Debunking Protectionist Myths: Free Trade, the Developing World, and Prosperity
By Arvind Panagariya
More than 170 years ago, Frédéric Bastiat noted in his masterly work Economic Sophisms that the “opposition to free trade rests upon errors, or, if you prefer, upon half-truths.”1 Ever since Adam Smith successfully replaced mercantilist orthodoxy with free trade doctrine in his celebrated book The Wealth of Nations, free trade critics have repeatedly challenged the doctrine, offering half-truths to bolster their case. In each instance, free trade advocates have successfully exposed the falsehood of arguments made by critics. Although free trade has gained increasing acceptance among policymakers over time, challenges to it have remained omnipresent.
The latest of these challenges has manifested itself in increased tariffs on steel and aluminum in the United States and on a number of selected products in India. At the heart of these tariff hikes has been the belief that through targeted protection and industrial policy, governments can produce outcomes that are superior to those that free trade and competition would produce.2 Intellectual inspiration for this belief in recent decades has come from writings of a group of influential scholars who have interpreted the experiences of the highly successful East Asian “tiger” economies — Hong Kong, Singapore, South Korea, and Taiwan — during the early decades following the Second World War and of China during more recent decades as being the result of selective protection and industrial targeting.
Systematic evidence, however, demonstrates that free trade rather than selective protection and industrial policy must be credited with propelling these economies to miracle-level growth. Just as Bastiat observed, the case made by free trade critics in favor of industrial policy and selective protection is based on half-truths. Contrary to the assertions by these critics, a logical case for infant industry protection does not exist. Moreover, compelling empirical evidence linking trade openness causally to higher per capita incomes is now available.
A Quick Historical Perspective
In the immediate aftermath of the Second World War, there was consensus among economists and policymakers that economic recovery in industrial countries required progressive opening of trade among them. Simultaneously, it was agreed that newly independent developing countries needed protection so that they could industrialize by substituting domestic output for imported manufactures. The former idea led to the signing of the General Agreement on Tariffs and Trade (GATT), which became the vehicle for progressive liberalization of trade among industrial countries. The latter idea led to the grant of special and differential treatment to developing countries within the GATT framework. During the early decades following the Second World War, these countries got full freedom to protect their industries.
The idea that import substitution industrialization (ISI) was the right policy for the newly independent developing countries had its origins in the assumption that their comparative advantage lay in primary products and that exports of these products could not serve as the engine of growth. The reason was that both income and price elasticities of demand for these products were low. Low income elasticity meant that over time, rising incomes in industrial countries would shift global demand away from these products and thus shift the terms of trade against them. Low price elasticity meant that any efforts by developing countries themselves to expand exports through increased investment or enhanced productivity would lead to a sharp decline in primary product prices, resulting in reduced export revenues.
These logically correct arguments led economists and policymakers to the conclusion that faster growth required industrialization, and hence, protection. The “infant industry” argument was then invoked to impart intellectual legitimacy to protection for industry. Thus, in the immediate aftermath of the Second World War, virtually all developing economies wound up embracing import substitution. Only Hong Kong, which the British had owned and maintained as a free port, remained a free trading entity.
Interestingly, however, by the early 1960s, Singapore, Taiwan, and South Korea broke away from this consensus. Having completed the substitution of domestic output for imports of labor-intensive products, they were faced with choosing between extending import substitution to more capital-intensive products or expanding further into labor-intensive products by switching to export expansion. Recognizing the small size of the domestic market, they opted for the latter strategy and became progressively outward-oriented. The strategy proved an unqualified success. They could achieve increases in per capita incomes and reductions in poverty in three decades what Western industrial economies had taken more than a century to achieve.
The success of these economies exposed a key flaw in the model on which the original consensus in favor of ISI was based. By conceptualizing the economy as consisting of only two sectors — primary products and industry — the model ended up arguing that ISI offered the only road to industrialization. What the experience of the East Asian tigers revealed was that the industrial sector was not a monolith but a collection of many products, of which some were labor-intensive and others capital-intensive. It was therefore possible for labor-abundant developing countries to industrialize by specializing in and exporting labor-intensive industrial products while importing capital-intensive ones.3
Comparative studies of East Asian tiger economies and economies that remained wedded to ISI, such as India, Mexico, and Egypt, were carried out during the 1970s and early 1980s, which led to a complete turnaround in the conventional wisdom. Economists such as Bela Balassa, Jagdish Bhagwati, Anne Krueger, and Ian Little concluded that openness to trade was as desirable for developing countries as for developed countries. In the following years, these economists emerged as influential proponents of industrialization and development through outward-oriented policies.
