Arvind Panagariya and Shruti Rajagopalan Talking Trade

Episode 1: Openness, Growth and Prosperity

RAJAGOPALAN: Welcome to a new discussion series on free trade and liberalization as part of The 1991 Project at the Mercatus Center. I’m Shruti Rajagopalan and in this conversation I am going to be talking trade with Arvind Panagariya who is a professor of economics and the Jagdish Bhagwati Professor of Indian Political Economy at Columbia University. In the past, Arvind had served as the first vice chairman of NITI Aayog for the government of India and as the chief economist of the Asian Development Bank. He is the author of a number of books, but for today’s conversation, we’ll focus on the recent 2019 book “Free Trade and Prosperity” published by the Oxford University Press. Welcome, Arvind. It’s a pleasure to have you here.

PANAGARIYA: Shruti, pleasure as always.

What Is the Relationship Between Free Trade and Growth?

RAJAGOPALAN: Arvind, you serve as the director of the Deepak and Neera Raj Center on Indian Economic Policies at Columbia University. The center's mission is to promote economic prosperity in India, by improving the understanding of the Indian economy through research and by sustaining ongoing dialogues on major policy issues.

Here, we have a shared interest. The 1991 Project at the Mercatus Center commemorates the liberalization of the Indian economy and its consequences on its 30th anniversary. Our main goal is to revive the discourse on economic growth and the reforms that are centered around economic ideas in India.

This series on trade is a joint effort by Mercatus’ 1991 Project and Columbia's Deepak and Neera Raj Center. Arvind, thank you for sharing your insights based on decades of research on trade.

But before we talk about the reforms in India and future episodes, I want to start at a more fundamental point, which is the relationship between liberal trade policies and economic development and wellbeing in society. This is a theme that you have explored in the bulk of your academic research and writing throughout your career. More specifically, in this book, you have argued that sustained growth and prosperity almost always require low or declining trade barriers.

What is the relationship between free trade and economic growth? And what is the mechanism by which lowering trade barriers leads to more economic growth in developing countries?

PANAGARIYA: Good, a very broad question, Shruti. Probably the answer is going to be a bit long. Let me begin. This really goes to the heart of the entire field of international trade. There are a number of ways to answer the question, and each of them gives you some aspect of why trade matters so much.

In the classroom setting, of course, we begin with the principle of comparative advantage. That simply tells you that you should specialize in what you’re best at. It doesn’t matter, even if you’re not good at any one of the tasks—but still, as long as you’ve specialized in the tasks within your abilities, the one that you do the best, go for it.

It’s like in a hospital, for example, where you have many different doctors. There are neurosurgeons, neurologists, pediatricians and then there are physicians, and then there are nurses, et cetera. It’s possible that the neurosurgeon is both a top-class neurosurgeon but also a top-class physician. But, would we say that this neurosurgeon should also, therefore, do some duties of the physician? Normally, we’d say no because there are other physicians who can’t do what a neurosurgeon can do, but they can do the job of the physician.

The neurosurgeon, even if he is a better physician than all the other physicians in the hospital, ought to really specialize in neurosurgery. It’s the same kind of principle that applies to countries: Do what you do the best, and specialize in it. And then you export in return for what other countries are better than you at. That’s the fundamental principle of comparative advantage. That’s the starting point, but now we can go to other aspects, why specialization actually is beneficial.

One of the very important reasons, of course, that we also now teach in our international trade classes, is the economies of scale, that there is one reason, which is what we call the comparative advantage. I should also point out, actually, that if a country like China or India has a lot of abundance of labor, then it will actually produce more cheaply the products that use labor more intensively, relative to capital, let’s say. Countries like the United States, European Union, et cetera, which have a lot more capital relative to labor, they are labor-scarce countries, they can produce more cheaply capital-intensive products and export those. This is all a part of the principle of comparative advantage, where specialization is driven by the differences in the comparative costs.

Now, economies of scale is the second source, a very important one. There, again, we see, particularly, if you look at the success of countries like China, scale has been one of the very important keys. When you produce a billion iPhones, your costs are going to be a lot lower than a country that’s producing, let’s say, only a few million, because you’re able to actually spread your fixed costs of production. An even more dramatic example is the aircraft industry, right? The first aircraft, it is very costly because you have to set up the entire machinery and so forth. Once you have set up the machinery, the second aircraft becomes much cheaper, the third becomes even cheaper and so forth. Again, that’s the economies of scale.

In practice, there are other reasons why specialization in trade, and particularly being a part of the global economy turns out to be so important. Your question was posed in terms of trade liberalization, which really means, “Don’t just operate on your home turf. Go out to the global markets, compete there.” If you’d want to draw an analogy from sports—if you’re playing cricket, and you really are playing cricket only within your home state, and not even national, you’re not going to produce a whole lot of world-class players like Tendulkar and Kohli and so forth.

Because you’re competing against the best, you learn from them, they learn from you. When the best in the world go against each other, they have to figure out all kinds of strategies. Entrepreneurship is something very similar as well; that when you’re in the global marketplace, you go and play against one another. Each of you is the best in your field, you learn from each other. That is yet another reason why openness to trade liberalization helps.

Again, technology is another very important reason why openness matters. Technological diffusion—innovation can happen in one country, but in the end, that innovation can become available through either patents, or ultimately, the fusion of technology, reverse engineering of products and so forth. That’s another source of gains from trade. Again, if we go back to Indian history, we know that we lost out so badly simply because we didn’t even allow the products to come into our country. We didn’t have any idea, in those days, that the quality of the product was so much more superior in other parts of the world than it was in India.

Once we opened up in 1991, we began to see a big difference. And that of course forces you to either through reverse engineering or through whatever, you say, “This is the best.” The customer also forces you. When the customer goes for what is available from abroad—better quality at a lower price—then of course, the domestic producers are forced to shape up as well. That as well as the technological diffusion and all.

