Any author whose work covers the whole of Latin America faces a series of problems. These problems are compounded when the period covers nearly two centuries. It is no surprise, therefore, that relatively few studies purport to survey the economic history of Latin America since independence, despite the rapidly expanding literature on the progress of individual countries and provinces. It is, however, the advance at the subregional level that makes necessary and feasible a new economic history for the whole region. From Chile to Mexico, a new generation of scholars has used advanced techniques to mine the primary sources and advance our knowledge across a broad range of issues.

Any economic history of Latin America involves a multidisciplinary approach, which runs the risk of offending the sensibilities of those scholars who prefer to work within a single disciplinary boundary. As a representative of the last generation to be encouraged to stray across disciplines, I have enjoyed the opportunity to draw on a huge literature covering economics, economic history, history, politics, sociology, anthropology, and international relations. As an editor of the multidisciplinary Journal of Latin American Studies since 1986, I have been uniquely privileged to gain access to new research in this area before it becomes widely disseminated.

The growth of exports was a necessary but not sufficient condition for successful export-led growth. Yet, as we saw in Chapter 3, only a small number of countries can be said to have met even this basic condition. The problem was not, in general, a shortage of demand; far more important were the constraints on the expansion of export supply. Countries with fast growth of exports had usually overcome the obstacles to export expansion on the supply side, whereas those with slow growth were unable to break down the often formidable problems that confronted the export sector throughout the nineteenth century.

Export expansion, whether fast or slow, could produce one of three export-led models: additive, destructive, or transformative. In the additive model the export sector was grafted onto the existing structure of production with very little change in the nonexport economy. Resources were attracted into the export sector without reducing output elsewhere, and factor productivity in the nonexport economy was unaffected by the growth of the export sector. An example of additive growth is provided by the expansion of banana exports from Honduras at the start of the twentieth century. The land – previously unutilized – had a zero opportunity cost, the capital was foreign, and the labor was provided in large part by migrant workers from the British West Indies and El Salvador. The impact on the rest of the economy was minor.

Since the first edition of this book was published in 1995, there have been major advances in our knowledge of the economic history of Latin America as a result of new data series. This has made it possible in this edition to estimate foreign trade, fiscal accounts, and even gross domestic product (GDP) per head for all the main Latin American countries back to the time of independence 200 years ago. As a result, we can compare with much more confidence the region's long-run economic performance not only among the Latin American republics themselves, but also against other countries.

It has long been recognized that Latin America had “fallen behind” the United States, at least as measured by GDP per head, even before independence. New empirical research for this edition, however, suggests that the gap was much bigger than previously thought. Thus, Latin America had fallen well behind the United States even before it became “Latin America.” Indeed, the gap was almost certainly substantial even before the United States became the “United States.” Thus, two-thirds of the gap in current living standards had emerged two centuries ago, and only one-third can be attributed to the independence period.

When war broke out in Europe on August 2, 1914, it was not just the international balance of power that was shattered; the global trade and payments system, which had slowly evolved since the end of the Napoleonic wars, was also thrown into disarray. With the signing of the armistice in 1919, a brave face was put on attempts to reconstruct the prewar system. The old international economic order had perished, however, and the new one was dangerously unstable. This instability was scarcely perceived at the time, leaving peripheral regions, such as Latin America, extremely vulnerable to the collapse of international trade and capital flows at the end of the 1920s.

The main feature of the old order had been the existence of relatively unrestricted international trade – a reflection of the interests of the dominant economic power (Great Britain) in the nineteenth century; the limited restrictions in force generally took the form of tariffs, which had the advantage for all concerned of being transparent. Both capital and labor were free to move across international boundaries, and passports were the exception rather than the rule. The gold standard, adopted first by Great Britain, had spread to all the main industrial countries by the end of the century and provided a well-established mechanism for balance-of-payments adjustment. Internal equilibrium (full employment and zero inflation) was regarded as less important than external equilibrium, so the burden of adjustment to adverse shocks was usually achieved through price deflation and underemployment.

The expression “Latin America,” whose origin goes back to the mid-nineteenth century, at first had little more than geographical significance – it referred to all those independent countries south of the Río Grande in which a language derived from Latin (i.e., Spanish, Portuguese, and French) was predominantly spoken. In this original meaning, the only characteristics common to the countries of Latin America were their location in the Western Hemisphere and the origins of their language. In many respects the differences between the countries were considered to be as important – if not more so – as what they shared.

