WORLD BANK LATIN AMERICAN 96467 v2 AND CARIBBEAN STUDIES Latin America and the Rising South Changing World, Changing Priorities Augusto de la Torre, Tatiana Didier, Alain Ize, Daniel Lederman, Sergio L. Schmukler OVERVIEW OVERVIEW Latin America and the Rising South Changing World, Changing Priorities Augusto de la Torre, Tatiana Didier, Alain Ize, Daniel Lederman, and Sergio L. Schmukler This booklet contains the Overview and a brief list of contents from the forthcoming book, Lat- in America and the Rising South: Changing World, Changing Priorities (doi: 10.1596/978-1- 4648-0355-0). A PDF of the final, full-length book, once published, will be available at openknowledge .worldbank.org, and print copies can be ordered at www.amazon.com. Please use the final version of the book for citation, reproduction, and adaptation purposes. © 2015 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 18 17 16 15 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Exec- utive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immuni- ties of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: De la Torre, Augusto, Tatiana Didier, Alain Ize, Daniel Leder- man, and Sergio L. Schmukler. 2015. “Latin America and the Rising South: Changing World, Changing Priori- ties.” Overview booklet. World Bank, Washington, DC. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the at- tribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copy- right owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover image: © Inked Pixels/Bigstock.com. Used with permission. Further permission required for reuse. Cover design: Critical Stages. Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Contents of Latin America and the Rising South: Changing World, Changing Priorities . . . . . . . ix Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Changes at the center of the world economy . . . . . . . . . . . . . . . . . . . . . . . . . . .2 How the rise of the South conditioned development in Latin America and the Caribbean: An interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Changing world, new priorities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Annex A (table A.1): Country group composition . . . . . . . . . . . . . . . . . . . . . . . 35 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 iii Foreword The dynamics of the world economy have than 100 years, Latin America’s average income changed radically and the once immutable per capita has remained barely 30 percent of that assumptions of the global trade and financial of the United States. In other words, the region order no longer hold fast. In the last two decades has been unable to narrow a gaping income dis- alone, wealth has shifted so profoundly that the parity with its northern neighbor. simple, old North-South hierarchy—where the This is not to say that Latin America has been North were the rich few and the South were the unable to grow. In fact, during the commodity many poor countries of the world—is no longer a boom of the 2000s, average growth rates reached given. In fact, in 1990, the majority of the world nearly 5 percent. Moreover, income growth of the population, 62 percent, lived in poor countries. poorest 40 percent was higher in Latin America As of 2010, 72 percent of the world’s population and the Caribbean than in any other region of lived in middle-income countries. the world, relative to the total population, mak- Such tremendous transformation is the ing growth also equitable. inspiration for the World Bank’s latest regional Global economic activity, however, has flagship report for Latin America and the Carib- slowed and medium-term growth prospects have bean, Latin America and the Rising South: diminished. Latin America is now in its fourth Changing World, Changing Priorities. As an year of growth deceleration, and it is expected to in-depth look at the region’s expanding global grow below 1 percent in 2015. This poses brand connections in trade and finance, and a sober new challenges, particularly as the conditions assessment of its promise and challenges, the that led to the good years of the 2000s are not report is an important contribution in and of with us anymore. itself; at the same time, as a report that tracks Current global conditions pose similar chal- global trends, it also provides an invaluable anal- lenges to all middle-income countries, not only ysis that the World Bank is uniquely positioned those in Latin America. Indeed, disappointing to undertake. growth in major emerging economies around While these global trends were the inspira- the world raises important concerns, particu- tion, the motivation behind this report is the larly considering that two thirds of the extreme urgent need to disentangle the complicated knot poor in the world still live in middle-income of Latin America’s growth problem. For more countries. For the World Bank Group, a global v vi   F o r e w o r d institution committed to eradicate extreme past two decades. It is therefore our hope that a poverty by 2030 and to boost prosperity for the profound look at the way Latin America—and bottom 40 percent of the population, these are the world—have been integrating will help shed crucial challenges. a light on the way forward. In other words, our The web of connections that have multiplied expectation is that a clearer understanding of throughout the world from the North to the how the South has been rising—and how it has South, from the South to the North, and, per- not—will help those countries break out of their haps more significantly, from the South to the middle-income status and move closer to the South represents an important change over the group of rich nations. Jorge Familiar Vice President for Latin America and the Caribbean The World Bank Acknowledgments T his report was prepared by a core team from Paulo Bastos, Laura Chioda, Ana M. Fer- comprising Augusto de la Torre, Tatiana nandes, Eduardo Fernández-Arias, Michael Didier, Alain Ize, Daniel Lederman, and Ferrantino, Margaret Ellen Grosh, Jose Luis Sergio L. Schmukler. Additional contributions Irigoyen, Ayhan Kose, Gian Maria Milesi-Fer- were made by Erhan Artuç, Chad Bown, Fer- retti, Marc A. Muendler, Ana L. Revenga, Sergio nando Broner, Constantino Hevia, Ha Nguyen, Urzua, and other participants in the authors’ Samuel Pienknagura, Luis Servén, and Ganesh workshop that took place on February 27–28, Wignaraja. We thank Magali Pinat for invalu- 2014. While we are grateful for the guidance and able help in putting together and coordinating comments received, the core team is respon- the documents that constitute this 2015 Regional sible for all remaining errors, omissions, and Flagship Report. We are particularly grateful interpretations. for the truly outstanding research assistance Book design, editing, and production were provided by Matias Moretti (Chapter 4), Magali coordinated by the World Bank’s Publishing and Pinat (Chapters 1 and 2), and Diego Rojas (Chap- Knowledge department under the supervision ter 3), who in addition co-authored some of the of Patricia Katayama and Mark Ingebretsen. We background papers for this Report. We also ben- also appreciate the assistance provided by Mauro efitted from able research assistance at different Lopes Mendes de Azeredo, Sergio Jellinek, and stages of the project provided by Diego Barrot, Marcela Sanchez-Bender on the report’s pub- Julia Gottlieb, Lucas Rusconi, Martin Sasson, lication and dissemination activities. Finally, Tanya Taveras, and Shajuan Zhang. we thank Ruth Delgado Flynn and Jacqueline The team was fortunate to receive superb Larrabure Rivero for unfailing administrative advice and guidance from the following peer support. reviewers: Eduardo Cavallo, Tito Cordella, Barry Eichengreen, Caroline Freund, Aart Kraay, Wil- liam Maloney, Andrés Rodríguez-Clare, David Rosenblatt, and Shahid Yusuf. We are also grate- ful for valuable comments and insights received vii Contents of Latin America and the Rising South: Changing World, Changing Priorities Foreword Acknowledgments Abbreviations Overview 1 Three Global Trends That Shaped Latin American and Caribbean Development at the Dawn of the Twenty-First Century 2 The Structure of Trade Linkages and Economic Growth 3 Big Emerging Markets, Big Labor Market Dislocations? 4 The Changing Patterns of Financial Integration in Latin America and the Caribbean 5 Ascending with the South Winds: Will Low Saving in Latin America and the Caribbean Be a Drag? ix Abbreviations EAP East Asia and Pacific ECA Europe and Central Asia ER exchange rate FDI foreign direct investment FVA foreign value added GDP gross domestic product GVC global value chain G-7 Group of Seven IR interest rate LAC Latin America and the Caribbean M&A mergers and acquisitions MENA Middle East and North Africa SA South Asia SSA Sub-Saharan Africa xi Overview T he world economy is not what it used Made in China?” (De la Torre and others to be 30 or even 15 years ago. The 2011). While China was the sole focus then, rise of the South—that is, the growing the analysis here is deeper and broader, not economic influence of emerging economies— least because it covers the evolving role of has changed the global economic landscape.1 emerging economies more generally. The changes have been deep and most likely This report argues that as the world econ- permanent. They reflect not only the grow- omy has irreversibly changed, LAC has been ing economic heft of the South, given its sub- adjusting to the associated global economic stantially higher growth rates with respect to shocks, both commercial and financial. The the North (that is, the advanced economies), adjustment process has been conditioned by but also structural changes. The South has LAC’s trade and financial structures and become a driver of global economic trends reflected in the observed patterns of struc- by playing a role that is qualitatively different tural change. Key challenges have emerged from that of the North. At the epicenter of for the region, particularly because the these changes has been China. changes may not have improved the region’s This report focuses on the restructuring of prospects for long-term economic growth. the global economy and its implications for Simply put, economic policy priorities in the the development and policy priorities of Latin region have evolved in response to worldwide America and the Caribbean (LAC). It exam- changes even as these changes have exacer- ines how the global economy has changed, bated some of the region’s long-standing especially with regard to the patterns of development challenges, such as those associ- international trade and financial integration ated with its dependence on mineral and agri- as well as the differential roles played by the cultural commodities and its comparatively large emerging economies and the traditional low saving rates. The debate in the region economic powers. Some of these themes were over public policy priorities in the context of explored, in a preliminary fashion, in the a new global landscape will thus likely inten- September 2011 issue of the LAC Region’s sify, with the growth agenda at its core. semiannual report series, “Latin America The rest of this overview addresses the and the Caribbean’s Long-Term Growth: “what,” the “how,” and the “so what” 1 2   LATIN AMERIC A AND THE RISING SOUTH questions associated with the rise of the and foreign investment as potential drivers South and its implications for LAC. The of growth and productivity; labor market overview is organized in three main sec- frictions, which make economic adjustments tions. The first documents salient features of sluggish and thus reduce the potential gains the new global economic order by focusing from globalization; and the region’s notori- on the rising prominence of emerging econ- ously low national saving rates, which may omies. It characterizes the tectonic shifts hamper long-term growth by undermining in the global economy, including by look- external competitiveness. ing at the data through the lens of network analysis. It then examines the fundamental change in the role of the South in the global Changes at the center of the economy and highlights key dimensions of world economy heterogeneity within the South. To fully understand the implications of the The second section provides an economic economic rise of the South, it is helpful to interpretation of how the changes at the distinguish between the economic weight of heart of the global economy are conditioning emerging economies, the extent of trade and growth and employment prospects in LAC. financial integration of these countries, and This narrative posits that, from the point of the different roles played by the North and view of LAC, the rise of the South manifested South countries that are systemically import- itself as a set of economic shocks working ant for the world economy. through commercial and financial channels. The impacts of these shocks varied across the Tectonic shifts in the global region, depending on countries’ initial trade economic landscape structures, resource endowments, degree of financial globalization, and saving patterns, For most of the 20th century, global eco- among other factors. nomic activity was concentrated in the The third section assesses broad policy developed North (composed of Canada, the areas that, given the rising South phenome- United States, the Western Europe coun- non, should find their way to the top of the tries, and Japan, which joined the pack region’s growth-oriented reform agenda. only after World War II). Since the dawn of Among these areas are the structure of trade the 21st century, the South (defined as all FIGURE 1  The rise of the South a. World GDP b. World trade 100 100 90 90 80 80 Share of world trade (%) Share of world GDP (%) 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 19 1 19 4 19 7 19 0 19 3 19 6 19 9 19 2 19 5 19 8 19 1 19 4 20 7 20 0 20 3 20 6 20 9 12 19 1 19 4 19 7 19 0 19 3 19 6 19 9 19 2 19 5 19 8 19 1 19 4 20 7 20 0 20 3 20 6 20 9 12 6 6 6 7 7 7 7 8 8 8 9 9 9 0 0 0 0 6 6 6 7 7 7 7 8 8 8 9 9 9 0 0 0 0 19 19 South China North Sources: Calculations based on data from World Development Indicators (WDI) and Direction of Trade Statistics (DOTS). Note: The North includes the G-7 members and Western Europe countries. The South includes all other economies. G-7 = Group of Seven; GDP = gross domestic product. O V E R V I E W   3 developing economies not in the North), led 48 percent (figure 2, panel a), and its share of by China and other large emerging econ- global imports of primary (agricultural and omies, has risen with surprising speed. In mineral) goods expanded from 32 percent to fact, several South countries have become 47 percent (figure 2, panel b). An accelera- major, systemically important players in the tion of financial globalization accompanied global economy. The gross domestic prod- the rise of the South in commercial flows. uct (GDP) of the South, which represented The South’s share of global capital inflows about 20 percent of world GDP between (including foreign direct investment [FDI]) the early 1970s and the late 1990s, doubled rose from about 18 percent in the 1970s to to about 40 percent by 2012, with China 25 percent in the 1990s and to more than alone accounting for 12 percent of global 50 percent by 2012 (figure 3). GDP (figure 1, panel a). The increase in the economic weight of The rising share of the South in global the South is likely here to stay: it is probably GDP was accompanied by increasing influ- neither short lived nor reversible. Although ence in international trade and finance. long-term economic forecasts are notoriously Indeed, although the secular process of uncertain, current projections suggest that globalization of the South had long been the South will continue to gain importance in advancing, the 2000s saw a notable intensi- the world economy. According to the World fication of this process. The South’s partici- Bank’s 2013 Global Development Horizons, pation in global trade rose from 24 percent the share of the South in global GDP will in 1970 to 35 percent in 2000 and 51 percent reach 55 percent by 2025. A 2012 report by in 2012 (figure 1, panel b). This advance was the U.S. National Intelligence Council proj- associated with major transformations in the ects this share to reach 70 percent by 2030. structure of world trade, as the weight of the The Asian Development Bank forecasts South varied across sectors. Between 2000 that the share of exports from the South and 2012, the South’s share of global exports will rise to 64 percent of global exports by of manufactures increased from 32 percent to 2030 (Anderson and Strutt 2011). The 2013 FIGURE 2  The South’s share of global trade flows a. Share of exports of manufactured goods b. Share of imports of primary goods 50 50 45 Share of world imports of Share of world exports of manufactured goods (%) 40 40 primary goods (%) 35 30 30 25 20 20 15 10 10 5 0 0 00 01 02 03 04 05 06 07 08 09 10 11 12 00 01 02 03 04 05 06 07 08 09 10 11 12 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Other South China India Russian Federation Other South China India Korea, Rep. Korea, Rep. Poland Turkey Czech Republic Thailand Australia Lithuania Singapore Vietnam Malaysia Source: Calculations based on data from Comtrade database. Note: The eight South countries that gained the most in market share between 2000 and 2012 are shown separately from the rest of South countries. The North includes the G-7 mem- bers and Western Europe countries. The South includes all other economies. G-7= Group of Seven. 