Trade and Development

The Latin American Paradox

Canal de Panamá
Trade and Development : The Latin American Paradox - Antoine Cerisier


Latin America provides an interesting lesson about trade preferences and their complex link with economic development. This paper evaluates the claim that Latin America has adopted restrictive trade policies and highlights a double paradox. First, the region is particularly protectionist despite relatively high levels of GDP per capita and exports. Second, economic development and import dependence have little impact on trade openness in the region. The cross-country evidence for this paradox relies on a sample of industrialised and developing economies, including twelve Latin American states. The paper also identifies three factors that could explain protectionism in the region: political economy, recent economic history, and politics and ideology. It then briefly evaluates the impact of restrictive policies on recent economic performance. In particular, it demonstrates that high tariffs and economic growth can coexist.


Scholars worldwide have written extensively about trade policies and their impact on economic development. Most of the existing empirical literature focuses on the effects of trade liberalization or protectionism on economic growth, poverty, and inequality, often with diverging conclusions. The debate remains vivid within the academic sphere: while some defend liberalization and its impact on the economy, others criticize free trade and point to the possible benefits of protectionism. However, few scholars have studied specific regions’ experiences with trade and development. Moreover, the vast majority of empirical studies treats trade openness as an independent rather than a dependent variable—in other words, as a starting point rather than an outcome. Economic and political factors determining trade preferences often remain unaccounted for.

To fill these two gaps in the existing literature, this paper will seek to explain the high prevalence of protectionist trade policies in Latin America and assess whether these have been a rational—and successful—policy choice. To avoid confusion, I will use a broad definition of Latin America, namely all American territories where the Spanish or Portuguese language prevails; this includes Mexico, Central, and South America. After highlighting the Latin American “exception” of highly protectionist policies relative to other regions, the paper will address the economic, social, and political factors fuelling protectionism on the continent. It will then evaluate the effectiveness of this development strategy and its implications, if any, for other regions.

Empirical Analysis: The Latin American Exception

Research Design

In recent years, Latin American states have come under fire for being too protectionist. This pressure has only increased since the global financial crisis. The British newspaper The Economist, known for its strong advocacy of free trade, has insisted that “regional integration, not protectionism, is the right response to fears of deindustrialisation” in Latin America.1 In similar fashion, EU Trade Commissioner Karel de Gucht has pointed to “worrying signs of protectionism that are appearing in some Latin American countries.”2 However, one must verify these assertions empirically: is the region under study more prone to protectionism than others? To answer this question, I use recent data from the World Bank and the Central Intelligence Agency and compare Latin American states with their foreign counterparts. The diverse sample comprises 71 countries (twelve of which are in Latin America) from all continents. The European Union and Hong Kong are each counted as one country. The sample includes developing, middle-income, and advanced economies, as well as most major economic and demographic powers:

Several indicators help to establish the so-called Latin American exception: import dependence and GDP per capita as independent variables, and trade openness as the dependent variable. Import dependence is measured by dividing total imports by total exports for a given country; the higher the number, the more import-reliant the country is. GDP per capita is the market value of all goods and services of a country at a given time, divided by population size; values are adjusted with the Purchasing Power Parity (PPP) method to measure relative economic strength.

The Trade Tariff Restrictiveness Index (TTRI) published by the World Bank measures trade openness, the dependent variable. It reflects “the equivalent uniform tariff of a country’s tariff schedule that would maintain domestic import levels constant.”3 In other words, the TTRI score measures a country’s average applied tariff rate on all products. While previous studies have used trade flows and other indicators, the Harvard University economist Dani Rodrik insists that “the available indicators of tariff... averages are reasonably accurate in ranking countries in terms of trade policy openness.”4

One would expect that both GDP per capita and import dependence should impact trade restrictiveness. Indeed, a simple linear regression using the aforementioned variables shows that poorer economies tend to be more protectionist, since tariffs provide a valuable source of revenue. As John Coatsworth and Jeffrey Williamson have remarked, customs revenues are “essential to support central government expenditures on infrastructure and defence.”5Furthermore, import-reliant states’ industries usually struggle to compete on the international market. These states are often more restrictive than export-led ones, as import-competing industries seek protection through tariff barriers, which can also help offset the trade deficit.

