the fund's monitoring of the global economy and liaise with the international groups such as the G 20 G7 shaker. A warm welcome to Kobe time.
Thank you. Demo. Very nice to be here.
It's great to have you and Shekhar. I am really keen to talk about two of your recent research output. Uh early last year, you and your colleagues at the IMF released a note on geo economic fragmentation, which subsequently became several notes and papers. So let's begin with you explaining what geo economic fragmentation means.
Sure. Um So if you look at global flows of goods and capital, you'll find that they've plateaued since the global financial crisis services continue to rise from a small base but goods uh and FD I have plateaued at the same time when you look at trade restrictions, they've
been rising very steeply. And since well before the pandemic, if you look at the number of new trade restrictions that are being imposed in 2022 there were almost 10 times as many trade restrictions imposed as the flow of trade restrictions about one decade ago in 2012. Um This is happening at a time when geopolitical concerns are rising to the forefront all over the world on a number of different issues, whether it's Russia, Ukraine,
whether it's the Middle East. Um And this is, this can again be be seen tangibly in the economic sphere. For example, the IMF releases every year, a report on every single country's exchange rate restrictions. If you look at recent reports and see how many times the the the there is mention of the word national security, this has been rising very steeply. So you can see that this is coming to the forefront of policy discussions.
Um If you look at earnings calls from the corporate sector, we've done text mining on corporate sector earnings calls and you find that the number of times they, they use words like friend shoring, reshoring has been rising exponentially in recent years. So all of this suggested to us a working definition of geo economic fragmentation, which we propose in the note that you refer to and which I think has now become or gained some measure of common acceptance.
We define geo economic fragmentation as a policy driven reversal of global economic integration. It's a deliberately broad definition but not the policy driven part that's important. So obviously, le let's say that there's a slowdown in trade because consumer preferences shift from goods which tend to be tradable to services, which tend to be nontradable. Um As a result, you find some kind of shrinking of trade that's not geo economic fragmentation, that's simply consumer
preference is changing. Similarly, if technological shifts occur or if there are changes in transportation or communication costs and these affect trade, that's not geo economic fragmentation, it has to be policy driven. So, so that's a crucial part of the definition. And so obviously, when we do this, there are some prudential policies such as macro prudential controls on certain capital account transactions, which we think are valid
and legitimate. And you know, there's good economic reasoning. You don't want that to be defined as geo economic fragmentation. So we want to exclude prudential policies. But of course, when we do so, we recognize that there is no bright line between prudential policies and protectionist policies and often, you know, the latter can masquerade as the former. So, but this is the working definition and this is what we mean by the term good.
Um I want to sort of, you know, build the key messages of the paper, but I want to add a couple of more concepts to this this area. So I suppose the opposite of geo economic fragmentation shaker is globalization, which is a much maligned word these days in certain political circles. So how would you assess the impact of globalization? Say from the nineties to the mid 20 tens after which we began to see strong evidence of geo economic fragmentation.
Yeah, it's a, it's a very good question and a very good way of putting it and a very good way of thinking through it. So, you know, we've amassed immense amounts of evidence in the economics literature over the last half century about the great benefits that globalization brings through multiple channels. So that's an excellent starting point. Let's start with trade. That's the most obvious. There is a lot of evidence, for example, that international trade is strongly linked to economic growth.
Um And this is this is an important bridge especially for developing countries to use as a ladder of development. You know, if you if you look at, let's say the last 30 or 40 years of data and you divide developing countries into globalizer, those which opened up their trade regimes and integrated more with the world where the trade to GDP ratio went up and non globalizer which remained relatively OIC did not open up to the world, et cetera.
You find that over a long period of time, I'm talking about 30 or 40 years, there is an enormous difference between the growth rate of the globalizer and the non globalizer. Also, the globalizer have grown much faster than rich countries, thereby achieving convergence and catching up with Western living standards. This is what you and I studied when we looked at neoclassical growth models or solo models in graduate school.