Influenced by this new conventional wisdom, and also under pressure from U.S. president Ronald Reagan, who firmly believed in the efficacy of open markets, the International Monetary Fund and the World Bank went on to aggressively promote trade liberalization in developing countries during the 1980s. Predictably, the turnaround in academic opinion in favor of free trade and its embrace by influential international financial institutions produced a reaction from pro-protection forces. This reaction found expression in what has been called a revisionist interpretation of the experiences of the East Asian tiger economies. Political scientists Alice Amsden and Robert Wade argued that the success of South Korea and Taiwan, respectively, was the result of cleverly designed industrial policies and selective protection.4 Economists Dani Rodrik and Ha-Joon Chang later voiced their agreement with Amsden and Wade.5
Although pro-free trade economists such as Bhagwati and Little have challenged some of the arguments of revisionists, a systematic response to the latter and a full-scale defense of free trade as the engine of growth and poverty alleviation in developing countries has been lacking. This is the task I undertake in my book Free Trade and Prosperity.6 In the following, I offer some flavor of the book by exposing a number of myths spread by revisionists. The first myth relates to the superiority of the ISI approach to development, taking developing countries as a whole. The remaining myths relate to the experiences of fast-growing developing economies, most notably those of South Korea and Taiwan.
Myth 1: The Years 1960-1973 Represent the Golden Age of Growth in Developing Countries
Writing in 1999, Rodrik argued that taken together, developing countries grew the fastest during 1960-1973 when they followed inward-looking, import-substitution industrialization policies.7 Later, Chang forcefully echoed this argument in his 2007 book.8 But there are three serious problems with the thesis.
First, factually, developing countries as a group did not grow the fastest during 1960-1973. As Table 1 makes amply clear, developing countries have grown the fastest during the decades following 1990. This was the period during which these countries came to genuinely embrace and own liberal policies instead of being forced into accepting them by international financial institutions in return for access to financial resources. At the time Rodrik wrote, he may have lacked these data, but by 2007, when Chang published his book, available evidence was loud and clear.
Second, had Rodrik gone into individual-country details, he would have found that even during 1960-1973, the fastest-growing economies were those that had embraced outward-oriented policies. I have already mentioned the four East Asian tiger economies, which grew at rates ranging from 8 to 10 percent during this period. But even Brazil, a much larger country that saw its growth accelerate during this period, had brought down its tariffs and devalued its currency multiple times to correct for overvaluation of the latter.
Finally, the OECD countries had grown significantly faster during 1960-1973 than during post-1990 decades. As such, developing-country growth during the earlier period received a boost from OECD growth. Similar pull-up effect had been missing from the post-1990 period. Instead, the impetus for growth in developing countries during this period came from their own economic policies, including trade liberalization.
Myth 2: Industrial Policy, Including Selective Protection, Was behind the Success of East Asian Tiger Economies
This is the key claim made by free trade critics, which has given an edge to continued advocacy of protection by many. But consider the experience of South Korea. As Table 2 shows, the country grew at an annual average rate of 9.1 percent during the decade 1963-1973 compared with 4.2 percent during 1954-1962 and 6.9 percent during 1974-1982. Years 1954-1962 are identified with import substitution, while years 1963-1972 saw South Korea adopt an export-oriented strategy. This latter decade was characterized by policies that were sectorally neutral. Selective industry promotion was limited to cement, fertilizer, and petroleum refining in the early 1960s and to steel and petrochemicals in the late 1960s and early 1970s. Calculations by Larry Westphal show that when the economy-wide implications of all interventions are considered, the policy regime exhibited a slight bias in favor of exports relative to what would have prevailed under free trade.9 Among other things, neutrality gave rise to growth of sectors no one had predicted: wigs and human hair exports, entirely absent until 1963, came to account for 10.1 percent of Korean exports by 1970.
