Geographical Perspectives on International Trade (SpringerBriefs in Geography) 3319717308, 9783319717302

This book analyzes spatial and temporal patterns of international trade from a geographical perspective. Trade is an imp

117 56 7MB

English Pages 231 [225] Year 2021

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Foreword
Preface
Information Text
Acknowledgement
Contents
About the Author
List of Figures
List of Tables
Chapter 1: Introduction
1.1 The Background
1.2 International Trade: Theoretical Background
1.3 Recent Literature on International Trade
1.4 Geographic Perspectives on Globalization
1.5 Database and Analytical Framework
1.6 Presentation of the Book
References
Chapter 2: Volume and Growth of Trade: Merchandise and Services
2.1 Introduction
2.2 Database and Analytical Techniques
2.3 Trends in Merchandise Trade
2.4 Growth of Service Trade
2.5 Concluding Remarks
References
Chapter 3: Spatial Structure of Trade Flows
3.1 Introduction
3.2 Database and Methodology
3.3 The Factor Analysis of Flow Patterns (Q Mode Analysis)
3.4 The Factor Analysis of Flow Patterns (R Mode Analysis)
3.5 Conclusion
References
Chapter 4: An Analysis of the Commodity Composition of International Trade
4.1 Introduction
4.2 Methodology and Database
4.3 Composition of Merchandise Export
4.4 Concluding Remarks
References
Chapter 5: Output and Trade Relation
5.1 Introduction
5.2 Relation Between Economic Growth and Export: The Background
5.3 Sources of Data and Database
5.4 Methodology
5.5 An Analysis of the Relation Between Growth and Export
5.6 Concluding Remarks
References
Chapter 6: Regional Trade Blocks: A Case Study of Mercosur
6.1 Introduction
6.2 A Brief Analysis of the Trading Pattern of MERCOSUR Economies
6.3 An Analysis of the Relation Between Trade and GDP
6.4 Trade Structure by Major Commodity Groups: MERCOSUR
6.5 Concluding Remarks
References
Chapter 7: Concluding Observations
7.1 Introduction
7.2 Major Issues
References
Annexures
Annexure 2.1
Developed economies
Economies in Transition
Developing economies
America
Africa
Asia
Annexure 2.2
Annexure 2.3 Value of Merchandise Trade (US Million $)
Annexure 2.4 Annual Average Growth Rate of Merchandise Export and Merchandise Import (Values in Per Cent)
Annexure 2.5 Compound Annual Growth in Merchandise Trade
Annexure 2.6 Value of Service Trade
Annexure 2.7 Compound Annual Growth Rate of Service Trade
Annexure 3.1 Results of Q Mode Analysis, Communalities (1990)
Annexure 3.2 Results of Q Mode Analysis, Factor Loadings (1990)
Annexure 3.3 Results of Q Mode, Factor Scores (1990)
Annexure 3.4 Results of Q Mode, Communalities (2015)
Annexure 3.5: Results of Q Mode Analysis, Factor Loadings (2015)
Annexure 3.6. Results of Q Mode, Factor Scores, (2015)
Annexure 3.7 Results of Factor Analysis (R Mode), Communalities (1990)
Annexure 3.8 Results of R-Mode, Factor Loadings (1990)
Annexure 3.9 Results of R-Mode Analysis, Factor Scores (1990)
Annexure 3.10 Results of R-Mode Analysis, Communalities (2015)
Annexure 3.11 Results of R-Mode Analysis, Factor Loadings (2015)
Annexure 3.12 Results of R-Mode Analysis, Factor Scores (2015)
Annexure 4.1 Values of Location Quotient for Export of Major Commodity Groups (1995)
Annexure 4.2 Values of Location Quotient for Export of Major Commodity Groups (2005)
Annexure 4.3 Values of Location Quotient for Export of Major Commodity Groups (2015)
Annexure 4.4 Values of Location Quotient for Import of Major Commodity Groups (1995)
Annexure 4.5 Values of Location Quotient for Import of Major Commodity Groups (2005)
Annexure 4.6 Values of Location Quotient for Import of Major Commodity Groups (2015)
Annexure 5.1 Relation Between GDP and Merchandise Export (1990)
Annexure 5.2 Relation Between GDP and Merchandise Export (2015)
Annexure 5.3 Residuals from Regression of Merchandise Export on GDP (1990)
Annexure 5.4 Residuals from Regression of Merchandise Export on GDP (2015)
Annexure 6.1 Argentina: Standardized Residuals from Regression of GDP Ratio on Trade Ratio (1990 and 2015)
Annexure 6.2 Brazil: Standardized Residuals from Regression of GDP Ratio on Trade Ratio (1990 and 2015)
Annexure 6.3 Paraguay: Standardized Residuals from Regression of GDP Ratio on Trade Ratio (1990 and 2015)
Annexure 6.4 Uruguay: Standardized Residuals from Regression of GDP Ratio on Trade Ratio (1990 and 2015)
Annexure 6.5 Top Ten Trading Partners in Food and Live Animals of Argentina
Annexure 6.6 Top Ten Trading Partners in Food and Live Animals of Brazil
Annexure 6.7 Top Ten Trading Partners in Food and Live Animals of Paraguay
Annexure 6.8 Top Ten Trading Partners in Food and Live Animals of Uruguay
Annexure 6.9 Top Ten Trading Partners in Beverages and Tobacco of Argentina (1995 and 2015)
Annexure 6.10 Top Ten Trading Partners in Beverages and Tobacco of Brazil (1995 and 2015)
Annexure 6.11 Top Ten Trading Partners in Beverages and Tobacco of Paraguay (1995 and 2015)
Annexure 6.12 Top Ten Trading Partners in Beverages and Tobacco of Uruguay (1995 and 2015)
Annexure 6.13 Top Ten Trading Partners in Mineral Fuel of Argentina (1995 and 2015)
Annexure 6.14 Top ten Trading Partners in Mineral Fuel of Brazil (1995 and 2015)
Annexure 6.15 Top Ten Trading Partners in Mineral Fuel of Paraguay (1995 and 2015)
Annexure 6.16 Top Ten Trading Partners in Mineral Fuel of Uruguay (1995 and 2015)
Annexure 6.17 Top Ten Trading Partners in Manufactured Goods of Argentina (1995 and 2015)
Annexure 6.18 Top Ten Trading Partners in Manufactured Goods of Brazil (1995 and 2015)
Annexure 6.19 Top Ten Trading Partners in Manufactured Goods of Paraguay (1995 and 2015)
Annexure 6.20 Top Ten Trading Partners in Manufactured Goods of Uruguay (1995 and 2015)
Annexure 6.21 An analysis of the Relation Between the Actual and Expected Trade (Trade Ratio and Ratio of the Gravity Model Estimates)
Recommend Papers