Asserting the Latin American Paradox

The linear regression model — statistically significant and with good predictive power—seems to confirm our expectations. Both import dependence and GDP per capita are significantly correlated with trade restrictiveness; the latter has a particularly strong impact. In other words, poor and import-reliant economies tend to be more protectionist than rich, export-led ones.

In view of the results obtained with the regression model (see Figure 1), one would expect Latin America to be fairly open. Indeed, Latin American countries are all middle-income economies according to World Bank standards, and most of them are not particularly reliant on imports. However, a simple graphical analysis reveals that most—though not all—of the twelve Latin American states in the sample are quite restrictive compared to other middle-income countries. In other words, they are outliers. This is shown by the rounded data points in Figure 2 below. The graph illustrates this paradox: only Ecuador’s TTRI is lower than five. The three regional powers in terms of GDP and population size, Argentina, Brazil, and Mexico, remain very protectionist despite their economic dynamism. Argentina and Mexico are particularly restrictive—with TTRI of 11.4 and 12.7, respectively—even though they both rank among Latin America’s richest countries. Both are thus more restrictive than numerous poorer nations in Asia and Sub-Saharan Africa, such as Cambodia, Kenya, or Zambia.

Table 1 reveals a second paradox: while the region’s trade policies are generally quite restrictive, the correlation between economic development and import reliance on the one hand, and trade restrictiveness on the other, does not seem to hold. Indeed, as mentioned previously, the three largest regional powers, Argentina, Mexico, and Brazil, are also the most protectionist. Furthermore, some of the most open Latin American economies lie in poor regions of Central America. Guatemala and El Salvador are some of the poorest countries in the region, with GDP per capita well below Latin America’s average of $12,000. Surprisingly, they also have relatively low tariffs compared to their regional counterparts, with TTRI of 5.9 and 5.8,

respectively. Moreover, export-led states such as Argentina, Brazil, and Bolivia are more restrictive than three Central American countries—El Salvador, Guatemala and Honduras—that rely heavily on imports. Ecuador, the most open Latin American economy in the sample, also imports more than it exports.

Hence, Latin America exhibits a double exception. First, the region is more protectionist than Europe, Asia, and North America, with an average TTRI of 7.7, as opposed to 7.5 for the whole sample. Second, as illustrated by Table 1 below, import dependence and GDP per capita have a minimal impact on trade restrictiveness in the region.

Historical Roots of Latin American Protectionism

The empirical data above has demonstrated the Latin American exception in trade policies: states in the region tend to have relatively high tariffs despite high levels of exports and GDP per capita. Some experts have attributed the collapse of global trade and the surge in protectionism to the recent financial crisis, partly due to the lack of trade credit. However, Latin American protectionism existed long before the 2008 financial crisis. As Coatsworth and Williamson observe, “Latin America had the highest tariffs in the world as early as 1865, a leadership position it held until the 1930s.”6 Even during the so-called golden age of globalization, the region was extremely restrictive. In fact, tariffs kept rising until the outset of World War I, a period usually considered the belle époque for Latin America. International trade in the early 20th century was characterized by “an enormous variance in levels of protection between the regional club averages”: tariffs in Brazil and Colombia were over ten times those in India and China.7 Thus, Latin American protectionism is not a new phenomenon; neither is the double paradox highlighted above. The variance in tariffs within the region—between Colombia and Chile for instance—was already considerable in earlier periods of history.