4   LATIN AMERIC A AND THE RISING SOUTH FIGURE 3  The South’s share of global capital inflows have yet to be linked to a wide set of other countries, especially in terms of financial 100 connections. Indeed, only 18 percent of the 90 potential South-South connections related to portfolio flows were active in 2011.4 80 Share of world capital inflows (%) 70 The fundamental change in the global 60 role of the South 50 Changes in relative economic weight provide a bird’s-eye view of the rise of the South. But, 40 impressive as they are, they do not illustrate 30 the full scale of the economic shifts in the global landscape. Further insights into the 20 nature of the rise of the South emerge when 10 trade and financial connections are viewed from a global network perspective. Four key 0 stylized facts arise from this approach (for a 19 0 19 2 19 4 19 6 19 8 19 0 19 2 19 4 19 6 19 8 19 0 19 2 19 4 19 6 20 8 20 0 20 2 20 4 20 6 20 8 20 0 12 7 7 7 7 7 8 8 8 8 8 9 9 9 9 9 0 0 0 0 0 1 more detailed analysis, see chapter 1 of this 19 South China North report). First, the North is no longer the center of Source: Calculations based on data from Balance of Payments Statistics (BOPS). Note: Gross capital inflows include portfolio, banking, and foreign direct investment flows. The the global trade network and the South is no North includes the G-7 members and Western Europe countries. The South includes all other econ- longer its periphery. Indeed, several econo- omies. G-7= Group of Seven. mies from the South have become part of what can be empirically characterized as the “cen- Global Development Horizons projects that ter” of global trade. This momentous change by 2025 the South will account for 63 per- is highlighted in figure 4, which shows the cent of world capital inflows and 80 percent global trade network in 1980 and 2012. Each of world capital outflows. node in the graphs represents a country, and As the South gained weight in the global each link corresponds to exports from one economy, the number of its bilateral economic country to another (indicated by the arrows). connections proliferated. These ties increased Connections that are trivial in magnitude are in every direction, but new South-South con- not graphed, but once graphed, each con- nections rose more rapidly than North-South nection has the same weight. The greater the linkages in both trade and finance. In 1980, number of its connections to other countries, the number of active South-South trade con- the more centrally located a country is. nections was 40 percent of all possible con- The change has been remarkable. In 1980, nections (the number of connections that only a few North countries—the United would exist if every South country were con- States, some Western Europe countries, and nected to every other South country). This Japan—stood at the center of the global trade figure rose to 46 percent in 1990 and 70 per- network. In contrast, by 2012, several South cent in 2012. Trade linkages between North countries—including not only China but also and South countries expanded less rapidly Brazil, India, the Russian Federation, South (from 92 percent in 1980 to 96 percent in Africa, and Turkey—had moved to the center. 1990 and 98 percent in 2012), at least in part Second, at the center of the global trade because they had been almost fully exploited network, the role played by countries from since the 1980s.2 Similar trends are observed the South and countries from the North across different types of financial flows.3 To differs. This stylized fact is illustrated in be sure, this process is far from mature, as a figure 5, which shows the relative (rather significant number of countries in the South than absolute) importance of each country O V E R V I E W   5 FIGURE 4  The global trade network a. 1980 b. 2012 BRA North countries Latin America and the Caribbean Other South countries Source: Calculations based on data from DOTS. Note: Networks are drawn using the Kamada-Kawai algorithm. Each node represents a country. Each link corresponds to an active connection between a pair of countries. Arrows indicate the direction of these connections. The North includes the G-7 members and Western Europe countries. Other South includes all other economies except Latin America and Caribbean countries. Only trade flows (exports) greater than $10 million in 1980 or greater than $100 million in 2012 are shown. The figure thus ignores very small countries. It would show similar results if these connections were reported. G-7 = Group of Seven. FIGURE 5  Similarity and systemic importance in the global trade network a. 1980 a. 1980 structures structures in trade in trade Similarity Similarity Systemic relevance in global trade Systemic relevance in global trade b. 2012 b. 2012 structures structures in trade in trade Similarity Similarity Systemic relevance in global trade Source: Calculations based on data from DOTS. Systemic relevance in global trade Note: Each node represents a country. Each link corresponds to an active trade connection between a pair of countries. Arrows at the end of each link capture the direction of these connections. Trade connections are measured as exports as a share of total exports of the source country. Only shares greater than 1 percent are reported. The distance between countries reflects similarity in the structure of their trade connections: the closer countries are to one another, the more alike they are in terms of export shares. Countries capturing a larger share of other countries’ exports and connected with a larger number of trading partners appear on the right-hand side of the figure (more systemically relevant countries in global trade). The smaller the distance between two countries along the vertical dimension, the more similar the structure of their trade connections across other members of the network. O V E R V I E W   7 in the global trade network. The vertical dis- trading partners. The right side of the figure tance between countries in the figure reflects resembles a star, with small groups of central the degree of similarity in the structure of countries placed at a certain vertical distance their trade connections, whereby more sim- from one another. The Russian Federation ilar countries are grouped closer together. 5 and Turkey, for example, are not located The farther to the right of the figure a coun- near any North core country from Europe, try is located, the greater its importance to and Japan is not close to either China or the the global trade network.6 Republic of Korea. The implication is that Panel a of figure 5 shows that in 1980 systemically important South countries play only North countries were clustered toward a different role from the role played by North the right of the graph, thus indicating that countries in the global trade network. These they were of greatest systemic importance to different roles seem to be inherently linked the global trade network. In addition, these to fundamental differences in factor endow- countries were very close to one another ments, trade, production, and aggregate along the vertical dimension, reflecting a demand structures, as discussed below. high degree of similarity in the structure of Third, there is a notable asymmetry in the their trade connections with other countries patterns of change in global trade and finan- in the network. cial networks. In the sphere of trade, the tra- The global trade network in 2012 shifted ditional overlap between the North and the dramatically (figure 5, panel b). Several coun- “center” (and the South and the “periphery”) tries from the South appeared on the right no longer holds. In contrast, in the sphere of side of the figure, indicating their increased finance, countries from the North still stand systemic relevance to world trade. However, alone at the center, as illustrated in figure 6 they remained somewhat distant (along the for syndicated bank loans. A similar picture vertical dimension) from the other (North) emerges for portfolio investments, merg- countries on the right side of the figure, ers and acquisitions (M&A), and greenfield reflecting differences in trade shares across investment flows. Whether this asymmetry FIGURE 6  The global financial network for syndicated bank loans a. 1996 b. 2012 North Latin America and the Caribbean Other South Source: Calculations based on data from SDC Platinum. Note: Networks are drawn using the Kamada-Kawai algorithm. Each node represents a country. Each link corresponds to an active connection (a positive flow of investments) between a pair of countries. Arrows indicate the direction of these connections. The North includes the G-7 members and Western Europe countries. Other South includes all other economies except Latin America and Caribbean countries. G-7 = Group of Seven. 8   LATIN AMERIC A AND THE RISING SOUTH proves transitory is debatable, although most of production activities belonging to the observers agree that it is unlikely to be dis- same production processes across countries. lodged soon, for several reasons. For start- As GVCs have gained prominence on the ers, there is broad recognition that the U.S. international trading scene, exports of final dollar continues and will continue to have a products have become increasingly composed stronghold as both the privileged currency of imports of intermediate inputs. To date, for international contracts and the safe haven GVCs are mostly regional, not global. The in times of global risk aversion. In addition, foreign value added (FVA) content in exports the scale and network effects associated with typically originates in neighboring countries the dominance of the advanced financial cen- (figure 7).7 For example, about 56 percent of ters (including New York, London, Frank- the FVA in the exports of East Asian coun- furt, Tokyo) will not be easy for the South tries come from other East Asian economies, to overcome. This trade-finance asymme- and more than 72 percent of the FVA in the try in global networks stands in sharp con- exports of European countries come from trast to broad historical developments since other European economies. There is also the Industrial Revolution and throughout clustering—albeit less intense—across coun- most of the 20th century, when countries tries within LAC subregions. For instance, that became important trading powers also imports from other South American coun- became important international financial tries represent about 30 percent of the FVA in centers. the exports of South America. Fourth, despite an increase in the number of connections around the world, there is a The heterogeneity of the South significant degree of regional (geographic) clustering within global trade and financial The rise of the South in global economic networks. Underpinning these clustering pat- affairs conceals important differences across terns has arguably been the development of South countries. Four types of heterogeneity global value chains (GVCs)—the distribution are noteworthy. The first is differences in the FIGURE 7  Regional clustering in global value chains, 2011 45 40 35 Share of exports (%) 30 25 20 15 10 5 0 East Asia Europe North and Central South America America Intraregional East Asia Europe North and Central America South America Sources: Calculations based on data from Eora MRIO and WDI. Note: Figure shows the geographical composition of sources of foreign value added used in a country’s exports, scaled by the country’s exports. O V E R V I E W   9 changes in export and import shares of the India, Korea, Poland, and Turkey) are outside South (recall figure 2). The rise of the South LAC. As such, LAC gained global relevance implied a growing share of the South (as a as a major commodity exporting region even whole) in global manufacturing exports. But though it lost relevance as a manufacturing only a subgroup of South countries carried exporter. the load in this regard, with China the leader A second important dimension of het- by a wide margin. China’s share in global erogeneity within the South is the contrast manufacturing exports increased by more between LAC and the East Asian economies than 10 percentage points, from slightly less in terms of the density of their regional trade than 5 percent in 2000 to more than 15 per- networks. Figure 8 highlights this feature cent in 2012. In contrast, the other top 20 by providing snapshots of the regional trade South countries in terms of their increases networks of these two regions in 1980 and in global shares—a group that includes 2012. Each regional trade network includes Brazil and Chile—increased their share of (as nodes) all countries of the region plus the global manufacturing exports as a group by five countries from the rest the world that are only about 8 percentage points. The shares the largest trading partners for each regional of world manufacturing exports of several network.8 large South countries (for example, Malay- In 1980 the trade networks of LAC and sia, Mexico, and the Philippines) actually East Asia were similar: they were thin, declined. unbalanced, and centered on a few domi- The rise of the South also featured a sub- nant North economies. Japan and the United stantial increase in its share of trade (exports States were the only two dense nodes in the and imports) of primary (mineral and agricul- 1980 snapshot of the East Asian network, tural) products. But cross-country differences and the United States was the sole dense node within the South are stark. In particular, the in the 1980 LAC network. set of South countries whose shares in com- By 2012 the two regional networks had modity exports rose most significantly has diverged. The East Asian network had little overlap with the set of South countries become substantially denser and more bal- whose shares of commodity imports rose. anced, with high-density connections distrib- In contrast, the set of South countries whose uted rather evenly across numerous countries shares of manufacturing exports rose signifi- (nodes), including not just Japan, the United cantly (virtually all of which are outside LAC) States, and China but also Korea, Malaysia, has greater overlap with the set of South Singapore, and Thailand. In contrast, the countries whose shares of commodity imports 2012 snapshot of the LAC trade network rose. Australia, Brazil, and the Russian Fed- was almost as thin as it was in 1980, and it eration jointly accounted for the largest gains remained dominated by the United States, in the shares of global primary exports (their with Brazil a very distant second. A signif- share rose from 13 percent in 2000 to 23 icant change between 1980 and 2012 was percent in 2012). Other top 20 commodi- that China joined the LAC network, albeit at ty-exporting countries from the South include a comparatively low density.9 Azerbaijan, India, Kazakhstan, and several The large difference in regional network LAC countries (Bolivia, Chile, Colombia, densities in 2012 reflects trade connections Ecuador, Peru, and Uruguay). China stands within East Asia that became multidirec- out as a giant commodity importer: its share tional (that is, intense in the direction of of global imports of agricultural and mineral virtually every country within the network). commodities rose from less than 4 percent in In contrast, connections within the LAC net- 2000 to more than 15 percent in 2012. All work have remained largely bi-directional, other South countries with rising manufac- linking LAC countries mainly with the turing export shares that also increased their United States and secondarily with China shares of imports of commodities (such as (and, within the South America subregion, 10   LATIN AMERIC A AND THE RISING SOUTH FIGURE 8  Density maps of regional trade networks a. The Latin American network, 1980 b. The Asian network, 1980 (ccontinued) (continued) O V E R V I E W   11 FIGURE 8  Density maps of regional trade networks (continued) c. The Latin American network, 2012 d. The Asian network, 2012 Sources: De la Torre, Didier, and Pinat 2014 and DOTS. Note: Figure shows the density maps of two regional trade networks based on bilateral exports, measured as a share of total exports of the sending country in 1980 and 2012. The density of a country in these maps depends on the number of neighboring countries and the economic distance between countries. The node density is translated into colors using a red-green-blue scheme in which red indicates the highest density and blue the lowest. Each country is represented by its three-letter acronym. See box 1.1 in chapter 1 of this report for technical details. 