I think I’d say one last reason why this helps is also the fact that when you open up—and let me just take an extreme case—when you have free trade, you’re benchmarking your own domestic producers to the best in the world. Then you say, “Look, compete against those.” For the government, that also is a challenge that if they really commit to that, then they can come back and ask, “Why are my manufacturers, producers in certain sectors not able to compete? Something may be fundamentally wrong with my policies.” You then begin to fix the domestic policies.

But if you take the opposite approach, when you say that there are a lot of domestic policy problems, you don’t actually reform those policies, then the manufacturers come in and say, “Look, I got this disability, there’s 10% disability because of high electricity prices and another 10% because my labor laws are inflexible and all. So I’ve got about almost 25% disability relative to the foreign suppliers.” I come and say, “Okay, I’ll give you 25% tariff protection.” But I have not solved the fundamental problem. What I’ve done is to allow my high-cost producers to continue to produce. That’s the wrong thing to do.

I think even for policy reform, commitment to free trade or something very close to free trade, forces reforming the domestic policies that are an obstacle. These are some of the four or five very important reasons why free trade actually does help growth. Then, of course, we can come as we go forward to the empirical link between openness and growth, as well as the empirical link between openness and poverty.

RAJAGOPALAN: I would actually add one more aspect that I’ve learned from you and your work, if you permit me, which is that once we open up globally to free trade, all the inputs also become cheaper without tariffs and protectionism. This means that everything that is being produced—whether it is for the global economy or the domestic economy, automatically without any change in technology, without any additional changes in global productivity—you can still produce everything cheaper because now you have access to the cheapest and best inputs from abroad. This is going to make your domestic and global output better, and sometimes cheaper inputs may just give you larger margins. And sometimes they may help you compete better by competing better on price globally. That, I think, is one more important effect of opening up to free trade.

PANAGARIYA: Absolutely, I think that’s a very good point I missed out. But technology is not the only thing, but also the inputs themselves. You’re absolutely right. We’ve got fantastic examples from India. If you go back to the ’50s and ’60s, we wouldn’t allow, for example, the clothing manufacturers to import fabric, and the fabric we produced domestically was not world-class. When you use that fabric to manufacture clothing, you can’t export that because the global customer comes in and says, “This cloth is just not my kind of taste or my kind of quality.” Even on the export side, you are not able to compete. They are both factors, not just the input prices, but also the quality of imports.

Similarly, a lot of the machinery—because of India’s initial emphasis on heavy industry—a lot of machinery began to be produced domestically, but this machinery was not good quality. Not only was it more expensive, but it was also not the world-class quality. If machines are substandard, then the products that you produce with them also can’t compete. Absolutely, I think this is very, very important as well because technology, after all, is also embedded or embodied in machines.

Does Free Trade Benefit the Poor?

RAJAGOPALAN: Absolutely, I am obviously very persuaded by your arguments that free trade leads to rapid economic growth. And not just by the arguments, but also some of the evidence that I have seen. But there are often skeptics and those who are persuaded that free trade may increase the size of the market, may lead to economic growth, but they are not persuaded that it actually benefits the poor in developing countries.

First, can you tell us the link or rather the broad, empirical evidence that connects free trade to economic growth? And then whether actually it reduces poverty in developing countries? What has the track record been globally, say, in the last 50 or 100 years?

PANAGARIYA: Okay. Ultimately, the link of openness to poverty principally is through growth. One can draw, at least conceptually, some link more directly from openness to poverty, and we’ll perhaps come and discuss that as well. I want to start first with growth, because I think this is where the big impact on poverty comes from openness—if you can really engineer faster growth, growth in the range of 8-9% a year as some of the more successful countries were able to do. Then there’s no way the poor are left behind. And we can come to the evidence as well, to be sure that what I’m saying actually is supported by the data themselves.

Let’s get to the evidence. Now, of course, conceptually, all the factors that we have discussed, we have talked about just now—the specialization, economies of scale, technological diffusion, availability of inputs, the technology that is embodied in machinery—all sorts of factors are there, which can in principle contribute to faster growth. Of course, the theory on this is a little more ambivalent, depending on how you write your model. There are these endogenous growth models that one could look at, but in the end, they don’t provide you enough guidance in the sense that depending on how you write the model, you could show trade to be leading to faster or slower growth. Then you have to choose which model and so forth.

This is why I think the empirical evidence here becomes most critical. Let me just say that we can look at the empirical evidence, and let’s do that systematically at two or three different levels. This is a very vast subject among economists, especially among trade economists, and at least three levels at which we can look at it, one is at the very aggregate level. Over the years, if we look particularly at the developing countries, starting in the post-independence era largely how the developing country economies have performed during various periods.

The second we can look at this issue of causation—because a lot of the skeptics come in and say, “Look, fine, you see this correlation or association between faster growth and faster growth in trade, but how do you know that causation is going from growth to trade or from trade to growth?” There is also a good bit of literature on causation whether, actually, the faster growth in trade leads to faster growth in the per capita incomes and so forth. We can come to that.

And third is what I find most persuasive personally, I would say, which is the country case studies. We can look at countries that have grown very rapidly over certain periods of time, and try to see whether that growth is being driven by policies that largely rely on open markets. I should say from the beginning here that trade is one of the things which is very important for faster growth. There can certainly be factors that could prevent that benefit of openness, leading to faster growth failing to realize.

It’s very simple that you can have free trade policy, but if your transportation links to ports are not good, if your infrastructure is awful, then even if you open yourself in terms of the policies, you’re not going to get there because no trade response would happen. Likewise, if the rest of the world is closed, you are open, it still is not going to have much impact because the rest of the world is not trading.

They’re vectors, and then you can extend it to some of the domestic policies. If domestic policies are incredibly constraining—like India had all this investment licensing, all sorts of rigidities in the labor market, rigidities in the land markets, whatnot—then also the impact of the policies will not be realized. I’m perfectly aware that the lack of flexibility in many of the other policies can negate basically the benefits that could otherwise accrue from trade.