Uh Yes, it's happening but it's only happening for those developing countries which are globalizer, which are integrating with the global economy. It's not happening with the rest. So tremendous association between between trade and growth now intimately linked to that of course, is the success that the world has had with poverty reduction because we know that poverty reduction is intimately related to growth. So trade has had an enormous impact in kind of lifting up the developing world via
convergence and reducing poverty. Um Another thing I would mention about trade and this is more to do with advanced economies. There's a lot of evidence that trade reduces consumer prices uh for the poorest consumers in advanced economies. So we're talking about, we're not talking about sort of richer college educated people who anyway tend to consume larger shares of
services than goods. But when you look at low income consumers in advanced economies, they have benefited tremendously from trade. Um So that's, that's just trade. Then of course, let me, let me touch on technology diffusion, which is another enormous source for convergence and for kind of the, the uplift of the entire globe. Um We have very good evidence that FD I for example, and we'll get into this later when we discuss the second paper you
had in mind, um can diffuse technology across borders. So, you know, every country in the world doesn't have to re invent the uh developing countries can take advantage of progress that's been made at the technological frontier and it can catch up and during the catch up phase, it can enjoy faster rates of growth. But of course, this can only happen if there's some channel for technology diffusion and the channels for technology diffusion tend to be things like trade, foreign
direct investment participation in global value chains. Um Let me come to one more which is kind of close to my heart, which is migration. Uh I know that migration can be very controversial. Certainly it's a big political football in a lot of Western countries. But again, it is one of those things which and this is true actually for millennia, it's not just true of the last five decades. Migration is a primary uh mover of the diffusion of ideas and
technologies and knowledge throughout the world, right? It's, it's happened throughout history. It's, it's not a recent phenomenon. Um And you know, just to look at recent history, there's ample evidence if you look at Silicon Valley, for example, and if you look at the number of immigrants and the amount of dynamism that immigrants have brought uh to and, and that's, that's the leading edge of world technology, right? In the US, there's ample evidence that immigrants are more
likely to have college degrees. They are more likely to have stem certifications. They are more likely to start a new business, they are more likely to patent. Um There's this amazing statistic which might appeal to you Temur, if you just look at ethnic Chinese and ethnic Indians living in Silicon Valley, just that small group of ethnic Indians and Chinese, the amount they account for 11% of all patents in America, their patent output is greater than the combined patent output
of the bottom 28 states in the US. We're just talking about Chinese and Indian ethnic investors living in the San Francisco Bay area. So yes, I mean, you know, they're a tremendous source of dynamism and it works both ways. So when you have immigrants abroad, they send back remittances, we all know that Remittance, remittances tend to be countercyclical, remittances were much more stable
during the COVID-19 pandemic than other types of capital flows. So, so they're countercyclical, they help with macroeconomic stabilization um and they help with poverty reduction.
Um And then you have diaspora effect. So I think the success of, let's say the Taiwanese semiconductor industry can be traced to earlier waves of skilled immigrants who went to Silicon Valley and Wall Street and then brought this technological expertise back with them to this small, backward, agriculturally dominant island, which today is a global leader of high tech. So migration, I think is another channel which is a threat uh given geo economic fragmentation. Let, let me stop there.
I could go on on this forever. But that's, that's, that's three things, trade, um FD I migration, tremendous benefits of globalization. And of course, all of this could go into reverse with geo economic fragmentation.
OK. Very compelling Shekhar. You may or may not have noticed that I smiled when you talked about convergence and I'll take a personal detail to tell you why I smiled over two decades ago. I was at a party in Van Nest in Washington DC. And I heard a new economist in the IMF who just joined talk adly about conditional convergence. It was you. And so so when you mentioned that you really took me back to the very first days when I got to know you. All right, back to
the uh podcast. All right. So you made very compelling arguments in terms of the benefits from globalization and the fact that there is volumes of, there are volumes of empirical evidence to, to establish that point. Um We don't have a very large data set for geo economic fragmentation, maybe a decade, maybe 12 years. So how does one measure robustly the cost of geo economic fragmentation?