Numerous scholars have attempted to account for the high tariff levels in 19th- and early 20th-century Latin America. The sociologist Miguel Centeno has pointed to several historical factors. First, most states in the region were newly independent and lacked the bureaucratic resources to tax income, expenditure, or wealth.Tariffs were thus an easy source of revenue for central governments. Second, the region experienced over thirty major conflicts between 1819 and 1880. As a result,military expenditures rose to almost 90 percent of government spending in the region. This further encouraged Latin American governments to adopt very high tariffs to finance military spending.8

Even though Centeno’s arguments are historically valid, they cannot account for present-day protectionism in the region. Indeed, most Latin American states have been independent for over a century and maintain relatively low levels of military expenditures. The following section of this paper will address a number of other factors that could explain the region’s protectionist tendencies.

Explaining Latin American Protectionism

Political Economy & Social Structures

David Ricardo first contended that a country has a comparative advantage in a product if it is relatively abundant in that product. The economists Wolfgang Stolper and Paul Samuelson later applied Ricardo’s theory to the three main factors of production, namely land, capital and labor, and the effects of different trade policies.9 Stolper and Samuelson insisted that “protection benefits... owners of factors in which society is poorly endowed.”10 As a result, people will support free trade or protectionism depending on their factor’s abundance. Beneficiaries of free trade will attempt to accelerate trade liberalization to increase their wealth and influence. In other words, trade can shape political and societal cleavages. According to the political scientist Ronald Rogowski, increasing exposure to trade “must result in urban- rural...or class conflict” depending on the distribution of wealth among factor owners.11 For instance, poor countries with abundant land but scarce capital will often witness urban-rural cleavages between workers and landowners. In this model, the outcomes of such power struggles determine trade preferences.

The Stolper-Samuelson theorem, also known as the factors model, provides a compelling approach to international trade and can be applied to Latin America. Most countries in the region are land-rich and capital-scarce, with low population density. Consequently, countries such as Brazil, Bolivia, and Argentina have been experiencing urban-rural conflicts predicted by the factors model. Labor and urban middle classes often unite against rich landowners: the former support their government’s protectionist policies while the latter—owners of the abundant factor—often reject them and push for less restrictive trade policies.12 The 2008 food crisis exacerbated these tensions as governments in Argentina and Brazil used trade policy to increase domestic food supplies and reduce prices. For instance, the Argentinean government, led by President Cristina Kirchner, raised levies on soybeans and introduced export bans on crucial agricultural products such as wheat and flour to prevent staple- food shortages. This led to a major conflict between the government and the agricultural sector, resulting in mass protests, farmers’ strikes, and road blocks all across the country. After four months of confrontation, the Senate narrowly rejected the rise in export levies and cancelled the reforms.13

The Argentine case illustrates the centrality of trade issues to social and political cleavages in Latin America. The region’s tariff system reflects urban-rural tensions. In Brazil and Argentina, import restrictions are considerably higher on manufactured goods than on agricultural products. The trade regime thus reveals societal cleavages in both countries: landowners pushed for openness while the urban working and middle classes, as well as industrialists, obtained restrictions on non-agricultural products. However, as Rogowski has admitted himself, the factors model depends on “simplifying assumptions that are never achieved in the real world, among them perfect mobility of factors.”14 Hence, the model’s application to the Latin American trade regime contains several flaws. For example, Peru and Ecuador both have high tariffs on agricultural goods despite abundant agricultural resources and competitive exports.

Economic History

Political-economy models provide compelling analyses of the ways in which factorial distribution and societal cleavages shape trade policies in Latin America. However, as noted above, such theorems rely on assumptions that real-life examples often contradict. Furthermore, a multiplicity of factors must account for Latin American protectionism. Economic history is always useful to understand present trends. Two contemporary episodes in Latin American history help shed light on the issue at hand.

First, restrictive trade policies might result from the enduring influence of import substitution industrialisation (ISI) policies. Ha-Joon Chang, an institutional economist, summarizes this much debated development strategy: “A backward country produces industrial products that it used to import, thereby substituting industrial products with domestically produced equivalents.”15 Countries achieve this by providing home producers with temporary or “strategic” protection against imports. ISI originated in Latin America, where it was first implemented in the 1930s and remained influential on the continent, as well as in the global South as a whole, until the 1970s.