The point here is that trade by itself is not sufficient, but it is perhaps the most important necessary condition. And always in designing your policy reforms, it is a very good starting point. Once you begin to open and feel committed to opening the economy to trade, you are also going to do other policy reforms systematically and so forth. With that, let me now get to the evidence. In broad terms, does the evidence really support the link between openness and growth?

Now, some of the early work that came out particularly from Dani Rodrik—first, he did this short book in 1999, where he argued that if I look at the data, then the period during which the developing countries do the best is the one prior to the oil crisis. Because the data really are available starting from 1961 on a consistent basis for a large number of developing countries.

He takes those data sets, 1961 to, I think, 1973 or 1974, and finds that during that period, in aggregate, the growth rate for the developing countries was higher than the subsequent periods. You can take it from 1974 or 1975 to 1995 or so because he’s writing in 1999, and it turns out to be true that ’61 to ’73 or ’74, the GDP growth rate of the developing countries is higher. Therefore, he concludes that really the golden period of growth for the developing countries was when they were actually pursuing import substitution.

Now, broadly speaking, that observation is also perhaps correct. Broadly speaking, that was a period during which the developing countries, as an aggregate if you take it, were more into import substitution. Then later actually, the same sentiment is then expressed, very much echoed by Ha-Joon Chang, who is another relatively influential free trade critic. But Ha-Joon Chang was already writing in 2007. By this time, the evidence had changed. [chuckles] But either he did not do the figures right himself or relied on old evidence. So he has no excuse actually for missing out the data and being factually incorrect, because I’ve done the data.

What happens is that it is true that initially ’61 to ’73 or ’74, the developing countries do grow better, but then the oil crisis hits and all sorts of bad things happen, and then there’s also liberalization that is largely driven by the World Bank (WB) and the International Monetary Fund. There’s a debt crisis that happens in the 1980s, and that leads Latin American countries, the African countries debt traps, and that gives the International Monetary Fund (IMF) a lot of clout, so they really go in. And largely, you can say there was some bit of forced liberalization.

In that liberalization, in the end, at that time, at least, it was not so much owned by the countries themselves. Then of course, what that also means is that a lot of the other complementary reforms that ought to happen probably didn’t happen at the time. And then even generally the global economy doesn’t do that well during this period, so developing countries did do poorly. But of course, there was this impact of this drive by the IMF and the World Bank partially, of course, but also there was the diffusion of ideas of trade economists that was beginning to happen.

In the 1970s, some major research projects had been done, one by the OECD, another by NBER, a third by the World Bank. All these research projects were diffusing. The ideas, as well as evidence from these projects, led to the gradual acceptance actually on the part of many of the developing countries, and they began to own this process of liberalization. We know of course India was a big example, ’91, but China before that had started already. A lot of the Latin American countries also came on board, countries in Africa came on board and then, lo and behold, we begin to see the evidence completely turned around.

I’ve done the numbers, and I’ll give you some numbers. I’ve taken 1961 to 2013, which is a fairly large period, and I divided it into three different periods. Now, you can divide it any which way, these are close to five decades of data, and whichever way you want to divide it, it doesn’t matter. The evidence is so strong that once this liberalization took root and countries begin to own it, growth rate is significantly higher actually in the mid-1990s onward period.

So ’61 to ’75 in my taking is about 3.1% for the developing countries. Then ’76 to ’94, it does drop to 2.4%. This was the period, the World Bank pushing it out and the entire Soviet Union was breaking down, the Eastern European countries turning from inward policies to outward policies, but this is a transition period.

They drop to 2.4%, but ’95 to 2013, 4%. It’s significantly higher than the 3.1% that had been achieved in 1961 to ’75. That’s the one basic point. You can check, you can slice this data whichever way you want.

I do it by decades. By decades, if you do ’61 to ’70, 1960s decade, growth was 2.9%, ’71 to 1980 was 3.3%. But then you come to the later decades, and you have 2001 to 2010, 4.7%. It was really significantly almost 5% as an aggregate for the developing countries, quite unprecedented.

Not only that, actually, it also turns out that during the early decades when growth was happening in the 1960s and early 1970s, OECD countries were growing very rapidly. They grew much more rapidly. If you want to take the period 1961 to ’75, OECD countries are growing at 3.5%. They’re providing the kind of impetus—they’re serving as the engine of growth for the developing countries as well. If you take the third period 1995 to 2013, when the developing countries are growing 4% compared to 3.1% in the first period from ’61 to ’75, OECD countries are growing during this period from ’95 to 2013 only 1.4%.

It is the liberalization by the countries themselves that is making all the difference. It’s clearly their liberalization, not the rapid growth of OECD countries. Sometimes people say, “Oh, today the economic environment is not good enough, the global economy is not growing, and therefore, trade cannot be the engine of growth, and all.” But China, all the developing countries taken as a group, ’95 to 2013, OECD countries are growing only 1.4%. But because they have their own policies which were much more outward-oriented and admittedly some of the other complementary policies in place as well, they grew at 4%.

One last point here, which is also important to bring in, is that even for the period ’61 to ’75, even if one were to say this was a good period for the developing countries, which countries are driving that growth? You have Korea, you have Taiwan, you have Singapore, Hong Kong, some of the most open economies are the ones that during this period actually grew so rapidly, something like 8% to 10%. Even in that period, if you read Rodrik or Ha-Joon Chang, you’ll see no references. References to Korea go elsewhere, but not in this context. In the end, this is what matters.