Uh Very good question. Um Although I should say that this, you know, your previous remarks demonstrate the perils of doing podcasts with people who remember going to parties with you 20 years ago. Um Look, this is a new literature.
It's a field which is in its infancy, it's just developing uh in the and because we don't have so many years of geo economic fragmentation compared to the many decades of globalization, it will be a while before we can do you know, very good analytically rigorous work on this that said in the Sdn we review four recent papers which try to estimate the cost of fragmentation the way it's done these days is through modeling exercises. So essentially all the four papers we review model geo
economic fragmentation in some way. They set up blocks of countries, they introduce trade barriers or investment barriers between the blocks. They have very different assumptions about the severity of the barriers between blocks. They have different definitions of the blocks themselves. In some of the papers, non aligned countries are allowed which don't belong to either block. This is just a long way of saying that the four papers are extremely
disparate from each other. Nonetheless, all of them show significant costs, global costs of geo economic fragmentation. And depending on the severity of the assumptions, uh you can get very large results up to 8% global losses and much higher losses for individual countries up to 12% 15%. Uh depending on the scenario that you have in mind. Um There are some common lessons however, that seem to emerge from these studies. Let me just go through some
of them. First, the deeper the fragmentation, the larger the estimated losses. What do we, what do I mean by deeper the fragmentation, there's many parameters in these models which could act to deepen the fragmentation uh for one thing, greater barriers between the blocks, right? So typically in these models, you put some kind of tariff barriers or some kind of non tariff barriers between the two blocks, depending on how, how restrictive you make those uh
you get larger output losses for the globe. Um Second, do you allow non-aligned countries? So, you know, if you allow non-aligned countries, then it's a less fragmented world. If you force every country to choose one block, then that's a deeper level of fragmentation. So the deeper the fragmentation, the higher the loss is second, if you put technological diffusion on top of trade barriers, then the losses are much greater. You know, we talked
about this a bit earlier. There are modeling techniques which allow you to do this. And when you do this, you find that the losses are higher. Third, the losses vary widely among countries. They're not heterogenous. And in particular, the losses tend to be higher for emerging markets and
developing economies. And this is not surprising because as we discussed, they tend to be further back from the technological frontier and they can benefit a lot from trade and FD I in terms of technological diffusion, if you remove that, you remove an important source of growth, so they are very badly hurt. And then finally, um short term elastic cities of substitution are likely to be smaller than long run elastic cities of substitution.
And that means that even the cost that you see in the papers that we survey are likely to be an underestimate for the short run because these are all long run equilibrium numbers in the short run, you could have uh disruptions to output, which are many multiples of what we find in the long run. Um So that's, that's just, you know, some flavor of,
of where the literature has been going. I should also mention that, that I did some work with IMF co authors specifically on FD I fragmentation and FD I fragmentation is quite interesting because it's taking the literature in a new direction I think. And the basic idea is simple, what we do is we take voting patterns from the United Nations and there's a few academics at Harvard who have built up a database which, which show for any pair of countries how far apart
they are geopolitically based on un voting patterns. And the nice thing about this database is that it gives you a geopolitical distance of bilateral geopolitical distance which is time varying. So then this is like extremely nice for all kinds of econometric work. And the way we do it is we try to establish relationships, you know, we divide the world into quintiles.
And in the first quintiles are countries which are quite close together geopolitically according to this measure that I've talked about. And then the final quintile is countries which are quite far apart from each other. And when you follow these quintiles over time, you find that they are becoming very important in determining trade patterns.
In particular, the share of world trade that is taking place among countries which are similar to each other in terms of geopolitical alignment is going up steeply. So it's already more important than geographical distance. And over time, its importance seems to be increasing further. Then you can be more rigorous about it and put it into a gravity model, which we do. And indeed, we find that there's a strong correlation between FD patterns and geopolitical distance.
