Keywords

JEL Classification

Introduction

The countries in the Western Hemisphere that are former colonies of Spain and Portugal not only share a common heritage but also similar patterns of development. After a disastrous period of economic performance in the decades following independence in the early nineteenth century, Latin American countries pursued a primary export-led model of development at the turn of the twentieth century and then implemented an inward looking industrialisation strategy following the Second World War. After struggling with a severe crisis in the 1980s, the region embraced the market-driven Washington Consensus in the early 1990s. The current phase of development is being shaped by a renewed recognition of the importance of the state, increased attention to the high levels of poverty and inequality, and the emergence of China as a leading force in global trade. Concerns associated with the primary export-led model of development are again at the top of the agenda.

Common endowments and institutions explain remarkable similarities in terms of living standards, commodity export dependence and high levels of economic inequality. GDP per capita in PPP terms, for example, only varied by a factor of 5.4 between Nicaragua (the lowest) and Chile (the highest) in 2007, in comparison to Asia, where it varied by a factor of 32 between Nepal and Japan (Milanovic 2011). There are also important differences in the region. Size is a case in point: the seven largest countries in the region account for over 80 % of the population and GDP. Brazil, alone, has around one third of the region’s population and GDP. Geographical location is another differentiating factor. Mexico and the Central American and Caribbean countries are more economically dependent on the USA than countries in South America. This is in part due to trade and tourism, but also to labour migration and remittances. Ethnic differences are a third factor that have impacted economic and social outcomes. Slavery was much more important in the Caribbean and the northeast of Brazil, while late nineteenth and early twentieth century European immigration played a much more significant role in the Southern Cone countries. It is the similarities that make a study of Latin America relevant, and it is the differences that force analysts to search beneath the generalisations in order to understand the region’s heterogeneity.

Between 1870 and 1981, real GDP per capita increased by nearly eightfold in the region, which is faster than growth in any other region of the world, and comparable to growth in the USA (Table 1). However, during the subsequent 20 years, Latin America was among the slowest growing regions in the world. In the years immediately before and after the global recession of 2008–2009, Latin America once again experienced relatively fast growth, in part as a consequence of better macroeconomic policies, but also as a result of the economic tailwinds from China. A wave of optimism spread throughout the region in the new millennium as living standards increased at a rate not experienced since the 1970s.

Latin American Economic Development, Table 1 Real GDP per capita by region
Full size table

Two of the most daunting challenges that Latin America faces relate to the low quality and effectiveness of institutions, and the persistence of poverty and inequality. There is an increasing awareness that the quality of both state and market institutions needs to improve significantly. The Latin American state has many important functions that need to be carried out more effectively. Well-functioning markets also rely on institutions – legal, anti-trust, regulatory – which are still far below the standards of the developed world. It is also the case that Latin American countries remain among the most unequal in the world. Yet inequality has been falling for more than a decade in many countries, and the rate of poverty reduction has accelerated. Although there are grounds for optimism based on recent performance, Latin America’s future hinges crucially on continued progress in these two areas.

Primary Export-Led Growth: 1870–1929

Most Latin American countries gained independence in the first quarter of the nineteenth century. The subsequent 50 years were characterised by political instability and slow growth. Although data are scarce, between 1820 and 1870 income per capita appears to have been stagnant in the region (Madison 2008). In contrast, between 1870 and 1913, Latin America pursued a primary export model of development and grew faster than any other region in the world. Real income per capita rose at an annual rate of over 1.8 % for the eight Latin America countries shown in Table 1. Growth in Argentina was particularly impressive in this period, with 1913 income per capita surpassing the Western European average. Latin American growth continued at 1.5 % per year from 1913 to 1929.

Expansion was largely extensive, based on the incorporation of new land and additional labour. In some countries, especially in the Southern Cone, immigrants also contributed to population growth and an increase in human capital. The export boom was accompanied by capital inflow that helped finance investments in infrastructure. The length of railroad tracks in the region, for example, increased by a factor of 12 between 1870 and 1900, from about 50,000 to 60,000 km (Thorp 1998).

Exports were extremely concentrated in a small number of products, creating a vulnerability to commodity booms and busts that could last for decades. Brazil, El Salvador, Guatemala, Haiti and Nicaragua, for example, earned between 62 and 85 % of export earnings from coffee alone around 1913, while Bolivia, Chile, Cuba, Ecuador and Panama earned at least 64 % from a single product (Bulmer-Thomas 2003).

