Mapping Vertical Trade

Ferrarini, Benno Working Paper Mapping Vertical Trade ADB Economics Working Paper Series, No. 263 Provided in Cooperation with: Asian Development Bank (ADB), Manila Suggested Citation: Ferrarini, Benno (2011) : Mapping Vertical Trade, ADB Economics Working Paper Series, No. 263, Asian Development Bank (ADB), Manila, https://hdl.handle.net/11540/2032 This Version is available at: https://hdl.handle.net/10419/109390 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/3.0/igo ADB Economics Working Paper Series Mapping Vertical Trade Benno Ferrarini No. 263 | June 2011 ADB Economics Working Paper Series No. 263 Mapping Vertical Trade Benno Ferrarini June 2011 Benno Ferrarini is Economist at the Economics and Research Department of the Asian Development Bank. The author is grateful to Cindy Castillejos-Petalcorin and Elenita Pura for their assistance with the preparation of the paper for publication. The author accepts responsibility for any errors in the paper. Asian Development Bank 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org/economics ©2011 by Asian Development Bank June 2011 ISSN 1655-5252 Publication Stock No. WPS113785 The views expressed in this paper are those of the author(s) and do not necessarily reflect the views or policies of the Asian Development Bank. The ADB Economics Working Paper Series is a forum for stimulating discussion and eliciting feedback on ongoing and recently completed research and policy studies undertaken by the Asian Development Bank (ADB) staff, consultants, or resource persons. The series deals with key economic and development problems, particularly those facing the Asia and Pacific region; as well as conceptual, analytical, or methodological issues relating to project/program economic analysis, and statistical data and measurement. The series aims to enhance the knowledge on Asia’s development and policy challenges; strengthen analytical rigor and quality of ADB’s country partnership strategies, and its subregional and country operations; and improve the quality and availability of statistical data and development indicators for monitoring development effectiveness. The ADB Economics Working Paper Series is a quick-disseminating, informal publication whose titles could subsequently be revised for publication as articles in professional journals or chapters in books. The series is maintained by the Economics and Research Department. Contents Abstract v I. Introduction 1 II. The Data 3 III. The Network Trade Index 5 IV. Mapping Vertical Trade Networks 7 A. The Global Network of Vertical Trade 8 B. The Global Network of the Electric/Electronics Sector 15 C. The Global Network of the Automotive Sector 18 V. Value-Chain Dependence on Japan 22 VI. Conclusions 26 Appendix 27 References 36 Abstract The paper develops a method to map global networks of production sharing and processing trade. Relying on highly detailed bilateral trade data across a matrix of 75 countries, a network index gauges countries’ interdependence according to the extent of trade in parts and components for further processing and assembly of final export goods. The set of bilateral network relations is then subjected to an algorithm that lays it out for visualization as a world map of vertical trade networks. Maps are drawn in relation to processing trade across all industries, as well as for the electric/electronics and automotive industries, where such trade is most prominent. The analysis identifies three major hubs in the global networks: the People’s Republic of China in connection with Japan, Germany, and the United States. Apart from Mexico (mainly because of its maquiladoras network ties to the United States) the analysis highlights that outside Asia, developing countries are not yet involved in global production networks to any significant degree. Just before turning to the network maps, Tables 4 and 5 further clarify the structure of the NTI with respect to its constituent elements.8 Appendix Table 4 ranks the top network partners within the electric/electronics industries, and Appendix Table 5 does the same for the global automotive sector. Additional detail in relation to each country pair is provided with respect to the partner’s share in total parts imports (the main element of the NTI) and the industry’s share in total exports by the assembling country (the weight component of the index). Appendix Table 4, for example, shows that Mexico’s extreme network dependence on the US is determined by a 41.7% share of parts imports from that country, out of Mexico’s total imports of parts in the electric/electronics industries. For Mexico’s economy as a whole, its dependence