12   LATIN AMERIC A AND THE RISING SOUTH FIGURE 9  Composition of foreign assets and liabilities in the Brazil). The density of connectivity in the South, by region East Asian network also suggests strong feedback effects, whereby tighter trade con- a. LAC-7 nections within East Asian emerging econ- 10 omies boost trade with advanced countries 5 in the North and vice versa. In contrast, 0 LAC countries (with the possible exceptions –5 of Mexico and Costa Rica) seem to signifi- Percent of GDP –10 cantly underexploit the potential for comple- –15 mentarities and mutually reinforcing effects –20 between intraregional trade and global –25 trade. These different patterns may be linked –30 to the fact that East Asian countries partici- –35 pate much more actively in GVCs than LAC 1990 1993 1996 1999 2002 2005 2008 2011 countries do. Year A third salient dimension of heteroge- b. Asia-7 neity concerns the asymmetric shifts in the 20 net debtor-creditor positions with respect 10 to the rest of the world for different emerg- ing regions in the South. LAC and East Asia 0 followed a similar pattern in this respect, in Percent of GDP –10 sharp contrast with countries from Eastern Europe and Central Asia (figure 9). During –20 the 2000s, there was a major shift from debt –30 to equity in the external net liability posi- –40 tions of East Asia and LAC (in the context of the rise of the South). In contrast, Eastern –50 1990 1993 1996 1999 2002 2005 2008 2011 Europe and Central Asia shifted its position Year toward debt liabilities. Regarding debt contracts, East Asia and c. ECA-7 5 LAC went from being large net debtors with 0 respect to the rest of the world in the 1990s –5 to significant net creditors during the 2000s. This change reflected a strengthening of –15 Percent of GDP macrofinancial policy frameworks, which –25 entailed a process of external debt reduction by governments coupled with self-insurance –35 through accumulation of international –45 reserves by central banks.10 It also reflected the continued presence of large current –55 account surpluses, particularly among the 1990 1993 1996 1999 2002 2005 2008 2011 Year high-saving East Asian economies. Over the same period, both East Asia and Net debt position Net equity position LAC became more active users of foreign equity finance, which led to rising net debtor Source: Calculations based on updated and extended version of dataset constructed by Lane and positions in risk-sharing equity contracts Milesi-Ferretti 2007. Note: Ratios are calculated at the country level and then averaged across countries (simple average) (particularly FDI) with respect to the rest of between 1990 and 2011. LAC-7: Argentina, Brazil, Chile, Colombia, Mexico, Peru, and Uruguay. Asia-7: the world. The equity-laden position LAC China, India, Indonesia, the Republic of Korea, Malaysia, Philippines, and Thailand. ECA-7: Croatia, the Czech Republic, Hungary, Lithuania, Poland, the Russian Federation, and Turkey. GDP = gross and East Asia achieved in the 2000s arguably domestic product. represents a more resilient form of integrating O V E R V I E W   13 into often volatile international financial mar- FIGURE 10  Saving, investment, and the current account kets than the debt-laden external net liability position of Eastern Europe and Central Asia. a. LAC-7 A fourth dimension of heterogeneity that 40 is key to understanding the implications of 30 the rise of the South is the differences in 20 Percent of GDP the relative importance of domestic versus 10 external demand in macroeconomic aggre- gates. The contrast is sharpest between 0 LAC and East Asia. While in LAC domes- –10 tic demand largely drives the economy, in –20 East Asia external demand is a dominant –30 force. That LAC exhibits domestic demand– –40 driven macroeconomic patterns implies an 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 excess of aggregate demand over national income and, hence, typically low saving b. EAP-5 rates and a penchant for current account 40 deficits (figure 10). The external demand– 30 driven patterns of East Asia imply an excess 20 of national income over aggregate demand Percent of GDP and, hence, typically high domestic saving 10 rates and current account surpluses. The 0 macroeconomic patterns of the emerging –10 economies of Eastern Europe and Central –20 Asia are more similar to LAC than to East –30 Asia. As argued below, a macroeconomic pattern that relies on external demand, and –40 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 therefore high national saving rates, may 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 be more conducive to seizing the potential c. ECA-6 growth benefits associated with the rise of 40 the South. 30 20 How the rise of the South Percent of GDP 10 conditioned development in Latin America and the 0 Caribbean: An interpretation –10 –20 The rise of the South has left a noticeable mark upon the world economy. The pre- –30 ceding discussion highlights the heteroge- –40 neity of structural economic characteristics 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 within the South before and during its rise, Investment Saving Current account especially since 2000. This section interprets these global and regional trends, based on the Source: Calculations based on data from the IMF’s International Financial Statistics (IFS). evidence presented in this report. Note: Simple regional averages are presented. LAC-7 includes Argentina, Brazil, Chile, Colombia, Mexico, Peru, and Uruguay. EAP-5 includes Indonesia, the Republic of Korea, Malaysia, the Philip- From the viewpoint of small open- pines, and Thailand. ECA-6 includes Croatia, the Czech Republic, Hungary, Lithuania, Poland, and economies, including LAC countries, the Turkey. GDP = gross domestic product. rise of the South can be understood as having set three types of global shocks in motion: a supply shock, a demand shock, 14   LATIN AMERIC A AND THE RISING SOUTH and a financial shock. Both the demand and of manufactures, led by but not limited to supply shocks have been associated with China. This shock presumably lowered the the asymmetric rise of the South across (quality-adjusted) prices of manufactured industries and trade flows (exports ver- goods and thus dampened global inflationary sus imports). The financial shock has been pressures. The shock can be interpreted as related to the recycling of savings from the emanating from an increase in the number of emerging South. manufacturing workers engaged in interna- LAC countries responded differently to tional trade, whose labor services were previ- these shocks as a result of differences in ously not integrated into the global economy initial conditions, including factor endow- (arguably the case of China before it joined ments, initial trade structures, and macro- the World Trade Organization in 2001). economic frameworks. As it is difficult to For LAC economies, this shock implied precisely identify the direction of causality, increased international competition for var- this narrative provides an interpretation of ious manufacturing industries. It thus insti- the facts and statistical findings rather than gated structural changes across sectors as a model of how the world economy has been well as within LAC’s manufacturing sector. operating. The resulting decline in the relative prices of This section thus characterizes the rise manufactured goods was also associated with of the South from the viewpoint of LAC improved terms of trade for economies that as a combination of external shocks. Sub- were net importers of manufactured goods. sequently, it examines the heterogeneous A demand shock was associated with an responses to such shocks across countries in increase in global demand for primary goods. the region and discusses the potential impli- It reflected the relatively high commodity cations for LAC’s long-term growth and (to a intensity of imports of the larger rising South lesser extent) employment. countries, particularly China. The result was a rise in commodity prices—an unusu- ally vigorous upswing phase of a veritable The rise of the South as external shocks commodity supercycle.11 For commodity for Latin America and the Caribbean exporters, including in LAC, this shock was A global supply shock was related to the huge associated with terms of trade gains. expansion in South-originated production The effects of the global supply shock may have dominated the effects of the global demand shock to the extent that large cur- FIGURE 11  Real U.S. interest rates rent account surpluses were observed at the epicenter of the shock (China and other East 12 Asian economies). Consequently, the com- 10 bination of the global supply and demand 8 shocks engendered a global financial shock. Percentage points 6 This shock was associated with the inter- 4 national recycling of net savings from the 2 South, particularly from the Asian and Mid- 0 dle Eastern countries, and changes in relative –2 prices in financial markets around the world, –4 including exchange and interest rates. These –6 South countries integrated into the global economy with persistent current account 80 19 2 84 86 19 8 90 92 19 4 96 20 8 00 02 04 06 08 10 12 14 8 8 9 9 19 19 19 19 19 19 19 20 20 20 20 20 20 20 surpluses that were accumulated mainly in Sources: Calculations based on data from the Board of Governors of the U.S. Federal Reserve System the form of international reserves, most of and the Federal Reserve Bank of Cleveland databases. Note: Series was constructed by deflating the (effective) monthly federal funds rate by the inflation which were recycled through the North. The rate for the previous 12 months. result was a “global savings glut” that eased O V E R V I E W   15 financial constraints in countries with exter- a deterioration of their terms of trade because nal and fiscal deficits, particularly the United of their export dependence on light manufac- States, and exerted significant downward tures and high level of imports of commod- pressure on world interest rates.12 Accom- ities. In addition, in some LAC economies, modative monetary policy in the North con- low domestic saving rates further reduced the tributed to the maintenance of unusually competitiveness of the manufacturing sector, low global interest rates (figure 11). With and in economies with large agricultural and low interest rates in the North, a search for mining sectors, wages were pushed up, as yield among investors triggered capital flows explained below. to the South, including LAC, where borrow- Illustrative of the differences within LAC ing spreads fell to historically low levels and as a whole, figure 13 shows the evolution of currencies experienced strong appreciation indexes of manufacturing export similarity pressures. for Brazil and Mexico. Brazil’s highly diver- sified export structure (spanning from agri- cultural commodities to automobiles) has Heterogeneity of impacts as a result of been more similar to that of the United States initial sectoral trade weights and the European Union than that of China. The combination of these supply and demand In contrast, Mexico’s manufacturing export shocks affected the LAC countries’ patterns basket has been consistently more similar of trade differently, depending on their nat- ural endowments, geographical character- istics, economic size, and initial production FIGURE 12  Terms of trade within Latin America and the and trade structures. The shocks were chan- Caribbean neled through changes in the terms of trade starting in the early 2000s and reflected the 170 extent to which initial trade structures were similar to those of China, at the epicenter of 160 these shocks, and the United States. Only a few countries in the region—chiefly 150 Mexico and, to a lesser extent, countries in Central America—maintained an export 140 structure similar to that of China. The trade Index (2000 = 100) structures of most countries in the region 130 were quite different from that of China. For the economies of South America, where the 120 dominant resources are land and mining endowments, the combination of external 110 supply and demand shocks translated into unequivocal and significant improvements in 100 their terms of trade (figure 12). In contrast, Mexico’s diversified economy—which com- 90 bined an initially broad and relatively strong manufacturing base with substantial produc- 80 00 01 02 03 04 05 06 07 08 09 10 11 12 13 tive capacity in commodities (such as fossil 20 20 20 20 20 20 20 20 20 20 20 20 20 20 fuels, coffee, and iron ore)—experienced South America Mexico Central America stagnant terms of trade.13 In Mexico, the sup- Sources: Calculations based on data from the Economic Commission for Latin America and the ply shock that kept manufacturing prices in Caribbean (CEPAL). check was compensated for by the demand Note: Simple average across countries within each LAC subregion are presented. South America includes Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Paraguay, Peru, Uruguay, and República shock that increased commodity prices. Cen- Bolivariana de Venezuela. Central America and Caribbean includes Costa Rica, the Dominican tral America and the Caribbean experienced Republic, El Salvador, Guatemala, Haiti, Honduras, Nicaragua, and Panama. 