When critics claim success for industrial targeting, they entirely eschew the discussion of the crucial decade of 1963-1973. Instead, they focus on the following decade, in which Korea did engage in a heavy and chemical industry (HCI) drive. But the growth rate during 1974-1982 actually fell to 6.9 percent. Moreover, toward the end of this period, the economy faced serious macroeconomic instability, culminating in the abandonment of the HCI drive and the restoration of a neutral policy regime. That in turn returned the country to 8.7 percent growth during 1983-1995. Chang has claimed that the policy was nevertheless successful because industries promoted under the HCI drive eventually became profitable. But this amounts to a post hoc fallacy. After a decade of rapid growth and near double-digit annual increases in real wages, South Korea had been becoming more and more labor-scarce and capital-abundant. Therefore, capital-intensive sectors promoted under HCI would have emerged even absent the HCI drive. What the HCI drive did was to advance that process by a few years. To legitimately claim his case, Chang must demonstrate that the benefits of advancing the process exceeded its costs.10
Myth 3: Export Expansion Cannot Be Credited with Catalyzing Growth Because It Followed, Rather than Led, the Acceleration in GDP Growth
Rodrik has argued that expansion in exports in Korea and Taiwan actually followed acceleration in growth. Therefore, expansion could not have catalyzed growth. There are two counterarguments here. First, even if the catalyst to growth was domestic in nature, it is highly unlikely that these countries could have sustained 8 to 10 percent growth for several decades without a massive expansion of exports. For example, in South Korea, exports expanded from just 5 percent of the GDP in 1965 to more than 20 percent by 1972, and imports rose from 10 percent to more than 25 percent of GDP over the same period. By the time South Korea seriously got down to targeted promotion of HCI, it was already a highly open economy.
Second, and far more important, Rodrik is wrong about his claim that exports were not a catalyst to growth. His error lies in the failure to disaggregate the total exports into its components. The shift in GDP growth to more than 8 percent in 1963 from less than 5 percent in the prior years had been preceded by a gradual shift in policy toward reducing anti-export and pro-import-substitution bias in policy. The first major step in this direction in 1959 eliminated tariffs exporters paid on inputs contained in their exports. In the early 1960s, exporters also got exemption from indirect taxes. By the late 1950s, the exchange rate had also become considerably overvalued. Devaluation of domestic currency from 65 won per dollar to 100 won per dollar in January 1961 and to 130 won per dollar in February 1962 brought it closer to the market rate. The government also worked toward removing infrastructure-related barriers to trade, especially at ports.
These measures produced a salutary effect on the exports of manufactures during the early 1960s. Between 1961 and 1964, they grew at the average annual rate of 87.9 percent per annum. This rate was higher than in any other subsequent four-year period. Over the same period, the share of manufactures exports in the total exports rose from 21.9 percent to 62.3 percent. Total exports mask this major structural shift in exports. Moreover, because primary product exports performed poorly during the early 1960s, total exports also give the misleading impression that exports were unimportant to the shift in the growth rate beginning in 1963. This point applies equally to Taiwan.
Myth 4: Exports Were Too Tiny to Have Been the Engine of Growth
Rodrik has also argued that in the first half of 1960s, exports as a proportion of GDP were too small to serve as the engine of growth in South Korea and Taiwan. Although plausible on the surface, this argument, too, fails to withstand close scrutiny. There are two problems with the argument. First, even if export sales were small in relation to the GDP, the total sales of exportable products were not. The latter include domestic sales of export products. When profitability of exports rises and sales of export products are diverted from domestic to foreign markets, domestic prices of those products rise, making domestic sales profitable as well. Therefore, the pull effect of export incentives works not just on exports but on domestic sales of export products as well. Reinforcing this factor is the ability of efficient export firms to exploit scale economies. Vastness of the export markets enables these firms to rapidly expand and lower production costs, which in turn enables them to expand domestic sales.