Import substitution lost prominence in the 1980s and has been largely deemed inefficient since, partly due to the economic crises experienced by Latin American states during that period. However, Rodrik insists that “trade and industrial policies had very little to do with bringing on the crisis.”16 Instead, poor monetary and fiscal policies, as well as a global economic downturn, caused the hardship experienced by developing countries. In fact, ISI strategies have contributed to positive economic developments in the global South. In the 1960s, over 40 developing nations enjoyed annual growth rates exceeding 2.5 percent per capita. Latin American states were particularly successful in the 1960s and ‘70s, with per capita income growing at over 3 percent a year. It is thus no surprise that they should seek to reproduce ISI strategies, in spite of the consensus against such policies.

A second factor behind Latin American protectionism may be the perceived failure of trade liberalization programs over the past thirty years, particularly as part of the so-called Washington Consensus (as coined by American economist John Williamson). The principles of this neoliberal reform package heavily influenced the relationship between international financial institutions and developing countries in the 1990s. The Consensus contained ten broad policy recommendations, including fiscal discipline and deregulation. Trade liberalization was an essential component. As Latin America witnessed a series of financial crises in the 1980s, it became the primary target of the neoliberal agenda. Governments in Peru, Bolivia, Argentina, and other Latin American nations undertook radical reforms—often called Structural Adjustment Programs (SAPs) by the World Bank and the International Monetary Fund—that included a sharp reduction in import tariffs.

Such reforms were far from successful: regional economies stagnated and poverty increased in some countries. As Rodrik observes, “The 1990s as a whole saw less growth in Latin America than in 1950-80, despite the dismantling of the state-led, populist and protectionist policy regimes of the region.”17 Brazil and Bolivia, in particular, experienced a troubling stagnation in GDP and living standards following the implementation of SAPs. The Argentine crisis, which occurred from 1999 to 2002, has often been described as the ultimate failure of the Washington Consensus. Following the devaluation of the Brazilian real— which harmed Argentinean exports— Argentina suffered an enduring recession, hyperinflation, widespread unemployment, and increasing social unrest. The Argentinean government finally defaulted on its external debt and engaged in a slow recovery in 2003; import substitution and high tariffs were among the tools the Kirchner administration used to redress the country. While the Washington Consensus was not responsible for the crisis per se, the IMF faced heavy criticism for its slow reaction and short-sighted praise of the Argentine economy just months before the crash.

Critics of trade liberalisation also point to the Mexican crisis of the 1980s as an example of failed liberalization. Neoliberal policies undertaken by President Miguel de la Madrid from 1982 onward led to rising unemployment and poverty, as well as economic stagnation. Mexico’s GDP per capita only grew by an average 0.1 percent a year between 1985 and 1995. Rapid trade liberalization wiped out whole swathes of Mexican industry and arguably undermined the country’s agricultural sector.18 In sum, the perceived—and actual—failures of trade liberalization policies in the 1980s and ‘90s still resound in Latin America and may explain current protectionist tendencies.

Politics & Ideology

Trade preferences are strongly correlated with people’s ideology and political opinions. For instance, opinions on free trade are often linked to political views on consumerism, free markets, and state intervention in the economy.19 Fiscal conservatives and libertarians tend to support free trade as an integral part of the neoliberal worldview. Left-wing voters and parties, on the other hand, are often more critical of globalization and highlight the negative effects of free trade. Thus, left-wing governments can be expected to implement more restrictive trade policies than right-wing and neoliberal ones. This may help to explain the prevalence of protectionism in Latin America. Indeed, the vast majority of countries in the region are governed by left-wing or “populist” parties. The most prominent heads of state in this camp include Dilma Rousseff and her predecessor Lula da Silva in Brazil, Cristina Kirchner in Argentina, and Uruguay’s José Mujica, also known as the world’s “poorest president” for his inexpensive lifestyle. Some Latin American presidents, including Venezuela’s late Hugo Chavez and Bolivia’s Evo Morales, even claim to follow Marxist principles. Of the 12 Latin American states in the sample, only two (Chile and Honduras) are governed by right-wing parties. Both have relatively low tariffs.