16   LATIN AMERIC A AND THE RISING SOUTH to China’s. Approximately 60 percent of the Caribbean, Central America, and Mexico, Mexico’s exports of manufactures were sim- where initial export structures were similar ilar to those of China, compared with only to China’s (panel a). In contrast, the negative 30 percent in the case of Brazil.14 The global impact of the rise of China on manufacturing manufacturing supply shock dampened exports was significantly weaker for South the potential growth of LAC’s manufactur- American economies. The positive impact ing exports in general, with the effect most on their exports of agricultural and mineral acute in countries whose export structures commodities was substantial (panels b and were most similar to China’s at the outset (in c).15 In fact, South American countries repre- 2000). LAC countries that benefited the most sent all the observations in the three panels of from the Asia-led global commodity demand figure 14 that were above the LAC average. shock were countries that were rich in nat- ural resources and had a commodity-ori- Weak participation of Latin America and ented initial export structure that matched the Caribbean in global value chains the structure of commodity (agricultural and mineral) imports of China. The sectoral composition of trade conditioned Empirical attempts to gauge the impact the within-LAC heterogeneity of export and of the rise of the South on LAC exports are import responses to the global supply and consistent with differences in the evolution demand shocks. These shocks boosted LAC’s of the terms of trade and the variance in the share in world commodity exports while degree of similarity between the LAC region’s undercutting the region’s share in global initial trade structures and the trade structure manufacturing exports. Financial flows to of China. Figure 14 illustrates these patterns LAC countries seem to have reinforced these by presenting indexes of the quantitative trends. Specifically, LAC’s cross-border finan- impact of the rise of China on the growth rate cial inflows from the South have been more of manufacturing, mineral, and agricultural biased toward the primary sector than flows exports for a large sample of LAC countries from North countries. For example, during between 2000 and 2011. The heterogeneity of the 2000s, 92 percent of the total cross-border the estimated impacts across countries in the M&A investments from the South in LAC region is pronounced. The negative impact on went to the primary sector, whereas only the exports of manufactures was stronger for 48 percent of the same type of investments FIGURE 13  Export similarity indexes in manufacturing in Brazil and Mexico a. Brazil b. Mexico 0.7 0.7 Manufacturing exports similarity index Manufacturing exports similarity index 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0 0 1999 2003 2006 2009 2011 1999 2003 2006 2009 2011 China World United States European Union Korea, Rep. Japan Russian Federation South Africa Sources: Calculations based on data from World Integrated Trade Solution (WITS) and Comtrade; index proposed by Finger and Kreinin 1979. Note: The higher the index, the greater the similarity between the manufacturing export baskets of two economies. FIGURE 14  Effects of the rise of China on gross exports from Latin America and the Caribbean, by sector, 2001–11 average a. Manufacturing exports Haiti Honduras El Salvador Mexico Dominican Republic Costa Rica Guatemala St. Lucia Panama Nicaragua LAC St. Kitts and Nevis St. Vincent and the Grenadines Dominica Grenada Colombia Brazil Peru Jamaica Belize Ecuador Uruguay Chile Argentina Bolivia Suriname Venezuela, RB Paraguay Guyana Cuba Ϫ20 Ϫ18 Ϫ16 Ϫ14 Ϫ12 Ϫ10 Ϫ8 Ϫ6 Ϫ4 Ϫ2 0 Percentage change b. Agricultural exports c. Mining exports Paraguay Brazil Argentina Chile Guyana Honduras Brazil Peru Uruguay Cuba Bolivia Jamaica LAC Guyana Cuba LAC Nicaragua Bolivia Venezuela, RB Uruguay St. Kitts and Nevis Haiti Dominican Republic Grenada Suriname Dominica Jamaica Belize Mexico Paraguay Chile Colombia Guatemala Argentina Peru Dominican Republic Belize Venezuela, RB Costa Rica Mexico El Salvador Suriname St. Vincent and the Grenadines Guatemala Haiti St. Lucia Panama Ecuador Honduras Nicaragua Dominica Panama Grenada St. Vincent and the Grenadines Ecuador Costa Rica St. Lucia El Salvador Colombia St. Kitts and Nevis 0 2 4 6 8 10 12 14 16 0 5 10 15 20 25 30 35 Percentage change Percentage change Source: Artuç, Lederman, and Rojas 2015, based on data from WITS and Comtrade. Note: Sectoral classification of trade flows is based on the ISIC classification, Revision 3. Agriculture corresponds to ISIC codes 0111–0500, mining to ISIC codes 1010–1429, and manufacturing to ISIC codes 1511–3699. See box 3.1 in chapter 3 of this report for technical details. LAC = Latin America and the Caribbean. 17 18   LATIN AMERIC A AND THE RISING SOUTH from the North in LAC went to the primary evidence to support a nuanced yet positive sector (figure 15). Large, albeit less striking, answer to the first question.17 The second differences are also observed in cross-border question is examined in a subsequent section. greenfield investments and syndicated loans.16 New forms of cross-border trading These trends suggest that the proliferation emerged alongside the rise of the South. One of LAC’s ties with the South was driven to manifestation of this phenomenon was the a larger extent by natural endowment–based proliferation of GVCs. These chains entail comparative advantages than by integration the offshoring and international distribution into manufacturing GVCs. Two key ques- of specialized activities that are part of an tions may be raised in this regard. First, is integrated production process. They typically LAC indeed characterized by weaker inte- involve a group of firms located in different gration into GVCs than other South regions? countries that operate at different stages of Second, are some types of trade structures the same production process in a coordi- (such as structures associated with partici- nated fashion, all under the aegis of a lead pation in GVCs) more conducive to growth firm, with the goal of enhancing the over- than others? The rest of this section provides all efficiency of the chain. The GVC-based FIGURE 15  Sectoral composition of cross-border flows in Latin America and the Caribbean, 2003–2011 average LAC’s exports North South Syndicated Financial flows to LAC loans from the North Mergers and acquisitions Greenfield investment Syndicated Financial flows to LAC loans from the South Mergers and acquisitions Greenfield investment 0 10 20 30 40 50 60 70 80 90 100 Percent of total cross-border flows Primary Light manufacturing Heavy manufacturing Source: Calculations based on data from Comtrade, SDC Platinum, and fDi Markets. Note: The primary sector includes agriculture, hunting, forestry, and fishing; mining; and crude petroleum and natural gas. The light manufacturing sector includes food, beverages, and tobacco; textiles and apparel (including leather); and wood and paper-related products. The heavy manufacturing sector includes refined petroleum and related products, chemicals and plastics, nonmetallic minerals, metals, machinery and equipment, and transport equip- ment. The North includes the G-7 members and Western Europe countries. The South includes all other economies. Figure excludes offshore centers. G-7 = Group of Seven; LAC = Latin America and the Caribbean. O V E R V I E W   19 globalization pattern is thus driven more by Within LAC, Central America and Mex- firms’ global strategies than by traditional ico gained relative importance during the country-based comparative advantages. The early 1990s, probably as a result of the North resulting multicountry production process American Free Trade Agreement (NAFTA). calls for a finer analysis of trade patterns that They peaked around 2000 and then lost goes beyond the traditional focus on broad ground, even as Eastern Europe rose, until sectors and skill categories (see, for instance, about 2009. Since then, Central America and Baldwin 2012). Mexico seem to have experienced a rebound. Measuring the intensity and quality of The contrast with South America is stark: it integration of a country into GVCs is a chal- did not experience a relative surge in terms of lenge. Given the paucity of suitable data, proxies must be used.18 One way to do so is to focus on exports of GVC-relevant inter- FIGURE 16  Exports of intermediate goods as share of mediate goods, as these fragmented pro- total exports in three global value chains duction processes require that parts and components cross borders before finished a. By North and South goods are shipped to final markets. Figure 16 Share of world exports in 3 GVCs (%) 14 documents the rise of exports of intermedi- 12 ate goods that are relevant for GVCs in three 10 industries: apparel and footwear, electronics, and automobiles and motorcycles. 8 The North started visibly losing its dom- 6 inance in the exports of these intermediates 4 (measured as share of total exports of GVCs 2 in the three industries) in the late 1980s, when the South’s activity appears to have 0 1962 1970 1980 1990 2000 2012 taken off (figure 16, panel a). This process From North countries From South countries accelerated in the 1990s; by 2009 the South’s exports of intermediate goods for these GVCs b. By region in the South had surpassed the exports of the North. The 2.5 North’s relative importance in GVC-relevant Share of world exports in 3 GVCs (%) intermediate exports began to decline around 2.0 2000—yet another piece of evidence that a major global restructuring broadly coincided 1.5 with China’s accession to the World Trade Organization. 1.0 Participation in GVC-relevant exports of intermediate goods varied widely across coun- 0.5 tries and regions within the South (figure 16, panel b). The first economies from the South 0 1962 1970 1980 1990 2000 2012 that picked up sizable shares of global trade East Asian Tigers Other East Asia MENA SSA China in intermediates were the East Asian Tigers ECA South Asia South America Central America, the (Hong Kong SAR, China; Korea; Singapore; Caribbean, and Mexico and Taiwan, China), whose surge began in Sources: Calculations based on data from Comtrade; classification of intermediate goods the 1970s. They were followed by other Asian into three major global value chains (apparel and footwear, electronics, and automobiles countries (Indonesia, Malaysia, the Philip- and motorcycles) is from Sturgeon and Memevodic 2010. Note: The North includes the G-7 members and Western Europe countries. The South pines, and Thailand), which picked up sharply includes all other economies. East Asian Tigers include Hong Kong SAR, China; the in relative importance during the 1990s but Republic of Korea; and Singapore. Other East Asia includes Indonesia, Malaysia, the Phil- ippines, and Thailand. All other regions follow the World Bank classification of countries. then lost ground precipitously after 2000, ECA = Europe and Central Asia; G-7 = Group of Seven; GVC = global value chain; MENA = when China rose to a dominant position. Middle East and North Africa; SSA = Sub-Saharan Africa. 20   LATIN AMERIC A AND THE RISING SOUTH exports of GVC–relevant intermediates, and larger countries and China? The conse- it never had as large a share as many other quences were indeed asymmetric across LAC South regions. This evidence suggests that countries and tradable industries, as could be geography (that is, proximity to the United expected. States and distance from East Asian coun- In Argentina, Brazil, and Mexico, the tries) played a key role within LAC as a con- share of manufacturing employment, espe- ditioning factor for the region’s participation cially formal employment, has declined since in GVCs. roughly 2000 (figure 18). The fact that it was Another way of gauging a country’s inte- most apparent in Mexico—one of the coun- gration into GVCs is to focus on GVC-related tries in the region hardest hit by the rise of forward and backward linkages. From this China in global markets of manufactured perspective, even raw commodity exporters products—suggests that the employment can participate in GVCs, albeit in the forward impact of China was particularly intense linkage space, by, for instance, exporting where the trade effects were largest. inputs (such as crude oil) for the manufacture Evidence from the simulation models pre- of intermediate goods with greater degrees of sented in this report indicates that the impact processing or final goods (such as gasoline of China on labor market dynamics in Argen- and other oil derivatives). Figure 17 shows the tina, Brazil, and Mexico (through global mar- differences between regions and subregions kets of manufactured goods, agriculture, and around the world in terms of their backward- mining) was substantial in the short run but, and forward-linkage participation in GVCs. perhaps contrary to expectations, relatively Mexico and Central America relate to weak in the longer run (for technical details, GVCs mainly as manufacturers of final see chapter 3 of this report). Labor market goods, hence predominantly in the backward frictions appear to have significantly increased linkage part of GVCs. Moreover, they have the short-run pain of the adjustment for work- integrated toward the final stages of GVCs ers in the manufacturing industry. However, with North countries, particularly the United these effects were counterbalanced in Argen- States. South American countries, by con- tina and Brazil by the positive employment trast, being net commodity exporters, are effects of rapidly rising agriculture and min- inserted mainly in the forward-linkage seg- ing imports from China. Mexico fared a bit ments of GVCs. worse: the simulation estimates suggest that The East Asian countries show equal par- the negative effects on labor demand in man- ticipation in the forward and backward seg- ufacturing were too large to be compensated ments of GVCs, implying that about half of for by the relatively small positive effects on their GVC-related trade is from imports of Mexico’s labor demand in agriculture and intermediate goods and half from exports of mining. This China-led rise of the South can final goods. This benchmark of 50 percent thus plausibly and at least partially explain may be relevant for growth, as it could be a why wages (adjusted for purchasing power sweet spot for the maximization of certain parity) rose faster in Brazil than in Mexico learning spillovers, as, for instance, produc- since the early 2000s (figure 19). The evidence ers of tradables can learn as much from their on the seemingly small longer-run employ- suppliers of imported goods as from the buy- ment impacts should be interpreted cau- ers of their exports. tiously, however. Evidence from other sources discussed in this report suggests that labor market frictions that inhibit labor migration Differential employment effects within countries may result in significant How did the economic shocks emanating long-term losses in areas that had high levels from the restructuring of global trade affect of manufacturing employment before the rise employment in LAC, especially given the sim- of China (see, for instance, Autor, Dorn, and ilarity in the trade structures of the region’s Hanson 2013; Chiquiar 2014). O V E R V I E W   21 FIGURE 17  Backward and forward participation in global value chains, 2011 a. Across regions b. By country in Latin America and the Caribbean EAP Mexico El Salvador North Panama ECA Aruba SA Barbados Honduras MENA Bahamas, The Dominican SSA Republic Mexico and Uruguay Central America Belize The Caribbean Nicaragua South Costa Rica America 0 10 20 30 40 50 60 70 80 90 100 Guatemala Share of GVC participation (%) Jamaica Argentina Chile Haiti Antigua and Barbuda Cuba Ecuador Paraguay Bolivia Colombia Brazil Trinidad and Tobago Peru Venezuela, RB 0 10 20 30 40 50 60 70 80 90 100 Share of GVC participation (%) Backward GVC linkages Forward GVC linkages Sources: Calculations based on data from Eora-MRIO and WDI. Note: Participation in global value chains (GVCs) is proxied by the share of a country’s export that is part of a multistage trade process. This measure is constructed by adding the foreign value added used in a country’s own exports (backward GVC linkages) to the value added supplied to other countries’ exports (forward GVC linkages) and scaling the total by the country’s total exports of goods and services. Panel a reports cross-country averages. The North includes the G-7 members and Western Europe countries. The South includes all other economies. All other regions follow World Bank classification of countries. EAP = East Asia and Pacific; ECA = Europe and Central Asia; G-7 = Group of Seven; GVC = global value chain; MENA = Middle East and North Africa; SA = South Asia; SSA = Sub-Saharan Africa. into the world economy. This seldom explored Low saving rates in Latin America structural dimension of globalization is based and the Caribbean on the composition of demand—that is, the LAC’s response to the global shocks was also relative importance of domestic versus exter- conditioned by the net integration of countries nal demand relative to the country’s income. 22   LATIN AMERIC A AND THE RISING SOUTH FIGURE 18  Employment shares in the formal and national saving could be related to external informal manufacturing sectors of Argentina, Brazil, competitiveness, balance of payments sustain- and Mexico ability, investment, and growth, among other factors. This section documents key relevant a. Argentina facts regarding the patterns of saving, invest- 1991–99 14.7 ment, and real exchange rates in LAC relative to other middle-income South regions. The Formal 2000–05 8.8 effects of (low) saving on growth are dis- 2006–12 9.9 cussed further below. 1991–99 5.8 Figures 20 and 21, which come from an Informal 2000–05 5.5 econometric model discussed in this report, show the comparative dynamics of saving, 2006–12 4.5 investment, the current account, and the real 0 2 4 6 8 10 12 14 16 exchange rate resulting from global shocks for Share of employment in manufacturing sectors (%) LAC and non-LAC emerging economies.19 As b. Brazil discussed earlier, the supply shock in the first decade of the 2000s seems to have dominated 1990–99 10.2 the demand shock. Hence, the focus is on the Formal 2001–05 9.3 response to an increase in global supply and to a decline in world interest rates (equivalent 2006–11 9.9 to a shock from monetary easing). 