Munich Personal RePEc Archive Global Value Chains and Export Sophistication in Latin America Manuel, Flores and Marcel, Vaillant Departamento de Economía FCS-UdelaR January 2011 Online at https://mpra.ub.uni-muenchen.de/47310/ MPRA Paper No. 47310, posted 01 Jun 2013 13:39 UTC Nº 32 Volume 15 January-June 2011 Inter-American Development Bank Integration and Trade Sector Institute for the Integration of Latin America and the Caribbean @journal E-Biannual Publication Nº 32 // Volume 15 // January-June 2011 RNP: 926191 ISSN: 1995-9532 Publication property of the Institute for the Integration of Latin America and the Caribbean, Inter-American Development Bank (IDB-INTAL). All rights reserved. Steering Committee Ricardo Carciofi Antoni Estevadeordal Editorial Committee Gala Gómez Minujín Erika Francescon Juan Martin Christian Volpe Coordination-Editing Susana M. Filippa Julieta S. Tarquini Ideas and opinions expressed in the articles published in this Journal are those of the authors and do not necessarily reflect the official policies and views of IDB-INTAL. Quotations from the published articles are allowed as long as the source is acknowledged. Where quotations are used, a copy of the publication will be greatly appreciated. Institute for the Integration of Latin America and the Caribbean Esmeralda 130, 16th Floor C1035ABD Buenos Aires, Argentina Tel: (54 11) 4323-2350 Fax: (54 11) 4323-2365 e-mail: intal@iadb.org http://www.iadb.org/intal CONTENTS T RADE AND THE INTERNATIONAL ORGANIZATION OF PRODUCTION: PROSPECTS FOR LATIN AMERICA AND THE CARIBBEAN 1 Juan Blyde and Christian Volpe Martincus A RTICLES International Fragmentation of Production: Its Effects on the Labour Market 5 7 Sascha O. Becker and Karolina Ekholm Regional Production Sharing Networks and Hub-Ness in Latin America and East Asia: A Long-Term Perspective 17 Lurong Chen and Philippe De Lombaerde Global Value Chains and Export Sophistication in Latin America 35 Manuel Flores and Marcel Vaillant Global Value Chains in the Services Sector: Business Strategies and Latin American Insertion Andrés López, Andrés Niembro and Daniela Ramos 49 I NTERVIEWS 61 Panel Interviews about Global Value Chains Javier Martínez Álvarez (TENARIS) Daniel Herrero (TOYOTA) Jorge E. Sequeira Picado (PROCOMER) 63 67 71 S TATISTICS 77 B OOKS AND ARTICLES REVIEWS 87 Power and Plenty Trade, War, and the World Economy in the Second Millennium 89 Ricardo Carciofi El procedimiento de opinión consultiva en el Protocolo de Olivos sobre resolución de Controversias 93 Luis Fernando Castillo Argañarás Unclogging the Arteries: The Impact of Transport Costs on LatinAmerican and the Caribbean Trade María M. Supervielle 95 GLOBAL VALUE CHAINS AND EXPORT SOPHISTICATION IN LATIN AMERICA MANUEL FLORES Graduate in Economics, Universidad de la República, Uruguay; and Master candidate in International Economics. Researcher-professor (2nd year) Economics Department, School of Social Sciences, Universidad de la República. MARCEL VAILLANT Doctor in Economics UFSIA University (Antwerpen, Belgium). Professor of International Trade, Department of Economics, School of Social Sciences, Universidad de la República, Uruguay. Researcher (level III) for Sistema Nacional de Investigadores de Uruguay. Nº 32 // Volume 15 // January-June 2011 @ journal Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. The process of globalization has wrought a major change in the international economy, and the scope of feasible trade between countries has broadened. Global value chains (GVCs) combine participation by a large number of suppliers across the world in the production of modern manufactured goods characterized by an acceleration of technical progress. Unlike other developing countries, the Latin American countries have not played a leading role in this dynamic, with consistently low levels of export sophistication. This paper evaluates the level of Latin America’s export sophistication from a dynamic and comparative perspective, using a modern product sophistication indicator. Its results confirm the starting hypothesis, but reveal different situations and trajectories. An order according to levels of export sophistication was established in the countries selected: Mexico, Brazil, Argentina, Colombia, Costa Rica, Uruguay, and Peru. The work takes in the dimension of the dynamic of the export specialization pattern by product type (e.g. capital goods, primary inputs, processed inputs, consumer goods) in the period 2000-2007. In products with permanent specialization, which channel the bulk of exports and reflect the pattern of specialization, Mexico and Brazil are always around the Latin American average for all types of goods. Mexico reaches levels of sophistication similar to the OECD average in capital goods, processed inputs, and consumer goods, but with notably lower levels of sophistication in primary inputs. Brazil is below Mexico except in primary inputs, where it has an average similar to the OECD countries. In the products that describe the dynamics (i.e. that lose and gain specialization), the level of sophistication is always higher than goods with permanent specialization. The order within the countries analyzed does not hold and, in some cases, even has almost the reverse pattern. In all cases, the countries most notable for their levels of sophistication in goods that gain specialization are Peru, Uruguay, Costa Rica, and Colombia. In other words, in the margin, there may be evidence that these countries are participating in a recent process of export modernization, particularly in processed intermediate goods. 