So let me, let me give you, let me try to distill so you know, consider the geopolitical distance between France and the UK. They, they're quite close together, obviously, um if you move from that geopolitical distance to the distance between France and India, which are much further apart geopolitically, you would, you can, you can relate that to a 17% decline in FD I. So that just gives you an idea of how economically
powerful geopolitical distance can be. Um And then, yeah, then so once you have those parameters about, you know how important geo economic fragmentation can be for FD I, then you can feed that into a model. So we use the IMF SGM model to divide the world into different blocks and then to put barriers between blocks and then use these gravity estimates to parameterize the model and calculate global losses. And we find that you can get global losses from FD I fragmentation of 2% but the losses
are much higher for some countries than for others. In particular, they are much higher for developing countries than for advanced economies.
Um Shekar two follow up questions. One is if I were to rank all the economies in the world by say trade GDP ratio, so you know, a proxy for openness or trade intensity, uh would it be fair to say that that ranking would be congruent with the vulnerability of those economies that the most integrated are the most vulnerable to economic fragmentation?
Yes, although I think there are some subtleties. So, so yes, I mean, what what you can do and what we do in the in the Rio chapter. Uh for example, is to develop an index of geo economic vulnerability and it's quite easy to do that because you have this index of um geopolitical distance. So what you can do is you can wait geopolitical distance by trade shares.
And if you're trading a lot with countries that are very far away from you geopolitically, then mechanically you are more vulnerable, right? Um Now, the reason I'm saying is this is a little more subtle is that more recently, people are talking about
connector countries. And the idea here is that if you have a diversified trading base, then in some sense that builds resilience because if there's a geopolitical spat with one particular set of, of global factors because of your diversification and the fact that you're trading with a lot of people, there's some insulation from that shock. Uh People are also talking about connectors in
terms of linking hostile blocks to each other. So people like Laura Pau at Harvard have been looking at countries like Mexico, which is getting a lot of FT I from China at the same time that it's increasing its exports to the US. So in some sense, you could think of Mexico as being a connected country between China and the US. So,
so lots of work going on in this area. Um Yes, if you trade a lot with a lot of diverse countries, uh you may be vulnerable to the extent that you're trading with with people who are geopolitically distant from you. On the other hand, if you're doing this very diversely and you're doing it with a lot of people, then you can be a connector as well. Uh So, so there's some pluses and some minuses,
right? I'm compelled by the Mexico example because I think Mexico has been a big beneficiary of some of these restriction measures that the US has been putting in and countries like China uh have thousands of companies if not more uh in Mexico trading with the US under US MC A. But I think this has been noticed by the proponent of socio-economic fragmentation and therefore, we could see an undermining of us MC A uh if say Trump were to come to power because I think there's also
been some sort of promise made that if Chinese made EVs from Mexico or have their way to the US, there will be like 100% tired from them US MC A be damned. Um So uh so yeah, uh Mexico is indeed an interesting example of how at least in the short term it can be a buffer, but uh it's a dynamic game,
right? Let me, let me jump in. I think that's an excellent observation and I think this kind of brings us to the cutting edge of the literature. So uh so yes, so there are these papers talking about Mexico as a great connector. I share your skepticism if you saw Donald Trump's speech at the Republican National Convention, he explicitly mentioned we're not going to allow the Chinese to build automobile factories in Mexico and then export them
to the US. Uh That's on the Republican side. But if you look at the trade representative Catherine Sai under the Biden administration, she has made similar pronouncements in more guarded language. So, so clearly, this is something that's been noticed and it's something that may come to a stop at some point. So there's that concept of connector which is kind of a supply chain type of definition, right? You get investment from one block and then you do exports to another block.
And like you, I'm skeptical that that kind of connector can survive the hostile geopolitical environment. So I'm actually more attracted to a definition of connected countries that emphasizes the diversity of the links that you have. And there the paradigm is actually not so much Mexico
but a country like Vietnam or Cambodia. When you look at the diversification of their exports, at the amount of balance they have in their exports both to countries to which they are geopolitically closed and to countries which are geopolitically distant from them, it's very balanced.