1990–99 3.6 Assuming no major institutional or struc- Informal 2001–05 4.7 tural change during the entire period, a positive supply shock (an increase in global 2006–11 4.3 supply) boosts LAC’s investment, appreciates 0 2 4 6 8 10 12 its real exchange rate, and widens its current Share of employment in manufacturing sectors (%) account deficit more and more persistently c. Mexico than in other emerging economies (figure 20). At the same time, such a shock depresses 2000–04 13.4 LAC’s saving rates for a prolonged period (in Formal 2006–12 9.8 contrast with other emerging economies). Consistent with the earlier discussion, a 2000–04 Informal 8.2 favorable global monetary shock that took 2006–12 8.3 place over the same period accentuated the macroeconomic effects of the global supply 0 2 4 6 8 10 12 14 16 Share of employment in manufacturing sectors (%) shock in LAC. In fact, the econometric exer- cise finds that a decline in the U.S. interest Sources: Calculations based on data from Encuesta Permanente de Hogares-Continua rate led to a rise in LAC’s investment rate, an (EPHC) surveys in Argentina, Pesquita Nacional por Amostra de Domicilios (PNAD) surveys appreciation of its exchange rate, and a fall in in Brazil, and Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH) surveys in Mexico. its saving rate (figure 21). These effects were Note: Informal workers are defined as workers without social security benefits. also more durable than in other emerging economies. The patterns of low saving rates and The patterns of net integration of LAC coun- appreciating real exchange rates that pre- tries are undisputedly related to the region’s vailed in many LAC countries over the past historically low savings rates. Indeed, the dif- decade can thus be at least partially explained ference between aggregate domestic demand as region-specific responses to global shocks and income is the external current account, emanating from the rising South. The dif- which is also equal to the difference between ferences in macroeconomic responses to domestic saving and investment. For its part, the global shocks between LAC and other O V E R V I E W   23 emerging South regions seem to have dimin- FIGURE 19  Evolution of wages in Brazil relative to wages ished during the past decade, however, at in Mexico least in part thanks to improvements in 0.86 macroeconomic policy management. In par- 0.84 ticular, evidence from the econometric exer- 0.82 cise suggests that the adoption of inflation- targeting-cum-exchange-rate-flexibility and 0.80 improved fiscal rules in several LAC countries 0.78 Ratio appears to have led to significantly smoother 0.76 responses of output, consumption (hence sav- 0.74 ing), and investment to global shocks. This 0.72 smoothing was counterbalanced, at least 0.70 in inflation-targeting countries, by larger 0.68 responses in the real exchange rate. 0.66 L AC ’s pat ter ns of macroeconom ic 2001 2002 2003 2004 2005 2006 2007 2008 2009 responses to the global shocks, and the Source: National average wages in local currency are from the International Labour Office. They change in such patterns over the past decade, were converted to international purchasing power parity constant 2005 U.S. dollars using the con- are arguably influenced by LAC’s reliance on version factor from World Development Indicators (WDI). FIGURE 20  Responses to a positive global supply shock in Latin America and the Caribbean and other emerging market regions a. Investment b. Real exchange rate 1.5 0.002 1.0 Percentage points Percentage points 0 0.5 –0.002 0 –0.5 –0.006 1 11 21 31 41 1 11 21 31 41 Quarters Quarters c. Saving rate d. Current account 0.20 0.05 0.15 0.10 0 Percentage points Percentage points 0.05 0 –0.05 –0.05 –0.10 –0.10 –0.15 –0.20 –0.15 1 11 21 31 41 1 11 21 31 41 Quarters Quarters LAC Non-LAC emerging market economies Source: Hevia and Servén 2014. Note: Lines represent the accepted median model deviation from the trend from a global demand shock, in terms of the sign restrictions defined in Hevia and Servén (2014). See table 5A.4 in chapter 5 of this report for technical details on the sign restrictions. Non-LAC emerging market economies include Hungary, India, Indonesia, the Republic of Korea, the Philippines, Poland, the Russian Federation, South Africa, Thailand, and Turkey. LAC = Latin America and the Caribbean. 24   LATIN AMERIC A AND THE RISING SOUTH FIGURE 21  Responses to a global monetary easing in Latin America and the Caribbean and other emerging market regions a. Investment b. Real exchange rate 1.5 0.000 –0.001 1.0 Percentage points Percentage points –0.002 0.5 –0.003 –0.004 0 –0.005 –0.5 –0.006 1 11 21 31 41 1 11 21 31 41 Quarters Quarters c. Saving rate d. Current account 0.1 0.05 0 0 Percentage points Percentage points –0.1 –0.05 –0.2 –0.10 –0.3 –0.15 1 11 21 31 41 1 11 21 31 41 Quarters Quarters LAC Non-LAC emerging market economies Source: Hevia and Servén 2014. Note: Solid lines represent accepted model median deviation from the trend from a global demand shock , in terms of the sign restrictions defined in Hevia and Servén (2014). See table 5A.4 in chapter 5 of this report for technical details on the sign restrictions. Non-LAC emerging market economies include Hungary, India, Indonesia, the Republic of Korea, the Philippines, Poland, the Russian Federation, South Africa, Thailand, and Turkey. LAC = Latin America and the Caribbean. domestic demand (associated with low saving of payment vulnerability effect, which can rates and a penchant for current account defi- also hinder growth. 21 Where the ER chan- cits). Some evidence to back this statement nel dominates, one would expect to observe was provided earlier, in connection with fig- a pattern in which countries that save less ure 10, which shows that current account grow less and have appreciated real exchange deficits tend to emerge systematically in LAC, rates. Where the IR channel dominates, one even during the recent times of favorable would also expect to see that countries that terms of trade. save less grow less. Yet, real exchange rates Low saving rates arguably condition mac- would be undervalued in this case, reflecting roeconomic outcomes and responses to exter- low sovereign ratings and vulnerable balance nal shocks through one of two channels. The of payments trajectories. first is a real exchange rate (ER) channel—a The patterns observed in figure 22 are competitiveness-reducing effect caused by consistent with these expectations. The vari- appreciating real exchange rates that can ables of interest in the scatter plots reflect hinder growth. 20 The second is an interest medium-term equilibrium relations that are rate (IR) channel, associated with a balance presented in the form of deviations from the O V E R V I E W   25 benchmark.22 The size of the deviations can FIGURE 22  Domestic saving, real exchange rates, and sovereign be attributed largely to differences in policies risk ratings, 1990–2012 average and policy-driven institutions.23 a. Domestic saving and real exchange rate gaps Panel a of figure 22 shows that an ER 2 pattern is consistent with the entire analyzed sample: on average countries that save more have more competitive real exchange rates, 1 relative to benchmark. However, LAC coun- National saving EAP tries (divided into two groups, higher-income MENA countries [LAC1] and lower-income coun- 0 HI ECA tries [LAC2]) tend to be located in the SSA LAC2 lower-left quadrant, where exchange rates LAC1 are undervalued. In contrast, East Asia and –1 Pacific countries tend to occupy the upper- left quadrant, where oversaving is associated with undervaluation. These patterns sug- –2 –1 –0.5 0 0.5 1 gest that low saving rates have historically Real exchange rate influenced macroeconomic outcomes in LAC mainly through the IR channel—that b. Domestic saving and sovereign risk rating gaps is, through adverse balance of payments 2 vulnerability effects reflected in low coun- try ratings. This finding is consistent with the scatter diagram in panel b of figure 22, 1 which shows that worldwide data also sup- National saving EAP port an IR pattern (countries that save less MENA ECA tend to have lower sovereign risk ratings). 0 SSA LAC1 HI LAC is located closer to the fitted line, LAC2 although it still appears as an undersaving and underrated region. –1 Two key caveats have to be made in this regard. First, there has been consider- able heterogeneity within LAC, as shown –2 –1 –0.5 0 0.5 1 in panel a of figure 23. Between 1990 and Sovereign risk rating 2012, the region started to break free from LAC1 countries per period the spell of low sovereign ratings (figure 24) LAC1 countries 1990–2012 average and hence started to transition from an IR LAC2 countries per period to an ER pattern. Chile, Mexico, Panama, LAC2 countries 1990–2012 average and Peru appear as oversavers with under- Other countries per period valued real exchange rates (all relative to Other groups of countries 1990–2012 average benchmark), whereas the Bahamas, Barba- dos, Brazil, Costa Rica, and Uruguay appear Sources: Calculations based on data from United Nations (UNSTAT), WDI, and Institutional Investor as undersavers with overvalued exchange database. Note: The linear fit was calculated for the period version of the complete country sample for rates. These country cases thus conform 1990–2012. LAC1 countries are countries in Latin America and the Caribbean (LAC) with annual per to the ER pattern. In contrast, Colombia, capita GDP of more than $5,000; LAC2 countries are those with annual per capita GDP of $5,000 or less; see table A.1 for list of countries in all groups). GDP = gross domestic product; EAP = East Asia Ecuador, and Trinidad and Tobago sit in the and Pacific; ECA = Europe and Central Asia, HI = high income; MENA = Middle East and North Africa; lower-left quadrant, with low domestic sav- SSA = Sub-Saharan Africa. See annex 5A in the main report for details on how the benchmarks are calculated. ing and undervalued exchange rates. These patterns suggest that these latter countries have remained more persistently under the grip of the IR channel. Perhaps surprisingly, 26   LATIN AMERIC A AND THE RISING SOUTH FIGURE 23  Saving and real exchange rate gaps for higher-income were generated to effect the transfer of capi- countries in Latin America and the Caribbean tal abroad.24 a. 1990–2012 Second, consistent with the suggestion 0.5 stemming from the dynamic analysis referred ER high savers IR high savers to earlier, the benchmarking exercise identi- CHL PAN VEN fies an accelerated migration of LAC1 coun- ARG PER MEX tries toward the ER pattern during the first 0 decade of the 2000s, as real exchange rates National saving ECU COL BHS appreciated substantially and sovereign risk BRA URY ratings rose steeply. Country ratings actu- TTO ally converged in this period to those of the –0.5 BRB middle-income countries of Southeast Asia IR low savers (figure 24), with several countries in LAC CRI joining the investment-grade asset class. 25 ER low savers This migration reflected improvements in –1 –0.4 –0.2 0 0.2 macrofinancial policy frameworks and, at Real exchange rate least in South America, the powerful forces of the global shocks associated with the rise b. 2011–12 of the South. In fact, as shown in panel b 0.5 PAN* of figure 23, many LAC1 countries moved ARG* significantly closer to the ER pattern that VEN is observed for the entire sample (the fitted CHL MEX PER line) during the 2011–12 period. Particularly 0 TTO* National saving ECU strong real appreciations took place in Brazil, BHS* CRI COL Colombia, Costa Rica, and Uruguay. URY BRA –0.5 Implications for growth: Trade structure, foreign direct investment, BRB* and the composition of aggregate –1 demand –0.4 –0.2 0 0.2 0.4 Real exchange rate Do the LAC-specific trade and aggregate demand structures really matter for growth? Sources: Calculations based on data from UNSTAT and WDI. This section summarizes the main findings Note: The linear fit (shown in both panels) was calculated for the complete country sample for 1990–2012. Higher-income countries in Latin America and the Caribbean are countries with annual of a battery of econometric tests conducted per capita GDP of more than $5,000 (see annex table A.1 for list of countries). See annex 5A in chap- to shed light on this question, with special ter 5 of this report for technical details on the calculation of the benchmarks. Three-letter country groupings correspond to ISO 3166 standard. * = due to missing data for the 2011–12 period, the attention on the relevance for growth of trade latest available period was used. IR=countries affected by the interest rate channel. ER= countries structure, FDI, and domestic saving. The key affected by the real exchange rate channel. GDP = gross domestic product. message is that economic structures matter for growth. A reassessment of the region’s growth- and productivity-oriented reform Argentina and República Bolivariana de agenda from the angle of structure would Venezuela appear as high savers with over- therefore be useful. valued currencies. As these countries have The role of trade structure. The litera- had sovereign ratings well below the average ture supports the notion that trade openness of the LAC1 group, a plausible explanation can raise growth rates, at least temporarily, for their location in the figure is the repeated during the transition to a higher steady-state occurrence of exchange controls and epi- path of GDP per capita. 26 There is much sodes of massive capital flight, during which debate, however, regarding the channels excess saving and current account surpluses through which this transition may operate. O V E R V I E W   27 The traditional answer, dating back to the FIGURE 24  Country ratings for selected country groups neoclassical theories of trade, has been that 80 trade lifts growth (at least transitionally) 75 through the efficiency gains of specializa- 70 Sovereign risk rating index 65 tion based on comparative advantage. This 60 channel hinges on differences in either fac- 55 tor endowments (labor, capital, land, natural 50 resources) or average productivities across 45 27 40 countries. 35 More recently, the focus has been on a dif- 30 ferent (and arguably complementary) mecha- 25 nism, whereby trade boosts growth by serving 20 1980 1984 1988 1992 1996 2000 2004 2008 2012 as a conduit for learning spillovers and tech- nology diffusion (see Keller 2004 for an early LAC1 Argentina and Venezuela, RB Southeast Asian middle-income countries review of the literature). One implication is that when it comes to its impact on growth, Source: Calculations based on data from Institutional Investor database. not all trade is created equal. The question is Note: Middle-income countries in Southeast Asia include Indonesia, the Republic of Korea, Malaysia, less about whether and how much an econ- the Philippines, and Thailand. LAC1 countries are countries in Latin America and the Caribbean (LAC) with annual per capita GDP of more than $5,000 (see annex table A.1 for a list of countries). GDP = gross omy trades but rather how much it learns domestic product. from its international trade. This realization naturally shifts the debate toward questions such as how and with which partners a coun- try trades. Empirically, these questions point can thus surmise that to the extent that trade to measurable dimensions that can be used flows embody technology and knowledge, as proxies for learning-intensive trade. 