35 Manuel Flores and Marcel Vaillant INTRODUCTION Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. T 36 he process of globalization has, in recent decades, wrought a dramatic change in the international economy. One of the sources of this new dynamic is the swift process of technical change in the circulation of information and in the reduction of transport costs in general. There has been a significant increase in the levels of trade, and broadening of the basket of economic activities (goods and services) traded at the level of the international economy. Furthermore, the mobility of factors is also important, although this has an asymmetric pattern concentrated in capital (physical and financial) and in labor with a high human capital endowment. From the specialization point of view, this phenomenon is expressed in the process of fragmentation of production at the planetary level. The various stages making up the production of a given amount of economic activity —whether in the production of goods and/or services— have been distributed across the planet through numerous national jurisdictions in a vast array of organizational forms or modes of governance that are known, among other names, as global value chains (GVCs). This fragmentation of production is the modality that characterizes the internationalization of the production processes in many modern manufactured goods, but it has been spreading to a varied group of economic activities, notably the service sector. Unlike other regions (particularly Southeast Asia), Latin America has not played a leading role in this recent dynamic. The Latin American economies, which display low levels of export quality (commodities with low levels of change, low unit value, and low differentiation), might be able to modify this pattern if dynamically inserted in GVCs. This work aims to contribute to the analysis of some Latin American countries’ insertion in the process in recent years. As there are so few precedents in terms of quantitative approaches to the phenomenon, we have turned to the empirical identification of two stylized facts peculiar to insertion in GVCs: the emergence of new products in the export baskets and the presence among them of intermediate products. In addition, a measure of the products’ sophistication is used to incorporate a notion of quality in the different subbaskets identified. After this introduction, the work is structured in five sections. The Second Section, which defines the phenomenon of GVCs, illustrates the fundamental trends in terms of trade structure and discusses the main literature of international trade that has identified it (Baldwin, 2006a & 2006b). The Third Section links GVCs to the structure of exports and analyzes the dynamic effects of specialization, which suggest that economies’ capacity for growth is bound up with the quality of the products they export (their technology density). The Fourth Section presents and discusses the results, and the last highlights the major results. NEW WAVES OF GLOBALIZATION AND GVCS T he increased intensity of international trade is associated with the broadening of the scope of feasible trade at the international level. Trade specialization first began with a wave of fragmentation, signaling the separation between production and consumption in national markets. This opened up opportunities for trading and its resulting mutual benefits. The first type of international trading was based on a kind of trade determined by the differences between countries, both in the technology they use to produce and in the allocation of production factors. The essential principle is that of comparative advantages (different comparative costs in non-trade conditions), characterized by a pronounced interindustrial trade pattern: countries sell very different types of goods than the ones they buy. In the last two decades of last century, new phenomena emerged in international trade and new grounds for specialization were identified. The change is primarily grounded in technologies characterized by increasing returns to scale, which encourage specialization and trade in other areas and directions, and make it necessary to consider aspects such as noncompetitive market structures, product differentiation, and the existence of trade costs. This last dimension is where another pronounced trend is processed, consisting of a sharp drop in trade costs that stimulated a new geography in the production of manufactured goods on a global scale. Trade costs, strictly speaking, include transport costs, costs associated with adaptation to access to different Global Value Chains and Export Sophistication in Latin America markets (including distribution costs), and the cost of enforcing trade policy. Various different factors have combined to make a significant reduction in trade costs possible, most notably the processes of cargo unitization and increase in the scale of transport, standardization of production processes on a global scale, and trade liberalization. The lower the trade costs, the more relevant small differences become in production costs between different origins as determinants in the countries’ capacity to place global activities. A new trend in international trade incorporating new patterns of specialization has taken hold during the last decade. Although the process has been under way for longer, it has only come to maturity in the last few years, when the wave of technical progress has again played a central role. There are two complementary levels of change fuelling the socalled new reasons for specialization: there has been intense development of information technologies and universalization of computer networks; simultaneously telecommunications costs have fallen drastically. Production that was traditionally not internationally traded can now feasibly be turned into a globallytraded commodity. The impact of this is felt especially keenly in services. Trade in services was traditionally confined to the headings of “travel” and “tourism” (consumers move and purchase services from suppliers from elsewhere in the world) and to services