on the US gains further weight by the fact that the electric/ electronics sector in 2006–2007 represented 15.7% of the country’s total exports.9 In the opposite direction, Mexico’s share in the electronics parts imports of the US is a still sizeable 21.5%, but the industry constitutes only 4.1% of the merchandise exports basket of the US. IV. Mapping Vertical Trade Networks The NTI measures the intensity of vertical trade between country pairs. For a full identification of processing trade networks, the set bilateral connections needs to be set in relation to each other. To do so, we apply a suitable algorithm to sort through the data.10 A spring-embedded algorithm works by the assumption that the nodes (countries) are connected by springs (NTI) that attract or repulse each other. Just like a physical system in which Hooke’s law of elasticity applies, the nodes are assumed to exert a force on each other through the connecting springs, akin to magnetic repulsion or gravitational attraction. Equilibrium is reached by iteration, whereby the rings connecting the nodes are projected on to a plane while being subjected to an acceleration proportional to the various forces exerted on the edges as a whole. A state of equilibrium is reached when the total sum of forces in the system is minimized.11 8 Note that shares of parts imports and final goods exports are not listed in Appendix Table 1, since these are not applicable when the NTI is averaged across industries. 9 Note that both shares relate to totals computed among the 75 countries populating the data set, which are highly representative of but do not fully comprise world trade involving all the countries. 10 Force-directed algorithms are applied broadly in network analysis. Examples are graph clustering techniques, ubiquitous to the field of biotechnology, or the minimum cost spanning tree, a classic algorithmic solution to a vast class of optimization problems. 11 Essentially, Hooke’s law states that the displacement or size of the deformation of an object is directly proportional to the deforming force. Within the logic of a spring-embedded algorithm, this is roughly equivalent to assuming that the forces in the system are proportional to the difference between the distance of the nodes and the length of the springs. Mapping Vertical Trade | 7 The algorithm is applied to the set of NTI by country-pairs outlined in the previous section. To facilitate the visualization of the NTI–based network, for the implementation of the algorithm we rely on Cytoscape, a software environment developed for visualizing molecular interaction networks.12 More specifically, we apply a spring-directed algorithm to the NTI scores across country pairs to draw maps of vertical trade for all the sectors combined, as well as for the electric/electronics and the automotive sectors individually taken. These are now discussed in turn. A. The Global Network of Vertical Trade The most comprehensive map of vertical trade is drawn to represent the entire set of NTI values, computed across all the industries in which processing trade is observed. The NTI is averaged in relation to each of the 5,162 country pairs, to gauge the intensity of vertical trade links in both directions.13 For a great majority of country pairs, vertical trade is very small, with the index taking a value close to 0. To improve the intelligibility of the network map, only the main network connections are retained. We thus drop all network relations with NTI smaller than 0.05, which reduces to 192 the total number of connections retained.14 Figure1 shows that each country in the network is presented by a circle, the coloring of which indicates whether a country pertains to developing Asia (red), the group of high-income countries (green), or developing countries outside Asia (blue). The circles’ position within the network and their proximity to each other is proportional to the force of attraction countries exert on each other through the various network relations of processing trade that run directly between any pair of countries, and indirectly via third countries or country-clusters. The strength of bilateral network relations determines the width of the arcs connecting the countries. 12 Hidalgo et al. (2007) were among the first to use the open-source platform Cytoscape within the field of economics, in their case to produce a network representation of their ”product space“. For more information about this software, see www.cytoscape.org. 13 For the case of some country pairs only one country imports parts and components from the other. Therefore, at 2,693 the total number of network relations identified by the average NTI is slightly higher than half the number of country pairs. 