28 As producers can benefit more from exports and such, this report adds to the growing evi- imports that are part of the same industry dence that suggests that certain features of a or a GVC than they can from exports and nation’s trade structure matter for economic imports that correspond to unrelated or dis- development and growth. Some of these connected activities. The composition of features include the degree of intraindustry trading partners may also play an important trade, participation in GVCs, the composition role in how much countries learn and how of trading partners, and the degree of export quickly they adopt new technologies. concentration. These features shed light on The econometric evidence in this report the extent to which technology diffusion and suggests that trade linkages with the North the learning intensity of trade can positively could indeed yield higher growth payoffs affect growth and other economic outcomes, than trade with the South. The results, based such as macroeconomic volatility (see, for on data for 1960–2010, indicate that a 1 per- instance, Lederman and Maloney 2007; Alva- centage point increase in the degree of trade rez, Buera, and Lucas 2013; and Pinat 2015). openness with North countries is associated This report analyzed the relationship with a 1.6 percent increase in GDP per capita between several characteristics of trade per year over a five-year period, followed by structure and growth, given that there is no potentially longer-lasting effects. In contrast, overarching consensus in the literature as to the estimated effect of trade with the South which ones are most influential. Two partic- is much lower: a 1 percentage point increase ularly interesting characteristics—intrain- in the degree of trade openness with South dustry trade and participation in GVCs—are countries is associated with an increase in likely to be related to international technol- GDP per capita of only about 0.3 percent. ogy and knowledge flows because they tightly The difference in the estimated effects link trade to domestic factor and input mar- when trading with the North versus the South kets, logistics, and production processes. One seems to be associated with differences in the 28   LATIN AMERIC A AND THE RISING SOUTH structure of trade along several dimensions, M&A flows are positively (and significantly) arguably including the extent and manner associated with the recipient country’s labor in which countries participate in GVCs (see productivity within manufacturing indus- chapter 2 of this report and Didier and Pinat tries, North-South, South-North, and South- 2015 for technical details and a deeper analy- South flows are not (for technical details, see sis of the structure of trade linkages and eco- chapter 4 of this report and Didier, Nguyen, nomic growth). Controlling for the overall and Pienknagura 2015). These findings sug- volume of trade flows, increases in participa- gest that LAC and other South economies tion in GVCs, especially the middle segments have yet to benefit in terms of labor produc- of these chains, yield additional gains in GDP tivity increases within manufacturing indus- per capita. An increase in the share of total tries from their flourishing connections with trade that comes from intraindustry trade has the rest of the South or the North. a positive and statistically significant associa- Other evidence, however, suggests that tion with income growth. Trading with coun- LAC has benefited from the presence of mul- tries at the center of the global trade network tinational corporations through different is associated with higher growth, arguably channels, including by accelerating the exit of because these types of connections expose the low-productivity domestic firms and enhanc- country to the frontier of ideas and technol- ing the productivity of domestic firms across ogies. The econometric results also suggest all industries (see, for instance, Lederman that countries benefit more from interna- and others 2014). tional trade connections when they have a The new evidence on FDI presented more educated labor force, which points to in this report suggests that aggregate the importance of human capital formation industry-specific labor productivities in the for the absorption of foreign technology and South so far appear to be unaffected by knowledge. foreign firms’ mergers with or acquisitions Intraindustry trade and insertion into the of domestic firms. Future research could core of GVCs thus appear to be more condu- attempt to ascertain the features in North cive to higher long-term growth rates. Except economies that allow them to benefit from possibly in Mexico, Costa Rica, and Uru- M&A flows within industries, with an eye guay, the rise of the South has not systemat- toward understanding whether these positive ically yielded these types of growth-inducing effects depend on public policies (as imped- changes in trade structures in LAC. iments to or propagators of learning spill- The role of foreign direct investment. The overs), the quality of institutions, the quality increase in financial flows across countries, of human capital, or other factors. The sec- especially FDI, could be driven by com- tion on policy priorities below addresses panies seeking to capitalize on efficiency these issues. improvements made possible through the The role of the composition of aggregate fragmentation of production stages across demand. Do low national saving rates—a countries. Therefore, the rising participation trademark of LAC economies—hamper of the South in global financial flows could growth? Mainstream open-economy growth be a potential driver of economic growth. models typically assume that foreign and Such flows may not only ease financing con- domestic saving are perfect substitutes. straints in recipient economies but also be a Implicit in these models is the notion that conduit for technology diffusion and learn- what really matters for growth are invest- ing spillovers. Indeed, policymakers from the ment (and profit) prospects, but not how South, including LAC, see the attraction of investment is financed. This view is con- FDI and multinational corporations as a pol- sistent with the assumption that factors of icy priority. production (particularly capital) respond The empirical findings presented in this to small differences in relative returns by report indicate that although North-North flowing into their most productive uses, O V E R V I E W   29 both across countries and industries or surprising, given that factor mobility is lower firms within countries. The implication is (and factor misallocation higher) in emerg- that domestic saving, and more broadly the ing than advanced economies. Foreign and composition of aggregate demand, is not a domestic saving are thus less perfect substi- determinant of the equilibrium real exchange tutes, as far as growth is concerned, in these rate. Rather, the latter would be determined emerging economies. The result also suggests only by productivity differentials across trad- that saving rates can in some sense com- able and nontradable industries driven by pensate for market imperfections and pol- supply-side characteristics, such as the capital icy obstacles that get in the way of efficient intensity of production. Consequently, saving resource allocation. As the allocative func- and the real exchange rate would not affect tion of markets improves, saving should be growth, as small increases in returns to cap- less of a constraint on growth. ital would immediately attract capital to the Finally, important asymmetries seem to countries, industries, or firms that temporar- characterize the effects of saving on growth. ily offer higher returns. The real exchange In particular, a higher domestic saving rate rate would adjust back to its equilibrium level has a greater positive impact on growth when accordingly. countries experience current account deficits. This view clashes with certain well- This finding should not be surprising, as it established stylized facts. For example, coun- stands to reason that the benefits of a saving tries that rely on foreign saving grow less effort that help to avoid unviable balance of (see, for instance, Prasad, Rajan, and Subra- payments trajectories outweigh the benefits manian 2007); countries whose productivity of a saving effort that increase an already falls behind are countries that “tax” saving strong current account surplus. (see, for instance, Gourinchas and Jeanne When the data are explored in ways that 2012); and there is considerable misalloca- identify the underlying mechanisms, the rel- tion of factors of production, which shows up evance of both the external competitiveness in large and persistent dispersion of produc- (ER channel) and the balance of payment tivities across firms, sectors, and countries. vulnerability (IR channel) effects of saving is This report provides evidence in support borne out. Figure 25, which uses the entire of the alternative hypothesis that national sample, shows deviations from benchmark saving matters for growth, implying that in the domestic saving–real exchange rate domestic and foreign saving are imperfect space. For all observations in each quadrant substitutes. Econometric evidence suggests (that is, for all the dots plotted in figure 22, that national saving rates have an impact on panel a), figure 25 shows the average of the growth (for technical details, see chapter 5 of corresponding deviations from benchmark this report and De la Torre and Ize 2015). It for other key variables (namely, sovereign rat- shows that, on average, a 10 percentage point ings, growth rates, and investment rates). increase in the saving rate (which would Four key messages emerge from figure bring the average LAC saving rate to the level 25. First, countries with undervalued real in Southeast Asia) would increase GDP per exchange rates grow faster than countries capita by 1–2 percentage points a year for with overvalued currencies. This finding at least three years, followed by potentially is a restatement of the well-known finding long-lasting effects of similar magnitudes of Rodrik (2008). Second, the ER pattern thereafter. The evidence is preliminary and strongly emerges from the world data: coun- thus should be interpreted with caution. tries that oversave typically have under- However, it does strengthen the argument valued real exchange rates and grow faster that saving matters for long-term growth. than other countries, whereas countries that The findings also suggest that the sav- undersave typically have overvalued curren- ing-to-growth link is stronger for middle-in- cies and grow more slowly. Third, the IR come countries. This result should not be pattern also emerges from the data: countries 30   LATIN AMERIC A AND THE RISING SOUTH FIGURE 25  Sovereign risk rating, growth, and investment gaps, 1990–2012 1.0 1.1 0.7 0.2 Oversaving –0.1 –0.5 National saving gap Undervalued Overvalued Undersaving 0.2 0.1 –0.7 –1.1 –0.3 –1.3 Real exchange rate gap Sovereign risk rating Growth Investment Sources: Calculations based on data from UNSTAT and WDI. Note: Each bar in the figure represents the simple average of sovereign risk rating, growth, or investment gaps for the observations located in each quadrant of the scatter plot. The scatter plot is a reproduction of figure 22, panel a. Each point represents a country for a given time period. See chapter 5 for addi- tional details. that undersave and face balance of payments ratings for much of LAC, even as the region viability problems (that is, countries in which boosted growth and reduced systemic vul- sovereign risk ratings are well below bench- nerabilities. However, LAC adapted and mark) also have undervalued real exchange responded to these shocks with its traditional rates. Fourth, saving affects future growth domestic demand–reliant (low saving) mac- through investment: countries that oversave roeconomic structure, which led to strong relative to benchmark typically outperform real appreciations, especially in countries their peers in terms of investment rates, that save less.29 The force of the external tail- especially where the real exchange rate is winds was such that they more than offset undervalued. (and actually concealed) the adverse growth During the past decade or so, LAC was effects of low saving. Now that the tailwinds caught up in the forces of real and mone- of commodity prices no longer blow, one can tary global shocks precisely at a time when hypothesize that, given the vastly improved significant improvements in macrofinancial country ratings, low saving rates in LAC may policy frameworks were materializing. The hinder growth less through balance of pay- confluence of these external and internal fac- ments vulnerability effects and more through tors promoted rapid improvement in country external competitiveness effects. O V E R V I E W   31 Changing world, new priorities The root causes of such labor market fric- tions remain unclear. The policy agenda is The rise of the South has affected at least therefore far from obvious. Regulatory rigid- three major policy areas, all of which have ities, which are often bypassed by voluntary implications for employment and growth. In shifts to informality, are unlikely the only some respects, the global shocks may have source of friction (although they are undoubt- temporarily dimmed the urgency of such old edly important). Other sources could include policy challenges as commodity dependence, skills mismatches (including mismatches aris- labor market frictions, and low saving rates. ing from information asymmetries or limited However, as the pull of the rise of the South skill portability) and transport costs within tapers off and the tailwinds recede, the pol- countries. icy agenda should turn even more forcefully The role of skills mismatches is evidenced toward the issues highlighted below. by the well-known finding that the estimated costs of moving to a new job varies signifi- cantly across industries, which implies that Reducing labor market frictions skills are to a large extent industry or firm Labor market frictions made the process of specific. LAC’s experience over the past adjustment to the global supply and demand decade, as well as the powerful forces of tech- shocks unnecessarily costly, especially for the nical change, calls for a policy agenda aimed net commodity importing countries in LAC. at facilitating and enhancing skills develop- They explain why China was once the scape- ment, skills matching, and the formation of goat of choice for LAC policymakers.30 more flexible human capital, so that workers Especially since 2001, when China accel- can more easily adjust to production innova- erated its pace of growth in global trade, tions and shifting market realities by chang- workers in LAC could have benefited from ing jobs and careers over their working lives the declining prices of manufactures and the at lower personal (and social) costs. This pol- employment opportunities in agriculture, icy agenda naturally puts a premium on suit- mining, and nontraded domestic industries if able reforms to educational systems, labor they had been able to switch jobs easily. How- market rules and contracts, social protection ever, the evidence in this report, as well as the benefits (to make them more portable and public’s tendency to worry about competition compatible with labor mobility), and training from China, suggests that labor market fric- and retraining programs. tions prevent workers from easily transition- The potential role of transport costs (and ing to industries where they could be most hence transport-related policies) in interin- productive. The evidence indicates that work- dustry labor mobility has received little atten- ers behave as if they have “sticky feet,” the tion to date. The costs of moving labor across title of a recent World Bank report on trade industries may reflect the concentration of and jobs (Hollweg and others 2014). As Chi- industries across territories. In Brazil, for nese competition in manufactured goods mar- example, most manufacturing is concentrated kets became tough, manufacturing industries around São Paulo and the southeastern coast, had to adjust, partly by shedding workers whereas agriculture is located in the interior and partly by retooling to regain competitive- of the country. The costs of moving workers ness. Workers stuck in “senescent” (declining) and their families across vast geographical manufacturing industries bore a heavy price, regions may help explain the sluggishness of in the form of unemployment or informality. labor market adjustments within countries. They would have been better off had they In fact, a growing body of academic litera- been able to adapt their skills and more easily ture argues that transport costs may play an move within countries to take advantage of inhibiting role in the integration of domestic better employment opportunities. labor markets. 