So actually, you know, Francisca and I have a new paper which is not yet issued where we try to define connectedness in terms of the standard deviation of the geopolitical distance that you have from all your trade partners. The idea being that the more balanced you are the better connector you would make. So, so there are currently these debates about how should we define a connector?
Oh Brilliant. I look forward to the paper and I'm hoping my country Singapore scores high in that that balanced aspect. Um Shekhar, you are a staff of the IMF you're on leave writing a book that I'm also looking forward to. But let's talk about the International Monetary System a little bit that related to this issue of geo economic recognition. What are the consequences if we have persistent geo economic recommendation?
What does it mean for the international monetary system and the global financial safety net, which this international monetary system offers?
So I guess the first obvious consequence is that we will probably move towards regionalization of financial arrangements. So to the extent that there's currently a global safety net that may become a series of sort of block safety nets or regional safety nets depending on the configuration of those blocks. And of course, given that the blocks are smaller and uh you know, often may have oo often may be economically correlated more closely than outside the block.
It means that the amount of insurance that it can give you is suboptimal compared to a truly global safety net. So that's one concern. I I think one way to to sort of highlight the the the issue may actually be to look at debt restructuring, you know, that's a big topic right now because a lot of countries have elevated levels of debt and need some kind of debt restructuring as you and I both know because both of us worked at the
IMF in the early two thousands. Um The last time that this was tackled in the in the hippic debt relief initiatives, uh a very leading role was taken by the Paris Club. So the Paris Club is a group of of mainly rich country creditors who traditionally have been responsible for delivering most of the bilateral credit in the world. So when you had a debt restructuring, it was relatively easy to get all the Paris Club creditors in one room, get them to agree a deal and then you would
have the debt restructuring. Now if you look at how the debt structure in em Ds has been evolving. Non Paris Club, creditors are accounting for a larger and larger share of total debt. So when you look at countries like India, well, China, especially China and India, who were not important players in in in as as global creditors
earlier now they are. So if if you have hostile blocks and China is in one block and the Paris clubs in another block, it's going to be much more difficult to for creditors to agree on things like debt restructuring. So that's just one example to highlight ways in which the International Monetary system could start fragmenting and could suffer consequences as a result of geo economic fragmentation.
Sure. Um Again, uh really interesting research and I'm sure you'll be following up with this with uh more work including with Francisca. So look forward to that. Uh She, I want to talk about the other research that you've been doing uh earlier this month. Uh Bruegel published a paper co-authored by you. The title was a productivity spillovers from f dia firm level cross country analysis. We've touched upon this a little bit, but I'm really interested in
this subject sitting here in Singapore. So let's start with, you know, walk us through how you find this robust evidence of productivity spillovers from FD I.
OK. Um So, you know, a lot of work has been done in this area as you know, but a lot of it tends to be kind of country specific studies. What we do here is we collect a really large database with three sources of data. So we've got data on Greenfield investment on uh on a firm sector country level from FD I markets, which is a private database, um which, which uh which licenses its data out.