These differences – whether of size, population, ethnicity, natural resources, climate, or level of development – are still very important, but it has also become clear that the republics are held together by much more than geography and language. The shared colonial experience, as divisions above all of the Spanish or Portuguese empires, was crucial in shaping the economic and political destinies of the new republics after independence. The pattern of development in the nineteenth century, based on the export of natural resources to the industrialized countries, reinforced this sense of a shared past.

Population

The quality of population census data in nineteenth-century Latin America leaves much to be desired. Enumeration was often incomplete, the indigenous population was occasionally ignored, and the margin of error was substantial. In Guatemala, for example, it was standard practice until the 1950 census for the president to “request” the jefes políticos in the provinces to inflate the results. All estimates of total population must therefore be regarded with caution. However, recent scholarship has done much to remedy the deficiencies of pre-1914 statistics. As a result, figures exist for Latin American countries that can be combined to give estimates for all of Latin America at various intervals. For this Appendix, I chose the years 1850, 1870, 1890, and 1912.

For most of the period before the First World War, Puerto Rico was part of the Latin American community, and Panama did not exist as an independent country. In the interests of consistency, I therefore include figures for Puerto Rico after 1898 (following its annexation by the United States), and I continue to include Panama in the Colombian figures after 1903 (following its declaration of independence from Colombia).

Before the 1980s, there were very few estimates for gross domestic product (GDP) in the nineteenth century for any Latin American country. Celso Furtado's classic economic history of the region contained no data for GDP per head before the First World War. Those authors that did venture a number usually admitted that they were little more than educated guesses. Although quantitative studies on the economic history of the United States and other countries multiplied rapidly, students of Latin America had to rely heavily on qualitative indicators for the region's development in the first century after independence.

This started to change in 1981 when Paul Bairoch published an article giving very precise figures, but without details on sources or methods, for GDP per head in 1800 for different regions of the world. These suggested that at that time, GDP per head was slightly higher in “Third World America,” that is, Latin America, when compared with “North America,” that is, the United States and Canada. Because GDP per head in Latin America was much lower than in the United States by the early twentieth century, it was widely assumed that Latin America must have “fallen behind” after independence. This has given rise to a substantial literature on why this might have happened, much of which emphasizes the importance of the quality of institutions.

Since the first edition of this book was published in 1994, there have been considerable advances in our knowledge of GDP per head in Latin America – especially for the years before 1950. At the same time, information is now available electronically through the Internet in a form that makes comparisons across space and over time much easier. This Appendix 4 brings this information together in one convenient place.

The starting point is World Development Indicators (WDI) on the website of the World Bank, which has data for all twenty Latin American republics on population and GDP at 2000 US dollars starting in 1960. The only gaps are Cuba from 1960 to 1969 and Haiti from 1960 to 1990, which I filled using Bulmer-Thomas (2012). WDI has comparable data for the United States, so the ratio of Latin American to US GDP per head can easily be constructed.

From 1900 to 1960, the basic source is MoXLAD, which is based on a variety of official and unofficial secondary sources. MoXLAD has data on population and GDP at 1970 domestic prices, together with nominal exchange rates, so this information (subject to qualifications – see below) can be used to construct an index for each country and splice it to the data in WDI. The result is GDP per head at 2000 US dollars from 1900 to 2011. When three-year averages are taken, this gives 1901 to 2010.

The first decade after the First World War brought about further resource shifts in the major Latin American economies in the direction of structural change, industrialization, and diversification of the nonexport economy. In addition, financial and monetary systems were overhauled in many republics – prompted in some cases by the missions led by E. W. Kemmerer – as governments returned to exchange-rate orthodoxy and the gold standard in the postwar years. Without exception, however, economic performance remained heavily dependent on the fortunes of the export sector. By the end of the 1920s (see Table 7.1), exports still accounted for a high proportion of the gross domestic product (GDP), and the openness of the economy – measured by the ratio of the sum of exports and imports to GDP – varied from nearly 40 percent in Brazil to more than 100 percent in Costa Rica.

Structural change in the 1920s brought no diversification within the export sector. On the contrary, the composition of exports by the end of the decade was very similar to what it had been on the eve of the First World War, with a high degree of concentration. The three leading export products accounted for at least 50 percent of foreign-exchange earnings in all republics, and one product accounted for more than 50 percent of exports in ten countries (Bolivia, Brazil, Colombia, Cuba, Dominican Republic, El Salvador, Honduras, Guatemala, Nicaragua, and Venezuela). Virtually all export earnings came from primary products, and nearly 70 percent of external trade was conducted with only four countries (United States, Great Britain, France, and Germany).