Geographical Perspectives on International Trade (SpringerBriefs in Geography)
 3319717308, 9783319717302

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

SPRINGER BRIEFS IN GEOGRAPHY

Purva Yadav

Geographical Perspectives on International Trade 123

SpringerBriefs in Geography

SpringerBriefs in Geography presents concise summaries of cutting-edge research and practical applications across the fields of physical, environmental and human geography. It publishes compact refereed monographs under the editorial supervision of an international advisory board with the aim to publish 8 to 12 weeks after acceptance. Volumes are compact, 50 to 125 pages, with a clear focus. The series covers a range of content from professional to academic such as: timely reports of state-of-the art analytical techniques, bridges between new research results, snapshots of hot and/or emerging topics, elaborated thesis, literature reviews, and in-depth case studies. The scope of the series spans the entire field of geography, with a view to significantly advance research. The character of the series is international and multidisciplinary and will include research areas such as: GIS/cartography, remote sensing, geographical education, geospatial analysis, techniques and modeling, landscape/regional and urban planning, economic geography, housing and the built environment, and quantitative geography. Volumes in this series may analyze past, present and/or future trends, as well as their determinants and consequences. Both solicited and unsolicited manuscripts are considered for publication in this series. SpringerBriefs in Geography will be of interest to a wide range of individuals with interests in physical, environmental and human geography as well as for researchers from allied disciplines. More information about this series at http://www.springer.com/series/10050

Purva Yadav

Geographical Perspectives on International Trade

Purva Yadav Centre for Study of Regional Development, School of Social Sciences Jawaharlal Nehru University Delhi, India

ISSN 2211-4165     ISSN 2211-4173 (electronic) SpringerBriefs in Geography ISBN 978-3-319-71730-2    ISBN 978-3-319-71731-9 (eBook) https://doi.org/10.1007/978-3-319-71731-9 © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Dedicated to My Parents

Foreword

It is rare for Indian geographers to take up studies on International Trade. A study published in the journal Economic Geography that I recall was by Edison Dayal in the early 1970’s. To my knowledge, the study by Dr. Purva Yadav is one among such rarity. This is one of the reasons, other than the fact that Purva undertook this research for a doctoral degree in Geography under my supervision at the University of Delhi that I agreed to write a foreword for the book. The book is being published under the Springer Brief series. I recall a discussion, at the meeting of the Board of Research in Humanities of the University of Delhi, when considering Purva’s research proposal. The discussion was centred on the question - whether the study of International Trade falls under the ambit of Geography, Economics or Commerce. One of my colleagues at the meeting remarked that since the proposed research is based on data, it would be at least five years behind by the time the dissertation comes for examination, and therefore, may as well be classified under contemporary history. While many research themes fall under a single discipline, there are several that cut across the boundaries of academic disciplines. Despite wide ranging advocacy for the promotion of interdisciplinary research, the rigidity of the departmental structure in universities poses a challenge. Some universities started using the term ‘Centre’ instead of ‘Department’ with the hope that this would help overcome disciplinary boundaries. Coming back to the discussion alluded to earlier in this paragraph - the decision was that the research proposal could be taken up under the Department of Geography, since both the scholar and the proposed supervisor have master’s degrees in geography. Purva’s study also defies an unwritten rule that research for a doctoral degree in geography must be based on field work. However, there are several examples of doctoral studies in geography that defy this dictum. It is this background that led one of the examiners to comment that Dr. Yadav’s study is a bold attempt! Much of geographical inquiry can be viewed in a paradigmatic framework of matrices. The study uses commodity flow matrices as the base for a number of statistical manipulations (not used in a pejorative sense) to understand the spatial