And then from definitive icon, we've got data on mergers and acquisitions, which are an alternative way of doing FT I. And then finally, from the World Bank Enterprise Survey, we've got some data on the labor productivity of individual firms. So we put this together in a, in a big uh cross country cross firm panel study and what we are looking for specifically are spillovers. So our left hand side variable is labor productivity in a firm. And we're asking, you know, what are the sources of
that labor productivity? And there are three possible sources of labor productivity. On the right hand side, the first is what are called intra Indust industry spillovers. So supposing that there's FT I in your industry, are there spillovers to firms which didn't receive that FT I just because they happen to be in that industry. So intra industry spillovers. Second, are there spillovers to the firm because it's a supplier
firm to somebody downstream who is getting FD I? And third, um you know, are you getting spillovers because you are the user of products um of some firms upstream which are getting FD I. So essentially you can use this data to measure what are called intra industry spillovers, forward linkages and backward linkages. And when we do this using this big uh database, we, we find some striking results. First, there are positive intra industries spillovers, but those are restricted
to advanced economies. That's interesting to begin with. Why here's how we rationalize that result in an advanced economy. When you get FD I inside an industry, the other firms in that industry are normally at a similar technological level. So when one firm, let's say gets FD I, this induces the others to become more competitive, we call this the pro
competition effect, right? They have to up their game in order to kind of keep up with the firm that's getting the FD I. So this is a pro competition effect, which is a positive spillover. We do not find this effect. Paul EMD, why that, why might that be? Because we think in EMD, the market stealing effect could be stronger than the pro competition effect.
So let's say that the other firms in the industry are not quite at the same technological level as the favored firm, which is getting the FD I. Well, in that case, the FD to the favored firm could cause the other firms to fall further behind and lose market share to the firm that's getting FD I. And if the market stealing effect is stronger than the pro competition effect, then you won't find any, any positive
coefficient there. And that's what we find for em Ds. Now, what we do find very power powerfully for em DS and, and not for advanced economies is backward linkages. What does this mean? It, it means that if you have FD I in a, in a particular firm or a particular sector, there are positive productivity spillovers to supply of firms which are upstream. So all those firms which are supplying inputs to the firms which are getting FD I seem to get productivity spillovers. And this is just
like a classic channel in the development literature. This is what we expect. This is what we see. So not only does FD I help the firms that you're directly investing in. But it means that the firm now becomes a more exacting demander of skilled inputs. And so it forces all the input suppliers to up their technology, to up their productivity, to up the sophistication of what they're supplying.
So there is this this positive restraint. So those are the, those are the kind of main um results that we find um you know, happy to elaborate more.
So Shekhar you said that uh in the intra industry spillover, it's more uh robust in the case, or you find such a significant coefficient in the case of developed markets, but not for emerging markets. But what about the uh download linkage and upper linkage those areas E MD M similar effects.
So we find these very strong backward linkages for E MD ES not for advanced economies.
OK. Fascinating. So exactly what a development literature would suggest.
Absolutely, I mean, because there's they are further from the technological frontier. So when they get the FD I, there are these strong incentives for firms which are, you know, at different parts of the value chain to kind of up their game to participate fully in it,
right? And if we are talking about types of FD I, because you touched upon the issue that you know, your database also allows you to look at FD I in the form of cross border MN A. Um is that more potent or less potent than a typical rec congestion of capital in the company.
So we actually find that the MN A data is a lot noisier and our estimates are far less precise when it comes to M and A compared to Greenfield. Um As far as we're aware, we are the first study to actually put them side by side. So most of the studies are on Greenfield and don't consider MN
A at all. We also find some some. So, so we can confirm for MN A, the positive intra industries spillovers result for a a for for advanced economies, right, which I talked about for Greenfield, that result is still there. Uh But the backward linkages for em DS they disappear. So we do not find that for MN A, um we can only speculate on why that might be um
one speculation is the following. You know, when you have Greenfield investment in an emerging economy, almost by definition, you are expanding the market for local suppliers, right? Because we are you are you are adding new business activities. So there's more business for the local suppliers if you've got ma what you're doing is essentially you're taking over something that already exists. So you're not necessarily adding to the demand for local suppliers in the same
way that Greenfield is. And in fact, if you're taking over an existing firm, you may disrupt um you may disrupt existing local supplier relationships. For example, you may to import your inputs. Um you know, compared to the old system where you were looking to local suppliers. So these may be some reasons why we do not find for MN A, the same backward linkages that we found
for Greenfield. But I think this is an area that's going to take a lot of further research to actually sort out and explain uh some of these, some of these results
here, I'm thinking aloud if the types of industry make a difference like FD I going into oil and gas, which has, you know, all sorts of negative external its and may not be the passion of, you know, improvement in productivity given that they are largely mechanized industries. Anyway, maybe we don't get as much productivity bang or spill over there. As opposed to say FD I going into high tech industries which would immediately diffuse into the rest of the sector. What's your thought on that?