The benefits for long-term growth derived from building institutions, infrastructure, and a local market varied by product and country size. In terms of products, locations where factor endowments had been favorable to products with economies of scale (such as sugar) developed more extractive institutions that contributed to the persistence of high economic inequality (Sokoloff and Engerman 1997). In large countries, exporters had greater incentives to invest their profits in activities geared toward the domestic market, thus stimulating the growth of manufactured products (as in Colombia and Brazil). But the size of the country and the product that predominated were not the only factors that mattered. Differences in the composition of elites and the resultant patterns of land occupation help to explain the much more favorable twentieth century outcomes in Colombia and Costa Rica relative to El Salvador and Guatemala – all countries in which coffee was extremely important (Nugent and Robinson 2010).

The global disruptions that took place between 1914 and 1945 threatened the sustainability of the primary export-led model of growth. The depression of the 1930s reduced demand and prices for Latin American exports. The two world wars interrupted trade routes with Europe and made manufactured products difficult to import. These events created incentives and opportunities to industrialise. Large countries, and those with more autonomous public sectors (such as Uruguay and Costa Rica), took advantage of these opportunities in the 1930s. They abandoned the gold standard earlier and experimented with trade, credit and other policies that actively encouraged local manufacturing. Small countries, and those with more dependent governments (such as Cuba), remained on the gold standard longer and responded much more passively to the new external environment (Diaz-Alejandro 1984). The experiences of state-led industrialisation in the USSR and the New Deal in the USA provided examples of a much more active economic role of the state and became important references for governments in the region. Thus, the period between 1929 and 1945 was one of transition between different models of development. But it would only be after the Second World War that an inward looking model of industrialisation would emerge as the predominant strategy of development in the region (Baer 1972).

Import Substitution Industrialisation: 1945–1982

Import substitution industrialisation (ISI) emerged as a consequence of the disruption in global trade between 1929 and 1945 and at a time when there was growing dissatisfaction with the primary export-led model. An influential group of Latin American ‘structuralist’ economists led by Raul Prebisch at the UN Economic Commission for Latin America (ECLA) argued that the primary export-led model was unable to provide sustained improvements in living standards. Two key tenets of the structuralist school were the centre–periphery model of the world economy and a hypothesis regarding the secular decline in the terms of trade of primary exporters (Prebisch 1950). Latin American countries, as well as other ‘peripheral’ countries, depended on the centre not just as a market for exports of primary products, but also as a source of capital and technology. The periphery’s form of insertion into the world economy had a number of negative consequences for their socioeconomic structures, including slow productivity growth, low wages and small domestic markets. The decline in the terms of trade, resulting from the slow growth in the demand for primary products and less competitive markets for industrial products in the centre countries, was regarded as an additional obstacle to development.

According to the structuralist school, economic development in Latin America required industrialisation, which involved protecting the domestic market from external competition and an active role of the state in promoting strategic sectors. At roughly the same time as the ECLA economists were writing about ISI, advances in economic theory and the emergence of ‘development economics’ provided many of the economic arguments that justified state interventions to deal with market imperfections. Latin American intellectuals continued to contribute to theories of development through the dependency school that flourished in the 1960s and 1970s (Kay 1989). The reformist branch of the dependency school (e.g., Cardoso and Faletto 1969) criticised ISI, but believed, like the ECLA economists, that capitalist industrial development was possible in Latin America. The radical branch sought to develop a Marxist theory of dependency and believed that a socialist revolution was necessary (e.g., Dos Santos 1970).

The ISI policy toolbox included trade protection through tariff and non-tariff instruments, multiple exchange rates that were typically lower for imports of capital and intermediate goods, active industrial policy, and supportive fiscal and monetary policies (Franko 2007). The early stages of ISI focused on creating industries to produce basic consumer and durable goods. Tariffs and import quotas increased the price of these goods and restricted their quantity, thus increasing the incentives for local production. By 1960, nominal protection rates on durable goods were over 90 % in Chile, Colombia and Mexico, and over 250 % in Argentina and Brazil (Bulmer-Thomas 2003). Overvalued currencies lowered the price of the imported capital goods that were necessary for industrialisation. Tax breaks and subsidies further increased incentives for domestic production, and national development banks, such as CORFO in Chile and BNDES in Brazil, provided equity and long-term credit for key industrial projects.

The growth of local industry increased the demand for inputs, such as metals, chemicals and electricity, as well as for transportation and other crucial services. Because of the scale of these projects, the time horizon necessary to recoup the investments and shallow domestic capital markets, governments often undertook these projects through state-owned enterprises (SOEs). By the 1970s and 1980s there were more than 500 SOEs in Chile and Brazil, and over 1000 in Mexico (Edwards 1995). In many cases the SOEs dominated the sectors in which they operated. In Brazil, for example, over 95 % of assets in railways, ports, telegraph and telephone, and water, gas and sewers were held by SOEs. In chemicals, mining and electricity, the share was over 50 % (Evans 1979). The domestic private sector in many Latin American countries also found it difficult to move beyond consumer durables and compete in the production of capital goods or in sectors that required more advanced technology. Thus, many of the most dynamic sectors were dominated by multinational corporations (MNCs). Around 1970, for example, in Chile, Colombia, Mexico and Peru, the foreign share of domestic production in chemicals, transport equipment, and electrical machinery varied between 49 and 80 % (Jenkins 1984).