vii

viii

Foreword

dynamics of International trade. International trade involves the movement of commodities between countries. For such a trade to take place between two countries three geographic principles have to be met – (a) in a given commodity one country has a surplus and the second country has a deficit, (b) there are no intervening opportunities and (c) transferability – transportation costs are sustainable with the underlying assumption that there are no political barriers to trade. These conditions unfold a large variety of dynamic options in spatial patterns of international trade. It is a challenge to unravel meaningful patterns from such a complex set of commodity flows - a task under taken by Dr. Yadav in her study. Globalisation processes usher in diversification of consumption and specialization in production. These in turn promote structural changes in the economy. It is important to map the changing patterns of trade, since trade has multiplied and benefits accruing from free trade policies vary considerably across economies. The book while focussing on the contemporary global trading system from a geographical perspective, includes an intensive review of related literature on international trade and outlines the major issues, followed by an empirical study and finally, the results of the analysis. International trade is a partial reflection of spatial differences and therefore, brings out the underlying structure of the regional interdependencies. The increasing trade is a consequence of a paradigm shift in the economic policies of many countries. A number of countries introduced economic policy reforms from the early 80’s onwards, with the objective of the removal of barriers to international flow of trade, capital, and technology to promote better utilization of resources and reduce cost of production. The concept comparative advantage earlier formulated by David Ricardo is an influential trade theory of the classical period and has proved to be dynamic and relevant in the context of economic reforms, even in modern times. As observed by Chisholm, the idea of comparative advantage provides the conceptual structure that helps examine the regional specialisation or centralisation of production, and also links production with trade patterns. Some believe that the term ‘competitive’ is a macroeconomic phenomenon, whereas others perceive it as a function of abundant and cheap labour. Others consider abundance of natural resources as a mainstay of competitiveness. Lately many have viewed the strong impact of government policies on competitiveness. The classical theories presumed that trade would take place either because of the differences in technological advancement or factor endowments or both. Traditional theories attempt to predict trade when countries that are different from each other engage in exchange. Several empirical studies in recent years note the increasing intra-industry trade in the modern global trading system, and indicate the significant trade between countries that are similar. Therefore, to understand intra-industry trade, new trade theories emphasize the role of economies of scale and imperfect competition, and constant returns to scale  - key assumptions in explaining inter-­ industry trade.

Foreword

ix

National trade policies do shape the patterns of trade flows. International trade is one of the key drivers of globalization which itself is an uneven process across time and space. Almost all countries are affected by it; however, its impact varies. There are multilateral and regional institutions as well as bi-­lateral agreements that act as barriers or promoters of trade depending on how they are perceived. It has been observed that ‘the global’ is claimed to be a natural order, an inevitable state of affairs, in which time-space has been compressed, the ‘end of geography’ has arrived and everywhere is becoming the same. However, geography still matters, despite a shrinking space due to innovations in transportation. While the WTO may be perceived as an institution that promotes world trade, Regional Trade Blocks may be perceived as either promoters of regional trade or a barrier to Global Trade. Similarly, bi-lateral agreements may merely help diverting regional and global trade to bilateral flows of goods. Regional Trade Blocks and bi-lateral agreements remind us that distance matters. This study builds on the existing methodology related to trade analysis. The first step required is to build data matrices reflecting the structure of international trade across countries for which data can be collated for different time periods. In the present study, the time period from 1990 to 2015 is covered, and data matrices have been created. This includes trade related variables, such as the value of exports and imports by region, structure of imports by origin and exports by destination, structure of international trade by products, export concentration and import diversification indices; and growth-related variables like gross domestic product. When dealing with time series data there are problems of comparability of countries (with changes in national boundaries), thus the analyses is limited to countries for which time wise data is available. In selected cases, the three-digit commodity classification is used particularly when dealing with concentration and diversification of commodity composition. Data reduction techniques such as factor analysis and regression models for relating the economic structures and trade pattern have been employed. These models are used for both aggregate and disaggregated data. Similarly, the regionalization process also captures the underlying structural spatial dependency via trade flows in the globalised era. Regionalization of trade reflects dependence of less developed countries on trade with developed economies that are either geographically or geopolitically close. While attempting to realise the objectives of the inquiry, the study examines a set of hypotheses: (i) Countries have opened up by undertaking reforms in export and import policy in order to increase their trade, however, such expected increase is confined to few countries, (ii) South-­ south trade has significantly improved, (iii) Commodity composition, particularly, of the developing countries has also changed from low value raw material to high value manufactured goods, (iv) Competitiveness leads to specialization in production. Consequently, one expects an increase in the concentration in commodity composition of exports, and (v) the process of commodity concentration of exports would lead to diversification of imports.