Relationship Between Economic Growth and GDP Per Capita

RAJAGOPALAN: I think the last point you said really is important, and I know that’s one of the reasons both you and I think specific country studies become very, very important in this context. The 3.1% and the 4% in the later decades that you point out, that’s an aggregate across all countries which are put in the bucket of developing countries. But it doesn’t tell you very much about which are the countries which are growing fast, which are the countries growing slowly and which are the countries which are in negative rates of growth, which also happen in certain years for certain countries.

When you look from ’61 to ’75, India is in what is pejoratively dubbed as the Hindu rate of growth, and it’s the same years when India becomes more and more inward. The first attempt at devaluation fails in 1966, Mrs. Gandhi takes a turn to be even more socialist and inward-looking. That is exactly the period when they see the emergence of what we call the Asian tigers. You have South Korea leading the group and you have Singapore, you have Hong Kong, you have Taiwan, and then, of course, followed by China which is a different scale in terms of the size of the economy. The 3.1% and the 4% masks this difference even within developing countries. There is a separation even within that pool.

PANAGARIYA: Yes, absolutely right. There is a point that you have to disaggregate. You have to disaggregate. In fact, I’ve written on this, and this is reported in the relevant chapter in my book as well, that what I did, actually, for each of these periods, I separated the countries. The earliest, the first thing I wrote was a paper which was titled “Miracles and Debacles.” What I said was any country that grows on a sustained basis, which could be about one and a half decades, let’s say, in per capita terms at 3%, let’s call that miracle growth. It is arbitrary but you can choose another, 3.5 or you can choose 2.5. The basic thrust of the argument doesn’t change. Then I call debacles countries which in per capita terms either don’t grow or grow at negative rates. Actually, per capita income declines.

I look at both because there’s also this complaint that free trade causes decimation. Suddenly the domestic industry is destroyed by free trade, so it leads to this very negative kind of growth. You look at the debacles also—is trade behind that kind of debacle? I basically try to look at, at least the association, and when you do that, overwhelmingly countries that are growing at 3% or more in per capita terms, they overwhelmingly actually are also experiencing at that time rising trade to GDP ratio, rising exports to GDP ratio.

What that means is that the per capita GDP is growing 3% and the population is growing another 1-2%, then the GDP is growing at 4% to 5%. Even then, when you say exports to GDP ratio is rising, that of course means that your trade is rising even faster than the GDP. That’s the miracles part of the story. The debacles part of the story turns out that you almost rarely find that the negative growth is correlated with fast surges of imports. You just don’t see imports growing at any fast pace during the period when countries do very poorly in per capita terms.

RAJAGOPALAN: I’m going to ask you to elaborate more on per capita income and the relationship with economic growth. But before that, for the listeners, especially sometimes it seems like, 1%, 2%, these seem like such small numbers. But we’re talking about rates of growth. When you start compounding, the impact is really, really large. There are different estimates in different countries, but a very, very rough rule of thumb estimate for India is that an increase in 1% of GDP per capita growth leads to an additional 3 million people getting lifted out of poverty.

If you’re talking an additional 3% per capita growth, you’re talking about 9 to 10 million people being lifted out of poverty. Now, when you see that over, say, 15 years or 30 years, you’re basically eliminating the extreme poverty in a country in a relatively short period of time in global history. Fifteen years, 20 years is a very short time in global history to actually eliminate the poverty problem. On the face of it, it seems like 1% here, 2% here, it seems like you’re haggling over such small numbers, but the impact on real lives of the poor is quite extraordinary. Can you take us through the nexus between free trade economic growth now translating to improvements in per capita GDP?

PANAGARIYA: We can look at the data. We got some data here. I’ll give you some. We can look at this. The World Bank actually has compiled most of these data. Do some dramatic comparison here. Let’s do some dramatic comparison. Unfortunately, we don’t have the 1950s data. At least, the World Bank on a consistent basis doesn’t give the data for the 1950s. Although we know from other sources—not the World Bank, but other sources—where the scholars have tried to estimate the GDP per capita incomes in the 1950s.

You look at India, you look at China, you look at South Korea, those three roughly started in the same place around the early 1950s. Actually, at least for South Korea, there’s a period of civil war—not civil war—but the South Korea-North Korea war, which had decimated it. In 1954, it probably started below where India was because of the devastation of the war. Some of the other countries which did better, did well also—Singapore, Taiwan—they were probably a little bit higher. But by 1961, this is what it looks like.

I’ll give you the numbers. This is in real terms, so it’s in 2010 constant U.S. dollars. China’s per capita income was $141. India was actually significantly more according to the World Bank numbers, $336. South Korea is about three times of India at that time, $968. South Korea is about—you can say $1,000, a little less than that. China, about $140, $141 to be precise, and India $336. India is in the middle.

Come to, let’s say, 2000. You come to 2000. India gets to about $827. China has now surpassed India by a good margin. It’s almost twice of India: $1,768. Of course, South Korea is on a different scale by now: $15,400—$15,400, $827 and $1,768 for China. India is $827, China, $1,768, South Korea is $15,414. Come to 2019, this is roughly the current period. Now, where do we stand today? 2019, China is $8,242 per capita income, India is one-fourth of that, $2,152, and South Korea is about $28,676. You can see it’s totally—

Now, also very interestingly—we can look at the numbers later—but it turns out that in South Korea and China both—and then you can look at Taiwan, Singapore as well—poverty got eliminated pretty much by this kind of phenomenal growth they experienced, and it got eliminated by growth itself. They hardly had any social programs in the way India has actually at a very low level of income started these social expenditure programs like the National Rural Employment Guarantee scheme, or the public distribution system of food and so forth.

These countries really ran none of that actually on any significant scale whatsoever. You may find tiny programs here and there, but nothing significant. Much of the poverty reduction happened purely by growth. You’ll see when we look at South Korea, particularly as a country case study, we’ll find that will be the case.

I just wanted to briefly mention that this whole issue about causation, whether it is trade causing growth and growth causing trade. Now, economists have actually looked at this. Let me explain that and then we can come to the discussion of South Korea. I think that’s where a country case study gives you a lot more feel for what is happening.