Criticism of ISI’s shortcomings began to emerge from many quarters in the 1960s (see Hirschman 1968). Yet it was only in retrospect, as Hirschman (1987) emphasises, that the achievements of this period – characterised by ISI, increased urbanisation and greater labour market participation – could be fully appreciated. Although the 1950s, 1960s and 1970s, were the decades with the highest rates of population growth in the twentieth century, real income per capita rose by 2.6 % per year between 1945 and 1981 in the Latin American 8 (Table 1). These 36 years compare quite favourably with the first 45 years of the century when income per capita rose by 1.4 % per year, and with the subsequent 27 years shown in Table 1 when income per capita rose by only 1.0 % per year. Total lifetime income rose by much more than income per capita in this period, as life expectancy rose from 40 years in 1940 to 65 years in 1980. (These data are from Thorp (1998) and are based on Argentina, Brazil, Chile, Colombia, Mexico and Venezuela. Data from these six countries are almost identical to the broader set of 20 countries for which data are available for a more limited period of time. The data on illiteracy in this paragraph refer to 20 countries and also come from Thorp (1998)) Improvements in education, as measured by the illiteracy rate, tell a similar story. Illiteracy among the population age 15 and over fell from 49 to 21 % between 1940 and 1980. There is no question that the middle class expanded and living standards improved dramatically for a large portion of the population in the region.

The ISI model in Latin America did, nevertheless, have many shortcomings. First, it generated considerable inefficiency. The ISI strategy was supposed to protect infant industries for a limited period of time while domestic companies learned to work with new technologies, increased their scale and lowered production costs. As a result of rent seeking, however, protection and subsidies often continued indefinitely. Second, unlike with East Asian countries such as Japan and South Korea, the transition from ISI to an export-oriented development strategy was not successful in Latin America. This was particularly harmful to the smaller countries with little opportunity to achieve economies of scale. The failure to move past ISI contributed to slower growth in total factor productivity relative to the Asian Tigers, and this had long-term consequences for GDP growth and living standards. Third, the policies that were intended to subsidise industry implicitly taxed agriculture, thus exacerbating rural poverty and contributing to excessively rapid urbanisation. Fourth, by reducing the price of imported capital goods, the development model of this period favoured capital-intensive sectors that were unable to create enough employment to keep pace with a rapidly increasing urban labour force. The result was high informality, with negative consequences for the distribution of income.

The Debt Crisis and the Lost Decade of the 1980s

One of the principal weaknesses of ISI in the 1950s and 1960s was an insufficiency of export earnings and, thus, the pervasiveness of current account deficits which resulted in frequent balance of payments crises. The perennial shortage of foreign exchange was temporarily eased in the 1970s as a result of the large current account surpluses in oil exporting countries after the first oil price shock in 1973. While country experiences differed, rather than fully adjust to the new international terms of trade, most Latin American governments continued to target high growth rates by taking on massive amounts of debt from private international banks at low, yet adjustable, interest rates (Fishlow 1986). Between 1970 and 1982, medium-and long-term debt in Mexico rose from US$7 billion to US$88 billion. In Brazil and Argentina, debt rose respectively from US$5 billion to US$83 billion, and from US $4 billion to US$47 billion (Sachs 1990).

The real difficulties came as a result of the second oil shock in 1979 which again drove OECD countries into recession and reduced the demand for Latin American exports. As part of the fight against stagflation in the USA, interest rates were raised above 8 % in real terms in 1981. Thus, the attractiveness of negative real interest rates in the 1970s quickly became a liability. Mexico declared a moratorium on its debt payments the next year, and the crisis quickly spread throughout the region. Latin American countries that had become accustomed to net resource inflows of around US$10 billion per year in the late 1970s suddenly had to generate net outflows in excess of US$30 billion by 1983 (Edwards 1995). Thus, while unsustainable policies had created vulnerabilities, the external shocks triggered the crisis.

