Theophile POUGET-ABADIE, (2016), London School of Economics and Political Sciences, Msc Economic History.
Latin America is a fascinating case for anyone studying modern economic development. How does one explain the economic divergence of North and South America? How do you account for the East Asian miracle on one side of the Pacific, and the “lost decade” of the 1980s on the other? When placed in perspective, Latin America’s current level of development is perplexing at best, shameful at worse. In the late 19th century, it was a land of opportunity. Thousands of Europeans migrated to the continent to profit from the abundant natural resources and nascent manufacturing sector. In 1890, Argentinian GDP per capita was twice that of South Korea, and higher than France or Germany. Today, the situation is reversed.
Over the course of the 20th century, Latin America oscillated from free-market capitalism to state-led dirigisme. The latter was epitomized in the import substitution industrialization (ISI) strategy adopted in the mid-century. The question we ask ourselves is: how successful was ISI policy in promoting economic development in Latin America?
By definition, ISI is an attempt by economically less-developed countries to break out of the world division of labor. It consists of establishing domestic production facilities to manufacture goods that were formerly imported. The textbook history of Latin America can be summarized as follows. From the 1870s to the 1930s, Latin Americans played by neoclassical rules, fully integrating in the world economy. But the Great Depression and World War II forced autarky upon them. Faced with shortages and deteriorating terms of trade, they adopted ISI policies. However, in the 1980s, a series of debt crisis and subsequent IMF structural adjustment programs ended the ISI experience. The fact that it was abandoned in the 1980s and 1990s seems to point to its failure. Or does it? Does the data corroborate the failure of ISI? If so, was it a failure of policy, or of policy implementation? Does this help us understand the divergent experiences of Latin America on the hand, and East Asia on the other?
The thesis I will be defending is that it was not so much ISI in itself but its implementation that failed. This failure stems from the fact that it did not address the fundamental causes of economic growth: physical and human capital accumulation.
In the first section, we will begin by looking at the standard story of ISI in Latin America. In the second section, we will outline three important qualifications, drawing from Rodrik, Haber and Baer. These nuances will enable us to elaborate our thesis in the third section.
Import substitution in Latin America: the story of policy failure
We begin with a brief historical overview of the implementation of ISI in Latin America. From the late 19th century to the mid-20th century, Latin American development was consistent with the neoclassical ideas of comparative advantage and free trade. Export-oriented growth, primarily of primary products, made the continent rich.
|GDP pc||Argentina||Brazil||Japan||South Korea|
Figure 1: GDP per capita in Latin America and Asia (1990 international dollars)
This was due to a fall in transportation costs, waves of foreign investments and emigration from Europe to Latin America. However, the Great Depression and World War II stand as a watershed. The decline in international trade and finance cut the continent off from foreign capital, and the terms of trade deteriorated. Thus, trade volumes collapsed, as did foreign borrowing. Lacking important imports, the recovery was built around ISI. This strategy found its theoretical justification in the 1950s with the works of Raùl Prebisch and the ECLAC. The main thesis is the following: in international trade, the terms of trade will always turn against the periphery, favoring the center. To break this center-periphery system, an autarkic response through protectionism, trade and capital controls, is required. In the case of Latin America, what were the instruments used? Werner Baer provides a non-comprehensive list1:
- Protective tariffs and/or exchange controls,
- Preferential import exchange rates for industrial raw materials,
- Cheap loans by government development banks,
- Construction by governments of infrastructure especially designed to complement industries.
Thus, ISI was a policy implemented in Latin America from the 1930s to the 1980s as a response to the international environment.
What are the results of ISI on Latin American development? At first sight, they appear dismal. Indeed, the GDP pc numbers show a clear reversal of fortunes between Latin America and Asia in the 20th century.
|GDP pc||Argentina||Brazil||Japan||South Korea|
Figure 2: GDP per capita in Latin America and Asia (1990 international dollars)
After World War II, Latin America’s share of world exports declined sharply: from 11% in 1950 to 4% in 1975. As early as 1972, Baer sensed the failure of ISI to achieve its goals2. He notes that in the late 1960s, over 90% of Argentinian and Brazilian exports still consisted of primary products. In his introduction, he writes that by the late 1960s, “industrial growth had slowed, job opportunities in industry for Latin America’s rapidly growing urban population were scarce, income distribution had in many countries either remained unchanged or had become more concentrated than in the early post-World War II years, and most industrial goods produced within the region were priced so high that export possibilities were severely limited.” When contrasted with the East Asian experience, typically characterized as export or outward oriented, ISI policies in Latin America seem to have failed.