14 Such cut-off is arbitrary yet necessary to facilitate the graphical rendition of what would otherwise be a picture blurred by thousands of near-zero connections among all those countries that happen to have some sort of marginal relationship with each other. Depending on whether the focus of analysis is on the central or marginal features of a network, the range of NTI values to be retained for mapping can be adjusted quite arbitrarily to achieve the desired results. Here, we aim at highlighting the central features of the networks, for which a 0.05 cut-off appears to be roughly appropriate, after several trials and errors. For easier comparability of the networks discussed, the 0.05 cut-off is applied throughout the paper. 8 | ADB Economics Working Paper Series No. 263 Figure 1: Network Trade Index—All industries—Global TUN MUS HRV SWI POR SVN SPA BGR MAR DEN UKG KOR HUN TUR CZE NET SWE ZAF NOR EST ISR LVA LTU FIN SIN VIE NZL HKG THA EGY INO PHI AUS MAL IND CRI IRE CAN COL BRA PRC USA MEX JPN SRI URY ARG ROM FRA AUT ITA SVK GER BEL UKR BLR GRC POL MKD RUS Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. The main characteristics of global processing trade accruing from Figure 1 are highlighted as follows. Vertical trade is seen to concentrate around three global hubs, namely the US, PRC–Japan, and Germany, respectively. Although the sphere of influence of these hubs tends to be strongest within regions, it extends globally through network connections that involve hubs both directly (e.g., PRC–US) and indirectly, through a third country (e.g., PRC–Republic of Korea–US).15 Vertical trade with the US at its center is dominated by the country’s production sharing arrangements with the other NAFTA members, Canada, and Mexico. This should come as no surprise, considering the long-standing production sharing arrangements of the US with both countries: the US–Canada Auto Pact and Mexico’s extensive maquiladoras factories along the US borders, as will be discussed further below, in relation to the automotive and electric/electronics trade networks. 15 See Appendix Table 1 for a correspondence of country names with three-letter ISO codes. Mapping Vertical Trade | 9 Beyond its sphere of influence within North America, US vertical trade is strongest with Japan and the PRC, as well as with the Republic of Korea and the other countries pertaining to the tightly intertwined Asian networks (see Figure 2). Figure 2: Network Trade Index—All Industries—United States FIN NET SWE EST BEL GER UKG HUN DEN TUR CZE IRE ISR BRA SVK FRA ROM ITA AUT KOR MEX JPN USA THA SIN MAL PRC SPA HKG PHI VIE INO SRI CRI NET AUS COL CAN Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. The US’ network links with Germany and the rest of Europe appear to be far less pronounced. The same is true for US vertical trade with countries in Latin America. Colombia and Costa Rica can be seen at the outer margins of the global network, while Brazil is located somewhat more centrally, in between the US and European hubs and triangulating also with Argentina and Uruguay. Except for Mexico, which has direct ties with the US and also maintains network connections with Asian and European hubs, Latin America appears not to be strongly involved in global network trade, nor does the extent and intensity of vertical trade among its economies resemble anything close to the Asian networks. 10 | ADB Economics Working Paper Series No. 263 South of the US on the network map, the extensive Asian network is shown to extend around the PRC–Japan axis, involving a web of tightly connected East and Southeast Asian economies (see Figures 3 and 4). Often referred to as “factory Asia”, fragmented production activities scattered across the region typically involve the provision of high value-added parts and components by leading economies, such as Japan and the Republic of Korea, further processing in countries such as Malaysia and the Philippines, and final assembly in countries involving low labor costs and value added, predominantly in the PRC.16 Figure 3: Network Trade Index—All Industries—Japan GER BEL SVK NET DEN UKG USA JPN PRC KOR MEX SRI ITA AUT ROM CZE HUN FRA TUR MAL INO THA PHI SIN HKG VIE Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. 