32   LATIN AMERIC A AND THE RISING SOUTH There is, however, persistent, albeit rel- and are undistorted. From this viewpoint, atively low-level, rural-to-urban migration removing policy distortions that get in the within LAC countries, including Brazil and way of market-driven resource allocation and Mexico. It is thus also plausible that the reducing the costs of doing business will nat- choice of migration by workers across vast urally attract corporations from around the distances is driven not just by transport costs world. Trade structures would then special- but also by workers’ specific circumstances ize and respond endogenously to comparative and preferences, some of which may be unre- advantages and a business-friendly environ- lated to market signals. For instance, being ment. Whether the efficient outcome is a close to family may be an overriding consid- knowledge-intensive type of export growth eration for workers unless they face extreme will depend on factor endowments and rela- circumstances (shocks) or belong to commu- tive returns, but the outcome would move the nities with a historical inclination for migrat- economy to its production possibilities fron- ing to specific destinations. tier. This paradigm emphasizes public policy The objective here is not to prescribe spe- failures that hinder market forces rather than cific policies but rather to argue that policy market failures. It thus puts a premium on makers need to rethink broad priorities. reforms that seek to maximize the operation Infrastructure is one area that may be prime of the Invisible Hand. for reconsideration, not just because of its The alternative view is that by itself, the relationship with competitiveness (through market may not automatically bring knowl- its impact on firms’ cost structures) but edge from abroad and will thus underexploit also because poor infrastructure may make opportunities for boosting technology-driven domestic labor markets less nimble and thus endogenous growth dynamics. From this less able to absorb permanent shocks. perspective, some form of industrial policy will be required to induce market players to internalize the positive externalities associ- Fostering trade, foreign investment, ated with the exploitation of knowledge spill- and knowledge spillovers overs. A 2014 report by the Inter-American For some LAC countries, the rise of the South Development Bank, Rethinking Productive brought some benefits, such as lower bor- Development, provides a set of organizing rowing costs and better terms of trade for net principles to discipline thinking about choos- exporters of agriculture and mining prod- ing industrial policy interventions to target ucts. However, the structure of trade between specific types of market failures. LAC and the South seems to be less growth Looking through the prism of the rising inducing than its trade with the North. Like- South phenomenon, this debate boils down wise, FDI into LAC (in the form of M&A) to a balancing act. On the one hand are the that originates in other South countries does potential benefits of improvements in the not seem to be raising labor productivity market-enabling environment that reduce within industries in the region. Labor pro- trade costs for domestic agents, who in turn ductivity appears to more clearly benefit from are guided by competition and relative price North-North M&A activity. Both sets of signals in enhancing their trade and financial results suggest that some rethinking is called linkages with both the South and North. On for in the area of structural change and the the other hand are the coordinating roles of scope for learning and technology diffusion the state, including through the provision through ties with global partners. of specific tax or subsidy incentives, or tar- There have been two extreme paradigms geted loans and loan guarantees, for firms about policy challenges in this area. One is and workers to move into preselected activ- the laissez-faire view, which posits that learn- ities that have a good chance of becoming ing from foreign knowledge will take place part of GVCs or fostering intraindustry trade as long as domestic markets function well patterns. O V E R V I E W   33 A safe approach is one that strikes a Fourth, both vertical and horizontal sensible balance between the laissez-faire industrial policies need to be put on the table, and industrial policy approaches. First particularly for countries that have advanced and foremost, policy should do no harm: on the laissez-faire front, so that old policy policy-induced distortions that get in the distortions do not get in the way of the poten- way of efficient resource allocation and tial success of new industrial policies. Coun- unnecessarily raise the costs of international tries throughout LAC already have some transactions should be reduced. The report industrial policies in place, such as invest- highlights one such distortion: the region’s ment and trade promotion that targets cer- increasing reliance on temporary trade barri- tain types of firms and industries over others. ers (such as antidumping, countervailing, and An extension of this debate could encompass safeguard import duties), which appear to be policy-based incentives, including tax and overused, especially against China and other expenditure policies, with an eye on helping South economies. Many other actions can markets internalize large positive external- be considered in this regard, including elim- ities associated with research and develop- inating or redesigning government programs ment (R&D) and technology adoption and that unintentionally subsidize informality or adaptation. Given that industrial policies can unduly encourage firms to remain small. have significant downsides, it is important Second, there is plenty of room for pos- that they be designed and implemented in itive policy actions aimed at improving ways that generate information and learning the market-enabling environment—by, for (so that impacts can be assessed and mistakes instance, raising information transparency corrected promptly along the way) and com- and disclosure standards and strengthening plement and crowd in market forces (in order contract rights. In general, horizontal pol- to widen the scope for efficiency gains). icies of this nature can only help, although they may not necessarily remove the most Raising national saving rates binding constraints to the development of growth-friendly globalization patterns. Pol- A reform agenda in LAC focused exclusively icies aimed at improving the functioning of on the sorely needed enabling environment labor markets while maintaining adequate and supply-side reforms may not be suffi- labor protections are worthy of special atten- cient to avoid the downsides of globalization tion in this regard. while fully reaping its upsides. A demand- Third, it is time to get serious about side component focused on raising national assessing deficits in the formation of human saving rates, intended to prevent persistent and physical capital (particularly transport, currency overvaluations and balance of pay- energy, and telecommunications infrastruc- ments vulnerabilities, is also a crucial ele- ture), which may be constraining the abil- ment of the growth-oriented reform agenda. ity of individuals and firms to engage in This demand-side component is particularly cross-border transactions efficiently. On the important for LAC countries that exhibit human capital side, educational systems need chronic low saving rates. It is also key in the upgrading, particularly in ways that allow context of market imperfections that limit them to foster the type of skills modern econ- the scope of factors to quickly and smoothly omies demand. Workers need to be trained move to their more productive uses. and retrained, on and off the job, through- Keeping these considerations on the out their working lives. On the infrastruc- policy radar screen may not be easy, given ture side, closing gaps is essential to reducing that the region’s historical low-saving/low- international trade costs, a key determinant growth syndrome may be shifting in the of the emergence of, and incorporation into, context of the rising South and the region’s GVCs and other types of international com- more resilient macrofinancial policy frame- mercial relations. works. The greatly improved sovereign risk 34   LATIN AMERIC A AND THE RISING SOUTH ratings that now characterize much of LAC Third, careful social safety net reforms can may facilitate external borrowing, which (in strengthen domestic saving. The region made the best of cases) can conceal the adverse progress in the past decade in mainstreaming growth consequences of uncompetitive real and targeting social assistance to the poorer exchange rates or (in the worst of cases) and most vulnerable segments of the popu- rekindle LAC’s traditional tendency to suf- lation, including through highly successful fer from balance of payments sustainability conditional cash transfer programs. Several problems. LAC countries complemented these efforts Although economists often resist treating with improvements in noncontributory social saving as a policy variable, a saving-boosting benefits, especially through minimum pen- reform agenda is within reach, although it sion pillars (so-called social pensions) and will require patience and persistence and is the provision of health services at very low likely to be fraught with tensions. There are or no cost to poor households and informal at least four entry points for a comprehensive workers. Given the social benefits of higher policy approach. saving rates, however, as the region consid- First, raising public sector saving can raise ers second-generation reforms to the health, national saving, because it is unlikely that pensions, and unemployment safety nets, it the private sector will completely offset such should ensure that such reforms should not efforts by reducing its saving. Raising public only improve fairness and financial sustain- saving through fiscal tightening (by raising ability but also promote self-reliance (instead revenues, reducing expenditures, or both) of excessive reliance on the state), especially would not be easy in the current global eco- among the elites and upper social echelons. nomic environment. Fiscal reforms that boost Fourth, in designing short-run mac- public saving, and hence tilt public outlays in roeconomic interventions, policy makers favor of investment, would have to confront should take more explicit account of the the difficult and sensitive question of who growth-boosting saving agenda. Doing so would consume less today. Tensions would militates in favor of shifting toward a tighter thus arise over the distribution of taxes and fiscal, looser monetary macroeconomic policy expenditures across space, households, and mix—something that is politically difficult to firms as well as between current and future achieve, especially in the current environment generations. Deft political leadership would of weak world demand, which puts a pre- be needed to increase frugality and foster mium on spending rather than saving. The asset building (which implies a sacrifice of current international financial environment, some consumption today) in a way that pro- characterized as it is by low interest rates and tects the basic consumption needs of the poor. abundant liquidity, could encourage policy Second, there may be openings for imple- makers to borrow imprudently and hence risk menting saving-enhancing policies in the fiscal and balance of payments sustainability financial sector. Since the late 1990s, finan- problems in the future. To reconcile short-run cial development in LAC has been strongly aggregate demand management with lon- biased in favor of consumer finance when ger-run growth objectives, it is crucial that contrasted with other regions, as De la Torre, LAC maintain robust saving rates. Ize, and Schmukler (2011) show. Reforms of The rise of the South has deeply changed financial regulations could help promote sav- the global economy, and irreversibly so. Poli- ing, investment, and production rather than cies and reform agendas have to adapt to this consumption. Financial inclusion could be momentous change. The challenge is great, expanded on the deposit-taking and payment but it provides LAC’s political leadership with side rather than the lending side. Macropru- an opportunity to shine. It is time for cold- dential regulatory policy aimed at prevent- headed rethinking of policy priorities that can ing credit-fueled consumption booms is also unleash growth potential of an immensely called for. diverse and in many ways rich region. O V E R V I E W   35 Annex A TABLE A.1  Country group composition Region Countries Higher-income countries in Latin America Argentina, the Bahamas, Barbados, Brazil, Chile, Colombia, Costa Rica, Ecua- and the Caribbean (LAC1) dor, Mexico, Panama, Peru, Trinidad, Uruguay, Venezuela, RB. Lower-income countries in Latin America Belize, Bolivia, the Dominican Republic, El Salvador, Guatemala, Guyana, Hon- and the Caribbean (LAC2) duras, Nicaragua, Paraguay East Asia and Pacific (EAP) Bangladesh; Bhutan; Cambodia; China; Fiji; Hong Kong SAR; China; India; Indonesia; the Republic of Korea; Malaysia; Pakistan; Papua New Guinea; the Philippines; Sri Lanka; Thailand; Tonga; Vietnam Europe and Central Asia (ECA) Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Cro- atia, the Czech Republic, Estonia, Georgia, Greece, Hungary, Kazakhstan, the Kyrgyz Republic, Latvia, Lithuania, the former Yugoslav Republic of Macedo- nia, Moldova, Mongolia, Romania, Slovenia, Tajikistan, Turkmenistan, Ukraine High income Australia, Belgium, Canada, Cyprus, Denmark, Finland, France, Germany, Ice- land, Ireland, Israel, Italy, Japan, New Zealand, Norway, Portugal, Spain, Swe- den, Switzerland, the United Kingdom, the United States Middle East and North Africa (MENA) Algeria, the Islamic Republic of Iran, Jordan, Lebanon, Morocco, Syria, Tunisia, Turkey Sub-Saharan Africa (SSA) Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Chad, Côte d’Ivoire, Equatorial Guinea, Ethiopia, Gabon, The Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Madagascar, Malawi, Mali, Mauritania, Mau- ritius, Mozambique, Namibia, Niger, Rwanda, Senegal, South Africa, Sudan, Swaziland, Togo, Uganda, Zambia Note: The dividing line between LAC1 and LAC2 countries is per capita income of $5,000 a year. Notes of the degree of financial connectivity of LAC countries with the North and the South, see 1. In this report, the North includes the Group of Didier, Moretti, and Schmukler (2015) and Seven (G-7) members (Canada, France, Ger- chapter 4 of this report. many, Italy, Japan, the United Kingdom, and 4. The share of active financial connections the United States) plus the following Western within the South in 2011 was even smaller Europe countries: Andorra, Austria, Belgium, for mergers and acquisitions (1.4 percent), Denmark, Finland, Greece, Iceland, Ireland, syndicated loans (2.0 percent), and greenfield Liechtenstein, Luxembourg, Monaco, Nether- investments (3.6 percent). lands, Norway, Portugal, San Marino, Spain, 5. This similarity in export shares captures two Sweden, and Switzerland. The South includes distinct dimensions: the relative importance a all other economies, including all countries in given country has in other countries’ exports Latin America and the Caribbean (LAC). and the relative importance that other coun- 2. As a share of the total number of possible con- tries have in a given country’s exports. nections, the number of LAC trade connec- 6. The importance of a country to the global tions with North countries remained almost trade network rises with its share in other stable, at about 98 percent, between 1990 and countries’ exports and the number of its bilat- 2012, whereas the number of LAC-South con- eral trade connections. nections increased from about 40 percent in 7. The measure of FVA of exports captures only 1990 to 62 percent in 2012. backward linkages (the imports a country uses 3. LAC’s financial connections with other South in producing its exports). It does not capture countries also grew faster than its connections forward linkages (the exports of a country that with North countries, especially during the are used by other countries as inputs to pro- second half of the 2000s. For a deeper analysis duce their exports). The patterns of regional 36   LATIN AMERIC A AND THE RISING SOUTH clustering in forward linkages are qualitatively economies for an extended period of time. In similar to the ones reported here. Chapter 2 this sense, it was a supercycle (see Sinnott, of this report provides a detailed analysis of Nash, and De la Torre 2010). GVCs. 12. Bernanke (2005) argues that a confluence of 8. The algorithm underlying figure 8 is similar to factors led to the emergence of a global saving that of figure 5, in that it takes into account glut, including policy interventions to boost the relative (rather than the absolute) impor- exports in Asia, higher oil prices in the Middle tance of each country in its regional trade net- East, and a dearth of investment opportunities work. The distance between countries reflects and an aging population in advanced indus- the degree of similarity in the structure of trial countries. Mendoza, Quadrini, and Rios- their trade connections (“similarity” is mea- Rull (2007) attribute high saving in emerging sured in terms of the relative importance that market countries to relatively low levels of a country has in other countries’ exports and financial development, which generate greater the relative importance that countries have precautionary saving. Caballero, Farhi, and in a given country’s exports). Countries with Gourinchas (2008) instead emphasize the lack similar trade structures are clustered together of investment opportunities in these countries in figure 8. Unlike figure 5, however, figure 8 and the associated shortage of financial assets depicts the density of connections, hence the as the main source of the global saving glut. systemic importance of countries in their Similarly, the IMF (2005) stresses low invest- respective regional network, in terms of colors. ment rates following the Asian crisis rather The systemic importance of countries increases than an increase in saving rates. as colors shift from green to yellow to red. Dis- 13. In fact, between 2000 and 2011 Mexico was tance between countries is defined by the sum a net exporter of mining products every year, in absolute value of the differences in trade a net exporter of agricultural commodities in shares between countries for a given destina- some years, and a net importer of manufac- tion. The density captures the average distance tured products every year. Its gross exports of per number of connections; the smaller the dis- manufactured goods faced stiff competition tance, the higher the density (see De la Torre, from China, however, as discussed later in the Didier, and Pinat 2014 and Van Eck and Walt- overview. man 2010 for more technical details). 