What began as a debt problem in 1982 eventually became a much broader crisis that extended into the 1990s. Real GDP per capita fell 3 years in a row starting in 1981, and repeated the terrible performance in the years 1988–1990. Between 1981 and 1990, real income per capita fell at an annual rate of –0.62 % (Table 1). Some countries did not recover their pre-crisis levels of income per capita until the mid-1990s, thus giving rise to the term the ‘lost decade’ in Latin America. Average inflation was close to 100 % per year between 1982 and 1986, rising to over 200 % per year between 1987 and 1992 (Edwards 1995). In 1990, four countries had inflation rates over 1000 %. As a result of slow growth and contracting government expenditures, poverty increased sharply in the 1980s. The number of people living on less than two dollars a day rose from 91 to 131 million between 1980 and 1989 (Morley 1995).

International lenders initially diagnosed the problem as one of liquidity, not solvency, and thus were unwilling to forgive any significant amount of debt, especially for the larger countries. Orthodox stabilisation plans based on demand repression were initially adopted throughout the region in order to address the main symptoms of the crisis – fiscal and balance of payments deficits, and inflation. When these plans failed to control inflation, some countries experimented with heterodox plans that incorporated wage and price freezes, and introduced new currencies, as in the cases of the Austral plan in Argentina and the Cruzado plan in Brazil. Structural adjustment policies eventually began to complement, or replace, stabilisation plans in the 1980s as the focus shifted from demand management to supply.

While a number of creative solutions were attempted to reduce the debt burden, including debt-for-equity and debt-for-nature swaps, and making use of a secondary market that had emerged to trade discounted sovereign debt, it was not until 1989 that US Secretary of Treasury Brady promoted a plan that would eventually restructure Latin America’s debt. There was a menu of options that countries could choose from – including extended maturity dates, reduced face value of loans, and lower interest rates – that were all intended to reduce the burden of debt servicing. Multilateral institutions provided loan guarantees on the new debt in order to induce private banks to participate. The Brady plan, and many of the Washington Consensus reforms that will be discussed in the next section, contributed to restarting growth. In the final analysis, it was only through growth that the debt crisis faded.

The Washington Consensus of the 1990s

The need for a new development paradigm was a natural consequence of the discontent with the inward-looking growth strategy of the post-WWII period and the disastrous economic outcomes of the 1980s. Latin America was emerging from a severe episode of debt-overhang in the early 1990s, and consequently began to emphasise macroeconomic stability as a prerequisite for growth. At the same time, a shift towards a market-based development approach was favoured by the International Financial Institutions.

Chile was the first country in the region to transition to an export-oriented, market-based strategy. The transformation began in the mid-1970s, but was not accompanied by macroeconomic stability in its first decade. The country experienced several deep recessions and, for somewhat different reasons, also suffered from the 1980s debt crisis that hit the rest of the region. Poverty rose sharply in this period, and the liberalising strategy was tainted by having been imposed by an authoritarian regime. Yet important changes took place in this period that contributed to future success. These included the growth and diversification of exports, improvements in government budgeting, and modernisation of the business community (Ffrench-Davis 2002). Chile only became a model for many other countries in the region as of the 1990s. This coincided with the take-off of its economy, a return to democracy, and a more explicit policy focus on growth with equity. Income per capita grew by 4.9 % per year in the 1990s (Madison 2008), or three times the Latin American average, and poverty fell by half (ECLAC 2004). In addition to the successes of the 1990s, Chile began to reform the reforms about a decade before most other countries. This contributed to its role as a policy leader.

Summarising the convergence of views between officials in Washington and technocrats in Latin America in the late 1980s, Williamson (1990) listed a Decalogue of policies that became known as the Washington Consensus. Trade and foreign direct investment liberalisation, the so-called apertura, as well as the deregulation of key markets and the privatisation of state-owned enterprises were its core components. Fiercely challenged in the political and ideological arena, the Washington Consensus became an expression synonymous with ‘neo-liberalism’.

The new paradigm also underscored the importance of policies that resulted in fiscal discipline, market-determined interest rates and competitive exchange rates. Although not directly stating the means to achieve these ends, the Washington Consensus was soon associated with the need to provide independence to central banks. This was seen as a prerequisite to curb inflationary financing of the fiscal deficit and the use of preferential interest rates for specific groups. Budget deficits and overly rigid exchange rates, however, continued to plague many countries and would contribute to a new round of crises at the end of the 1990s.

Trade and capital account liberalisation, financial deregulation, privatisation and central bank independence, were widely adopted in the region during the late 1980s and early 1990s through the so-called ‘first generation’ wave of structural reforms (Edwards 1995; Lora 2001). By 1995, Latin American countries had lowered their tariffs from an average of 50 % in the mid-1980s to 12 % in the mid-1990s, and dismantled quantitative restrictions and other forms of non-tariff protection (Franko 2007). Contrary to the advice of those who supported a gradual and sequential approach, particularly in relation to the liberalisation of the capital account, the political economy of the process often led to the bundling of the reforms. But countries pursued the reform process at different speeds. ‘Aggressive reformers’ (Argentina, Bolivia, Peru, and Chile in an earlier period) often opted for shock therapies, while ‘cautious reformers’ (Brazil, Colombia, Costa Rica and Mexico) proceeded much more gradually (Stallings and Peres 2000).