associated with the international transport of goods. In recent years, in association with this new wave of fragmentation in global economic activity, other commercial services have grown, most notably corporate services. In other words, fragmentation not only occurs within the plant producing the goods, but also within the Two bodies of original ideas in economics converge in the new division of labor unfolding at the scale of the international economy: on the one hand, the link between division of labor and productivity as discussed by Adam Smith; on the other, the Ricardian determinants of countries’ productive and commercial specialization, which, in modern parlance, is based on relative productivities in producing goods or performing tasks. In conventional models of international trade based on comparative advantages, movements of goods and factors of production are considered as substitutes (the effects of the movement of factors is analogous to those of the movement of products to which these factors have been applied). Taking into account the new reasons for trade, the current trends include a high degree of complementarity between the movement of factors and trade, particularly with trade in intermediate goods and task trade. However, the enhanced mobility of factors of production —a typical feature of globalization— follows an asymmetric pattern almost exclusively affecting physical capital through foreign direct investment (FDI), financial capital, and highly skilled workers. To summarize, the distinctive feature of the current wave of globalization is the widespread increase in the degree of internationalization of economic activity. Figure 1 shows growth in trade in “other commercial services” and “intermediate goods” over the last three decades. This growth was faster than that of general trade in goods, which grew sixfold during the reference period, with trade in inputs and parts growing tenfold and that of other commercial services (including corporate services), fourteenfold. These components associated with the process of fragmentation of economic activity stood out as the most dynamic sectors in international trade: whereas in 1980 they represented 14% of world trade (goods and services), by 2009 they accounted for a quarter. 1 Another term for the process is “task trade”. A set of tasks needs to be performed for the production of a certain good or service. The new technologies allow these tasks to be relocated internationally and to enjoy the advantages of specialization. Nº 32 // Volume 15 // January-June 2011 @ journal Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. The combination of the above trends underpinned the development of GVCs, which combine participation by a large number of suppliers across the world in the production of modern technologyintensive manufactured goods. The greater buoyancy of trade associated with this new wave of specialization is intraindustrial in nature (close substitute goods in production and/or consumption are traded), both horizontally and —primarily— vertically. High growth in trade in intermediate goods is the most prominent stylized fact in the evolution of late-twentieth-century trade. management structure that produces corporate services. This phenomenon has been called building the “global office”.1 37 Manuel Flores and Marcel Vaillant Figure 1 EVOLUTION OF GOODS AND OTHER COMMERCIAL SERVICES GLOBAL EXPORTS 1980=100 indexes 1800 1600 1400 1200 1000 800 600 400 200 38 Goods Inputs & parts 2008 2006 2004 2002 2000 1998 1996 1994 1992 1990 1988 1986 1984 1982 1980 Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. 0 Other commecial services Source: UNCOMTRADE and WTO. The new international division of labor involves a reduction of the importance of the sectors specializing solely in import substitution (Baldwin, 2006a). It is difficult today to find sectors solely aligned to domestic market conditions without, at the same time, being linked to external conditions (either importing and/or exporting), and the global sectors that emerge are at the same time importing and exporting. The above set of changes, therefore, has a substantial impact on the political economy of trade policy. specialization-based strategies enable companies to maintain their global competitiveness, and any reversions would involve declines in their productivity and their capacity to penetrate global markets. A reversion would, therefore, have an impact on demand and employment levels contrary to the desired one. In a recent work for the World Bank, Cattaneo, Gereffi, & Staritz (2010) have argued that trade has withstood the crisis rather well and GVCs are partially responsible for this result.2 This phenomenon has been clearly expressed in the international crisis and post-crisis periods. The economic crisis, which fundamentally played itself out in the industrialized economies, raised new question marks over the risk of increased protectionism and a possible process of “deglobalization”. In such an event, the structural changes seen at the level of international trade specialization could slow or even reverse. However, as the global structure of economic activity registers such high levels of interdependence, this is extremely difficult to alter. Fragmentation and 2 According to Cattaneo, Gereffi, & Staritz (2010), the redeployment and localization of world production that has emerged over the past 20 years are so far-reaching in their scope that they withstood the tests and temptations of protectionism. World production is now geographically dispersed, organized by networks and chains with companies of global size and reach that structure the process. There seems to be no turning back from the phenomenon. Global Value Chains and Export Sophistication in Latin America EVOLUTION OF TRADE BY COMPARATIVE ADVANTAGE COMPARATIVE ADVANTAGES, EXPORT