By Martin Wolf
In the summer of 1972, as a “young professional” at the World Bank, I went on a mission to South Korea. It was my first experience of something extraordinary: a country that was developing at a breathtaking rate. The country had already enjoyed a decade of economic growth at close to 10 per cent a year. It continued to grow at close to that rate for another quarter of a century.
What struck me about Korea was the determination of its policy-makers to sustain rapid industrialisation. I saw the construction from scratch of the vast Hyundai shipyard at Ulsan that was soon to join the first rank of ship-builders. That bet itself demonstrated something even more remarkable: Koreans’ belief in their country’s ability to achieve global competitiveness.
For the Koreans, exports were both a tool of development and a test of its success. How different this was from east Africa and India, on which I was to work for the following five years. India was almost as sealed from the world economy as it was possible to be. Its annual growth in income per head had fallen in the 1970s to about 1 per cent a year, while industrial productivity seemed to be declining, despite its desperately low level.
The contrast between South Korea’s success and India’s failure was striking. Both used protection and other tools of industrial policy. Yet the orientation of India’s policies was inward-looking and anti-competitive, while that of South Korea was the opposite. In the literature on development and trade, the Korean strategy came to be called “export promotion”, because its economy did not have an overall bias towards the home market.
The contrast between South Korea and India raised the biggest questions in economics: why have some countries succeeded with development and others failed? Why has Korea jumped from poverty to prosperity in a lifetime? Why did India do badly then, but much better recently?
The broad question is the one Erik Reinert states in his title: How Rich Countries Got Rich... and Why Poor Countries Stay Poor. Reinert is a Norwegian professor who now teaches at Tallinn, Estonia. Ha-Joon Chang, a well-known Korean development economist, teaches at Cambridge. But both give strikingly similar answers to this question.
L. Alan Winters: Martin's review of these books hits all the right notes. He notes Reinert's advocacy of the nineteenth century model and Chang also frequently makes this argument. Such advocacy needs to be supported by at least a little evidence that the parallels hold for the twenety-first century - after all we do not generally advocate nineteenth century medicine or engineering. One failed parallel is that growth rates then were much slower than governments and their people expect these days. The following data (from Angus Maddison)compare growth in China since 1978 with the most successful industrialisations of large countries in the past. Both the world economy and world leaders now perform far better than they did then.
Annual growth rates over periods of history:
1700-1820 world 0.5%, UK 1.0%
1820-1870 world 0.9%, UK 2.1%, USA 4.2%
1870-1913 world 2.1%, USA 3.9%
1978-2004 world 3.1%, China 7.5%
alternative view of China:
1978-2004 world 3.3%, China 9.6% (World Bank)
Source: L Alan Winters and Shahid Yusuf, 2007, Dancing with Giants: China, India and the World Economy, World Bank, Washington and IPS, Singapore.
Posted by: L Alan Winters | July 24, 2007 at 04:07 PM
Arvind Panagariya (guest commenter): Martin Wolf is to be congratulated for his incisive “two-in-one” review of “How Rich Countries Got Rich...and Why Poor Countries Stay Poor” by Erik S. Reinert and “Bad Samaritans: Rich Nations, Poor Policies and the Threat to the Developing World” by Ha-Joon Chang. Wolf gives the authors a fair hearing but rejects their core arguments favoring protection in one form or the other. Indeed, Wolf is gentler to the authors than their poorly conceived arguments warrant.
Thus, consider the recommendation by Reinert that countries in Africa (or Latin America) must practice free trade among themselves with high barriers to trade against outsiders. Wolf offers devastating conceptual criticisms of such an approach. But since Reinert prefers practical experience to a priori reasoning, let us ask whether the last sixty years of history offers a single example of a developing country achieving sustained rapid growth under the strategy advocated by him. The answer is a resounding no.
Virtually every developing country that has achieved sustained rapid growth in the last sixty years has done so while rapidly expanding its trade on a largely non-discriminatory basis, with the bulk of the trade expansion accounted for by developed country trading partners. It is delusional to think that an African country can rely on other African countries as engines of growth.