16 The Asian network and the individual network relations involved are listed in Appendix Table 3, for country pairs with NTI >0.05. Mapping Vertical Trade | 11 Figure 4: Network Trade Index—All Industries—People’s Republic of China TUN HRV CHE POR SVN SPA MAR DEN UKG KOR HUN TUR CZE NET SWE ISR FIN SIN VIE NZL HKG THA EGY INO PHI AUS MAL IND CRI IRE CAN COL PRC USA MEX JPN SRI ROM FRA AUT ITA SVK GER BEL POL Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. The Asian network as a whole and its countries individually are clearly drawn to the US through a range of network connections. By contrast, Asian networks appear to be largely disconnected from Australia and New Zealand. This reflects the nature of trade between Asia’s networked economies and the Pacific, involving mostly final consumer goods in exchange for primary commodities. Within the Asian networks, countries’ position in relation to each other sheds further light on the configuration and intensity of vertical trade links. For example, the proximity of the PRC and Japan testifies to the exceptionally strong vertical linkages between these two economies. Both countries are tightly linked to the main actors of factory Asia. Together, they lay at the crossroads of the major network axes spanning from East and Southeast Asia to Europe and the US. Contrary to Japan, however, the position of the PRC needs to be interpreted also in the light of Hong Kong, China’s role as an entrepôt for the mainland. Consequently, much of the PRC’s network trade with Hong Kong, China represents transshipments, including mainland processing trade with the region and the rest of the world. Hong Kong, China’s relative position within the Asian network may thus be interpreted as partly representing 12 | ADB Economics Working Paper Series No. 263 the PRC’s, which could thus be thought of as being located closer to the cluster of Asian countries it relates to as the region’s assembly hub.17 The Republic of Korea, by contrast, stands out as being less broadly connected to the Asian network, which it interacts with mainly through the PRC and, to a lesser extent, Japan. Whereas the Republic of Korea’s location on the map shows a close relation with its East Asian neighbors and with the US, its looser ties with the other countries of Asia explain its position at the outskirts of the regional networks. Among the Asian countries represented on the map, India appears most marginalized in the global trade networks, as it relates exclusively to the PRC and certainly is not yet part of factory Asia. Sri Lanka, on the other hand, is positioned rather well, with connections to the PRC; Hong Kong, China; and Japan, and a foothold into the European networks through Italy. Moving on to Europe (Figure 5), Germany is seen dominating the region’s processing trade, much in line with its status as one of the world’s top industrial and trading powers. Facilitated by the ease with which parts and components are allowed to cross national borders within the European Union’s single market, vertical trade can be seen forming a tight web of links among many of its member countries. Network trade is strongest between Germany and its neighbors, Austria, France, and Italy. The latter two countries themselves are closely networked within Europe and the major world hubs, as well as with the developing world. France, for example, displays strong ties to francophone Africa, including Tunisia and Mauritius.18 The United Kingdom (UK), by contrast, figures rather marginally within the network, broadly reflecting its fading status as an industrial power within Europe. Moreover, the country’s proximity to the US on the network map is reflective of its deep Atlantic alliance, including its industrial and foreign direct investment structures. Indeed, the NTI reckons the UK’s ties to the US to be nearly as strong as those it has with Germany. 17 To emphasize the special connection between the PRC and Hong Kong, China, the data panels in relation to the two countries could simply be collapsed into one, as frequently is done in empirical analysis taking focus on the PRC’s external trade. 18 It will be recalled that the index underlying the maps averages the two unidirectional values of the NTI characterizing any pair of countries. As a result, countries such as Tunisia are drawn onto the map mainly because of their strong dependence on a stronger network partner, which is almost never reciprocated beyond a certain degree. For example, the NTI of Tunisia toward France is 0.700, whereas it is only 0.019 in the opposite direction. Mapping Vertical Trade | 13 Figure 5: Network Trade Index—All Industries—Germany LVA EST UKR LTU BLR NOR RUS SWE ZAF MAR SVN HRV SWI BGR POR MKD GER GRC BEL POL SPA SVK NET IRE ISR FIN DEN UKG USA JPN PRC KOR MEX BRA SRI ITA AUT ROM CZE HUN FRA TUR INO THA HKG Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. Further testimony to Germany’s central role as Europe’s hub is its outreach to providers of parts and components well outside the region, such as