x

Foreword

I believe that this book would be useful for two kinds of readers: those that have interest in learning about the contemporary changes in the spatial patterns of international trade; and those interested in learning about analytical frameworks and techniques in studying flow matrices. Besides complementing Dr. Purva Yadav and the publisher Springer Nature for their endeavour, it is a pleasure to have been associated with the study. Prof. Dr. Hariharan Ramachandran, ICSSR National Fellow

Preface

The geographical perspectives on trade is an attempt to capture the effects of globalization via trade flows in the spatial context, and how developments in the global and regional economies influence the trading landscape. The response to fast paced globalization via flows differs considerably across countries depending upon their geographical location, evolving geopolitical relations, resources, logistics, reforms etc. The book attempts to shed light on the North-North and South-South interdependencies via trade flows, and how the emergence of Asian tigers has brought the tectonic shift in the global trading system. This book further explores the structural linkages via functional regionalisation noting the persistence of the significant traditional North-North trade, and also the evolving South-South trade reflecting upon the changing dynamics of the spaces of flows in the globalisation process. For example, the rise of China and India is noticeable in this context. The similarities and dissimilarities in the trade and welfare gains between and across global north and south is also influenced by the nature of the commodities and services they trade. On the similar lines, it is believed that the globalisation has infused competitiveness that in turn has led to the specialisation in production (export) and diversification of consumption (import). Hence, the current book also attempts to capture the degree of the concentration/diversification of merchandise export and import across different set of countries. Similarly, the other noticeable trend in the global trading system is the proliferating web of RTAs in the multilateral trading system. More recently, developing countries have joined developed countries under new varieties of regional arrangements, such as North-South RTAs and North-South-­ South RTAs, highlighting increasing North-South interdependence. Deepening integration among developing economies via South-South trade is also on the rise. Some of these arrangements, such as MERCOSUR, is touched upon in the current book. Delhi, India

Purva Yadav

xi

Information Text

This book analyzes spatial and temporal patterns of international trade from a geographical perspective. Trade is an important key to understand the changing dynamics of economic spaces over time. However, studies by geographers are largely confined to case studies, whereas the spatial dimension is often missing from the approach of economists. This study highlights spatial patterns and commodity composition of global trade and the nature of relationship between trade and other economic attributes. A case study of the MERCOSUR trade block examines inter-regional and intra-regional trade flows. The book captures a comprehensive picture of the structure of international exchange by using ample maps and illustrations as supporting features. Many different methods are applied such as the location quotient to capture concentration and diversification of commodity composition, data reduction techniques such as factor analysis and regression models for relating the economic structures and trade patterns as well as residual mapping among others. This book is a significant contribution to geographical, economical and social sciences research and very useful to graduate and post-graduate students as well as scientists of all related areas who have interest in exploring the changing dynamics of the global economy via trade flows. It provides a road map to further explore different dimensions of international trade and its role in understanding the transforming global economy.

xiii

Acknowledgement

I would like to put on record my appreciation for the help and assistance received during my research for the book, especially from my mentor Prof. Hariharan Ramachandran, for his generous time and commitment. He has read through my initial draft, listened to my queries patiently and supported me every step of the way. Throughout my research he has encouraged me to inculcate independent thinking. His contributions, detailed comments and vision have been of great value to me and has been instrumental in completing the book. I am also thankful to Springer Nature Switzerland AG for publishing my research work in the form of this book. I would further like to thank anonymous reviewer whose comments substantially improved this manuscript. I would also like to acknowledge Ms. Poonam Chandel for providing assistance in preparing maps presented in the book. Last, but not the least, I am also thankful to my mother, sister and brothers who were a constant source of encouragement, and without their support this book would not have been possible.

xv

Contents

1 Introduction����������������������������������������������������������������������������������������������    1 2 Volume and Growth of Trade: Merchandise and Services������������������   21 3 Spatial Structure of Trade Flows������������������������������������������������������������   43 4 An Analysis of the Commodity Composition of International Trade������������������������������������������������������������������������������   59 5 Output and Trade Relation ��������������������������������������������������������������������   83 6 Regional Trade Blocks: A Case Study of Mercosur������������������������������   99 7 Concluding Observations������������������������������������������������������������������������  121 Annexures ��������������������������������������������������������������������������������������������������������  125

xvii

About the Author

Purva Yadav  is an Assistant Professor of Economic Geography at the Centre for the Study of Regional Development (CSRD), School of Social Sciences, Jawaharlal Nehru University (JNU), New Delhi, India. Prior to joining JNU, she has taught in Miranda House, University of Delhi, and had also worked in the National Institute of Urban Affairs (NIUA). She has obtained her Masters degree in Geography from CSRD, JNU, New Delhi; M.Phil and Ph.D from Department of Geography, Delhi School of Economics, University of Delhi, India. She has written on issues pertaining to trade, capital flows and transport geographies.