Both Reinert and Chang rely on the experience of South Korea among others to advocate infant industry protection. But they misread the Korea economic history. During 1963-72, Korea had registered the average annual growth rate of 9.5 per cent. With rare exceptions, its policies during this period were industry-neutral. The so-called Heavy and Chemical Industry (HCI) drive to which the proponents of infant-industry protection like to allude did not begin until the early 1970s. And the growth rate during the following decade, 1973-82, actually fell to 7.2 percent.
Additionally, various policy measures forced on Korea to sustain the HCI drive culminated in a macroeconomic crisis in the late 1970s. The crisis led Korea to finally abandon the HCI drive and undertake systematic liberalisation in the early 1980s. These measures eventually restored Korea to the higher growth path with the growth rate returning to 9.9 per cent during 1983-90.
If infant industry protection is supposed to be as good as these authors advocate, one must explain why India did so poorly during the 1960s and 1970s. During the 1950s and early 1960s, India and South Korea had both grown at the annual average rate of 4 per cent. In the early 1960s, Korea abandoned the import-substitution policy in favour of outward orientation and accelerated to near-double digit rate. After a brief, failed experiment with liberalisation in the mid 1960s, India turned progressively protectionist and interventionist. It tried to promote the production of every conceivable product by banning its imports. During 1965-81, India grew a paltry 3.2 per cent per annum.
Critics of free-trade advocates like to argue that free trade alone is not the answer to the development problem. This is like knocking down a straw man: few free-trade advocates argue that free trade by itself is enough to launch a country into high-growth orbit. My own forthcoming book India: An Emerging Giant (OUP, NY) is 500-pages long precisely because it carefully spells out the reforms needed in various areas to achieve and sustain a double-digit growth in India. If free trade was all that was required, I could have finished the book in less than hundred pages! But like other advocates of free trade, I am aware that an outward-oriented trade policy regime by itself is not sufficient to guarantee sustained rapid growth. I am also aware, however, that outward-oriented policy regime is usually necessary so that I do devote considerable space to explaining the case for it.
(Arvind Panagariya is a professor at Columbia University)
Posted by: Arvind Panagariya | July 25, 2007 at 01:32 PM
Anne Krueger: In the 1950s and 1960s, economists and policy makers thought that industrialization was “the key” to economic growth. While I agree that rapid economic growth should surely be a major objective of economic policy in poor countries, that does not necessarily imply that industrialization is the only way to achieve it; more importantly, focusing on industrialization as a policy objective is almost certainly wrong. Mechanization and increasing productivity in all sectors usually leads to more rapid growth of industry than of other sectors, but that is the outcome of appropriate policy. While it is highly likely that growth of agricultural productivity – a necessary part of overall economic growth – will shift returns to induce movement of workers to industrial (and service) activities, a focus on industrialization as an instrument, rather than an outcome, can lead to low growth if not stagnation.
Further, as Reinert and Chang apparently do, economists and policy makers regarded trade protection as a major policy instrument for achieving rapid growth through industrialization. Much sad experience has shown that not to be the case. Many countries adopted policies of “import-substitution”, protecting new industries (indefinitely) with import prohibitions or very high walls of protection. There is enough experience to show why that strategy does not work. Martin Wolf’s arguments are valid, but there is much more.
It was not neoclassical economic theory, but primarily the failures of the “import-substitution” strategies that resulted in unsatisfactorily low growth rates. In country after country, high-cost domestic monopolies or duopolies were developed in industry after industry. They achieved little total factor productivity growth, and remained high cost behind high walls of protection. Over time, growth rates slowed as each wave of new industries was more capital-using and higher-cost than the last one. But once established, politicians could not or did not reduce protection and the political power of the industrialists increased. Interestingly, there was little growth in employment, as most producers chose capital-intensive techniques of productin.
It is certainly true that an outward orientation in trade policy will not by itself lead to a satisfactory growth performance. But an outward orientation will, if policy makers are committed to it, lead to policy decisions regarding infrastructure, low tariffs (to enable imports of needed inputs), relatively flexible labor markets, and other spheres of economic activity that are conducive to growth.
I would also note that some countries, pre-China, grew rapidly through opening up their economies. Martin Wolf notes the cases of Korea and Taiwan, whose spectacular growth preceded the Chinese. While it is certainly true that Korea did not abandon all protection during the rapid growth years (l960 to 1990), tariff rates were dropped substantially, quantitative restrictions were abandoned, and the degree of protection to import-competing producers was no greater than the incentives for exporters. And, it is important to note, export incentives were uniform and applicable to all who exported. In that regard, it is also questionable whether 19th century protectionism caused success (in the U.S.) or whether success occurred in spite of protection.