Yeah, I think that's, that's a fascinating area for research. We don't have the kind of finely grained sector decompositions that you would need to answer that type of question. But I think that this is very relevant. You know, a lot of people are also talking about strategic sectors and you know, whether, you know, you're, you're aware that in things like trade disputes and FD I restrictions, there's a lot of concern that
you know about strategic sectors in pa in particular. And these strategic sectors are often the kind of high tech sectors that you mentioned. So, so yeah, I think, I think the profession will need to try to discriminate between these different sectors and, and look at how important spillovers are from a heterogeneous point of view.
Oh, that's a very good segue to my final question. Trigger on industrial policy because the moment you said, you know, strategic sectors and so on, I mean, look, it's so much in vogue now it was something that East Asian countries did and now everybody wants to do it. So what's your thought on industrial policy?
So, so look, I I think first of all, like, you know, everybody uses industrial policy in a different way and they mean different things and the conversation can often get derailed because the participants in it mean fundamentally different things by industrial policy. So let me perhaps start by saying that, you know, I would be in favor of industrial policy if you defined it in a certain way, as I'm
sure would be 99.9% of professional economists. So if the idea was that sometimes there are externalities and governments need to have public policy measures to cure these externalities. Let me give you an example. Climate change is perhaps the biggest externality of our generation to the extent that governments try to have public incentives to channel resources into industries, which would combat climate change. I think an overwhelming majority of economists would say that
that's fine. That's good. That's what the econ 11 textbook suggests that you should be doing. So, you know, if it's simply incentivizing the private sector to tackle something to, to overcome an externality, I think we can all be in favor of that. I think what's much more insidious is trying to pick winners and losers among particular firms and particular industries. Um when there's no obvious externality when it's just the government, you know, thinking that this is the sector of the
future and we have to support it. First of all, we know that governments have a terrible track record with picking winners and losers. And it's not obvious that they have a better crystal ball um than Wall Street and financiers. So if there's no obvious externality, it's not clear why
the government should be involved. Um And, you know, my personal history is, I'm from India, I lived through the era of Nehru and socialism where, you know Indian planners were constantly trying to channel resources to favored sectors. You had government monopolies in the commanding heights of the economy and it wasn't such a successful experiment, right? It took the liberalization of the mid 19 eighties and the mid 19 nineties to get away from that mindset.
And once that was in the rearview mirror, you had a step change in economic growth, you had a massive reduction in poverty, you had all kinds of things that weren't going on. Uh you know, during what you might call the industrial policy era. Now, you know, people have pointed to successful examples of industrial policy. Uh South Korea is often mentioned, but uh again, I think one needs
to go into the nitty gritty of it. Um First of all, I think one needs to have an honest effort at the counterfactual, you know, South Korea had a highly educated population, it had clean standards of governance, it had excellent rule of law. Um So it's not clear to me that in a counterfactual where they didn't do industrial policy, they wouldn't have done equally well, maybe they would have done
even better, right? Given that they had all these other conditions for for growth which we know are very important for growth. Secondly, to the extent that they subsidize the table, it should be noted that the c were export oriented firms which faced market discipline in international markets and global
competition from other world leading technology firms. So I think that's one important lesson like if you must do industrial policy, at least do it in an export oriented sector which faces natural competition from outside, you know, in India, the industrial policy was for was for companies which were supposed to service the domestic
market and essentially had no competition. And then of course, there's the danger that you're just throwing bad money off to good and there's no competitive pressure to make sure that there's any kind of actual advancement. So Yeah, it's a subtle issue. Um, I think it depends on what exactly you mean by industrial policy. If it's a matter of getting the basics right. Um, incentivizing, uh, public externalities to be, to be combated, then I think
it's an excellent idea. If it's about picking winners and losers, individual firms in individual industries, then no.