xix

List of Figures

Fig. 2.1 Trend in the value of world merchandise trade. (Source: Based on Annexure 2.3)����������������������������������������������������   24 Fig. 2.2 Growth of merchandise export. (Source: Based on Annexure 2.4 (UNCTAD Handbook of Statistics 2016))����������������   26 Fig. 2.3 Growth of merchandise import. (Source: Based on Annexure 2.4)����������������������������������������������������   26 Fig. 2.4 Growth of merchandise trade, 1990–1995. (Source: Based on Annexure 2.5)����������������������������������������������������   28 Fig. 2.5 Growth of merchandise trade, 2000–2005. (Source: Based on Annexure 2.5)������������������������������������������������������������������������������   28 Fig. 2.6 Figure 2.5: Growth of merchandise trade, 2010–2015. (Source: Based on Annexure 2.5)����������������������������������������������������   29 Fig. 2.7 Trade in services. (Source: Based on Annexure 2.6)�����������������������   35 Fig. 2.8 Major categories of services. (Source: Based on UNCTAD Handbook of statistics, 2006–07)����������������������������������������������������   37 Fig. 2.9 Growth rate of service export (values in percent). (Source: Calculation based on Annexure 2.6)����������������������������������   37 Fig. 2.10 Growth rate of service import. (Source: Calculation based on Annexure 2.6)������������������������������������������������������������������������������   38 Fig. 2.11 Growth of service trade, 1990–1995. (Source: Based on Annexure 2.7)������������������������������������������������������������������������������   39 Fig. 2.12 Growth of service trade, 2000–2005. (Source: Based on Annexure 2.7)������������������������������������������������������������������������������   39 Fig. 2.13 Growth of Service trade, 2010–2013. (Source: Based on Annexure 2.7)������������������������������������������������������������������������������   40 Fig. 3.1 Fig. 3.2

Intra- and inter-regional trade among developing countries (2015). (Source: UNCTAD 2016)������������������������������������   44 Dominant shippers with their Hinterland (1990). (Source: Based on author’s calculations)�����������������������������������������   50

xxi

xxii

Fig. 3.3 Fig. 3.4 Fig. 3.5 Fig. 4.1

Fig. 4.2 Fig. 5.1 Fig. 5.2 Fig. 5.3 Fig. 5.4 Fig. 5.5 Fig. 5.6 Fig. 6.1 Fig. 6.2 Fig. 6.3

List of Figures

Dominant shippers with their Hinterland (2015). (Source: Based on author’s calculations)�����������������������������������������   51 Dominant receivers with their major Shippers/Origins(1990). (Source: Based on author’s calculations)�����������������������������������������   53 Dominant receivers with their major shippers/origins (2015). (Source: Based on author’s calculations)�����������������������������������������   54 Commodities, at the 3-digit level or by broad product group, classified in accordance with the United Nations Standard International Trade Classification (SITC). (Source: UNCTAD Handbook of Statistics, 2006–07; see Annexure 4.2 for detailed commodity composition)����������������������������������������������������   61 World merchandise export of major commodity groups. (Source: Based on UNCTADSTAT database)����������������������������������   63 Association between merchandise export and GDP (1990) (arithmetic scale). (Source: Based on Annexure 5.1)����������������������   90 Association between merchandise export and GDP (2015) (arithmetic scale). (Source: Based on Annexure 5.2)����������������������   91 Association between merchandise export and GDP (1990) (logarithmic scale). (Source: Based on Annexure 5.1)��������������������   91 Association between merchandise export and GDP (2015) (logarithmic scale). (Source: Based on Annexure 5.2)��������������������   92 Residuals from regression of merchandise export on GDP, 1990. (Source: Based on Annexure 5.3)���������������������������������������������� 94 Residuals from regression of merchandise export on GDP, 2015. (Source: Based on Annexure 5.4)������������������������������������������   95

MERCOSUR: Merchandise export. (Source: UNCTADstat)����������  103 MERCOSUR: Merchandise import. (Source: UNCTADstat)���������  104 Argentina: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.1)�������������������  106 Fig. 6.4 Argentina: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.1)�������������������  107 Fig. 6.5 Brazil: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.2)���������������������������������  107 Fig. 6.6 Brazil: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.2)���������������������������������  108 Fig. 6.7 Paraguay: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.3)���������������������������������  108 Fig. 6.8 Paraguay: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.3)���������������������������������  109 Fig. 6.9 Uruguay: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.4)���������������������������������  109 Fig. 6.10 Uruguay: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.4)���������������������������������  110

List of Figures

xxiii

Fig. 6.11 Top Ten Trading partners of Argentina, 1995. (Source: Based on Annexure 6.5, 6.9, 6.13, 6.17)���������������������������  113 Fig. 6.12 Top Ten Trading partners or Argentina, 2015. (Source: Based on Annexure 6.5, 6.9, 6.13, 6.17)���������������������������  114 Fig. 6.13 Top Ten Trading partners of Brazil, 1995. (Source: Based on Annexure 6.6, 6.10, 6.14, 6.18)������������������������  114 Fig. 6.14 Top Ten Trading partners of Brazil, 2015. (Source: Based on Annexure 6.6, 6.10, 6.14, 6.18)�������������������������  115 Fig. 6.15 Top Ten Trading partners of Paraguay, 1995. (Source: Based on Annexure 6.7, 6.11,6.15, 6.19)��������������������������  115 Fig. 6.16 Top Ten Trading partners of Paraguay, 2015. (Source: Based on Annexure 6.7, 6.11,6.15, 6.19)��������������������������  116 Fig. 6.17 Top Ten Trading partners of Uruguay, 1995. (Source: Based on Annexure 6.8, 6.12, 6.16, 6.20)�������������������������  116 Fig. 6.18 Top Ten Trading partners of Uruguay, 2015. (Source: Based on Annexure 6.8, 6.12, 6.16, 6.20)�������������������������  117