Finally, free trade agreements (or customs unions) among a group of countries with similar factor endowments and commodity prices have been tried and have been found wanting. While there are certainly parts of the international economic system that are not as conducive to poor countries’ rapid growth as they might be, the answer ought to be to recognize the potentially powerful support a rapidly growing international economy can provide for a poor country, and to support those changes in the WTO that will further improve the functioning of the international economy. Even if the current international economy is not ideally organized, it offers a far better hope for poor countries than would an autarkic world or one in which most trade was conducted among preferential trading blocs.
Posted by: Anne Krueger | July 30, 2007 at 10:19 AM
Martin Wolf: I would like to thank Alan Winters, Arvind Panagariya and Anne Krueger for stiffening my spine. I agree with virtually everything they say.
I agree, of course, on the importance of outward orientation for success in development, on the irrelevance of US experience in the 19th century to contemporary development, on the folly of preferential trade agreements among developing countries and on the diminishing returns to import substitution.
That said, I think it is also important for those on “our side” of the argument – the right side, of course – to recognise some of the difficulties.
First, industrialisation has indeed been the only successful route to mass prosperity. The reason is that this is the sector characterised by increasing returns and so to high rates of growth of productivity. High productivity in manufacturing then raises real wages everywhere else. Moreover, those increasing returns are in substantial part externalities: they are the result of positive feedback effects not taken into account by individual industrialists. At the same time, industrialisation cannot work without a great deal of competitive pressure. That is the reason why the import substitution route is ultimately unsuccessful. So, as Anne Krueger remarks, a country does not industrialise successfully by focusing only on industry and certainly not by focusing on a protected home market. But it may well still be interested in policies that promote efficient industrialisation.
Second, world-class businesses do not normally come from nowhere. I recognise that some businesses have emerged in developing countries without any initial push from government: Indian software seems to be a good recent example. But that was not the case for Toyota, Hyundai or Samsung. Is it plausible that Japan and South Korea would have been able to produce world-class manufacturing companies without initial government help? Should countries just rely on inward foreign direct investment, instead?
Third, virtually all the big successes have been countries that started off labour-abundant and began with labour-intensive manufactured exports. India seems to be starting with labour-intensive services, instead. But what should we say about countries with good resource endowments and relatively high initial wages? How should they best start off a true development process?
Finally, we need to admit that, along with the huge recent successes from progressive liberalisation (notably China and India), there have been some disappointments. I am thinking of Brazil and Mexico here. What lessons should sensible people learn from these experiences?
I recognise the extreme importance of not going back to the failed import substitution policies of the past – a point to which I intend to turn to in future columns. But we do know that free trade is not enough. So let’s consider a few of the things needed in addition.
Posted by: Martin Wolf | July 30, 2007 at 05:23 PM
Edmund Phelps: Martin Wolf's July 21/22 review of two books on development was thought-provoking. It made me wonder, what may we economists reasonably say to the underdeveloped aspiring to embark on economic development?
My antennae picked up Martin's growing feeling that economic culture is awfully important. A few years ago I would not have known whether to believe in the importance of culture or not. However, research of mine a year ago (written up for the Venice conference of CESifo/Center on Capitalism and Society in July 2006 and available at www.earth.columbia.edu/ccs) found that various cultural values (the value of competition, the value of work) and especially various workplace attitudes (toward initiative, taking orders, accepting responsibility, etc.) are surprisingly significant in explaining differences in economic performance among OECD countries. But I agree that economic culture is mutable and perhaps can be molded for the better. Maybe globalisation will pull up attitudes where they are wanting.
I appreciated the authors' willingness to tackle big questions. But we should not let pass their argument that manufacturing is always the better bet - because productivity growth will be mainlyin that sector, as it always has. This argument overlooks that some countries are going to produce services and, along an equilibrium trajectory, the relative price appreciation occurring in the services sector will exactly compensate for the lack or deficiency of productivity growth in those areas. If some people didn't think that, there would be no capital invested in services.