Yes. Also there is a issue related to survival bias. We've looked at certain industries in Japan and Korea that succeeded ignoring the fact that many, many billions of dollars were wasted in corruption or poor investment ideas. Because to your point, bureaucrats can't really pick winners and losers that efficiently Shekhar. What about the lesson from China, which of course has also been,
you know, designating industrial champions. But I think one thing that characters somewhat differently from others, it's a big economy and which allows it to have substantial interprovincial competition. Uh so capital can go to certain industries here and there. But then it's not just for one company and many companies fight each other, which we're seeing with Chinese EVs
for example. So would that be a kind of a lesson that you would take that if somebody wants to pursue industrial policy, that they would, they should ensure that competition stays?
Absolutely. I mean, II I think you couldn't have said it better. Let me first say that the point you made about survival bias I think is a very important one and I think it's kind of related to the point that I'm making about counterfactual, right? One doesn't know what the counterfactual would have been in the absence of industrial policy.
Um Yeah, so I indeed, so you're pointing to alternative mechanisms, so competition can occur at the export level where you're competing with other countries and other global firms, but it can occur internally as well in terms of interprovincial competition.
However, whatever the source of the competition, I think it's extremely important to keep competition alive if you've got industrial policy because industrial policy without the competition is almost a guarantee that, you know, you're just sending bad, good money after bad and there's no accountability and no sense in which an inefficient firm will ever exit the market. So, so one must try to avoid that at all costs.
Um I've been thinking about this issue of tech related industrial policy lately that the US and its allies are very keen to make sure certain leading as technology does not fall into the hands of the Chinese that they don't have the capability of replicating some of the, you know, tiniest chips that's out there. Uh And uh and, and I hear these arguments that it's really not possible to fragment the world that way anymore. Uh chips are available in open markets, they can leak into China even
if the West is keen on preventing it. And therefore this just creates inefficiencies, but ultimately, it is futile. Uh Any thoughts.
Yeah, I I think that there's two different things going on here. So one is kind of the technological element that you refer to. So is it really possible when you have, let's say dual use technology for chips to kind of distinguish between chips that have military applications and chips, which are high tech without being military, maybe technologically, it's just not possible to do that division but the the danger that I see as even greater
than that is the political economy danger. Because once you, once you tell a politician that yeah, you know, you can, you can ban this or you can put non tariff barriers uh by appealing to national security. Then very quickly, the definition of national national security becomes extremely elastic. So I would remind you of, of the Trump tariffs on steel and aluminum,
these were imposed on national security grounds, right? I mean, this is something that hadn't been used, I think for 20 or 30 years before that national security grounds, but steel and aluminum across the board, including from the Eu and Japan. So that's a very elastic definition of national security. Recently, the Joe Biden administration was raising questions about Nippon
steel investing in the steel sector in, in, in the US. So, you know, when you're talking about, about things like security in the context of Japan, there's hardly a closer ally that the US has, it becomes very elastic. So my concern is not just the tech logical point you're raising about whether you can distinguish national
security related trade from, from other types of trade. But also the political economy dimension is very tempting for politicians to just become protectionist under the guise of national security. So I think it's very dangerous path to go down.
Ok. That's a very, very apt note to end on Shekhar. I know you've been nursing a call. So I'm really grateful that you made the time to come to this podcast and talk about these two very important subjects. So thank you very much for your time and insights.
Thank you, Temur. Thanks for inviting me. Great to be with you.
Fantastic. And thanks to our listeners and viewers as well. Copy time was produced by Ken Delbridge at Spy Studios, Violet Le and Daisy Sharma provided additional assistance. It is for information only and does not represent any trade recommendations. All 130 episodes of the podcast are available on youtube and on all major podcast platforms including Apple and Spotify. As for our research publications, webinars and live streams, you
can find them all by Googling devious research library. Have a great day.