SOPHISTICATION, AND GVCS We chose three types of countries for the analysis and presentation of the results: (i) selected developed countries (Germany, the Benelux countries, South Korea, United States, and Japan); (ii) Brazil, Russia, India, and China (BRIC), i.e. new emerging countries with large markets; and (iii) selected Latin American countries (Argentina, Brazil, Colombia, Costa Rica, Mexico, Peru, and Uruguay). If the data by country is analyzed, we can see that, the weight of exports with advantage in the more developed countries is lower, at three-quarters of the total, whereas in the developing countries, this percentage is usually above 90%, at least in the smaller countries. Table 1 sets out the percentages for the selected countries and, as can be seen, there are no major changes in the period under consideration. F Figure 2a Figure 2b EVOLUTION OF INTERNATIONAL TRADE WITH AND WITHOUT RCA EVOLUTION OF INTERNATIONAL TRADE WITH AND WITHOUT RCA Number of products Value of imports 12000.0 250,000 10000.0 200,000 8000.0 US$ Millons 300,000 150,000 100,000 6000.0 4000.0 50,000 2000.0 0 0 2000 Not Exported 2001 2002 2003 2004 Exported without RCA 2005 2006 2007 Exported with RCA 2000 2001 2002 Exported without RCA 2003 2004 2005 2006 2007 Exported with RCA Source: Based on trade data from UNCOMTRADE. Nº 32 // Volume 15 // January-June 2011 @ journal Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. or its empirical approach to the phenomenon of GVCs, this work uses the UNCOMTRADE database for annual FOB exports by country of origin for the period 2000-2007. Countries registering no information in one of the eight years have been eliminated, leaving a total sample of 121 countries. In terms of products, we used the 2002 6-digit Harmonized System (HS) classification, and we applied the criterion of maintaining those products for which trade was registered in one or other year of the period analyzed (4,913 products of the existing 5,224). This produces a database with 594,473 observations. Figure 2a and Figure 2b represents the evolution of trade in the 4,913 products across the 121 countries. There are three types of situations for country-product combinations in each year: the good is not exported in that country; it is exported, but without revealed comparative advantage (RCA); or it is exported with RCA. As Figure 2a shows, the number of products in each situation considered remains stable, with approximately 47% of products not exported, 42% of products exported without RCA, and just the remaining 11% of products in the HS exported with RCA each year. While the value of trade has risen steadily over the past few years, as shown in Figure 2b, the proportion of the value of exports with RCA is stable, at around 81%. 39 Manuel Flores and Marcel Vaillant Table 1 EXPORT COMPOSITION CONSIDERING RCA % of total exports for selected countries, 2000 & 2007 Country 2000 2007 Developed Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. 2000 2007 Germany 75.4 77.0 Mexico 81.7 81.8 Benelux countries 76.4 77.9 Brazil 81.3 83.7 South Korea 84.5 84.1 Argentina 85.6 84.9 United States 74.9 74.1 Colombia 91.8 91.1 Japan 86.0 82.5 Costa Rica 94.9 93.9 Uruguay 92.3 92.3 Peru 93.7 93.8 BRIC (exc. Brazil) 40 Country Latin American China 81.0 84.1 India 88.3 84.9 Russia 91.2 91.3 Source: Based on trade data from UNCOMTRADE. EVOLUTION OF EXPORT SOPHISTICATION A methodology has recently been developed that provides an indicator of the levels of product sophistication and overcomes the critiques of previous proposals that registered a high degree of endogeneity.3 The only information used to calculate it is a countryproduct grid each cell of which indicates whether the country (column) has RCA in the product (row). An economy with low levels of sophistication is an economy that specializes in few products (i.e. it lacks the capacity to produce most goods) that many countries, also with low levels of sophistication, specialize in. This is what in common parlance is called a “banana republic”: specialization in a product that many other similar economies specialize exclusively in. A highly sophisticated economy is the opposite: there is specialization in many products that other also highly sophisticated countries specialize in. The method of reflections implements this idea. The Annex Figure presents the resultant ranking from the calculation of the indicator of levels of export sophistication for the 3 The new indicator is known as the level of products sophistication of the “method or reflections” (Hausmann & Hidalgo, 2009). 121 countries considered in each year of the period 2000-2007.4 As can be seen, the countries of the region display relative stability in export sophistication. Among the selected countries, Costa Rica and, to a lesser extent, Colombia have achieved clearer improvement whereas Argentina and, to a lesser extent, Brazil have lost sophistication. This, together with the decline in the Indian and Russian baskets, accounts for a simplification of three of the BRICs’ export baskets. The evolution of China’s exports contrasts with this pattern, being characterized by a marked improvement. In the developed countries, there is the usual stability in their relative position, with the exception of South Korea, which belongs to the group of countries with greatest export sophistication. 