The neoclassical prescription is to forget about the question of the best industrial directions to take and leave that to the magic of the market: It will work out in short order what a country's comparative advantages are. This strikes me as way too deterministic and mechanistic. I have discussed in a couple of papers a theory of Chinese development in which that China will be transferring technologies in the areas where the West leaped ahead in previous decades while exporting garments and other goods where the West advanced little. (One message was that as China comes nearer and nearer to the "frontier," wages there will be rising to such heights that its exports of those manufactures will be disappearing - in the end, replicating the US more or less.) The message pertinent here is that a China's comparative advantages of
the current moment lie where the West has to date made few technical gains. More broadly, a country's comparative advantages are entirely artificial and, in particular, depend greatly on the history of innovation overseas.
But, going deeper, it's immediately clear that a country - its "market," if it is given the task - has to think about the directions of the economic advances abroad over a long future. Those prospects will, to take one example, weigh heavily on what values the market decided to put on the various sorts of business assets - commercial space, equipment, employees of various types, etc. There isn't any objective answer. Perfect foresight is not an answer, since there is none. So people are going to have to think where it is best to invest, acquire skills, gain knowledge. There is no deus ex machina to make those calculations for a country.
Last point: When a country innovates -- creating or at any rate adopting NEW goods and/or NEW methods - it may (with luck) be CREATING a comparative advantage that it did not have before. (I was stimulated by the paper of Dani Rodrik and Ricardo Hausmann around 2003 in which they urge Brazil to find its comparative advantages the way one might urge someone to find his inner self.) A Brazilian film maker contemplating making a new movie for the world market does not stop to ask whether it is a manufacture or a service. So it would seem that we have found the solution to development: let everyone who cares to innovate for the global marketplace in whatever directions appear to have the best prospects of financial gain and personal growth.
The trouble is that the people of a country might lack savvy about markets in countries unlike their own. Their backwardness is holding them back from development. So maybe the two authors are right: Manufacturing - presumably of a traditional, standardised, sure-fire kind - will be done will be done by the seriously underdeveloped and as long as that is the best they can do they will remain underdeveloped. The developed world can send money and experts over to the underdeveloped countries to raise their efficiency and maybe their technological practice to a higher level; and that would reduce disease. But unless and until people in a country have an experienced sense or gifted intuition of what might be good to invest in, it will stay underdeveloped.
Posted by: Edmund Phelps | August 02, 2007 at 08:35 AM
Ha-Joon Chang: I am delighted that Martin has taken my book seriously despite our differences and written a considered review. However, I am saddened by the fact that his presentation of my argument in relation to trade has left an unfortunate room for misunderstanding.
Let me first of all make it clear that I do not advocate "across-the-board import substitution", as Martin seems to imply at places. I don’t think he meant it, but the result was to create a mistaken impression of my argument, as evidenced by the comments by Alan Winters, Arvind Panagariya, and Anne Krueger. I go to a great length in the book to explain why trade is essential for economic development. I daresay that I am even more pro-trade than those mentioned above. But that is not the same as saying that free trade is the best way to promote trade and economic development.
Country after successful country, from the 18th-century Britain to the late-20th-century South Korea and Taiwan, have first used various policy measures (trade protection included in most, although not all, cases) to create the space in which their producers can build up their productive capabilities before they can compete with better producers from abroad, either in the domestic market or in the export market In some cases the protection was brief, while in others it was long – the Japanese car industry needed nearly four decades of protection and subsidies before it could successfully compete in the world market (but then those were totally worth it for everyone, not just the Japanese, because those policies eventually created the most efficient and environment-friendly car industry in the world). But the intention behind successful protectionism was always to fully join the world economy.
There is huge difference between "across-the-board import substitution" that was practised in some countries and the kind of purposeful (although not necessarily less protectionist, if you just look at things like average tariff rates) protectionism that I advocate. I feel like a man being accused of promoting a copious consumption of vodka when all I have done is to recommend moderate amount of red wine as a part of balanced diet.