The shift to outward orientation in the 1980s brought the ISI experience to an end. In the 1970s, the failure of ISI had become clear to economists and policy makers. Thus, Latin American governments were faced with a dilemma. They could put an end to market distortions by opening up to foreign trade (which meant tearing down trade barriers and devaluing their currencies). Or, they could benefit from the boom in international capital markets and borrow to subsidize a system that had run into decreasing marginal returns. This is what they chose to do. In the words of Stephen Haber, they “bought time by borrowing abroad”3. This ended in numerous debt crises in the 1980s, and the IMF stepped in and brought an end to ISI through structural adjustment programs that liberalized trade and capital flows. Overall, the story seems to be that, in a world of globalization, inward looking development is very costly. Latin America missed out on the benefits of openness, championed in the Washington Consensus. These are: imports of investment and intermediate goods that are not available domestically at comparable costs, access to foreign savings, and the transfer of technology and ideas. In contrast, ISI was distortionary and prevented an efficient allocation of resources along the lines of comparative advantage. Critics point to the indiscriminate way in which ISI was carried out. Given small domestic markets, limited capital and a dearth of skilled manpower, autarkic industrial growth led to the development of inefficient and high cost industries, export structures lacking diversity and an agricultural sector bereft of investments. In contrast, the arguments in favor of openness were laid out in an article by Sachs and Warner4. They argued that open economies grew faster than closed economies over the course of the 20th century.
Under this light, the ISI experience was a complete failure in Latin America. But this textbook version has been contested and nuanced by different economic historians.
In the following section, we will qualify the textbook story of ISI in three different ways.
A critical review of the standard ISI-story: three vital nuances from Baer, Rodrik and Haber
To begin with, one could argue that all countries must go through some form of ISI. This is the argument Baer makes in his 1972 essay. He argues that Western Europe and the United States went through a stage where investments in industries were made to replace imports from Britain. Protectionism was used to shield nascent manufacturing sectors. He also points to the decisive role of the State in the industrialization of these two regions over the course of the 19th century. In an 1892 speech, US president McKinley even gave a moral justification for protectionism: “Under free trade the trader is the master and the producer the slave. Protection is but the law of nature, the law of self-preservation, of self-development, of securing the highest and best destiny of the race of man”5. In the same vein, Stephen Broadberry describes how European states led the way in infrastructure, and in railroads in particular6. For example, the French state provided hefty subsidies to private investors. Even in Latin America prior to World War I, the region had very high tariffs on a select group of goods. Moreover, East Asian countries applied ISI policies during their “miracle”, and the Chinese economist Zhu argues that the success story of China and Taiwan is that of export-orientation supporting (but not displacing) import substitution7. Thus, why didn’t Latin America follow Western Europe’s path in the 19th and 20th century? Baer points to the socio-economic structures in Latin America. Elites controlling attractive external markets for the region’s primary exports had little desire to change the status quo. Moreover, Latin America did not have the entrepreneurial classes, labor force, infrastructure, market size, or administrative capacity to cope with an extensive industrialization process. Here, the benefits of outward orientation over import substitution, as they appear in the standard story, are not as straightforward as they first seemed. Interestingly, Baer underlines the importance of institutions in the failure of Latin America.
To add to this complexity, Rodrik raises interesting issues in his book Making Openness Work8. He points to the fact that ISI actually worked in the 1960s and 1970s, with high GDP growth rates. Moreover, as the argument goes, when Latin America did end up liberalizing trade, reforming labor markets and opening up to globalization, the rate of industrial growth and TFP expansion was way below the levels experienced under the ISI period. Therefore, the developmental failure was not of ISI policy but of the external shocks that wrecked Latin economies in the 1980s. Essentially, the point is that openness has not worked much better for Latin America. This is easily borne out by the evidence in the “lost decade”, the debts crises in the 1990s, and the Argentinian debacle of the early 2000s. Consequently, Rodrik makes three propositions about openness, stimulating for the purpose of our argument. First, openness by itself is not a reliable mechanism to generate sustained economic growth. There are potential benefits to be reaped, but the fundamental determinants of economic development are the accumulation of physical and human capital, and technological development. Secondly, openness will likely exert pressures that widen income and wealth disparities within countries. Again, this is borne out by the evidence, in the lost decade.