2007 was notable for the progress that several of the region’s countries (e.g. Uruguay, Costa Rica, Colombia, Peru, and Argentina) made, despite a fall in all cases over the previous years. If this evolution were analyzed by clustering countries, Argentina’s decline would have moved it 4 For a detailed description of the methodology, see Flores & Vaillant (2011). Global Value Chains and Export Sophistication in Latin America from the best-positioned group in the region (along with Mexico and Brazil) to the top of the group with medium sophistication (along with Colombia, Costa Rica, Uruguay, and Chile). Peru remained stable throughout the period, and, alongside Bolivia and Ecuador, formed part of the group with least export sophistication in the South American region.5 EVALUATING THE CHANGES For the purposes of this work, we wish to investigate the dynamics through which countries have acquired advantages in certain products in the period analyzed. While the indicator of the method of reflections considers exports with RCA, and brings together products exported without RCA and products not exported under a single category, we propose here to differentiate the latter two situations. As shown in Table 2, observations corresponding to country-product cases of no exportation stand at 45.2% in the initial period, and 38.4% maintains this feature in the final period (stable not exported). For exports without RCA, the figure is 36.9% in the initial period, 31.1% of which maintain the characteristic. Finally, the cases of export with RCA were 9.5% in the initial period and 7.2% maintained such a condition. 5 It should be remembered that Venezuela is one of the countries that have been excluded due to a lack of information for 2007 in the UNCOMTRADE database. 6 The dominance criterion in the four years of the subperiod considers dominant any situation arising in a number of years strictly higher than any of the possible alternative situations. Table 2 (Panel A) shows that the largest number of products that change situation do so either from a situation of non-dominance (ND) to being exported without advantage (NA), or from not being exported (NE) to a non-dominant (ND) behavior or to being exported without advantage (NA). Panels B and C measure the value coverage of exports in the initial and final subperiods. The table allows us to see the distribution of different types of traffic identified, the importance of which can be quantified either in the initial or final period. It is interesting to compare the distribution of importance in the initial classification as measured by exports in the initial period against the distribution in the final period as measured by exports in the final period (values highlighted in bold in Panels B and C). Table 2 identifies 16 patterns in the export specialization dynamic. Five categories can be identified in order to simplify the variable and consider the most relevant patterns: those that have an advantage in the initial and final periods (A,A); those that acquire some degree of specialization (NE,A), (ND,A), (NA,A), plus (NE,NA) and (ND,NA); those that lose specialization (A,NE), (A,ND), (A,NA), plus (NA,ND) and (NA,NE); those exported without specialization (NA,NA) plus (ND,ND); (NE,ND), (ND;NE); and last those that are never exported (NE,NE). Table 3 presents the five patterns of the dynamic and products are further distinguished by type of good (capital goods, primary inputs, processed inputs, and consumer goods).8 7 These figures do not coincide exactly with the ones presented above because, in this case, the products have been classified in light of the situation prevailing in each subperiod; in the four years comprising it as a whole, however, there may arise nondominant situations. So, for example, the value of positive trade seen in products classified as non-exported corresponds to exports in a year of products that register no exports in most years of the subperiod. 8 We have used the classification by broad economic categories (BEC). Nº 32 // Volume 15 // January-June 2011 @ journal Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. As a way of incorporating a dynamic view, the analysis is divided into in an initial subperiod (2000-2003) and a final subperiod (2004-2007). Four possible scenarios for product-country observation are identified in each subperiod, paying attention to a dominance criterion in each four-year group: no dominant behavior (ND); dominance of periods of no exportation (NE); dominance of periods of export with no RCA (NA); and dominance of periods of export with RCA (A).6 With these elements, it is possible to simply appreciate the relevance of the different evolutionary patterns of countries’ export structures: both the stable behavior across subperiods and the changes that reflect a country’s situations of gain or loss of specialization in a product. The weight of stable cases (exports in 2004-2007 faced a situation similar to that of 2000-2003) reaches 78% (values highlighted in the diagonal of each panel). These products channeled 88% of the value of world trade in the initial period and 86.5% of it in the final period.7 However, Table 2’s interest does not lie in stable situations, but in its description of the changing situation. 41 Manuel Flores and Marcel Vaillant Table 2 EVOLUTION OF EXPORTS BY SITUATION BETWEEN 2000-2003 & 2004-2007 % of total products and of world export value A) Distribution of total of products situations NE ND NA A Total No Exportation (NE) 38.4 3.3 3.2 0.3 45.2 No Dominance (ND) 2.7 1.4 3.4 0.8 8.4 No Advantage (NA) 1.8 2.6 31.1 1.4 36.9 Ini \ Fin Advantage (A) 0.2 0.7 1.4 7.2 9.5 Total 43.2 8.1 39.0 9.7 100.0 NE ND NA A Total No Exportation (NE) 0.03 0.00 0.07 0.08 0.20 No Dominance (ND) 0.15 0.34 1.84 2.04 4.37 No Advantage (NA) 0.03 1.06 15.42 1.58 18.09 Advantage (A) 0.15 2.17 2.78 72.24 77.35 Total 0.4 3.6 20.1 75.9 100.0 ND NA A Total B) Distribution of products in exports in 2000-2003 Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. Ini \ Fin 42 C) Distribution of products in exports in 2004-2007 Ini \ Fin NE No Exportation (NE) 0.02 0.05 0.17 0.49 0.7 No Dominance (ND) 0.01 0.31 1.37 3.41 5.1 No Advantage (NA) 0.01 1.82 15.00 3.41 20.2 Advantage (A) 0.02 1.38 1.32 71.21 73.9 Total 0.06 3.56 17.86 78.52 100.0 Source: Based on data from UNCOMTRADE. 