List of Tables

Table 2.1 Table 2.2 Table 2.3 Table 2.4

Share of merchandise export (shares in percent by country group)�������������������������������������������������������������������������   25 Top five players in international merchandise trade (percent share to world merchandise trade)���������������������������������   27 Share of service trade (World export and import of service in billions of dollars and share in percent)�����������������   35 Top five players in international service trade (percent share to world service trade)������������������������������������������   38

Table 4.1

Distribution of export of primary commodities and manufactured goods (percent to total merchandise export)���������������������������������������������������������������������   64 Table 4.2 Distribution of world merchandise export of all food items by destination (values in percent)���������������������������������������   64 Table 4.3 Distribution of world merchandise export of raw materials from agricultural sources by destination (values in percent)�������   64 Table 4.4 Distribution of world merchandise export of ores and metals by destination (values in percent)������������������������������   65 Table 4.5 Distribution of world merchandise export of fuels by destination (values in percent)�������������������������������������������������   65 Table 4.6 Distribution of world merchandise export of manufactured goods by destination (values in percent)��������������������������������������   65 Table 4.7 Composition of export of all food items by destination (Share in percent)�������������������������������������������������������������������������   66 Table 4.8 Composition of export of raw materials from agricultural sources by destination (Share in percent)�������������������������������������   66 Table 4.9 Composition of export of ores and metals by destination (share in percent)��������������������������������������������������������������������������   66 Table 4.10 Composition of export of fuels by destination (share in percent)��������������������������������������������������������������������������   67

xxv

xxvi

List of Tables

Table 4.11 Composition of export of manufactured goods by destination (Share in percent)��������������������������������������������������   67 Table 4.12 Commodity composition of global export basket (1995)�������������   74 Table 4.13 Commodity composition of global export basket (2005)�������������   75 Table 4.14 Commodity composition of global export basket (2015)�������������   75 Table 4.15 Commodity composition of global import basket (1995)������������   79 Table 4.16 Commodity composition of global import basket (2005)������������   79 Table 4.17 Commodity composition of global import basket (2015)������������   80 Table 6.1 Table 6.2 Table 6.3 Table 6.4

Share of merchandise export from MERCOSUR partners����������  104 Share of merchandise import from MERCOSUR partners����������  105 Merchandise export of MERCOSUR economies by major commodity groups (percent share in total export)��������  112 Merchandise import of MERCOSUR economies by major commodity groups (per cent share in total export)�������  112

Chapter 1

Introduction

Abstract  The globalization process integrates global spaces and places via free flow of commodities, capital, people, innovation etc. The structural changes that are occurring due to the differential nature of organization as well as integration of the world economy have varying spatial trajectories. It is important to map the changing contours of trade in the globalized era because world trade has significantly multiplied over the years, and the gains accruing from free trade policies vary considerably across economies. It is popularly believed that globalization has created homogeneity; however, in the case of trade, there is visible heterogeneity in the development experience as well the structure and pace of integration. The primary focus of this book is to understand and present the global trading system from the geographical perspectives. In order to understand the global integration process via trade from the geographical perspectives, it is important to understand the trends, patterns and structure of flows embedded in space. The current chapter outlines the related literature on international trade and the major issues and defines the major objectives, followed by an explanation of the database and the methodological approach. It also outlines the rest of the book. Keywords  Comparative advantage · Geography · Globalisation · Trade and growth

1.1  The Background Globalization over the decades has accentuated the degree and extensity of economic interdependencies and integration. According to Aoyama et al. (2011), ‘economic globalisation is being driven by the geographical dispersal of markets, the functional integration of production activities, and the increasing interconnections and interdependencies between people and places in the world economy’. It is worth noting that the spatial character is inherent in the process. In this context, Coe and Yeung (2001) have written, ‘Any international economic relationship-be it © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9_1

1

2

1 Introduction

associated with trade, investment, finance, labour, or technology- is a reflection of, and partly constituted by, spatial difference...’ Hence, trade being one of the key drivers of globalization, an attempt is made through this book to map the changing underlying structure of the interdependencies via trade flows in the global economy over the past decades. World trade has grown rapidly during the last two decades of the twentieth century. It has grown from 73% during 1980–1990 to 84% during 1990–2000. The increasing trade is a consequence of a paradigm shift in economic policies of many countries. The growing global trade and flow of foreign capital have infused efficiency in some economies and may have also adversely affected others. Krugman (1995) asks, ‘Why has world trade grown, and what are the consequences of that growth?’ and observes that ‘these are surprisingly disputed issues’. Both developed and developing countries are affected by the integration of the market economy, but in varying degree. The research interest in international trade shows a spurt subsequent to liberal policies introduced in many countries since the 1980s. A number of countries introduced economic policy reforms from early 1980s, with the following objectives: (a) removal of barriers in flow of trade, capital, technology etc. to promote rapid economic growth and (b) better utilization of resources as well as reduction of cost of production.