Brazil, Mexico, and South Africa. Moreover, jointly with Sweden, Germany acts as a hub to the Scandinavian and northern European networks, spanning from Denmark, to Estonia, Latvia, Lithuania, and Poland. To the East, Germany is shown to link the Russian Federation and Ukraine through Belarus, albeit the marginal position of these countries suggests that none of them fully pertains to the European, let alone global, processing networks as such. Developing countries outside Asia are poorly represented on the network map. Of Latin America, only Mexico and Brazil are involved with global value chains. African countries are largely cut off from global networks, apart from sporadic connections to individual countries within the network’s core. This is true for South Africa, the continent’s most industrialized country, as much as for Egypt and for Mauritius, Morocco, and Tunisia, the only African countries with a strong enough NTI to make it onto the map. Also the CIS countries are relegated to the outer borders of the vertical trade networks. This is because their trade baskets tend to be heavily concentrated in natural resources, 14 | ADB Economics Working Paper Series No. 263 rather than networked manufacturing, and also because the Russian Federation, the regional hub, only weakly integrates with the world trading system. In sum, network analysis based on the NTI across industries indicates that the global distribution of vertical trade is heavily concentrated in Northern America, East and Southeast Asia, and Europe. By this measure, developing countries outside Asia are not yet integrated with the international production networks to any substantial degree. B. The Global Network of the Electric/Electronics Sector Figure 6 shows the map of vertical trade within the electric and electronics sectors, defined by the broad category No. 85 of the HS nomenclature. Again, only country pairs with the strongest links (NTI >0.05) enter the algorithmic transformation underlying the map. Out of the 2,599 country pairs identified within the sector, only 93 are retained. Appendix Table 4 ranks the top 15 among these country pairs, which correspond to the countries connected by the boldest arcs in Figure 6. Figure 6: Network Trade Index—Electric/Electronics Industries—Global MAL TUR SVK THA MEX CAN FRA GER FIN SWI UKG SPA IRE KOR HUN PRC PHI CZE NET POL SWE AUT ITA SPA POR SVN DEN TUN INO HKG CRI JPN USA SIN ISR Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. Compared to the map across industries discussed in the previous section, the electronics network is more sparsely populated, as several of the developing countries outside Asia fall below the 0.05 threshold value of the NTI. Mexico is the notable exception, which, paired with the US, ranks at the top of the entire network (Appendix Table 4) and also has extensive connections to the PRC, Japan, and other countries within the Asian networks (Figure 7). Mapping Vertical Trade | 15 Figure 7: Network Trade Index—Electric/Electronics Industries—United States MAL TUR SVK THA MEX CAN GER CHE SPA IRE KOR HUN PRC PHI CZE NET POL INO HKG CRI JPN USA SIN ISR Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. Figure 8: Network Trade Index—Electric/Electronics Industries—Japan MAL TUR SVK THA MEX CAN GER FIN SWI UKG SPA IRE KOR HUN PRC PHI CZE NET POL AUT SVN TUN INO HKG CRI JPN USA SIN ISR Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. 16 | ADB Economics Working Paper Series No. 263 Figure 15: Network Trade Index—Uni-Directed toward Japan—Top 30 Dependencies— All Industries TUR NET FRA CZE CAN SPA IND ISR IRE CRI ITA SVK BEL BRA HUN UKG NZL GER SIN SRI MEX JPN VIE MAL USA HKG PHI KOR INO PRC THA Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. Figure 16 and Appendix Table 7 rank countries according to their dependence on Japan’s supply of electric and electronics parts and components. Unsurprisingly, Japan’s central position within the regional production networks makes Asian buyers the most vulnerable to potential disruptions in the supply chain. Comparably less exposed is the US, which has the benefit of greater reliance on sources other than Japan for its electronic parts and components, most notably Mexico (Figure 7). Mapping Vertical Trade | 23 Figure 16: Network Trade Index—Uni-directed toward Japan—Top 30 dependencies— Electric/electronics industries CRI UKG GER NET VIE FIN SVK INO SIN USA EST TUR ISR SPA IRE POL BEL SWE AUT CAN FRA HKG PRC MEX KOR THA CZE PHI HUN MAL JPN Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. More vulnerable to disruptions in Japan appears the US’ automotive industry, as can be evinced from Figure 17 and Appendix Table 8. More generally, vulnerability seems to extend beyond the Asian region, to countries such as Estonia, Mexico, and Turkey. However, generally lower average NTI values indicate that supply dependency tends to be generally weaker in the auto industry compared to the electric/electronics supply chains. 