70% of world trade is in products where the countries’ advantages remain stable in the period. The weight of this situation for capital goods is lower (62%), which is perhaps an indicator that the pace of technical progress is faster in these types of goods. It is also apparent that almost 10% of trade in 2007 was in products where the countries recently acquired specialization (these products accounted for less than 5% of world exports in 2000). Among the types of goods comprising this set of dynamic products, primary inputs stand out, representing 18% of exports of products where there is a gain in specialization when they are worth just 8.7% of the value of total exports. Figure 3 evaluates the level of sophistication by type of good and dynamic.9 The highest degree of 9 The level of sophistication was measured at iteration 10 from indicator Kp (Hausmann & HIdalgo, 2009), and the order was then taken to percentiles. The averages of the percentiles by type of good and pattern of dynamic are presented. This simple average of the levels of sophistication can be seen to have an extremely high correlation (99.1%) with the country index used in Section Comparative Advantages, Export Sophistication, and GVCs. Using the sophistication of each product has the advantage of allowing us to work with subbaskets of the group of products exported by each country. Global Value Chains and Export Sophistication in Latin America Table 3 DYNAMICS OF EXPORTS AND TYPE OF GOOD As % A) Proportion of number of products exported in 2007 (NE,NE) No Spec. Specialization Lose Gain (V,V) Total Capital goods 0,6 9,7 1,0 2,2 1,5 14,9 Primary inputs 0,3 3,0 0,5 1,0 1,0 5,8 Processed inputs 2,3 30,1 3,5 7,6 6,4 49,9 Consumer goods 1,1 17,7 2,2 4,3 4,1 29,3 Total 4,4 60,5 7,1 15,1 12,9 100,0 (V,V) Total B) Proportion of value of exports in 2007 (NE,NE) No Spec. Specialization Lose Gain 0,0 3,9 1,1 2,6 12,5 20,1 0,0 1,1 0,3 1,7 5,6 8,7 Processed inputs 0,0 7,6 2,3 3,6 34,8 48,2 Consumer goods 0,0 3,6 1,0 1,8 16,7 23,0 Total 0,1 16,2 4,7 9,5 69,6 100,0 Source: Based on data from UNCOMTRADE. sophistication is seen in capital goods, followed by processed inputs, then consumer goods, and last, primary goods. Product-country combinations that acquire advantage at the global level do so by increasing levels of sophistication in relation to the group of products permanently exported with an advantage. Figure 4 shows the same structure, but in four regions: OECD (excluding Mexico and Chile); China; BRIC (excluding China), and Latin America (excluding Brazil). It is worth noting that the order by type of good holds good across all country clusters (capital, processed goods, consumption, and intermediate goods). In products with advantage, the OECD countries and China are undifferentiated in levels of sophistication. Whereas the OECD is somewhat better-off in terms of capital goods, China has higher levels of sophistication in primary products and consumption. They are extremely alike in terms of processed products. The other three BRIC countries always register lower levels of sophistication than China, and Latin America (excluding Brazil) always shows lower figures than the other regions. In terms of the dynamic, products where specialization is acquired have a similar level of sophistication across country clusters in all types of goods. This implies that the difference of sophistication among new and permanent products will have precisely the opposite order to the level of sophistication by region and type of good. Latin America is the region with the lowest levels of sophistication, but where new products with advantage have a major difference with the traditional pattern. The situation in Latin America is heterogeneous, as noted in previous sections. In products with permanent specialization, which channel the bulk of exports and reflect the pattern of specialization, Mexico and Brazil are always above the Latin American average in all types of goods (Figure 5). Mexico reaches levels of sophistication similar to the OECD average in capital Nº 32 // Volume 15 // January-June 2011 @ journal Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. Capital goods Primary inputs 43 Manuel Flores and Marcel Vaillant Figure 3 SOPHISTICATION (2007) BY TYPE OF GOOD AND DYNAMICS (2000-2007) Total global exports 70 60 50 40 30 20 Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. 10 44 0 Capital Goods Primary Inputs Lose Specialization Processed Inputs Gain Specialization Consumption Goods Permanent Advantage Note: Includes all 121 countries and 4913 products. Source: Based on data from UNCOMTRADE. goods, processed inputs, and consumer goods, and has notably lower levels of sophistication in primary products (Figure 4 and Figure 5). Brazil is below Mexico except in primary goods, where it registers an average similar to the OECD countries. Keeping the spotlight on products with permanent specialization, the order registered in intermediate processed goods is the same as the global order (i.e. the average of sophistication of the basket of products). In this type of inputs, third place in terms of sophistication is held by Argentina, followed by Colombia, Costa Rica, Uruguay, and Peru. This order has variants in the other types of goods, however. In capital goods, third place is held by Costa Rica, and Uruguay is in last place. In primary goods, Uruguay comes second, very close to Brazil and ahead of Mexico, while Peru comes last. Finally, in consumer goods, third place is held by Colombia, while Argentina and Peru come second to last and last respectively. In products that describe the dynamic (i.e. that lose and gain specialization), the levels of sophistication are always higher than goods with permanent specialization. The order among the countries analyzed does not hold good and is even almost the reverse in some cases (see Figure 5 for capital goods in cases that gain