Second, as Bhagwati has pointed out, improved export incentives such as duty-free entry of inputs used in exports, exemption from indirect taxes, and elimination of overvaluation of the exchange rate enhance the profitability of not just existing export products but also potential export products.11 Sufficiently large export incentives may turn many nontraded but tradable products — and even imported products — into export products. For example, wigs and human hair were entirely absent from South Korea’s export basket until 1963. But by 1970, they came to account for 10.1 percent of its total exports. Similarly, Taiwan exported no electrical machinery and appliances until 1959. They made their debut in 1960 and came to account for 12.3 percent of Taiwan’s vastly expanded total exports by 1970. Clothing and footwear had expanded from 0.8 to 2.6 percent of the total exports during the import-substitution phase from 1952 to 1960, but they shot up to 16.8 percent in 1970.
Myth 5: Success of Taiwan and South Korea Is Proof that Interventions Helped, Rather than Hurt, Growth
In his book on Taiwan, Wade offers a catalog of government interventions, big and small, without a coherent explanation of how they added up to the growth miracle and whether these interventions would have led to the miracle without the policies identified as important by advocates of outward-oriented strategy. He makes repeated references to the government acting strategically in specific contexts, but without articulating a “strategic action” model of economic development that he could recommend to other countries. The bottom line he offers is this:
The fact of big leadership or big followership does not mean that government intervention has been effective in promoting economic growth; it only means that government intervention cannot be dismissed as having made a negligible difference to outcomes. But the balance of presumption must be that government industrial policies, including sectoral ones, helped more than they hindered. To argue otherwise is to suggest that economic performance would have been still more exceptional with less intervention, which is simply less plausible than the converse.12
This statement illustrates in sharp relief how revisionists set a very low standard when it comes to providing the proof of their own thesis in comparison to what they demand from free trade advocates.13 More important, they shy away from asking critical questions that may lead them to an answer they may not like. This is the point Little made when he responded to Wade’s claim in these terms: “Since the less interventionist Hong Kong, Singapore, and Taiwan grew faster than Korea, it is unclear why Wade thinks it simply less plausible that less intervention would have been better, given also the widespread failure of government industrial policies elsewhere. I find it simply more plausible that Korea grew fast despite its industrial policies, than because of them.”14
Echoes of the argument made by Wade can also be heard in the argument made by Rodrik and Chang to explain the more recent successes of China and India. Like Wade, Rodrik argues that because numerous government interventions remain present in China, its experience does not support the case of trade liberalization. Chang goes a step further by arguing that China and India succeeded because they refused to wear a free trade straitjacket. But liberalization in the early 1980s had already placed China on a 10 percent growth trajectory. If protection and interventions that still remained were behind this success, further liberalization should have hurt its growth. But it was precisely through sustained liberalization, culminating in its entry into the World Trade Organization in 2001, that China sustained its high growth. Likewise, it took the dismantling of a large number of interventions for India to finally see its economy grow at an 8 percent rate beginning in 2003. Subsequently, as it suspended the process of import liberalization after 2007 and returned to more interventionist policies during 2009 to 2014, its growth suffered.
History forcefully demonstrates the power of openness to trade. Between 1960 and 1990, East Asian tiger economies succeeded in achieving increases in per capita income that Western industrial economies took a century to achieve. Their growth also led to the elimination of abject poverty despite no significant redistributive social programs. During 1980 to 2010, China has achieved the same success for its much larger population by shedding its Mao Zedong-era autarkic policies and giving greater play to markets. Today, India is poised to achieve something similar for its equally large population, provided it does not descend back into its failed illiberal external and internal policies. Lessons from the experiences of these countries apply equally to the developed world. The United States, in particular, must weigh the harmful long-term consequences of its recent turn to protectionism. It should not forget that in the medium to long term, a tax on imports is a tax on exports even when partner countries do not retaliate. When partner countries retaliate, the damage compounds.
1 Frédéric Bastiat, Economic Sophisms, trans. and ed. Arthur Goddard (Irvington-on-Hudson: Foundation for Economic Education, 1996), p. 3, http://www.econlib.org/library/Bastiat/basSoph1.html. Italics are as in the original.