24 | ADB Economics Working Paper Series No. 263 Figure 17: Network Trade Index—Uni-Directed toward Japan—Top 30 Dependencies— Automotive Industries AUS HUN CZE POL PRC HKG INO IND NET SWE ITA EST MEX KOR THA USA JPN GER BEL ECU COL SVK FRA ZAF POR UKG CAN BRA SPA TUR ARG Note: See Appendix Table 1 for definitions of country codes. Source: Author‘s calculation. In sum, the analysis here suggests that Japan’s position upstream in the global supply chains does create a certain degree of apprehension about the effects of potential disruptions affecting production downstream. Most likely to be affected are those countries relying most heavily on Japan’s provision of electronics parts, such as the PRC, Thailand, and the Philippines. However, it must be emphasized that the network dependencies highlighted by the NTI are defined as involving relative large volumes and shares of countries’ total parts imports. Therefore, for potential disruptions to affect a partner country’s processing industry, these would have to be sufficiently large to render it impossible for the importing country to source the shortfall of parts from alternative providers, including those in Asia. Disruptions of such entity appear to be a most unlikely scenario in the case of Japan, where fears about potential disruptions relate mostly to its continued capacity to provide some highly specialized, high-value added electronics parts, such as smart cards, and is more generally restricted to very few markets where Japan holds a disproportionate share of global production. Mapping Vertical Trade | 25 VI. Conclusions This paper develops a method that gauges the complex network relationships in world processing trade and leads to a plausible representation in the form of network maps. The paper finds that global processing trade centers on three major regional hubs. The first is the US, mainly through its strongly integrated automotive and electric/electronics production networks with the other NAFTA member states on one hand, and its close connections to the Asian electronics production networks on the other. The Asian network itself constitutes the second global hub, especially in relation to trade in parts and components within the electric and electronics industries surrounding the PRC–Japan axis, which also involves a number of economies in East and Southeast Asia. The third major hub is the European network with Germany at its center, broadly linking the single market’s value chains, most notably in relation to Europe’s strong automotive industry. Apart from Mexico—because of its maquiladoras network ties to the US—the analysis suggests that outside East and Southeast Asia, developing countries are not yet involved in global production networks to any substantial degree. There is no equivalent to factory Asia in Latin America, let alone Africa. Concerning the methodology underlying this paper, there is of course a margin for improvement of both the network index and the data set. As for the NTI, its properties and performance could be subjected to further evaluation by considering alternative weighting and standardization methods. Concerning the data set, the use of monthly data would allow identification of network effects of a more dynamic nature, which are particularly relevant for processing trade in relation to just-in-time production. Further insights could also be gained through the use of volume data and unit values for within-product classifications, to further distinguish partners’ share in parts and components imports according to different classes of unit values within 6-digit HS product lines. Finally, network maps can be drawn in relation to all the years available in the data underlying this analysis (1998–2007). Preliminary analysis shows that the evolution of processing trade since 1990 provides interesting aspects on the formation of production clusters in Asia. Due to space constraints, these maps will be the subject of discussion of a follow-up paper. 