Finally, some attribute the prevalence of protectionism in Latin America to the peculiarity of regional integration on the continent, which is dominated by two organizations: Mercosur, a political and customs union founded in 1991 by Argentina, Brazil, Paraguay, Uruguay, and Venezuela, and the Andean Community, a customs union created in 1969 by Bolivia, Colombia, Ecuador, and Peru. While both institutions have established free trade and flow of people among their members, a variety of import quotas on external products remain in place. The current levels of import tariffs have led some commentators to use the term “fortress Mercosur.”20 Intra-regional trade remains very limited in Latin America, making up only 25 percent of total exports in 2010 as opposed to 50 percent in Asia.21

A Successful Policy Choice?

Is protectionism a rational policy choice for Latin America? Has it been a successful development strategy? According to Coatsworth and Williamson, Latin American protectionism did not lead to economic growth in the 19th or early 20thcentury.22 Countries with high tariffs grew more slowly than open economies like Chile. Most mainstream scholars have come to the same conclusion for other regions, arguing that openness fosters economic growth. However, two empirical studies by Rodriguez and Rodrik and Halit Yanikkaya have demonstrated a positive and sometimes significant correlation between trade barriers and economic growth for developing countries. The latter study provides “considerable evidence for the hypothesis that restrictions on trade can promote growth, especially of developing countries, under certain conditions.”23 Table 2 offers an overview of trade restrictiveness and average annual growth for the period 2007-2011.

At first glance, Table 2 does not demonstrate a clear difference between open and protectionist economies in Latin America. Mexico, the most restrictive Latin American state in the sample, only grew at an average annual rate of 1.5 percent between 2007 and 2011. Ecuador and Chile, two of the most open economies in the region, enjoyed much higher growth rates in the same period. Nonetheless, three of the most restrictive Latin American countries, Argentina, Bolivia, and Brazil, grew by 6.9, 4.7, and 4.2 percent per year, respectively. By contrast, El Salvador was the slowest-growing state with a mere 1.1 percent average in 2007-11, despite its low tariffs. The case of Argentina is particularly striking: The country suffered a major economic crisis in 2002 and has some of the most restrictive trade policies in the region. Nevertheless, its GDP grew by an astonishing 6.9 percent annually between 2007 and 2011, in the midst of the global financial meltdown.24

Whether high tariffs have led to higher growth is debatable; however, the Latin American case demonstrates that both combinations can certainly coexist. Indeed, the region grew at an average of 4.6 percent annually since 2007, higher than the sample average (4.3 percent). Such growth highlights the region’s resilience to the recent financial crisis compared to some of its emerging counterparts; annual growth rates in Russia, South Africa, and Thailand did not surpass 3 percent over the same period. Other developing nations such as Bangladesh, India, and Ghana have grown very rapidly since 2007—at average rates of 6.2, 7.7, and 8.3 percent—with very high tariff barriers. Nonetheless, some more open Asian economies have also done well (especially China, Mongolia, and Indonesia). This illustrates there is no single recipe for economic success.


Latin America provides a compelling case for rethinking the relationship between protectionism and development. Empirical analysis illustrates the double paradox of trade policies in the region. First, while most states are export- led, middle-income economies, Latin American tariffs are higher than the sample average—especially in the three regional powers. Second, GDP per capita and import dependence are not major determinants of trade policy in the region. Indeed, poor, import-reliant countries in Central America tend to be more open. Latin America has numerous specificities that other regions do not necessarily share, among them the nature of its recent economic history and rural-urban conflicts. Political-economy structures, as described by the factors model, play a major role in creating rural-urban cleavages that result in very high tariffs on manufactured products. Economic history also determines trade policies: the enduring influence of ISI strategies and perceived failure of neoliberal liberalization programs might explain Latin America’s protectionist tendencies. Finally, trade preferences are strongly correlated with ideological orientation, and most Latin American states are governed by left-wing parties.