On causation, now what economists have done is to, at least the earliest studies that were done, this is a study by Jeff Frankel and his co-authors. What they did was they used the gravity model. Now, what the gravity model does is it shows that the distance between two countries offers a good measure of trade, meaning farther away countries, controlling for their GDP levels, countries that are located far apart trade far less, and countries that are closer to each other in distance trade a lot more. Distance turns out to be a very powerful explanation for bilateral trade, not total necessarily, but bilateral trade.

What they do is use this gravity model which uses distance as the determinant of trade. They estimate the part of the trade which is driven by the bilateral distance between the countries. You’re controlling for GDP. Therefore, this is the part of the trade that is not impacted by GDP. It is only impacted by bilateral distance. Then they run it, that if you take that part of the trade and see what impact it has on per capita income, they find that countries that are on balance trading more have higher per capita incomes than countries that are trading less.

This is a trade which is uncontaminated by GDP levels. After that, of course, now there is a long paper by Doug Irwin, which does a survey of a number of these studies that have been done, which try to find different clever ways econometrically to get around the causation issue. Now, there is plenty of evidence actually. After my book came out, a lot of studies have come out, actually which basically point out to this fact that starting from the mid-1990s, you’ll see very substantial increases in incomes of the countries that had opened up. They are able to actually also establish the causation running from opening up and expanding trade to GDP growth or per capita GDP growth. We have that evidence.

The South Korean Miracle

RAJAGOPALAN: I personally find that evidence quite persuasive, especially the post-’90s evidence after the collapse of the Soviet Union, as you point out, more and more countries are now embracing unilateral liberalization and opening up to free trade, which is the part of the transition economies and the Eastern Block, many developing countries, and over all, the questions people were asking changed.

Earlier, this was up for debate. And something that seems to have changed after the collapse of the Soviet Union is now instead of debating the fundamental question of whether free trade and growth are linked to each other, we start thinking about what kinds of mechanisms, exactly how are they linked to each other, exactly what kind of growth takes place and so on and so forth.

I want to dig deeper into the South Korean case. One of the miracles of the 20th century, especially the second half of the 20th century is South Korea. Of course, China is known better—the Chinese story—both because of the size and scale, the enormity of the success. The reforms initiated by Deng Xiaoping, it’s pretty well known across the world and the impact that it’s had on China. The South Korean miracle is, I think, very underrated and also underdiscussed and underrecognized. It comes in a different time period when this is not the prescription for most developing countries.

As you pointed out, in the 1950s, South Korea is devastated by civil war. In today’s terms, it would stand globally in the rankings of something like Syria or Yemen. Modern-day Syria and Yemen is where South Korea is starting in the 1950s. About 10% of its GDP is coming from American aid, so it’s really being propped up by aid and not as a trading economy. They have no real natural resources to speak of, no real big agricultural output to speak of.

Today when you look at South Korea, the story is dramatically different. Of course, if you just do a simple ranking of countries by GDP per capita, South Korea is 35th across the world. It ranks higher than even some European countries in some years, like Portugal. That’s one part of the success story. On other terms, it’s also a big power in terms of exporting what we call soft cultural exports. Whether you’re talking about Korean movies, Korean K-pop in popular culture, Korean fashion, cosmetics, Korea seems to have had a very different impact on the world for a country of its size. That’s a second part.

During COVID, we see that Korea seems to have managed and arrested the problem right from the beginning. They’ve had very unique solutions, and if there is a success story early on in COVID—not later on post-vaccinations, but very early on—I would say South Korea stands as a highlight. They’ve also improved their infrastructure, their state capacity, their social cooperation and cohesion, and so on. To me, it seems like South Korea is really the miracle story or the breakout story in the 20th century. Can you tell us a little more just about what happened in South Korea? Because I think the story is relatively unknown.

Then maybe we can delve into other parts of how the structural transformation took place in South Korea to take it from what was—they used to call South Korea “the global basket case.” How did it go from that kind of reputation to the reputation it has today?

PANAGARIYA: Yes. Very good. I think you summarized it beautifully. Two additional exports that are emerging from South Korea: Korean food is great.

RAJAGOPALAN: It’s fantastic.

PANAGARIYA: You mentioned Korean movies, but Korean dramas, and I’m a big fan myself of Korean dramas. [chuckles] I should also mention that actually, finally, I found the Korean masks to be the best. [chuckles] The designing, they really think about it. For teaching my classes, I’ve been now using Korean masks. Because for two hours, you are speaking with the mask on, so you need the mask to be such that nothing else comes in but at least air comes in. [crosstalk]

RAJAGOPALAN: I would actually add to your list of shopping Korean sunscreen for the summer. When you’re outdoors and mask-free, I have found Korean sunscreen to be the best of any sunscreen that I have used. It’s not sticky, it does the job really well and we can now get it in the United States.

PANAGARIYA: Yes. Recently, it started changing. We’ve been in this apartment for 17 years, so changing all the appliances, all of these are LG, they’re Korean. [laughs] It’s not just that they function well, but they think of the consumer conveniences, exactly how to—even small things like in the cooking range, what should be the size of the burners and what should be the distance between the burner itself and where the pot is going to sit, all sorts of things. They pay attention to these design issues. Just as I mentioned about the mask, designing is very important in almost any consumer item. We think of designing perhaps only on clothing, et cetera, but appliances also require proper designing.

Anyway, let me get to the key issue that you have raised on growth first. It is good to at least get some perspective on the numbers first. At least, the numbers that, hopefully, the audience can also remember as we go deeper.

1954 to ’62—’54 is when the Korean War ended, and Korea began to rebuild. 1954 to ’62, the growth rate in Korea—annual growth rate, this is the GDP not per capita, but GDP growth rate—is about 4.2%. That’s an eight-year period and 4.2% growth. Then a whole decade, 1963 to ’73, Korea grew 9.1%. Now, this was completely unprecedented, as far as I know.