26 | ADB Economics Working Paper Series No. 263 Appendix Appendix Table 1: List of Countries/Economies Code Country Code Country Code Country DZA Algeria ITA Italy ZAF South Africa ARG Argentina JPN Japan SPA Spain AUS Australia HKG Hong Kong, China SRI Sri Lanka AUT Austria KAZ Kazakhstan SWE Sweden AZE Azerbaijan KOR Korea, Rep. of SWI Switzerland BLR Belarus KWT Kuwait THA Thailand BEL Belgium LVA Latvia TUN Tunisia BOL Bolivia LBY Libya TUR Turkey BRA Brazil LTU Lithuania TKM Turkmenistan BGR Bulgaria MKD Macedonia UKR Ukraine CAN Canada MAL Malaysia UKG United Kingdom CHL Chile MUS Mauritius USA United States PRC China, People’s Rep. of MEX Mexico URY Uruguay COL Colombia MON Mongolia VEN Venezuela CRI Costa Rica MAR Morocco VIE Viet Nam HRV Croatia NET Netherlands CZE Czech Rep. NZL New Zealand DEN Denmark NGA Nigeria ECU Ecuador NOR Norway EGY Egypt PRY Paraguay EST Estonia PER Peru FIN Finland PHI Philippines FRA France POL Poland GER Germany POR Portugal GRC Greece ROM Romania HUN Hungary RUS Russian Federation IND India SAU Saudi Arabia INO Indonesia SIN Singapore IRE Ireland SVK Slovak Rep. ISR Israel SVN Slovenia Note: List of 75 economies underlying the network trade index computations. Three-digit International Organization for Standardization (ISO) codes are used, except for certain member economies of the Asian Development Bank, for which the Bank’s country codes or country names are used. Source: Author‘s listing. Mapping Vertical Trade | 27 Appendix Table 2: Network Trade Index—All Industries Code Country Code Country Network Trade Index Network Trade Index (average) JPN Japan PRC China, People’s Rep. of 0.707 0.646 PRC China, People’s Rep. of JPN Japan 0.585 0.646 MEX Mexico USA United States 1.000 0.611 USA United States MEX Mexico 0.221 0.611 CAN Canada USA United States 0.881 0.579 USA United States CAN Canada 0.277 0.579 AUT Austria GER Germany 0.892 0.507 GER Germany AUT Austria 0.122 0.507 CZE Czech Rep. GER Germany 0.813 0.489 GER Germany CZE Czech Rep. 0.164 0.489 HKG Hong Kong, China PRC China, People’s Rep. of 0.764 0.443 PRC China, People’s Rep. of HKG Hong Kong, China 0.122 0.443 HUN Hungary GER Germany 0.750 0.422 GER Germany HUN Hungary 0.094 0.422 THA Thailand JPN Japan 0.626 0.395 JPN Japan THA Thailand 0.164 0.395 PRC China, People’s Rep. of KOR Korea, Rep. of 0.393 0.375 KOR Korea, Rep. of PRC China, People’s Rep. of 0.356 0.375 KOR Korea, Rep. of JPN Japan 0.543 0.363 JPN Japan KOR Korea, Rep. of 0.182 0.363 TUN Tunisia FRA France 0.700 0.359 FRA France TUN Tunisia 0.019 0.359 SVK Slovak Rep. GER Germany 0.651 0.349 GER Germany SVK Slovak Rep. 0.047 0.349 POL Poland GER Germany 0.597 0.348 GER Germany POL Poland 0.099 0.348 JPN Japan USA United States 0.407 0.342 USA United States JPN Japan 0.278 0.342 ITA Italy GER Germany 0.409 0.288 GER Germany ITA Italy 0.167 0.288 Note: Top 15 country pairs ranked by decreasing network trade index. Source: Author‘s calculation. 28 | ADB Economics Working Paper Series No. 263 Appendix Table 3: Network Trade Index (All Industries)—Asian Economies Code Country Code Country Network Trade Index Network Trade Index (average) JPN Japan PRC China, People’s Rep. of 0.707 0.646 PRC China, People’s Rep. of JPN Japan 0.585 0.646 HKG Hong Kong, China PRC China, People’s Rep. of 0.764 0.443 PRC China, People’s Rep. of HKG Hong Kong, China 0.122 0.443 THA Thailand JPN Japan 0.626 0.395 JPN Japan THA Thailand 0.164 0.395 PRC China, People’s Rep. of KOR Korea, Rep. of 0.393 0.375 KOR Korea, Rep. of PRC China, People’s Rep. of 0.356 0.375 KOR Korea, Rep. of JPN Japan 0.543 0.363 JPN Japan KOR Korea, Rep. of 0.182 0.363 PHI Philippines JPN Japan 0.318 0.202 JPN Japan PHI Philippines 0.087 0.202 SRI Sri Lanka HKG Hong Kong, China 0.399 0.200 HKG Hong Kong, China SRI Sri Lanka 0.001 0.200 MAL Malaysia SIN Singapore 0.219 0.180 SIN Singapore MAL Malaysia 0.141 0.180 MAL Malaysia PRC China, People’s Rep. of 0.264 0.157 PRC China, People’s Rep. of MAL Malaysia 0.051 0.157 VIE Viet Nam PRC China, People’s Rep. of 0.298 0.155 PRC China, People’s Rep. of VIE Viet Nam 0.013 0.155 PHI Philippines HKG Hong Kong, China 0.276 0.151 HKG Hong Kong, China PHI Philippines 0.025 0.151 MAL Malaysia JPN Japan 0.232 0.144 JPN Japan MAL Malaysia 0.056 0.144 THA Thailand PRC China, People’s Rep. of 0.234 0.144 PRC China, People’s Rep. of THA Thailand 0.054 0.144 VIE Viet Nam JPN Japan 0.234 0.136 JPN Japan VIE Viet Nam 0.037 0.136 INO Indonesia JPN Japan 0.185 0.134 JPN Japan INO Indonesia 0.083 0.134 MAL Malaysia THA Thailand 0.117 0.111 continued. Mapping Vertical Trade | 29 Appendix Table 3: continued. Code Country Code Country Network Trade Index Network Trade Index (average) THA Thailand MAL Malaysia 0.106 0.111 HKG Hong Kong, China JPN Japan 0.207 0.108 JPN Japan HKG Hong Kong, China 0.010 0.108 SIN Singapore PRC China, People’s Rep. of 0.109 0.097 PRC China, People’s Rep. of SIN Singapore 0.085 0.097 SIN Singapore JPN Japan 0.131 0.086 JPN Japan SIN Singapore 0.042 0.086 PHI Philippines PRC China, People’s Rep. of 0.104 0.081 PRC China, People’s Rep. of PHI