It remains unclear whether high tariffs are detrimental for economic development, as the literature often assumes. Latin America achieved relatively high growth rates in the midst of the global financial crisis. Protectionist countries such as Argentina, Bolivia, and Brazil have been particularly successful. At the same time, more open economies like Chile and Ecuador have also grown quite rapidly. Moreover, numerous states in Asia and elsewhere have arguably benefited from low tariffs and open borders. This demonstrates the shortcomings of uniform, one-size-fits-all approaches to economic development.

Political institutions, social structures and economic history vary greatly among different regions. The coexistence of high tariffs and high growth rates during a major global economic downturn is particularly intriguing in view of the existing literature. It may be that strategic external shocks.

Notes & References

  1. The Economist, “Trade in Latin America: unity is strength,” March 10 2012, node/21549939.
  2. Martin Banks, “EU commissioner warns of worrying protectionism in Latin America”, The Parliament, June 27 2012, http://www.theparliament. com/latest-news/article/newsarticle/ eu-commissioner-warns-of-worrying- protectionism-in-latin-america/#. UO7VCPnm6-0
  3. The World Bank, World Trade Indicators (2010) and World Development Indicators (2011).
  4. Dani Rodrik, “Comments on Trade, Growth & Poverty by Dollar and Kraay,” (2000): 3.
  5. John Coatsworth and Jeffrey Williamson, “The Roots of Latin American Protectionism: Looking Before the Great Depression,” NBER Working Paper Series (2002): 10.
  6. Ibid.
  7. Ibid.
  8. Miguel Centeno, “Blood and Debt: War and Taxation in Nineteenth-Century Latin America,”American Journal of Sociology 102 (1997): 1565-1605.
  9. Wolfgang Stolper and Paul Samuelson, “Protection and Real Wages,” Review of Economic Studies9 (1941): 58-73.
  10. Ronald Rogowski, “Political Cleavages and Changing Exposure to Trade,” The American Political Science Review 81, no.4 (1987): 1121-1137.
  11. Ibid.
  12. Thomas Skidmore & Peter Smith, Modern Latin America (Oxford: Oxford University Press, 1984).
  13. Food and Agriculture Organization (FAO), “Policy Measures Taken by Governments to Reduce the Impact of Soaring Prices” (2009).
  14. Rogowski, “Political Cleavages,” 1134.
  15. Ha-Joon Chang, Bad Samaritans: The Guilty Secrets of Rich Nations and the Threat to Global Prosperity (London: Random House, 2007): 22.
  16. Dani Rodrik, “The Global Governance of Trade as if Development Really Mattered,” Report submitted to the UNDP (2001): 17.
  17. Dani Rodrik, “Goodbye Washington Consensus, Hello Washington Confusion?,” Journal of Economic Literature 44: 975.
  18. Chang, Bad Samaritans, 68.
  19. Martin Edwards, “Public Opinion Regarding Economic and Cultural Globalization: Evidence from a Cross-National Survey,” Review of International Political Economy 13, no.4 (2006): 587-608.
  20. Robert Plummer, “Protectionism: is it on the way back?,” BBC News, 17 September 2012, news/business-18104024.
  21. The Economist, “Trade.”
  22. Coatsworth and Williamson, “Roots.”
  23. Halit Yanikkaya, “Trade Openness and Economic Growth: A Cross-Country Empirical Investigation,” Journal of Development Economics 72 (2002): 57-89.
  24. This figure may be misleading, as Argentina’s actual inflation rate is much higher than the official estimate. See: Alberto Cavallo, “Online and Official Price Indexes: Measuring Argentina’s Inflation,” (M.I.T., 2012).
Antoine Cerisier graduated from University College London in 2011 and is currently enrolled in a double master’s programme in International Affairs at Sciences Po Paris and the University of St. Gallen in Switzerland. His research interests include political economy, international development, and global environmental policy. He is now completing his master’s thesis on international trade.