Well, there may have been in Taiwan, Singapore and Hong Kong alongside happening. But recognizing that fact, this is a completely unusual experience for anybody. No Western country, at least, all the prosperous countries, none of them had ever grown at 9.1% in a single decade. That was completely unprecedented. Then we see some setback to growth. This is the period ’74 to ’82, another eight-year period—6.9%, 7% if you wish—but clearly two percentage points below what had been achieved in ’63 to ’73. Then it resumes 1983 to ’95, the economy again grows about 8.7%. That’s another 12-year period, 8.7% fantastic growth.

Now, by this time Korea is fully transformed. It is no longer a poor country, transformation is practically complete. We’ll look at some of the actual changes that happened during this period. At least the way we think of poverty in Asia, in China or India, the abject poverty or of extreme poverty, that’s pretty much gone in Korea by 1995. Per capita incomes have really grown quite a bit. We looked at some of the numbers on per capita income already. That’s, in a nutshell, the growth story.

Now, let’s look at what happens to trade during this period. Now, trade, if you start out in South Korea around 1960s, till about 1965, exports are 5% of the GDP. They’re about 5% of the GDP. By early 1970s, ’72, ’73, somewhere there, it is already close to 30%, about 28% maybe. It’s already 28%. By 1987, you’ve gone to 38% of the GDP, what was 5%. Trade is exploding during this period, absolutely phenomenal.

Now, how is the transformation happening? Hong Kong, Singapore, Taiwan, South Korea and China—those are the countries which have grown at this pace of 8% to 10% on a sustained basis for almost two to three decades, almost three decades. In each of these cases, manufacturing is the first driver. It is the manufacturing growth that takes off. Then, as manufacturing grows, what happens is that it requires more and more workers. In each of these cases, you’ll see that the manufacturing that is growing very rapidly, and that is also operating in the export market, is highly labor-intensive.

It’s things like clothing, footwear, furniture, these kinds of products are what are being exported progressively more and more. It requires workers. Where do the workers come from? They draw them out of agriculture. These are also products which don’t require a whole lot of time to train the workers. The farmers can come out, and in six to eight weeks—Korea also was ahead of most of the countries in terms of literacy, whereas on higher education, they didn’t do very well for a long time. Today, of course, they’re well ahead of most other countries in the region. But on higher education, they didn’t spend much, but on primary education and literacy, they were very big.

In fact, in those days, anybody who acquired college education was required by law to teach in a primary school for five years, including actually, President Park Chung-hee. He’s the man behind the Korean miracle. Even he had actually taught in school for five years as a part of his mandate. He was an army general, of course, but he had also done that. Generally, you’ve got five or six-grade level primary education, then workers can be trained into these labor-intensive industries. That remains true today also.

That’s how they built up the workforce and they operated in the global economy. These are good jobs that got created and they kept growing, they’re growing. The share of manufacturing began to rise in the GDP. As these workers came in and decent wages, they got employed, they spent their income. As they spend, of course, a lot of non-traded services—because services tend to be non-traded, most of the services. In those days, particularly, the services also then begin to kick in as an engine. There is this connectedness between manufacturing growth and services growth.

Both services and manufacturing begin to draw workers out of agriculture. That process happens. The share in GDP of agriculture, which around 1965 was close to about 40%—by 1990, that dropped below 10%—that’s the share in GDP. Manufacturing rose during that period from about something like 13%, 14% in 1960 to about 30% in the 1980s throughout. Services, of course, particularly after 1975, services took off as well. There’s a sequencing that initially, it’s manufacturing that draws more workers, but then it becomes services.

Now workforce-wise, it is more dramatic. In 1965, about 60% of the workers are in agriculture. By 1990, that figure drops to below 20%. There’s 40 percentage points movement within a 30-year period. A dramatic transformation in terms of the profession.

What is most remarkable is that during this entire period, wages, real wages have been growing on average 9% to 10%. One can look at different periods. For instance, if you take ’65 to ’73, wages grew 9.3%; ’74 to 1982, they grew 9.6%; and 1983 to 1990, 10%. It’s between 9% and 10% throughout. This is a tremendous kind of boost to services demand, which have to be supplied locally and that creates jobs and services as well. It is quite remarkable.

One last part of the transformation of South Korea, you take 1960s until about 1965—again, middle of 1960s—it’s about 30% urban. By the way, in India, if you look at 2011 census, urbanization is still only 31%. Of course, Korea did start a little more urban than India did, initially. India has been obviously much more rural, but look at what happens to the rate of urbanization as this rapid growth happens. By 1990 already, urbanization was about 75%. It really kind of exploded. It’s not just that somehow all the rural labor force moved into urban areas. That was certainly part of the story, but it is also the fact that many of the rural areas turned urban.

This transformation always happens. There is a Mumbai-Shanghai model of urbanization, and then there is the Shenzhen model of urbanization. The Mumbai-Shanghai one is the one where cities already exist and more and more people come to these urban centers. That’s the Mumbai-Shanghai model, but Shenzhen is different.

Shenzhen, you go to 1980. This is in China. On the coast, Shenzhen, there was a bunch of fishing villages. Maximally the population was 300,000 in 1980. Today, it is one of the most urbanized spots on the face of earth. It’s incredibly urban. The entire robotics industry, et cetera, is located in Shenzhen and its population is something like 12, 13 million, much larger—from 300,000 to something like 12, 13 million. Per capita income is about $25,000 per year. It is rivaling Korea’s pretty much in Shenzhen itself.

This urbanization was the other kind of very, very important feature of this transformation. The transformation of Korea is not different from that of Taiwan, for instance, or before that of Japan—if you’d go want to go back in history, United Kingdom, United States, Germany, et cetera. The big difference is that what the earlier countries, the United States, U.K., Germany, et cetera, took over 100 years, South Korea compressed it all in 30. They really got to where these countries took 100 years to get to, they did it in 30. That was absolutely the big difference.