14. Exports are disaggregated at the four-digit 9. The contrast between the two regional net- level of the International Standard Industrial works in 2012 is captured by measures of Classification (ISIC). average node density, defined as the average 15. Brazil, Chile, and Peru were among the coun- across nodes of the number of links over the tries that benefited most from China’s rising total number of possible connections. The imports of mineral commodities. Some Cen- average node density in 2012 was 0.99 for tral American and Caribbean economies also East Asia and just 0.89 for LAC. The disper- seem to have received a boost in their mineral sion of node centrality (the standard deviation exports (such as zinc from Honduras and alu- of the node density) was 0.09 for the East Asia minum and bauxite from Jamaica), confirm- network and 0.31 for the LAC network. ing that natural resource endowments were 10. Various issues of the semiannual report series important determinants of the impact of the produced by the World Bank’s Chief Econo- rise of China. mist Office for LAC (http://go.worldbank.org/ 16. There are, however, significant differences WTVI133GT0) examine the improvement in within countries in LAC. Chapter 4 of this LAC’s macrofinancial policy management, report explores the link between trade and beginning with the April 2008 issue, enti- financial flows. tled “Latin America’s New Immune System: 17. The empirical literature on the extent of inte- How Is It Coping with the Changing External gration of LAC countries into GVCs is sparse, Environment?” but it has been expanding. Useful references 11. In contrast with other commodity cycles expe- are UNCTAD’s 2013 report Global Value rienced by LAC in the post–World War II era, Chains: Investment and Trade for Develop- the rise of the South was associated with the ment and the Inter-American Development simultaneous surge in the international prices Bank’s report Synchronized Factories (Blyde of virtually all commodities exported by LAC 2014). Chapter 2 of this report expands on O V E R V I E W   37 this literature by providing more detailed equilibrium relationships, each period is a evidence on LAC’s participation in GVCs, three-year average. including its integration into GVCs with 23. The benchmark is calculated based on regres- North and South countries. The general mes- sion analysis for the entire sample. It indicates sage of this literature is consistent with the where a country is expected to be, controlling message of this report—namely, that LAC’s for its stage of development (as proxied by participation in GVCs is lower than that of per capita income); structural features that other South regions, even though there has are largely beyond the control of policy (for been an increase in its participation since the example, demographic structure, natural 1990s. There is, however, considerable het- resource endowment, economic size); and the erogeneity across LAC countries. average policies of its peers (see De la Torre 18. FDI data, for instance, do not typically differ- and Ize 2015 for technical details). entiate between affiliates that provide inputs 24. Associated with capital flight episodes were to parent companies and affiliates that pro- multiple exchange rate regimes, which tend to duce the same good or service as its parent. show up in the data as overvaluations, given 19. Non-LAC emerging market economies include that the official exchange rate is typically used Hungary, India, Indonesia, Korea, the Philip- to measure the purchasing power parity index. pines, Poland, the Russian Federation, South 25. Chile, Colombia, Mexico, Peru, Trinidad and Africa, Thailand, and Turkey. The economet- Tobago, and Uruguay were in the elite group ric exercise entailed the estimation of struc- of “investment-grade” countries. tural vector auto-regressive models (SVARs) 26. See Frankel and Romer (1999) and Alcalá and (see Hevia and Servén 2014 and chapter 5 of Ciccone (2004), among many others. Singh this report for technical details). (2010) reviews this literature. 20. Low domestic saving implies an excess of 27. Endowments determine the structure of pro- domestic expenditure over income. For small duction, employment, and trade in neoclas- open economies, which cannot influence sical models of trade and development in the international prices, the excess expenditures tradition of Hecksher-Ohlin. Relative national that flow out of the country are satisfied by average productivities matter in Ricardian higher imports at unchanged international models. prices. The excess of expenditure that falls on 28. Some studies, notably Hausmann, Hwang, the nontradable sector of the economy raises and Rodrik (2007), put the emphasis on what domestic prices, particularly if the economy is a country trades as a means of identifying the near full employment. The rise in the prices productivity embedded in the traded good. of nontradables relative to tradables is a real However, the empirical approach in such exchange rate appreciation. It can become studies suffers from important limitations, as durable to the extent that factors (especially Lederman and Maloney (2012) note. capital) are sticky and reallocate sluggishly to 29. Both high and low savers in the LAC1 group more productive uses across sectors and bor- of countries had substantially undervalued ders, a fact that is borne out by the observed currencies in the 1980s and 1990s, and both large and persistent differentials in factor groups experienced a substantial appreciation productivities across firms, sectors, and coun- during the 2000s. However, the real apprecia- tries (see Hsieh and Klenow 2010; Svyerson tion was much more pronounced (and invest- 2011; and Artuç, Lederman, and Rojas 2015, ment and growth lower) among low savers, among others). which became significantly overvalued rela- 21. Low saving leads to a systematic tendency tive to benchmark by the end of the period. toward current account deficits, which imply In contrast, the high savers were able to retain a buildup of external liabilities over time. Such somewhat undervalued currencies by the end a buildup can make the balance of payments of the period. more vulnerable to shocks and raise the risk 30. As an example, at a summit meeting of the of default, which would be reflected in a bias Asia-Pacific Economic Cooperation in 2002, toward higher risk premiums. President Vicente Fox remarked, “It is not 22. Each point in the scatter plot represents a clear whether or not China is actually com- country for a given time period. As the aim petitive. Perhaps it is, but perhaps its current of this figure is to capture medium-term success is based on the fact that they do not 38   LATIN AMERIC A AND THE RISING SOUTH respect a series of rules that other countries, De la Torre, A., T. Didier, C. Calderón, T. Cord- such as Mexico, do respect” (cited in Leder- ella, and S. Pienknagura. 2011. “Latin America man, Olarreaga, and Perry 2009, 4). and the Caribbean’s Long-Term Growth: Made in China?” World Bank LAC Semi-Annual Report, September, Washington, DC. References De la Torre, A., T. Didier, and M. Pinat. 2014. Alcalá, F., and A. Ciccone. 2004. “Trade and Pro- “Can Latin America Tap the Globalization ductivity.” Quarterly Journal of Economics upside?” World Bank Policy Research Working 119 (2): 612–45. Paper 6837, Washington, DC. Alvarez, F., F. Buera, and R. Lucas. 2013. “Idea De la Torre, A., and A. Ize, 2015. “Saving for Flows, Economic Growth, and Trade.” NBER Growth: New Views on an Old Debate from Working Paper 19667, National Bureau of a Latin American Perspective.” World Bank, Economic Research, Cambridge, MA. background paper for this report, Washington, Anderson, K., and A. Strutt. 2011. “Asia’s Chang- DC. ing Role in World Trade: Prospects for South- De la Torre, A., A. Ize, and S. Schmukler. 2011. South Trade Growth to 2030.” ADB Working Financial Development in Latin America and Paper 264, Asian Development Bank, Manila. the Caribbean: The Road Ahead. Washington, Artuç, E., D. Lederman, and D. Rojas. 2015. DC: World Bank. “The Rise of China and Labor Market Adjust- Didier, T., M. Moretti, and S. Schmukler. 2015. ments in Latin America.” World Bank Policy “The Changing Patterns of Financial Integra- Research Working Paper No. 7155, Washing- tion in Latin America.” World Bank Policy ton, DC. Research Working Paper 7190, Washington, Autor, D. H., D. Dorn, and G. H. Hanson. 2013. DC. “The China Syndrome: Local Labor Market Didier, T., H. Nguyen, and S. Pienknagura. 2015. Effects of Import Competition in the United “FDI and Growth: Does the Source Country States.” American Economic Review 103 (6): Matter?” World Bank, background paper for 2121–68. this report, Washington, DC. Baldwin, R. 2012. “Trade and Industrialisation Didier, D., and M. Pinat. 2015. “The Nature of after Globalisation’s Second Unbundling: How Trade and Growth Linkages.” World Bank, Building and Joining a Supply Chain Are Dif- background paper for this report, Washington, ferent and Why It Matters.” In Globalization DC. in An Age of Crisis: Multilateral Economic Finger, J. M., and M. Kreinin. 1979. “A Measure Cooperation in the Twenty-First Century, ed. of Export Similarity and Its Possible Uses.” R. Feenstra and A. Taylor. Chicago: University Economic Journal 89 (356): 905–12. of Chicago Press. Frankel, J. A., and D. Romer. 1999. “Does Trade Bernanke, B. 2005. “The Global Saving Glut and Cause Growth?” American Economic Review the U.S. Current Account Deficit.” Speech 77, 89 (3): 379–99. Board of Governors of the Federal Reserve Sys- Gourinchas, P.-O., and O. Jeanne. 2012. “Global tem, Washington, DC. Safe Assets.” BIS Working Paper 399, Bank for Blyde, J., ed. 2014. Synchronized Factories: International Settlements, Basel. Latin America and the Caribbean in the Era Hausmann, R., J. Hwang, and D. Rodrik. 2007. of Global Value Chains. Washington, DC: “What You Export Matters.” Journal of Eco- Inter-American Development Bank. nomic Growth 12 (1): 1–25. Caballero, R.J., E. Farhi, and P.-O. Gourinchas. Hevia, C., and L. Servén. 2014. “The Macroeco- 2008. “Financial Crash, Commodity Prices nomic Impact of Global and Domestic Shocks and Global Imbalances.” NBER Working on Latin America.” Background paper for this Paper 14521, National Bureau of Economic report, World Bank, Office of the Chief Econ- Research, Cambridge, MA. omist for Latin America and the Caribbean, Chiquiar, D. 2014. Presentation at the conference Washington, DC. on “The Global Insertion of Asian Economies,” Hollweg, C. H., D. Lederman, D. Rojas, and Center for Latin American Monetary Studies E. Ruppert-Bulmer. 2014. Sticky Feet: How (CEMLA)–Central Bank of Mexico, June 5–6. Labor Market Frictions Shape the Impact of www.cemla.org/actividades/2014/2014-06 International Trade on Jobs and Wages. Wash- -InsertionAsianEconomies.html. ington, DC: World Bank. O V E R V I E W   39 Hsieh, C.-T, and P. J. Klenow. 2010. “Develop- Pinat, M. 2015. “Diffusion of Ideas and Central- ment Accounting.” American Economic Jour- ity in the Trade Network.” World Bank, Wash- nal: Macroeconomics 2 (1): 207–23. ington, DC. IADB (Inter-American Development Bank). 2014. Prasad, E., R. Rajan, and A. Subramanian. 2007. Rethinking Productive Development: Sound “Foreign Capital and Economic Growth.” Policies and Institutions for Economic Trans- Cornell University, Department of Economics, formation. Washington, DC. Ithaca, NY. IMF (International Monetary Fund). 2005. Rodrik, D. 2008. “The Real Exchange Rate and “Global Imbalances: A Savings and Investment Economic Growth.” Brookings Papers on Eco- Perspective.” World Economic Outlook, Sep- nomic Activity 39 (2): 365–439. tember, Washington, DC. Singh, T. 2010. “Does International Trade Cause Keller, W. 2004. “International Technology Dif- Economic Growth? A Survey.” World Econ- fusion.” Journal of Economic Literature 42 omy 33 (11): 1517–64. (3): 752–82. Sinnott, E., J. Nash, and A. de la Torre. 2010. Lane, P. R., and G. M. Milesi-Ferretti. 2007. “The Natural Resources in Latin America and the External Wealth of Nations Mark II: Revised Caribbean: Beyond Booms and Busts? Wash- and Extended Estimates of Foreign Assets And ington, DC: World Bank. Liabilities, 1970–2004.” Journal of Interna- Sturgeon, T., and O. Memevodic. 2010. “Map- tional Economics 73 (2): 223–50. ping Global Value Chains: Intermediate Goods Lederman, D., and W. F. Maloney. 2007. “Trade Trade and Structural Change in the World Structure and Growth.” In Natural Resources: Economy.” UNIDO Development Policy and Neither Curse nor Destiny, ed. D. Lederman Strategic Research Branch Working Paper and W. Maloney. Palo Alto, CA: Stanford Uni- 05/2010, United Nations Industrial Develop- versity Press. ment Organization, Vienna. ———. 2012. Does What You Export Matter? In U NC TA D (United Nations Conference on Search of Empirical Guidance for Industrial Trade and Development). 2013. Global Value Policies. Washington, DC: World Bank. Chains: Investment and Trade for Develop- Lederman, D., J. Messina, S. Pienknagura, and ment. World Investment Report, Geneva. J. Rigolini. 2014. Latin American Entrepre- U.S. National Intelligence Council. 2012. Global neurs: Many Firms, Little Innovation. Wash- Trends 2013: Alternative Worlds. Washing- ington, DC: World Bank. ton, DC. Lederman, D., M. Olarreaga, and G. Perry, eds. Van Eck, N. J., and L. Waltman. 2010. “Software 2009. China’s and India’s Challenge to Latin Survey: VOSviewer, a Computer Program for America: Opportunity or Threat? Washing- Bibliometric Mapping.” Scientometrics 84 (2): ton, DC: World Bank. 523–38. Mendoza, E. G., V. Quadrini, and J.-V. Ríos-Rull. World Bank. 2012. Global Development Hori- 2007. “Financial Integration, Financial Deep- zons. Washington, DC. ness and Global Imbalances.” CEPR Discus- sion Paper 6149, Centre for Economic Policy Research, London. “This report is an impressive piece of work, very clearly written and bristling with new facts and interpretations.” — Barry Eichengreen George C. Pardee and Helen N. Pardee Professor of Economics and Political Science University of California, Berkeley “Between the last decade of the past century and the first decade of this one, the share of the South in the world economy doubled, from 20 to 40 percent of global GDP. This excellent report argues, however, that the gains from such a momentous shift have not been equally shared by all countries in the South. The report highlights, correctly in my view, that long-standing weaknesses of the Latin American and Caribbean region associated with its low savings rates and distorted labor markets are impeding it from fully benefiting from the rise of the South. The implication is clear: this region of the world needs to go substantially beyond improved macroeconomic management if it wants to avoid being the world economy’s laggard in the decades ahead.” — Santiago Levy Vice President for Sectors and Knowledge Inter-American Development Bank “This report tackles interesting and important questions for development policy. It identifies a series of new stylized facts relevant for the connection between trade and growth that should stimulate lots of research. It also helps in moving the debate from ‘does trade cause growth?’ to ‘what type of trade causes more growth?’ The second question is of first order importance for development policy.” — Andrés Rodríguez-Clare Edward G. and Nancy S. Jordan Professor of Economics University of California, Berkeley SKU 32897