Philippines 0.059 0.081 VIE Viet Nam HKG Hong Kong, China 0.159 0.081 HKG Hong Kong, China VIE Viet Nam 0.003 0.081 INO Indonesia SIN Singapore 0.107 0.081 SIN Singapore INO Indonesia 0.054 0.081 INO Indonesia THA Thailand 0.093 0.075 THA Thailand INO Indonesia 0.058 0.075 SRI Sri Lanka JPN Japan 0.149 0.075 JPN Japan SRI Sri Lanka 0.001 0.075 SRI Sri Lanka PRC China, People’s Rep. of 0.146 0.073 PRC China, People’s Rep. of SRI Sri Lanka 0.000 0.073 INO Indonesia PRC China, People’s Rep. of 0.115 0.065 PRC China, People’s Rep. of INO Indonesia 0.014 0.065 IND India PRC China, People’s Rep. of 0.118 0.064 PRC China, People’s Rep. of IND India 0.010 0.064 THA Thailand KOR Korea, Rep. of 0.092 0.059 KOR Korea, Rep. of THA Thailand 0.027 0.059 PHI Philippines THA Thailand 0.069 0.058 THA Thailand PHI Philippines 0.047 0.058 HKG Hong Kong, China SIN Singapore 0.086 0.057 SIN Singapore HKG Hong Kong, China 0.027 0.057 VIE Viet Nam THA Thailand 0.094 0.054 THA Thailand VIE Viet Nam 0.014 0.054 INO Indonesia KOR Korea, Rep. of 0.099 0.054 KOR Korea, Rep. of INO Indonesia 0.008 0.054 MAL Malaysia HKG Hong Kong, China 0.068 0.053 HKG Hong Kong, China MAL Malaysia 0.037 0.053 PHI Philippines SIN Singapore 0.081 0.051 SIN Singapore PHI Philippines 0.021 0.051 THA Thailand SIN Singapore 0.063 0.050 SIN Singapore THA Thailand 0.037 0.050 Note: Network relations among Asian country pairs with NTI >0.05. Source: Author‘s calculation. 30 | ADB Economics Working Paper Series No. 263 Appendix Table 4: Network Trade Index—Electric/Electronics Industries Country 1 Country 2 Parts Imports from Country 2 (%) Industry Share of Country 1 Exports (%) Network Trade Index Network Trade Index (average) Mexico USA 41.7 15.7 1.000 0.567 USA Mexico 21.5 4.1 0.135 0.567 PRC Japan 26.2 15.5 0.619 0.506 Japan PRC 30.5 8.4 0.392 0.506 Hong Kong, China PRC 42.2 13.3 0.854 0.478 PRC Hong Kong, China 4.3 15.5 0.102 0.478 PRC Korea, Rep. of 19.4 15.5 0.459 0.347 Korea, Rep. of PRC 24.8 6.2 0.235 0.347 Malaysia Singapore 17.9 11.8 0.323 0.253 Singapore Malaysia 21.0 5.7 0.183 0.253 Hungary Germany 21.8 14.3 0.476 0.252 Germany Hungary 5.3 3.6 0.029 0.252 Slovak Rep. Germany 20.4 14.3 0.443 0.225 Germany Slovak Rep. 1.4 3.6 0.008 0.225 Thailand Japan 28.4 8.6 0.370 0.223 Japan Thailand 5.9 8.4 0.076 0.223 Mexico PRC 17.6 15.7 0.422 0.216 PRC Mexico 0.4 15.5 0.009 0.216 Korea, Rep. of Japan 22.8 6.2 0.216 0.211 Japan Korea, Rep. of 16.1 8.4 0.207 0.211 Philippines Japan 17.9 11.7 0.318 0.203 Japan Philippines 6.8 8.4 0.088 0.203 PRC USA 11.0 15.5 0.260 0.200 USA PRC 22.1 4.1 0.139 0.200 Czech Rep. Germany 28.9 7.6 0.336 0.184 Germany Czech Rep. 5.7 3.6 0.031 0.184 Malaysia PRC 13.1 11.8 0.237 0.179 PRC Malaysia 5.1 15.5 0.121 0.179 Slovak Rep. Korea, Rep. of 16.4 14.3 0.356 0.178 Korea, Rep. of Slovak Rep. 0.0 6.2 0.000 0.178 PRC = People‘s Republic of China, USA = United States. Note: Top 15 country pairs—Share of Country 1’s parts and components imports sourced from Country 2, and industry’s share of total exports of Country 2. Source: Author‘s calculation. Mapping Vertical Trade | 31 Appendix Table 5: Network Trade Index—Automotive Industries Country 1 Country 2 Parts Imports from Country 2 (%) Industry Share of Country 1 Exports (%) Network Trade Index Network Trade Index (average) Canada USA 81.0 13.1 1.000 0.559 USA Canada 22.5 5.6 0.119 0.559 Mexico USA 65.3 12.2 0.751 0.429 USA Mexico 20.4 5.6 0.107 0.429 Slovak Rep. Germany 39.0 18.0 0.659 0.354 Germany Slovak Rep. 3.6 14.0 0.048 0.354 Austria Germany 57.2 8.4 0.451 0.291 Germany Austria 10.0 14.0 0.132 0.291 Czech Rep. Germany 45.0 9.8 0.415 0.269 Germany Czech Rep. 9.3 14.0 0.123 0.269 Spain Germany 28.6 16.6 0.445 0.265 Germany Spain 6.4 14.0 0.084 0.265 Spain France 24.9 16.6 0.388 0.264 France Spain 16.4 9.1 0.140 0.264 Thailand Japan 67.5 5.3 0.339 0.253 Japan Thailand 9.9 18.0 0.167 0.253 Hungary Germany 62.5 6.4 0.374 0.249 Germany Hungary 9.5 14.0 0.125 0.249 Argentina Brazil 52.5 6.9 0.341 0.209 Brazil Argentina 13.3 6.2 0.078 0.209 Japan PRC 23.3 18.0 0.393 0.208 PRC Japan 36.0 0.7 0.023 0.208 Poland Germany 38.8 7.9 0.288 0.192 Germany Poland 7.3 14.0 0.096 0.192 Belarus Russian Federation 44.8 8.9 0.376 0.190 Russian Federation Belarus 7.2 0.7 0.005 0.190 France Germany 27.5 9.1 0.235 0.189 Germany France 10.9 14.0 0.143 0.189 Korea, Rep. of Japan 27.5 9.3 0.242 0.186 Japan Korea, Rep. of 7.8 18.0 0.131 0.186 PRC = People‘s Republic of China, USA = United States. Note: Top 15 country pairs—Share of Country 1’s parts and components imports sourced from Country 2, and industry’s share of total exports of Country 2. Source: Author‘s calculation. 32 | ADB Economics Working Paper Series No. 263