Figure 3: Gini coefficients for Latin America from 1980 to 2008 (Source: Gasparini et al. 2010)
Finally, openness will leave countries vulnerable to external shocks that can trigger domestic conflicts and political upheavals. Thus, our concerns with import or export strategies may be overstated: maybe economic development should focus on investment and not the degree of openness.
Lastly, ISI policy was not a clean break for Latin America but had deeper roots than suggested in the standard story. This is Haber’s main thesis, as he reinterprets the history of Latin American industrialization. For one, he points to the fact that Latin America did have a substantial industrial sector well before the 1930s. Given the unfavorable terms of trade, this means that there must have been policies favorable to industrialization during the so-called era of export-led growth. Indeed, he finds evidence of substantial protectionism before the 1930s. For example, before World War I, Latin American tariffs were five times those of Western Europe. Furthermore, Latin American governments did not adopt protectionism because of vague economic goals but because of political pressures. He does not question the highly protected and woefully inefficient characteristics of Latin American economies. But Haber makes the point that this was not the product of the “developmental state”, but of a historically longer political process that involved economic agents making specific demands on politicians not accountable to broad constituencies. Basically, politicians were not accountable to the group that pays the ultimate cost of protectionism: consumers, who were not unified and lacked electoral mechanisms to sanction public officials.
These qualifications shed a new light on ISI policy in Latin America. First, the dichotomy between export-oriented industrialization (EOI) versus ISI is not straightforward, and a certain form of ISI is necessary for development. As a result, the failure of Latin American development is not to be found in ISI policy in itself, but in its implementation. We posit that the reasons for the faulty implementation of ISI are of a deep-rooted institutional nature.
The implementation of ISI in Latin America failed to address the fundamentals of economic growth
We have seen that ISI was used in Western Europe, the United States and East Asia in the early stages of industrialization. In addition, openness did not succeed in Latin America. We come to the logical conclusion that ISI in itself did not cause Latin America’s relative backwardness, but the way it was achieved. Can this be proven? In fact, in Latin America, ISI policies created massive distortions, on a much larger scale than in East Asia. Alan Taylor argues that the main difference between East Asia and Latin America was that policy in the latter was very price distortionary and affected incentives, whereas in Asia, interventions were numerous but less distortionary and left incentives intact10. Notably, Latin American policies prevented long term investments in physical capital. Henry Bruton explains the mechanisms11. He first points to the underutilization of the capital stock, due to a lack of qualified labor, limited demand for specific output, and limited availability of intermediate goods (the last two being direct consequences of the policies implemented). Secondly, he argues that distortions prevented high investment rates. So ISI in Latin America failed to promote physical capital accumulation. Another major difference between Latin America and East Asia was the role the State played. In the former, policy was ad hoc, and responsive to the needs of industrial lobbies. As Haber points out, there never was a global policy, only individual industries lobbying for protection. In East Asia, from the 1960s onwards, there were carefully thought-out plans, or “growth strategies”. There, Pranab Bardhan argues that the State was very present, but acted in decisive and subtle ways12. Through credit allocation, the promotion of industrial investment, guaranteeing loans, and nudging firms to upgrade their technology, States enhanced the market instead of replacing it.
So how does one account for the difference in implementation? We posit that the institutions were very different in Latin America than in Western Europe or in East Asia. In Latin America, policy makers were unconstrained by electorates. Consumers had no political voice, so there was no need for a growth strategy and no political cost for tariff protection. Moreover, vested interest groups were much stronger in Latin America. In Economic Origins of Dictatorship and Democracy, Daron Acemoglu describes Argentina as a classic case of democracy failing to consolidate because of high inequality13. Kenneth Sokoloff and Stanley Engerman explain the presence of these traditional interests through factor endowments. These vested interest groups had no interest in promoting primary education, and human capital remained low in Latin America. On the contrary, Bruton shows how the institutional context in East Asia after World War II was very different. He argues that the Japanese occupation had three unintended but positive consequences: it raised the human capital level of the population, enabled the transfer of productive technologies, and more importantly, the Japanese broke vested interests. Therefore, land reform in Taiwan and South Korea was possible because the Japanese had stripped landowners of their land, and the absence of a powerful landowning class in Korea and Taiwan gave governments and policymakers more flexibility. On the contrary, the institutions in Latin America failed to address the fundamental causes of development: physical and human capital accumulation. These translated into faulty implementations of ISI. Meanwhile, the institutional endowment in East Asia enabled the “Tigers” to commit fully to economic development, to adapt to and adopt new technology (higher human capital), and to stimulate investment.