1.2  International Trade: Theoretical Background Theories pertaining to trade has a long economic pedigree. Why do economies trade? What are the gains from trade? There exist well-established theories that attempt to explain international trade over the years. The early-noted attempts in explaining gains from trade were theorized by Adam Smith in the late eighteenth century and by David Ricardo in the early nineteenth century. Poon and Rigby (2017) have written, ‘At the time, it was widely considered that wealth take the form of gold and silver and thus nation-states should engage in international trade only insofar as it increased their accumulation of these resources. This was the view of mercantilism and it led to policy that limited imports while encouraging exports’. Adam Smith in his pioneer work The Wealth of Nations (1776) discussed the bases of competition and economic growth, arguing that in order to accrue gains from trade, a country’s resource should be utilized for producing goods that it had an absolute advantage over and other commodities that are efficiently produced in foreign countries must be imported. David Ricardo propounded the theory of comparative advantage, and it is considered as the most influential trade theory of the classical period and has proved to be dynamic and relevant even in the modern times. Ricardo’s book titled On the Principles of Political Economy and Taxation gave insight into the concept of comparative advantage. This theory has been long established in the field of economics and has been evolved in the early debates on the question of whether to participate in international trade and what are the gains and for whom. It is well accepted that some trade is profitable from the basic fact that production does not all occur in one infinitely small spatial unit. So in order to

1.2  International Trade: Theoretical Background

3

have access to a variety of commodities, exchange is necessary. Moreover, the production costs vary across place due to multiple geographical and non-geographical factors. Such determining factors affect the location of production centres and also spatially confine them. The theory of comparative advantage provides the conceptual structure that helps to examine the regional specialization or centralization of production and also to link production with the trade patterns (Chisholm 1968). John Stuart Mill further refines Ricardo’s theory, as put forth by Thoman and Conkling (1967), ‘by taking into consideration comparative costs as setting limits of tolerance (in a two-commodity case) within which actual costs would range. These actual costs, in turn, would be determined by demand and supply’. Alfred Marshall further refined this concept. The absolute and comparative advantage doctrine was limited in scope because of restricting the argument to simple production process with only one input (Poon and Rigby 2017). Furthermore, the Heckscher-Ohlin(H-O) model asserts that in the real world, the factor endowments may also influence trade pattern of countries, at least in the initial stages of exchange. This model is considered significant among trade theorists as it also extends to analyze intra-country distribution of benefits accrued by participating in the trade. The model suggests that owners of surplus factor of production tend to gain more from trade as compared to owners of scarce factors of production. In fact, the notion was first developed as Stolper-Samuelson theorem by economists W. Stolper and P. Samuelson. It argued that free trade happening between different countries tended to equalize factor prices between trading countries. Similarly, W. Leontief (1953) proposed a test of the H-O model and noted that it is yet not certain the degree to which a particular factor of production influences international trade. His early work based on input-output analysis undertaken for the US foreign trade stressed that the exchange is mainly based on specialized labour rather than the generalized notion of heavy capital investment. This was popularly referred to as the Leontiff paradox. Of the several critical arguments on Ricardo’s comparative advantage theory, the one is given by Gunnar Mrydal, Raul Prebisch, H. Singer and others have criticized Ricardo’s comparative advantage theory on numerous grounds. For instance, the empirical analysis suggested that the gap between developed and developing countries is increasing, and among other factors, the deteriorating terms of trade are significantly contributing to this trend. This could be because developing economies (except fuel-based economies) predominantly trade in primary commodities and the developed ones in manufactures. The latter has a higher elasticity of demand as compared to the former, which in turn result in the varying gains from trade accruing to different countries across the world, and further highlights the volatility tied with the free trade. New trade theory: The classical theories presumed that trade will happen either because of the differences in the technological advancement or factor endowments or both. Traditional theories attempt to predict trade and gains from it when countries that are different from each other engage in exchange; in this backdrop, the new trade theory extends the argument further. Several empirical studies in recent years noted the increasing intra-industry trade in the modern global trading system