To mobilise political support, the benefits of the reform process were oversold (Kingstone 2011). Outcomes in terms of economic growth and social progress were generally disappointing: Although Latin America’s annual per capita GDP growth of 1.5 % during the 1990s was much better than the negative growth experienced in the 1980s, it was too low to reduce poverty in a significant way. By the end of the 1990s the rate of poverty still had not been reduced to its 1980 level (ECLAC 2006).

The demise of the Washington Consensus as a development model resulted from the crisis that hit the region in the aftermath of the 1997 Asian financial crisis. The reforms of the 1990s had made Latin America more vulnerable to external shocks. Rather than developing a framework to deal with greater exposure to risk, the region continued to opt for policies that resulted in low domestic savings, high levels of external debt (in sectors with little capacity to generate foreign exchange), overly rigid exchange rates and imprudent lending practices. A sudden reversal in the direction of capital flows following the Asian financial crisis forced an adjustment. The twin fiscal and current account deficits became unsustainable, and as a result currencies were depreciated or allowed to float, and private and public expenditures reduced. Average annual growth per capita was zero between 1998 and 2002 in the region, and unemployment rates rose in most countries, exceeding 20 % in the case of Colombia. The Washington Consensus was blamed and became a politically ‘damaged brand’.

In spite of the demonisation of the Consensus, many countries in the region responded to the crisis by strengthening fiscal discipline, which was a core item on Williamson’s original list. A number of countries passed ‘fiscal responsibility laws’ and improved their budgetary institutions. Many introduced new sources of revenue, with innovative (although inefficient) taxes such as those on financial transactions. Some countries, notably Chile, adopted ‘fiscal rules’ with targets in terms of structural budgets which exclude the transitory components of revenues and expenditures. This allows fiscal policy to operate in a countercyclical manner, running deficits whenever the economy performs below its medium-term trajectory and surpluses when it is above. This feature, together with the adoption of flexible exchange rates in most countries, turned out to be crucial to offset the negative impacts of the global recession of 2008–2009. The efforts that the region undertook to strengthen prudential regulation and de-dollarise financial markets after the banking crises of the late 1990s were similarly critical for navigating the subsequent crisis. (The chapters included in Lora (2007) provide details on each of the reforms adopted since the 1980s.)

Another item that received increased attention after the 1998–2002 crisis was the re-prioritisation of social expenditures, not only to accelerate progress in these areas, but also to offset the consequences of external shocks. Countries throughout the region introduced conditional cash transfer programs (CCTs), such as Bolsa Familia in Brazil and Progresa/Oportunidades in Mexico, targeted to low-income beneficiaries. These and other innovative programs have been effective at reducing current poverty, and have sought to break the intergenerational transmission of poverty by increasing levels of education, nutrition, and health among the young (Fiszbein and Schady 2009).

Despite these adjustments, the Washington Consensus is no longer regarded as a development model (Birdsall et al. 2010). At best, it is seen as a necessary but insufficient condition for development that highlights the importance of macroeconomic stability. At worst, it is seen as an obstacle to be removed. This is the case in some countries that have reversed trends in market deregulation, liberalisation of foreign direct investment and privatisation. For example, nationalisations of ‘strategic’ sectors have been a central element of the development agenda in Bolivia, Ecuador and Venezuela. Most countries in the region, though, remain relatively open to trade and capital flows. The most significant change relative to the original Washington Consensus is that the state is back on the development agenda through an active use of what are now called ‘productive development policies’ (Melo and Rodríguez-Clare 2007).

The emergence of China as a major trading partner – and competitor – is perhaps the most significant economic development in Latin America since 2000. It also helps to explain the re-emergence of productive development policies in the region. In 2010, China accounted for close to one quarter of total exports for Chile, 15 % for Peru and 13 % for Brazil. China’s exports to the region have also increased markedly (20-fold in the case of Brazil between 2000 and 2010). Exports from Latin American countries to China are heavily concentrated in primary products, while imports from China include a diverse set of goods, ranging from textiles to sophisticated manufactured products. While some countries have been left out of the expansion of exports to China, almost all have experienced the effects of greater manufacturing imports from China and greater competition in third-party markets, with a cost in terms of output and employment. Indeed, the re-emergence of productive development policies in many countries owes a great deal to the need to curb the process of deindustrialisation and the loss of export market share in products other than commodities.