2 Protection has also seen an upsurge in the form of a trade war between the United States and China. The initial intent of the United States behind the tariffs against China was to press the latter into opening its markets wider. But lately it seems to have shifted to taking the view that the tariffs are helping local industry grow. As such, even the trade war with China has come to have a protectionist angle to it. The shift in thinking on the part of the United States is also reflected in its desire to expand the scope of protection to auto imports.
3 Anne Krueger, “Trade Policy and Economic Development: How We Learn,” American Economic Review 81, no. 1 (1997): 1-22.
4 Alice Amsden, Asia’s Next Giant: South Korea and Late Industrialization (New York: Oxford University Press, 1989); and Robert Wade, Governing the Market: Economic Theory and the Role of the Government in East Asian Industrialization (Princeton: Princeton University Press, 2004).
5 Dani Rodrik, “Getting Interventions Right: How South Korea and Taiwan Grew Rich, ”Economic Policy 20 (1995): 55-107; Ha-Joon Chang, Bad Samaritans: Rich Nations, Poor Policies and the Threat to the Developing World (London: Random House Business Books, 2007); and Larry E. Westphal, “Industrial Policy in an Export-Propelled Economy: Lessons from South Korea’s Experience,” Journal of Economic Perspectives 4, no. 3 (1990): 41-59. Economist Larry Westphal and his coauthors were among the earliest to study the success of South Korea, but they took a more nuanced view of its experience. Westphal partially credited industrial policy for the success of South Korea but also saw openness as being critical to it. He also felt that effective use of industrial policy requires very able and effective leadership, which is usually lacking in most developing countries.
6 Arvind Panagariya, Free Trade and Prosperity: How Openness Helps the Developing Countries Grow Richer and Combat Poverty (New York: Oxford University Press, 2019).
7 Dani Rodrik, The New Global Economy and Developing Countries: Making Openness Work(Washington: Overseas Development Council, 1999).
8 Chang, Bad Samaritans.
9 Westphal, “Industrial Policy in an Export-Propelled Economy,” Table 1.
10 Economists David Dollar and Kenneth Sokoloff provide more direct evidence of relatively poor performance of the highly capital-intensive sectors supported by the HCI. They note, “It is interesting that, of the industries supported by the HCI program, it is the very capital-intensive ones that exhibit poor TFP [total factor productivity] growth, while those of medium and light intensity generally show high TFP growth” (p. 322). David Dollar and Kenneth Sokoloff, “Patterns of Productivity Growth in South Korean Manufacturing Industries, 1963-1979,” Journal of Development Economics 33 (1990): 309-327.
11 Jagdish Bhagwati, “The ‘Miracle’ That Did Happen,” in Erik Thorbecke and Henry Wan, eds., Taiwan’s Development Experience: Lessons on Roles of Government and Market (Boston: Kluwer Academic Publishers, 1999), pp. 21-39.
12 Wade, Governing the Market, pp. 305-6. At the beginning of this quote, Wade uses the term “big leadership” to describe a situation in which the government leads private entrepreneurs through initiatives that significantly alter their investment and production patterns. Analogously, he uses the term “followership” to capture a situation in which the government follows the lead of private entrepreneurs in designing its interventions.
13 When evaluating the thesis advanced by Ian Little and others, revisionists, including Wade, demand that they demonstrate that their policy package offers a sufficient explanation of the Taiwanese miracle and not merely a positive contribution to it on balance. But for his own thesis, Wade wants to get away with simply demonstrating that the government industrial policies “helped more than they hindered” the process of development. In a similar vein, in relating trade to higher per capita income, free trade critics demand a causal connection between the two according to the highest standards of econometrics. Yet they have not even made an attempt to show that high protection is positively correlated with high per capita incomes, let alone tried to establish causation between the two variables.
14 I. M. D. Little, “Trade and Industrialization Revisited,” Pakistan Development Review 33, no. 4 (1994): 365.
Arvind Panagariya is a professor of economics and the director of the Deepak and Neera Raj Center on Indian Economic Policies in the School of International and Public Affairs at Columbia University. He is the author of Free Trade and Prosperity: How Openness Helps Developing Countries Grow Richer and Combat Poverty (Oxford University Press, 2019), from which this bulletin draws.