RAJAGOPALAN: Also, can you tell us another aspect of structural transformation, which is, initially South Korea starts out as a very low-tech exporter? They’re really exporting garments, they’re exporting hair, I remember you had mentioned is a big part of the export industry. The composition of exports and imports has also changed over the years, right? Now imports are much more design-based, technology-based, innovation-based. Whereas once upon a time, it was more someone else’s innovating and South Korea is the cheap labor manufacturing hub, which can produce things at scale and sell them cheaply. That has also changed as Korea has gotten richer.

PANAGARIYA: Oh, absolutely. Yes. The whole idea here—and this is where I think India can really learn and what we are doing is only we try to skill people only from the top. Because the Indian history, our development has been different because it started with the heavy industry and all. So you did acquire some expertise in these very high-tech industries, early in the game, and that advantage translated eventually, I would say, in the IT industry, software industry revolution, and so forth. But the whole point is that you’re still left—the large bulk of hundreds of millions of worker population—completely untrained, completely unskilled.

You can’t train the population into this very high-level skill by bringing a farmer into IT industry, a farmer into finance industry. You can’t do that. You really have to begin at the level at which you can bring them in and give them certain training, certain skills, impart certain skills that they are able to absorb with their level of education and all. Then you build up and then you build up. Also, during this process that I described, South Korea began to become more and more capital-abundant.

We talked about wages rising almost 9.5%, 10% a year. That is referring to labor scarcity. Cheap labor is no more available. Labor is becoming more and more scarce factor. You then can also shift into more capital-intensive exports, where India wants to do it right away. That is, again, the problem you see. Then your limited capital is being absorbed by, let’s say, the petroleum refining industry or some machinery industry or something. Then you deprive the other labor-intensive sectors of the capital. There are these lessons to be drawn from South Korea that nobody immediately started off with these high-tech exports.

Korea today, of course, is fully transformed because its per capita income is close to $30,000. At that level, evidently, you can do much more. The workforce has also become more educated and you have some world-class universities in Korea now. In higher education, they have caught up, even though they didn’t start. Notwithstanding the disadvantage of the language and all that, we all revel into this thing about English is our language and we know English. But South Korea had no English and it could still develop and become so influential, and now, even actually, project its soft power. In spite of the lack of English, it’s able to project its soft power too. It’s quite interesting, but that all goes with wealth. You see, if you create wealth successfully, then it can be done.

RAJAGOPALAN: At so many points in our history, when we read literature about how Korea was talked about in the ’60s and ’70s and then China, there is this pejorative first-world commentary, right? “Oh, these are people who just know how to make cheap T-shirts.” It’s a very pejorative first-world way of looking at the problem. But now we realize that Taiwan and South Korea used to be the countries that made cheap T-shirts, and then it got substituted by China, then now even China is too rich to be making cheap T-shirts as cheaply as the world needs, so now Bangladesh is the new place that is making cheap T-shirts.

If India doesn’t catch up, it’s going to be some country in Africa next. The idea is if you’re cheap labor and you are competitive at it, the irony is you’re not going to be cheap labor for too long because you’re going to get rich rather quickly. I think that is one of the underappreciated lessons, that there is no shortcut to the structural transformation. We can talk about this also in the future episodes of our conversation on premature deindustrialization, a lot of the barriers to structural transformation for India and other developing countries.

Does Targeted Industrial Policy Help?

On South Korea, before we wrap that story up, I want to just ask you the question about a lot of the commentary on South Korea said that there was a success of South Korea that can be attributed to industrial policy and not just to free trade. So far, our conversation on South Korea has been how free trade transformed the South Korean miracle growth story. How much of it can be contributed to the industrial policy of South Korea, if at all? Is there any merit in those arguments?

PANAGARIYA: You see, when a success happens, everybody thinks that it is their favorite theory that caused that success. They manage to throw in enough to confuse the wider audience and make it look like it was some sort of fancy stuff that they think countries ought to be doing. And that’s what Korea did and that’s how Korea succeeded. Industrial policy, of course, is one of these favorite ones that they throw in. Various terms get used sometimes as industrial policies. They’ll say that it was targeting or industrial targeting. That is what led to the success of Korea.

The broad point, first of all here to understand is that governments always do something. Governments don’t sit back. In this sense of when we say, “The invisible hand, let it do the work, and the government should be completely hands-off and all,” that is not what the governments do. Governments want to show that they are doing something and they are producing the success. There are always these interventions that happen. Then they’ll say, “They were doing all these interventions.” Robert Wade, for example, writes a whole thick book about Taiwan where he finds, “Here is this intervention, here’s that intervention,” and all.

The real question is, can you connect the interventions to the success? In the end, you’re saying that, yes, we all agree that Korea has grown rapidly and Taiwan grew rapidly, Singapore grew rapidly, some interventions did happen. Just because the interventions are happening, does that mean that those interventions are the cause of the rapid growth? Or, is it the case that those countries would have grown even more rapidly if interventions were absent?

There was this very interesting exchange between Robert Wade in separate papers, of course. They’re not face-to-face. Ian Little, who was very much at the forefront of advocating the outward orientation model—Robert Wade kind of asked this question rhetorically, saying, “If you are going to argue that in South Korea, interventions actually were harmful to growth, then you’re saying that South Korea would have grown even more rapidly than it grew. It was already growing so rapidly that it is completely implausible that the absence of those interventions would have actually led to even faster growth. My story is more plausible because it grew that rapidly.”