Not all protectionist countries have succeeded. Many have done poorly. However, I present evidence, in the book and in my academic writings, that most countries that have succeeded used protectionism to one degree or another. Typically, countries have liberalised trade after, rather than before, they became rich. Most countries that adopted free trade prematurely have failed. And I am not talking only of the poorest countries. There are many middle-income countries that had similar experience – Mexico and Brazil in the recent period being the best examples that Martin himself cites in his response to comments by other people. Today’s developing countries had pitiful growth records (if they grew at all) when they were practising free trade under colonialism and unequal treaties for a century until the Second World War. In the recent period, developing countries in Latin America and Sub-Saharan Africa grew much faster during the "bad old days" of protectionism than they have in the last two decade of neo-liberal policies, including free trade. Per capita income in Latin America grew at 3.1 per cent in the "bad old days" of the 1960s and the 1970s. In the "brave new world" of neo-liberalism between 1980 and 2004, it grew at 0.5 per cent. Even if we leave out the 1980s as the decade of adjustment, growth rate in Latin America was only about 1.5 per cent between 1990 and 2005. Moreover, growth rate in Latin America has fallen recently (growth rate in 2000-2005 was only 0.6 per cent), when the continent should be finally reaping the gains from the reforms (that have lasted for nearly two decades in some countries).
When discussing the (alleged irrelevance of) lessons of history, Martin pays almost exclusive attention to the US case. This is understandable, as the fact that the country used to be the most protectionist in the world for over a century until the Second World War is such a shock, given its role as the standard bearer of free trade in the recent past. Martin, as Anne Krueger does in her comment, argues that the US was an exception that grew ‘despite’ protectionism. I do not agree with this interpretation, but even if it were the case, how does he propose to explain the fact that virtually all of today’s rich countries, and not just the US, have protectionist past? Perhaps the US succeeded despite protectionism because of its large market, but then how about Taiwan or Finland that had tiny markets? If large-scale, high-quality immigration offset the ill-effects of protectionism in the US, how about countries like Sweden or Germany that were losing qualified people and practising protectionism in the late-19th century? It may be possible to dismiss the US as an "exception", but if there are another two dozen countries that have to be dismissed as "exceptions", then the theory has simply too many holes (the exercise reminds me of the pre-Copernican practice of drawing "epi-circles" in order to square evidence with geo-centrism).
I agree with Martin that I do not say much on what very poor economies should do. I see it as a shortcoming of my work. However, let me remind Martin that South Korea’s per capita income in 1961, at $82, was less than half that of Ghana ($179) and the country was described as a "bottomless pit" by the USAID in the late 1950s. If Korea could do it, why not other countries? I agree that the challenges facing the poorest countries today may be bigger than the ones faced by Korea in the 1960s. But one important reason for that is the intolerance for those "nationalistic" policies used by Korea in the past, on the part of the rich countries and multilateral institutions. My plea in the book is that this has to change and countries should be allowed to choose what they need and want, as such choice usually produce better results, as I have pointed out earlier. Martin summarised this argument at the end of his review much better than I can, so I will not elaborate on this point.
As for the comments of Alan Winters and Arvind Panagaryia, I hope they had at least the decency of reading the book (or at least reading the review more carefully, for Martin does give me a fair hearing except in the area of trade) if they were to denounce it with strong languages ("ill-conceived", said Arvind, and Alan likens my argument to ‘nineteenth-century’ medicine). For I don’t recognize my work in their caricaturisation of my argument as solely trade-focused and based on some outdated model and no empirical evidence.
Alan’s assertion that my argument is not serious because it is based on the "nineteenth century model" is very unfortunate. If he read the book, he cannot possibly accuse me of advocating a nineteenth century model – for my discussion is right up to today’s practice, such as huge R&D subsidies that the US gives to its industries to harsh regulation of foreign direct investment by Japan and Finland until the 1980s. If there is an idea in this debate that is "nineteenth-century", it is Alan’s implicit belief that human knowledge progresses linearly. This idea has proven problematic even in natural sciences, but no philosopher of science would take it seriously these days when it comes to social sciences. The interesting thing is that Alan would be actually in greater trouble if human knowledge progressed linearly and therefore newer theories were always better – for protectionist theories are newer than free trade theories. Finally, I think it is wrong to dismiss one’s opponent’s theory by labelling them with negative words (‘nineteenth-century’). How would Alan feel if I described him and his colleagues as "defenders of free-trade theory that was so strongly advocated by American slave-owners and opium-trafficking British imperialists"?