At the beginning of the 1960s it was widely believed that regional integration would restore the dynamism of the inward-looking model of development in the larger republics and provide a platform for industrialization in the smaller countries. Yet by the end of the decade the mood had changed. Regional integration – at least in South America – had not brought the expected gains, and the inward-looking model seemed to be subject to diminishing returns. The prestige of CEPAL, which had nailed its colors to both inward-looking development and regional integration, declined despite the regional organization's best efforts to revise its approach to industrialization, and the Latin American policymaking elite began to pay more attention to alternative ideas on trade and development.

The Achilles’ heel of inward-looking development remained the balance-of-payments constraint. After 1929, persistent balance-of-payments problems had persuaded more and more countries to abandon export-led growth based on primary products in favor of a new model that was expected to lower their vulnerability to external shocks. Yet balance-of-payments problems continued under inward-looking development as the policies adopted to favor industry undermined the export sector and shifted the composition of imports in the direction of complementary goods – the demand for which expanded rapidly in line with industrial growth.

The criteria used in Chapter 3 for measuring the success or failure of export-led growth require information on the ratio of exports to GDP, the price of imports, and the purchasing power of exports before 1914 for all countries. This Appendix 2 explains the methods I used to calculate the statistics. It also includes tables on export prices, the net barter terms of trade (NBTT), and the volume of exports from 1850 to 1912 (unless otherwise specified) for twelve countries.

The Ratio of Exports to GDP

Calculation of the required rate of growth of exports in equation (3.2) involves estimation of the ratio of exports to GDP (w). Because w can be assumed to change over time, it is necessary to estimate it at the beginning (circa 1850) and at the end (circa 1912) of the period.

The ratio of exports to GDP is the same as the ratio of exports per head to GDP per head. Table 3.6 provides data on exports per head in current dollars for all countries at different time intervals, and Appendix 4 provides data on GDP per head circa 1912 for nine countries at constant (2000) dollars. For some of these countries there are also estimates of GDP at current dollars (see Appendix 3), and it is preferable to use these for consistency because the export data are also at current prices. In the case of Mexico, I used a three-year average for 1908 to avoid the impact of the revolution; in the case of Venezuela, I used a three-year average for 1921 because the current price series starts in 1920 (in both cases exports per head therefore need to refer to the same year rather than to 1912). Estimates for ԝ in 1912 for other countries with constant price estimates can be obtained by converting the GDP data to current dollars using a deflator obtained from US wholesale price indexes; see Mitchell (2007). Assuming that 1970 equals 100, the 1912 price index can be calculated as 35. Thus the GDP data for these other countries were converted to current dollars by multiplying by 0.35. This allows calculation of ԝ in 1912 for any country for which there are constant price GDP data only.

Independence, which most of Latin America had secured by the early 1820s, came at the end of a long period of economic turmoil and political upheaval during which, it can safely be assumed, living standards fell sharply. From the outbreak of hostilities in 1796 between Great Britain and Spain, Latin America's external trade – imports and exports – was severely disrupted. The invasion of the Iberian Peninsula by Napoleon in 1808 and the imposition of his brother Joseph as king of Spain drove the Portuguese royal family to Brazil and created a temporary alliance between Great Britain and anti-Napoleonic forces in Spain. Exports from Latin America were adversely affected and domestic trade undermined by the flood of imports into Latin America as British merchants searched for an alternative to the blocked continental market.

Napoleon's successful invasion of Spain undermined Spanish authority in Latin America and provided the independence movement – until then weak and inchoate – with the impetus it desperately needed. By the time Napoleon had been finally defeated in 1815, the movement had acquired a dynamic of its own, and the reassertion of Spanish and Portuguese authority over the Iberian Peninsula could not be extended to Latin America. Brazil, a separate kingdom since 1815, refused to recognize the demands of João VI in Lisbon and crowned his son, Dom Pedro, as emperor in 1822. New Spain became Mexico (an empire for nine months under Agustín de Iturbide) and extended briefly to the northern border of Colombia following the annexation of Central America. Spain's South American colonies opted for republican government from the beginning, and by the mid-1820s Spain controlled only Cuba and Puerto Rico. Even Santo Domingo, the eastern part of Hispaniola, had been seized from Spain in 1822 by Haiti, which had gained its independence from France in 1804.