specialization). In every case, the countries that stand out for their levels of sophistication in goods that gain specialization are Peru, Uruguay, Costa Rica, and Colombia. In other words, there is evidence in the margin that these countries are participating in a recent modernization process in their export baskets, particularly, in intermediate processed goods. Global Value Chains and Export Sophistication in Latin America Figure 4 SOPHISTICATION (2007) BY TYPE OF GOOD AND DYNAMICS (2000-2007) Groups of countries b) China a) OECD 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 Capital Goods Primary Inputs Lose Specialization Processed Inputs Gain Specialization Capital Goods Consumption Goods Primary Inputs Lose Specialization Permanent Advantage c) Brazil, Russia & India Processed Inputs Gain Specialization Consumption Goods Permanent Advantage d) Latin America 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Capital Goods Lose Specialization Primary Inputs Processed Inputs Gain Specialization Consumption Goods Permanent Advantage Capital Goods Primary Inputs Lose Specialization Processed Inputs Gain Specialization Consumption Goods Permanent Advantage Source: Based on data from UNCOMTRADE. Nº 32 // Volume 15 // January-June 2011 @ journal Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. 0 45 Manuel Flores and Marcel Vaillant Figure 5 SOPHISTICATION (2007) BY TYPE OF GOOD AND DYNAMICS (2000-2007) Selected countries in Latin America 80 70 60 50 40 30 20 Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. Capital Goods 46 Primary Inputs Lose Specialization Processed Inputs Gain Specialization PER CRI URY COL BRA ARG PER MEX CRI URY COL BRA ARG PER MEX CRI URY COL BRA ARG PER MEX CRI URY COL BRA ARG 0 MEX 10 Consumption Goods Permanent Advantage Source: Based on data from UNCOMTRADE. CONCLUSIONS 8 0% of the value of world exports is accounted for by exporting countries with a comparative advantage. Developing countries tend to have an even higher proportion of exports, which allows us to speculate that they have greater stability in their patterns of specialization. The pattern of specialization is evaluated in terms of the performance displayed by the countries in their levels of export sophistication (estimated by the method of reflections). The Latin American countries do not do well in this indicator, which shows a low level of exports sophistication. A group of countries was selected that show relatively stable evolution in these indicators over the period considered, allowing us to differentiate cases such as Brazil, which is close to the lower OECD levels, and Peru, which has the lowest levels of sophistication in the region. Though minor, the changes show that Costa Rica in first place and Colombia in second show an improvement, whereas Argentina, and to a lesser extent Brazil, lose export sophistication. The mode in which the region’s economies are inserted in GVCs can be a determinant both in their recent developments and their prospects of modifying their role as exporters of goods with low levels of change, low unit value, and low differentiation (commodities). The development of GVCs is associated with relocation of production and is reflected in participating countries’ beginning to export new products, which therefore requires the incorporation of a vision of the dynamic of the pattern of specialization. According to the methodology set forth in this work, the weight of new products accounted for 10% of world trade in 2007. Given that a strict criterion is being applied that takes into consideration new sectors, such as those emerging between 2004 and 2007, the importance of the phenomenon in the reference period is plain to see. To round off the analysis, a third perspective is incorporated by product type (capital goods, primary inputs, processed inputs, and consumer goods), which captures a second fact: the fragmentation of production. It is interesting here to analyze processed inputs and, possibly, primary inputs among the new Global Value Chains and Export Sophistication in Latin America products being traded. Trade in primary inputs in the countries selected accounts for a significant percentage of the group of products in which the countries have recently gained specialization. The final analysis was done by crossing the dynamic of the pattern of specialization (lose, gain, and maintain specialization with advantage) and the type of product, and the level of sophistication of each of these twelve subbaskets was evaluated. In the case of products with advantage, China reaches levels of sophistication similar to the OECD average across all categories, followed by BRIC, and Latin America (whose performance is similar to the rest of the world). The order of levels of sophistication by product type is similar across all regions: capital goods, processed inputs, consumer goods, and, last, primary inputs. However, the products in which specialization is acquired have a similar level of sophistication across clusters of countries in all types of goods, and the same is seen across the Latin American countries. This implies that the difference in sophistication between new and permanent products will have exactly the opposite order to levels of sophistication by region and type of good. Latin America is the region with the lowest sophistication, but where new products with advantage are more differentiated from the traditional pattern. The countries that stand out most for their levels of sophistication in goods that gain specialization are Peru, Uruguay, Costa Rica, and Colombia. In other words, there is evidence in the margin that these countries are participating in a recent modernization of their export baskets, particularly where intermediate processed goods are concerned.u REFERENCES BALDWIN, R. 2006a. “Globalization: The Great Unbundling(s). Paper prepared for the Secretariat of the Economic Council under the EU Presidency of Finland. -----. 