Institutions and State Capacity

The broad topic of institutional and state reforms, labelled as ‘second generation’ reforms, proceeded on a separate track and often predated the Washington Consensus (Lora 2007). This is the case, for example, of fiscal and administrative decentralisation and judicial reform that were triggered by the return to democracy and the wave of constitutional reforms which began during the 1980s.

The region has moved forward on a variety of fronts, such as the independence of the judiciary, the professionalisation of the bureaucracy and the quality of political institutions, although the personalisation of politics and the lack of institutionalised and programmatic political parties remains a critical weakness. Specific measures of the quality of the state, however, such as the ability to protect investors against expropriation risk and the World Bank’s governance and ease of doing business indicators, still lag behind the developed world and emerging Asia.

Few measures of state capacity are as relevant as the ability of governments to collect tax revenues. In this case, international comparisons are not favorable to Latin America either. Using IMF data for the 1980–2006 period, total tax revenues relative to GDP were 13.4 % in Latin America in contrast to 22.2 % for the world. When only other former colonies are used as a reference, Latin America falls 5.5 points behind. The comparisons are even more telling with income taxes. In this case, the region is percentage points behind the world average of 9.8 % of GDP (Cárdenas 2010). Interestingly, this is still true despite an active tax reform agenda in Latin America. New taxes have been introduced, but their effects have only offset the decline in revenues associated with lower tariffs and the reduction in corporate income tax rates forced by globalisation.

Of course, generalisations do not capture the varied experiences within the region. A few countries – including Chile and Brazil – have been effective in raising taxes. Consequently, higher fiscal capacity has allowed these countries to provide more public goods and to pursue developmental goals more aggressively. In contrast, the majority of countries in Central America and a number of them in South America, such as Paraguay, have very weak state capacity with tax revenues as a share of GDP in the single digits. As a result, these countries are particularly vulnerable to economic and other shocks, such as natural disasters or security challenges such as the ones confronted in recent years by a number of Central American countries. Weak fiscal capacity also reduces the ability of countries to break out of poverty traps.

Inequality, both economic and political, has been singled out as a crucial obstacle to investment in state capacity (Sokoloff and Zolt 2006). Groups in power prefer the status quo of low taxation, low provision of public goods and low redistribution, perpetuating the effects of extractive colonial institutions. As argued by several authors in the volume edited by Fukuyama (2008), breaking that cycle is one of Latin America’s biggest challenges. Progress in terms of democratic institutions is undoubtedly a sign of hope. However, the evidence suggests that the adoption of the formal architecture of democracy does not necessarily deliver the expected results in terms of building state capacity, in part because it is a slow process, but also because economic inequality prevents democratic governance from delivering its full potential.

The Persistence of Poverty and Inequality

High and persistent inequality is perhaps the most salient feature of Latin America’s development process. The distributions of income, land, education, health and access to basic services all show extremely high degrees of concentration. Inequality is at the centre of virtually all explanations concerning the region’s development problems: economic and social exclusion, limited intergenerational mobility, and weak institutions. However, specific channels and historical evidence are still a matter of controversy.

Although there is some debate about the degree of inequality in Latin America before 1900 relative to other pre-industrial societies or to industrialising Europe, inequality at the time of independence was much higher in Latin America than in North America, mainly as a result of the patterns of colonial land occupation (Engerman and Sokoloff 1997). Recent research suggests that the level of inequality in the region increased after 1870 as a result of the increase in land and mineral rents (relative to wages) during the first phase of commodity export-led growth (Williamson 2010). High inequality in Latin America persisted during much of the twentieth century, in contrast to the equalising trends observed in the industrialised world, explaining why the region has had relatively high levels of inequality compared to Western Europe and Asia (Deininger and Squire 1998). Latin America missed the ‘egalitarian revolution’ that characterised Western and Eastern Europe, as well as North America and Australia up until the 1980s. This is what distinguishes Latin America from the rest of the world. The fact that Latin America adopted similar labour market and social security institutions suggests that a more fundamental element, namely state capacity, was missing.