The institutional failure of Latin America sheds a new light on the question. In the end, what is important in driving economic development is not reducing imports or boosting exports. The fundamental focus is on boosting investment and human capital. To do so requires a domestic investment strategy. As we have seen briefly, the East Asian miracles were due to outstanding investment efforts. These efforts came from domestic investment incentives driven by a set of institutions. South Korea and Taiwan both heavily subsidized investment. Zhu argues that Taiwan and China combined ISI and EOI. Both strategies worked hand in hand: EOI provided foreign exchange to ISI, while ISI developed East Asian economic independence Behind this was a set of institutions, which played a key role (the government, public enterprises, and the banking sector in particular). On the contrary, Latin America failed to implement a domestic investment strategy. In Making Openness Work, Rodrik argues that these investment strategies need to be backed up by a set of strong domestic institutions of conflict management. In order to work, comparative advantage requires restructuring economies. These adjustments are painful. But successful countries were those that had social and political institutions that maintained overall macroeconomic stability. The “wild” liberalization of Latin America and the 1980s and 1990s and the failure of IMF programs did the contrary.
To briefly conclude, the standard literature depicts ISI as an inadequate strategy in a globalized world. Latin America has been the prime example, when compared to East Asian miracles. However, our findings show that it was not ISI per se that was at fault in Latin America, but its implementation. This was due to a poor institutional context, which constrained policy makers and economic actors. I began this essay by writing that Latin America is a fascinating case for anyone studying modern economic development. I believe that the Latin American and East Asian experiences show that policy makers should not focus on reducing imports or boosting exports, but on raising investment in a stable macroeconomic environment. However, political choices are heavily constrained by the institutional environment, and this explains the failure of structural adjustment programs in Latin America. This institutional failure continues to permeate economic life in the Southern Cone today. One needs only to read the headlines on the political situation in Brazil to understand how corrupt Brazilian political life is; and, in the case of the Brazilian oil company Petrobras, how this translates into inefficient economic development.
- Baer, W. (1972) Import Substitution and Industrialization in Latin America: Experiences and Interpretations
- Haber, S. (2006) The Political Economy of Industrialization (in Cambridge Economic History of Latin America)
- Sachs, J. and Warner, A. (1995) Economic Reform and the Process of Global Integration
- McKinley, W. (1892) Speech made in Boston, Library of Congress
- Broadberry, S. (2010) The Cambridge Economic History of Modern Europe
- Zhu, T. (2006) Rethinking ISI: Development Strategies and Institutions in Taiwan and China
- Rodrik, D. (1999) The New Global Economy and Developing Countries: Making Openness Work
- Panagariya, A. (2002) A Case for Import Substitution?
- Taylor, A. (1998) On the Costs of Inward-Looking Development: Price Distortions, Growth, and Divergence in Latin America
- Bruton, H. (1998) A Reconsideration of Import Substitution
- Bardhan, P. (2005) Institutions Matter, But Which Ones?
- Acemoglu, D. and Robinson, J. (2006) Economic Origins of Dictatorship and Democracy
- Sokoloff, K. and Engerman, S. (2000) Institutions, Factor Endowments, and Paths of Development in the New World
 For example, innovations in shipping enabled Argentina to ship frozen meat to Europe.
 In 1950, he published an essay outlining his thesis, entitled The Economic Development of Latin America and its Principal Problems.
 Economic Commission for Latin America and the Caribbean
 The exception being Great Britain
 This point will be developed further on in the essay.
 Although Arvind Panagariya points to the exceptional growth rates in the developed world as essentially pulling growth rates in Latin America8.
 This is crucial. As Krugman shows, East Asian growths in the second half of the 20th century were driven by physical capital accumulation.