2006b. “Offshoring and Globalisation: What is New about the New Paradigm?” Mimeo. Geneva: Graduate Institute. CATTANEO, O., G. GEREFFI, & C. STARITZ. 2010. Global Value Chains in a Postcrisis World. Washington, DC: The World Bank. FLORES, M. & M. VAILLANT. 2011. “Cadenas globales de valor y sofisticación de la canasta de exportación en América Latina”. Documento de Trabajo del Departamento de Economía - UdelaR, 7. Montevideo. HAUSMANN, R., J. HWANG & D. RODRIK. 2007. “What you exports Matters”, Journal of Economic Growth, 12(1), pp, 1-25. HAUSMANN, R. & C. HIDALGO. 2009. “The Building Blocks of Economic Complexity”, PNAS, 106(26), pp 10575-10575. HIDALGO, C. A., B. KLINGER, A. L. BARABÁSI & R. HAUSMANN. 2007. “The Product Space Conditions the Development of Nations”, Science, 317(5837), pp, 482-487. HIRSCHMAN, A. 1958. The Strategy of Economic Development. New Haven, Conn.: Yale Press. KRUGMAN, P. 1988. “La nueva teoría del comercio internacional y los países menos desarrollados”, El Trimestre Económico, LV(217). January-March. Nº 32 // Volume 15 // January-June 2011 @ journal Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. The results show that product-country combinations acquire advantage by increasing levels of sophistication in relation to the group of products permanently exported with advantage. This can be seen in both world trade and the different regions analyzed. Significant heterogeneity among the Latin American cases selected is seen when we look at the sophistication of the subbasket of products exported with permanent advantage. This order confirms the above order at the global level. 47 Manuel Flores and Marcel Vaillant Annex Figure EXPORT SOPHISTICATION - REFLEX METHOD (KC,18) 2000 – 2007 Institute for the Integration of Latin America and the Caribbean (IDB-INTAL). All rights reserved. 2000 48 JPN CHE DEU SWE FIN GBR USA AUT MSR SGP TWN FRA KOR BLX CZE IRL RUS NLD CAN NOR SVN ITA ARM ISR DNK UKR HKG ESP HUN SVK BHS MEX MYS BRA CHN SAU MLT POL BLR CPV MYT GEO IND JOR NZL ZAF ARG EST HRV LVA AUS LTU ROM BRB PYF THA KAZ PRT BGR ISL TUR AZE CHL COL GRC URY IDN MKD CYP KGZ PHL BHR MDA DZA KNA CRI EGY SLV COM SWZ TTO BWA OMN TUN GTM NCL ALB QAT KEN MUS SEN BOL NAM PER GRL JAM MAR ZMB PRY PAN CIV MDG GMB NIC HND MLI MNG NER DMA ECU MOZ MWI SYC UGA TZA BLZ GUY STP BDI MDV ETH 2001 JPN CHE DEU SWE FIN GBR USA TWN AUT KOR SGP BLX CZE FRA MSR IRL SVN NLD RUS ITA ISR CAN SVK HKG MYS NOR UKR MEX DNK HUN ESP CHN POL ARM BRA BLR BHS MLT EST GEO MYT IND HRV SAU LVA BRB ZAF ARG JOR NZL PRT ROM BGR THA LTU TUR AUS KAZ KNA GRC AZE COM URY COL IDN ISL PHL MKD PYF CYP CHL TUN CRI MDA SLV TTO KGZ GTM MUS SWZ DZA BHR EGY GRL ALB OMN SEN PER NCL PAN KEN MAR GMB ZMB BWA NAM BOL JAM NIC ECU HND DMA MNG PRY CPV CIV QAT MDG SYC UGA MWI GUY MLI MOZ TZA BDI NER MDV ETH BLZ STP Source: Based on trade data from UNCOMTRADE. 2002 JPN CHE DEU SWE FIN GBR USA AUT TWN KOR MSR CZE SGP FRA BLX IRL RUS SVN NLD ITA NOR HKG ISR SVK CAN DNK HUN MYS ESP MEX MLT BLR POL UKR BHS MYT CHN ARM BRA EST LVA HRV NZL IND ZAF GEO PRT THA ARG BRB JOR LTU ROM SAU AUS BGR TUR GRC CYP ISL PHL COL KNA MKD PYF GRL IDN URY CHL SLV AZE TUN BHR MDA CRI KAZ OMN KGZ MUS NAM BWA KEN TTO DZA GTM SEN EGY NCL ALB MAR QAT SWZ GMB NIC PER ZMB PAN BOL HND STP SYC DMA JAM MDG PRY CIV MNG MLI ECU MOZ UGA COM CPV GUY NER MWI TZA MDV BDI ETH BLZ 2003 JPN CHE DEU FIN SWE KOR USA TWN GBR AUT SGP MSR CZE BLX FRA SVN IRL RUS ITA NLD ISR HKG SVK NOR CAN HUN MEX MYS MLT DNK UKR ESP CHN POL BLR ARM BHS BRA EST IND LVA MYT HRV KNA THA ZAF NZL PRT ROM SAU ARG BGR LTU AUS TUR GEO BRB GRC CYP ISL COL IDN URY MKD CHL JOR KAZ PHL BHR CRI SLV AZE MDA MUS TUN GTM OMN ALB NAM PYF KGZ EGY KEN DZA GRL PER MAR SEN TTO GMB NCL SWZ COM CIV QAT ZMB BDI BOL BWA HND JAM STP MOZ UGA ECU NIC PAN MDG MNG PRY DMA SYC BLZ CPV GUY TZA MWI MDV MLI NER ETH 2004 JPN CHE DEU KOR FIN SWE TWN USA SGP AUT GBR CZE BLX SVN FRA HKG ITA NLD IRL RUS MLT SVK ISR NOR HUN MYS MSR CAN MEX CHN DNK ESP MYT POL UKR BLR BRA BHS ARM EST IND THA ZAF HRV LVA SAU LTU PRT ROM ARG BGR BRB NZL TUR CYP GRC KNA GEO AUS IDN COL JOR PHL CRI ISL KAZ MUS SLV MDA MKD TUN CHL GTM KGZ URY PYF BHR COM EGY OMN TTO AZE KEN MDV QAT ALB GRL MAR SEN PER SWZ DZA NAM NCL HND JAM BOL BWA MOZ CIV SYC PRY ZMB MDG PAN MLI STP TZA UGA ECU MNG DMA NIC BDI BLZ GMB CPV MWI ETH NER GUY 2005 JPN CHE DEU TWN KOR FIN SWE AUT GBR USA CZE SGP SVN BLX FRA HKG ITA MLT NLD HUN IRL MSR ISR MEX MYS RUS SVK CHN DNK NOR CAN ESP POL BLR EST UKR BRA KNA LVA MYT HRV IND THA ARM LTU ROM BGR NZL PRT TUR ZAF SAU CPV BRB ARG GRC IDN AUS COL CYP KAZ JOR SLV TUN BHR MUS PHL CRI MDV MKD GEO MDA GRL ISL URY KGZ NCL QAT CHL TTO GTM ALB AZE SWZ MAR KEN SEN EGY HND PER NAM OMN PYF MDG DZA JAM BHS PRY BOL DMA GMB ECU MNG STP TZA BWA CIV SYC UGA COM ZMB BDI GUY PAN NIC MWI NER BLZ MOZ ETH MLI 2006 JPN CHE DEU TWN KOR FIN SWE AUT CZE GBR SGP USA SVN BLX FRA HKG HUN ITA NLD SVK ISR MYS CHN IRL MLT RUS DNK MEX MSR POL BLR ESP NOR CAN MYT UKR EST BHS BRA HRV ROM THA LVA IND LTU TUR BGR BRB PRT KNA NZL ZAF SAU ARM CPV GEO GRC ARG QAT CYP IDN COL MDA SLV TUN AUS JOR PHL GTM CRI KGZ MUS BHR KAZ MKD ISL URY AZE CHL COM MAR PYF TTO ALB KEN BDI GRL NCL HND OMN DZA SWZ EGY PER SEN NAM MDG ECU JAM PRY BWA GMB DMA MNG BOL TZA NIC ZMB CIV UGA MWI SYC PAN STP NER GUY MOZ BLZ MLI ETH MDV 2007 JPN CHE DEU KOR SWE FIN AUT SGP GBR USA SVN TWN BLX FRA ITA SVK HKG HUN NLD ISR DNK IRL NOR CHN MLT RUS POL CAN ESP EST ROM BLR HRV LVA MEX THA LTU MYT ARM UKR BRA MYS SAU TUR BGR MSR PRT CYP ZAF NZL KNA GRC ARG BHS COL ISL JOR MDA SLV AUS CRI IND MKD URY GRL TUN CPV KAZ PHL GTM BRB GEO CHL MUS IDN BHR KGZ TTO MAR ALB HND OMN AZE CZE DZA KEN PRY SEN PER QAT STP SWZ BDI NAM NCL MDG PYF ZMB BOL MNG NIC EGY JAM PAN TZA ECU UGA DMA GMB CIV GUY BWA BLZ MOZ NER SYC ETH MWI MLI MDV COM