A systematic analysis of inequality trends in Latin America is limited by the availability of comparable household surveys. Surveys included in the 2008 World Income Inequality database of UNU-WIDER which cover the time period 1867–2006 differ along many dimensions (coverage area, surveyed population, unit of analysis and measure of welfare). Despite data limitations, the general view is that inequality in Latin America increased during the first half of the twentieth century, and then fell during the 1960s and 1970s, only to increase again during the crisis of the 1980s. By the mid-1990s, average inequality in the region was comparable to what it had been in the early 1970s (Londoño and Székely 2000). The more recent evidence suggests that inequality increased somewhat in the 1990s as a result of the reforms and then the crisis at the decade’s end, but then fell during the 2000s (Gasparini and Lustig 2011, and the country papers in López-Calva and Lustig 2010). This suggests that during the last 40 years inequality trends have often moved inversely to overall economic conditions, and that Latin America made no sustained progress in reducing income inequality. It also suggests that some interventions, such as the slow but important increases in educational achievements in recent decades, take time to produce social dividends. A number of authors have pointed to the expansion of human capital and the associated reductions in wage premia as one of the important reasons for falling inequality in the 2000s (Barros et al. 2010; Székely and Sámano 2011). The rise of conditional cash transfer programs in many countries since the late 1990s also contributed to this decline.

However, there are significant differences across countries. The Gini coefficient for the distribution of national household income per capita ranges between 0.45 in Uruguay and 0.60 in Bolivia. In the case of urban areas and narrower definitions of household income the range goes from 0.45 in El Salvador to 0.55 in Brazil, which is still substantial. Regardless of the measure, Uruguay, Venezuela, Argentina and Costa Rica have relatively low levels of inequality, while Bolivia, Brazil and Colombia are among the most unequal societies in the region. In terms of changes, inequality has fallen significantly in Brazil and Chile since the early 2000s. Mexico has made continued, albeit slow, progress in the reduction of inequality since the early 1990s.

From an analytical perspective, Latin America is characterised by ‘excess inequality’, meaning that the level of inequality is greater than what would be expected given the level of overall development. The Gini coefficient is around 10 points higher for the average country in the region relative to what would be predicted by a regression on per capita GDP. In terms of fundamental explanations, the dependence on primary activities, the institutions associated with this economic structure, as well as race and ethnic inequalities, are all interdependent forces difficult to isolate (De Ferranti et al.2004). By any standard, indigenous and afro-descendent groups in Latin America, representing in some countries large shares of the population, are at a disadvantage relative to whites. In contrast, gender gaps have generally narrowed, and in many countries women now obtain more schooling than men.

High levels of inequality contribute to high rates of poverty in Latin America. Some authors estimate that poverty would fall by half if there were no excess inequality in the region (Londoño and Székely 2000). Poverty fell during the period of growth and falling inequality in the 1970s, but then rose sharply during the crisis of the 1980s. Poverty reduction was disappointing in the 1990s, but once again accelerated as income grew and inequality fell in the 2000s. The expansion of conditional cash transfer programs since the late 1990s contributed to this outcome. Poverty fell from 44 to 33 % of the population in the region between 1999 and 2008, and extreme poverty fell from 19 to 13 % (ECLAC 2010).

Until the debt crisis of the 1980s, most of the poor in Latin America lived in rural areas. Since around 1990 poverty became predominantly an urban phenomenon. Poverty rates in rural areas, however, remain almost twice as high as in urban areas – 52 % vs. 27 % in 2008 – and the depth of poverty is more severe. Extreme poverty is three to four times as prevalent in rural areas, which implies that more of the extreme poor in Latin America live in rural areas. As the millennium development goals have gained importance, the first goal of halving extreme poverty and hunger has placed renewed policy attention on rural poverty. Renewed growth and the expansion of CCTs helped push extreme poverty in rural areas down from 38 % to under 30 % between 1999 and 2008.

The causes of rural poverty relate to inequality, institutional deficiencies and market failures. Efforts to reverse high degrees of concentration in land ownership have failed for the most part. Even in Brazil, where there has been an active land reform program in recent decades, the land Gini remained constant at 0.85 between 1985 and 2006 (Hoffmann and Ney 2010). Insecure property rights make Latin America one of the regions in the world with the lowest share of land rented or sharecropped, which impedes access to land for the landless (de Janvry et al.2001). Credit market failures create obstacles to buying land, and make it difficult for small farmers to purchase capital and technology that could enhance their productivity. Insufficient land and capital, combined with low levels of education, help explain the high levels of extreme poverty in rural Latin America. For a significant share of the rural poor, poverty reduction will require access to higher productivity wage labour, migration or anti-poverty programs such as CCTs.

Low levels of education are another important explanation for poverty in Latin America. While primary education practically has been universalised, and over 70 % of youth are enrolled in secondary school in many countries, the distribution of education continues to be very unequal. In a number of the poorer countries, such as Nicaragua and Guatemala, net enrolment rates in secondary school are still below 50 %. And in countries like Brazil and Mexico that have much better average outcomes, the distribution is problematic. In Brazil in the mid-1990s, for example, adults in the bottom 40 % of the income distribution had less than half the education of adults in the top 20 % (Székely and Montes 2006). Mexico was little different.