Obviously, the interventions were making a positive contribution. How is that an argument? Ian Little had a good response to that. He said, “Why does Mr. Wade think that Korea would not have grown even faster had there been no interventions? Because Taiwan, in fact, did have less intervention, lot less intervention than Korea did, and Taiwan did grow faster than South Korea.” [laughs] That was one, but also, if we go back to the growth rates that I had mentioned of South Korea, there’s a full period from rapid growth beginning in 1963 to 1990s. Let’s look at that: 1963 to 1973—full decade—Korea grew 9.1%. This is, of course, the beginning of the miracle. Was that beginning of the miracle in any way connected to this industrial targeting? No, in fact, this is a period during where there is no industrial targeting. Policies are all neutral, entrepreneurs are left to themselves to decide what they want to export. We have discussed this great example of human hair which nobody thought could be an export item.

If you look back in 1962 or ’61, in the Korean export basket, there is no human hair. Suddenly, they begin to appear in the mid-1960s, and by 1970 to ’73, they’re 10% of the total Korean exports. There is no targeting happening during this period. Targeting starts in 1973 and they were feeling some possible threat from the United States that they might start imposing restrictions on Korean exports and things like that.

Also, the U.S. was making some noises about withdrawing part of its forces out of Korea and all. There was some fear that this military aid will also decline and all, and therefore, you need to ramp up your more capital-intensive industry. This is where the industrial targeting started and it’s called the Heavy and Chemical Industry Drive, HCI.

Certainly, they intervene here. Now, one side story to that, of course, which Anne Kreuger tells very nicely, some of the failures that happened simply don’t get any mention because they disappeared. She refers to this ball bearing factory that they had started in South Korea during this 1970s period, they wanted industrial targeting, it was a complete white elephant. It was so costly that they couldn’t compete in the export market. And in the domestic market, there was not enough demand. At the most, they could run it for two or three days out of the week. Eventually, that factory, that massive white elephant had to be closed down.

It was a failure. Some of the industries that were given this targeting business eventually did succeed, but the question is, was it the intervention? If you look at the immediate impact, of course, of the intervention the period from 1974 to ’82, your growth rate dropped to 6.9%. From 9.1%, it has dropped to 6.9%. If I really correlate growth to your period of industry targeting, you’re certainly not doing that well.

There are also now studies that many of these industries that were targeted exhibited a lot lower productivity growth than the other industries which had not been targeted. There is that evidence there as well, but the fundamental point really is that when they say, “Oh, you targeted, let’s say, auto industry or something and it became such a big success.” World Bank was arguing against Korea doing auto industry, but they went ahead and did it anyway. But the question is it’s a post hoc fallacy. Just because you did it and then success happened doesn’t mean that because you intervened at that time, the success happened.

Eventually, South Korea was becoming more capital-abundant and so, this would have happened anyway. Maybe the industrial targeting hastened it and brought it a little sooner, but certainly, the success was not due to that. Gradually, Korea was becoming capital-abundant and this transformation would have happened. This HCI drive is abandoned by the late ’70s. By 1980, more or less, they’ve abandoned it, but then look at the growth rate. Growth rate—again from 1983 to ’95—when you’re back to liberalization and neutral policies, no targeting, then growth rate jumps back to 8.7%. Where is the success of industrial targeting other than post hoc fallacy—what is it called? Post hoc ergo—

RAJAGOPALAN: Post hoc ergo propter hoc.

PANAGARIYA: Propter hoc, right? After this; therefore, because of this. That is what’s happening. There is a connection here. Just because you intervene and eventually, some industry succeeds. First of all, not every industry that you intervene succeeded. Actually, there are some major failures, as I mentioned, about the ball bearing factory in South Korea, but the successes were because—again, you talk of Japan, for instance. Ha-Joon Chang refers to all these—the Japanese interventions in the mid-1930s, let’s say, on behalf of auto industry, or 1950s. And then in the 1980s, Japanese cars become so successful.

“Look, if the Japanese government had not done it, then—” How does it follow? It’s entirely possible that the car industry maybe had the Japanese government intervention, maybe it would have disappeared. Does it mean that it would have never reappeared? [laughs] Once you become more cost effective, your capital becomes cheaper still because you have more capital abundance. Eventually, the industry would come back up. There is a lot of this post hoc fallacy going on in making these arguments.

RAJAGOPALAN: I think just a broader point that you have always said, which is even if you can find individual success stories, in one case here or in one decade here, or in one industry in a particular country, and so on, the question we’re not asking is if someone picked the winners and losers correctly in one case. It’s a question of sustained economic growth over say, two, three, four decades, which leads to a particular structural transformation.

On that, I think the evidence is quite clear in favor of trade liberalization and against industrial policy, even if they’ve had a handful of successes here and there, which can be attributed to particular industries. I would actually go even one step further and say that let’s see, even if there is merit in arguing that the Japanese government intervened and helped the car industry, and then decades later, Japan is one of the biggest leaders in car exports, and so on, we also don’t know the counterfactual of what else Japan may have done even better at than cars maybe had that not been propped up earlier.

Both sides, even in the success stories, it’s very difficult to argue if resources went to their highest-valued, most productive use in that particular instance—or maybe without the government nudging capital in a particular direction‑that capital would have gone somewhere even better. And maybe Japan would have produced computers even cheaper two decades before instead of cars or something like that. Those things are very difficult to untangle.

We can just make these arguments, but I think in terms of the two, three, four-decade evidence that you provide, the South Korean story is just extraordinary. Since we’re going to do a series of conversations around this topic, I have a number of other questions. Starting with industrial policy, I want to learn more from you about infant industry, and protectionist policies of that sort which have taken place in various developing countries.

Of course, we need to talk about the history of Indian trade policy, the history of Indian trade growth within the economy up to liberalization and after liberalization, and what is happening currently in India, which is a U-turn, in one sense walking away from trade liberalization, and so many other themes that you’ve written about. I think we can end the episode here, and when you come back, we’ll talk about many more such themes. Thank you so much for doing this today, Arvind.

PANAGARIYA: Thank you, Shruti. Great pleasure, great conversation.