Arvind’s argument that the failure of Heavy and Chemical Industrialisation (HCI) in Korea proves the wisdom of free trade may have cut some ice when Ian Little first said it in the early 1980s, when many of the HCI industries were in trouble (teething problems complicated by world recession), but it is a hollow argument today. Let’s get some facts straight. Korea did go through a bad patch in the late 1970s and the early 1980s, but it was in part due to the Second Oil Shock and the ensuing world recession, not just because of the HCI programme. Moreover, Korea did not abandon the HCI programme after the shock, as Arvind suggests. It re-organised and closed some of the firms that had little hope of doing well, but it kept supporting the good ones. Finally, if the HCI programme was a failure, how does Arvind explain that most of the leading industries of the 1980s onwards – steel, shipbuilding, automobile, electronics – have been the ones that had been set up and promoted through HCI, often against the advice of free-trade economists? (This is in fact a question that Ian Little himself could not answer when I put it to him in the early 1990s).
As for Arvind’s invocation of the Indian case as the proof that infant industry protection is not good, I have no problem in admitting that not all protectionist policies are going to be successful and that India’s protectionism especially until the 1970s was not well managed. Whatever the strategy is, there are going to be failures. All I am showing in the book is that, contrary to the conventional wisdom, protectionist strategy has a much higher chance of success than free-trade strategy. Invoking the Indian example to dismiss the infant industry argument is like arguing for the abolition of parenting because there is a man in his fifties who is still living off his parents.
Anne’s languages are gentler than those of Alan’s and Arvind’s but the substantive disagreement is no less. She says that while Korea "did not abandon all protection during the rapid growth years (1960 to 1990), tariff rates were dropped substantially, quantitative restrictions were abandoned, and the degree of protection to import-competing producers was no greater than the incentives for exporters". This kind of characterisation has already been discredited by the early 1990s among those who work on East Asia. Let me remind her some of the evidence. Average tariff rates, at 20-30 per cent, were still quite high during the period. Moreover, the average conceals very high variation – the country had zero tariffs for some products while others had tariff rates around 100 per cent (so going against her advice for ‘non-discriminatory’ policies). Quantitative restrictions were widespread – there were a lot more explicit QRs than Anne and her people acknowledge, and many QRs were not very visible because they were practised through domestic laws (for example, laws governing the state-owned railway companies). Most importantly, Korea practised strict foreign exchange rationing by the government until the 1980s, so tariff figures do not give us the real picture of protectionism across sectors. There were many items whose tariffs were relatively low on paper but which simply could not be imported (so facing an effective tariff rate of infinity) because they were far down the government priority list.
Finally, I can only agree with most of the things that Edmund Phelps said in his comment. His view on economic development is, although expressed in a slightly different language, exactly the same as mine. As he puts succinctly, I believe that "a country’s comparative advantages are entirely artificial". In other words, economies develop in the particular ways they do because someone somewhere makes a conscious decision to "invest acquire skills, and gain knowledge" in particular areas. In the same way in which business managers often go deliberately against market signals in making such decision (Nokia ran losses in its electronics business for 17 years), governments can, and often should, go against market signals. Of course, if that is done to permanently insulate a country from international market forces, it would lead to a disaster like North Korea, but as far as the final aim is to fully join the world economy, protecting the car industry for 40 years (as Japan did until the 1970s) or setting up a state-owned steel monopoly against the advice of the World Bank (as Korea did in the early 1970s) may be justified.
Reading the comments by Alan, Anne, and Arvind, I feel like this. I am a man whose book recommending the Mediterranean diet has been reviewed by a well-known anti-fat dietician, who unintentionally misrepresented me as praising beneficial qualities of all fats, when I had only praised olive oil. This was bad enough, but then a few other anti-fat dieticians read the review, go into a Pavlovian reaction on seeing the word, "fat", and accuse me of promoting excessive consumption of all fats, brandishing American obesity figures and Scottish heart-attack statistics. I regretfully have come to conclusion that I was absolutely right to say what I said at the end of chapter three in the book – "Trade is simply too important for economic development to be left to free trade economists".
(Ha-Joon Chang teaches in Cambridge and is the author of Bad Samaritans)