In terms of Latin America’s ability to compete in higher value-added activities with East Asian and other developing countries, not only has the rate of improvement in educational attainment lagged, but the quality of education is insufficient. Learning outcomes are captured in the OECD Program for International Student Assessment (PISA) test scores. All Latin American countries that took the PISA exams in 2009 obtained scores that were statistically significantly below the OECD mean, and in some cases were among the four lowest performers in the group of 31 non-OECD countries/economies that participated. Equally problematic is that there are significant quality differences within countries. There is a close association between differences in the socioeconomic background of secondary school students of private vis à vis public institutions and the differences in average PISA test scores. The differences in test scores and socioeconomic background of students in Latin America are much greater than those of other developing nations as well as OECD countries. Students in the private system on average perform better than those in the public system. A student in the private system in Brazil, for example, has cognitive skills that are approximately comparable to almost three additional years of public education (OECD 2010). This is a very powerful force that reproduces inequality.

Conclusions

Latin American countries have made enormous strides since the 1870s in improving living standards and human development indicators. Income per capita has increased by a factor of ten, life expectancy has risen by 45 years since 1900, and illiteracy has been reduced from well over two thirds of the population to under 10 %. Latin America has been transformed from a largely rural, agricultural region into a place where 80 % of the population lives in urban areas, over 90 % has access to improved drinking water and 70 % of adolescents attend secondary school.

But the long view on absolute improvements in living standards hides several very different periods. From 1870 to 1981 – a period spanning outward and inward oriented development strategies – Latin America was among the fastest growing regions in the world. Income per capita rose from around 80 to 120 % of the world average, grew as quickly as income in the USA, and rose from about 125 to 280 % of average income in Asia. Yet by 2008, income ratios relative to Europe and Asia were almost identical to what they had been in 1870, and had retreated from 28 down to 22 % of the US level. Latin America has had a growth problem since the early 1980s.

The roots of Latin America’s growth problem began well before the debt crisis of the 1980s. In 1960, output per worker was more than one and a half times greater in Latin America than in East Asia; it is now 50 % smaller. Total factor productivity (TFP) for Latin America relative to the USA, the G8 and East Asia has been declining since the 1960s in most countries in the region. Low TFP can be the result of many forces. One factor that has been singled out is the structure and composition of output in Latin America, which is still characterised by dependence on primary commodities and relatively small domestic markets as a result of the high levels of poverty and inequality. But low productivity also reflects the high levels of informality, as well as low levels of expenditure on innovation, and research and development activities.

Long-term growth that can generate sustained improvements in living standards in Latin America will require gains on many fronts: more effective policies, improved state and market institutions, and a more educated labour force. Another fundamental issue is Latin America’s low saving rates. Fortunately, many countries in the region have already begun to make progress on these issues. A new awareness has spread throughout the region about the importance of fiscal responsibility, low inflation and external balance. Simultaneously, most counties have adopted flexible exchange rates, and foreign exchange reserves rose to unprecedented levels in the 2000s. These factors contribute to reducing the impact of external shocks, and were a key reason why the global recession of 2008–2009 was much less painful for Latin America than the impact of the Asian Crisis in the late 1990s.

Latin American countries missed the opportunity that a number of East Asian countries seized in the 1960s and 1970s to invest heavily in education, open their economies and shift into higher value-added manufactured exports. A new group of Asian countries, led by China, has embarked on this path more recently, making this a more contested strategy to pursue. The growth of manufacturing in Asia has created new opportunities by increasing the demand for commodities and food, but it is also prematurely de-industrialising Latin America. The period since 2002 was one of rapid growth in many Latin American countries, and this created a sense of optimism in the region. But it is still too soon to proclaim that a new period of sustained growth has begun.

Latin American countries have always had an abundance of natural resources, and they are well positioned to take advantage of the most recent commodity boom. But they need to do so wisely, lest they repeat past episodes of disequalising growth, with booms followed by painful busts. They also need to find ways to move up the value-added ladder and develop institutions that assist small and medium enterprises to participate in these markets in order to democratise the benefits of export growth.

There are grounds for optimism about the future of the region, but many challenges remain. Latin American countries have accelerated progress on poverty reduction in the past decade as a result of more rapid growth, falling inequality and social policy innovations that hold the promise of reducing the intergenerational transmission of poverty. Primary school has almost been universalised, and enrolments in secondary school have expanded rapidly. Yet the quality of Latin American schools remains extremely low. Educational attainment as measured by years of schooling is important, but if Latin American workers are going to compete successfully, governments in the region must simultaneously prioritise an improvement in learning outcomes.

See Also