4

1 Introduction

and indicated the visible existence of trade between countries with similarity. Therefore, to understand intra-industry trade, new trade theories emphasized on economies scale and imperfect competition, rather than perfect competition and constant returns to scale (key assumptions to explain inter-industry trade). Paul Krugman’s new trade theory is one of the most widely acceptable theorem. In this context, Grant (1994) stated that international trade theorists have ‘rediscovered economic geography and resumed incorporating locational decisions into trade models’. Furthermore, Poon and Rigby (2017) have also shed some light on the new, new trade theory. They presented a comprehensive discussion on global outsourcing and highlighted spatial and intra-industry fragmentation of the production process. The ‘new, new trade theorists’ basically explain how heterogeneous firms of an economy decide what activities in the production process they will undertake and what will be outsourced across countries in the world economic system. It is important to note that the brief overview of trade theories presented in the preceding paragraphs broadly covers the theoretical underpinning of international trade. Nevertheless, it does highlight the fact that trade theories have evolved over the years with the changing world context. New (Krugman) and classical trade theories (Ricardo, Samuelson and Ohlin) did consider geographical dimensions. According to Andresen (2010), ‘The new economic geography is a variation of the new trade theory that considers firm-level economies of scale, imperfect competition and product differentiation. The primary theoretical expectation from the new economic geography is that a decrease in trade barriers (tariff or nontariff reductions, for example) leads to increases in the agglomeration of production in the larger of the two economies. This occurs through migration of firms from one economy to the other with the larger of the two economies exporting to the smaller economy to minimize transportation costs. Thus it is the nature of agglomeration, not comparative advantage, that dictates trading patterns’.

1.3  Recent Literature on International Trade There is a large and growing body of literature on international trade and economic growth; it is convenient therefore to classify the literature on the basis of analytical framework and findings. This review includes publications both in the field of economic geography and economics. Panayotou (2000) highlights, ‘Globalization has been the defining trend in the closing decade of the 20th century and the dawn of new millennium heralding a new era of interaction among nations, economies and people’. Fischer (2003) found that in the 1990s, the term globalization captured the debate relating to the direction of international economic relations. It is the most-­ studied process by social scientists (Caselli 2006). It has been noted in an article (Anonymous 2004) that in the globalizing world, export constraints are diluting through the removal of trade barriers, falling transport costs, spread of new information and communication technologies and international migration and free flow of capital that have created a highly competitive global market. In this fiercer

1.3  Recent Literature on International Trade

5

competition, leaders are those who have a competitive edge over others. As per the study, being competitive rests on a country’s ability to produce competitive goods and services as well as to export them. In the same line of thought, Porter (1990) had also argued that competitiveness has become linchpin of every nation’s government and industry, and ‘The United States is an obvious example, with its growing public debate about the apparently greater economic success of other trading nations. But intense debate about competitiveness is also taking place today in such '“success story” nations as Japan and Korea. Socialist countries such as the Soviet Union and others in Eastern Europe and Asia are also asking this question as they fundamentally reappraise their economic systems’. This simply refers to the fact that globalization of competition has intensified over the years. Some believe that the term ‘competitive’ is a macro-economic phenomenon, whereas others perceive it as a function of abundant and cheap labour. Others consider abundance of natural resources as a mainstay of competitiveness. Lately, many viewed a strong impact of government policies on competitiveness. None of the analyses succeeded in floating a convincing explanation of national competiveness. Kay (1998) has also highlighted that national competitiveness is the one issue which has created more confusion than insight. He has also explored the adaptive capacity of the firm as a key to European competitiveness. A similar type of study was undertaken by Chaudhary and Saleem (2003), who analyzed Pakistan’s comparative advantage of exports, their complementarity and instability and market diversification. Sachs et. al. (1999) analysis of trade patterns and economic development extends to capture the scenario where endogenous and exogenous comparative advantages coexist. According to Krugman (1995), a nation’s emphasis on international competitiveness can be a risky venture, so the thrust should be in augmenting free trade. Panayotou (2000) says, ‘Globalization in general, and freer trade in particular, result in a shift in industrial structure more in line with a country’s comparative advantage’. The impact of trade on growth and income has been debated at length in economic history. Since the early classical contributions like that of Adam Smith (absolute advantage, 1776) and David Ricardo (comparative cost advantage, 1817), there have been several studies conducted till date to study the possible benefits accruing from free trade. Bidlingmaieri (2007) has highlighted that according to Ricardo and Heckscher, gains stem from specialization in production through international trade. In consonance with this, Harrison (1995) has also briefly reviewed growth models developed by Solow and others. According to them, technological change was considered exogenous to a country’s openness. In stark contrast, new growth theories highlighted the impact of trade policy on the long run growth via technological breakthrough. Grossman and Helpman (1989) largely neglect the effect of trade on growth but have tried to explore the ‘reverse causation from growth and accumulation to trade patterns’. They have highlighted the key role played by endogenous technological improvements in growth. An attempt was also made for the first time to explore the nexus between trade intervention and long run growth. Farole (2013) has also emphasized on the instrumental role played by trade in triggering growth both in the short and long run. From the regional perspective drawing example from

6

1 Introduction

North American Free Trade Agreement (NAFTA) and its impact on the Mexican economy, he stressed that with the enlargement of market accessibility via trade interaction there is a ‘quick and transformative impact on growth’. In this backdrop, Harrison points out that ‘Since theoretical literature does not provide a clear answer, empirical work is needed to help resolve the debate’. Hence, many scholars, like Rodriques (2000), Dollar and Kraay (2004), Sachs and Warner (1995) and Harrison (1995), have analyzed empirical data to measure the impact of trade on growth, using cross-country and panel data regression analysis. In these studies, indicators of openness are regressed on income or growth of income controlling for other important growth variables. Sachs and Warner (1995) have divided countries into closed and open categories on the basis of export policy, black market exchange rate premia, import tariffs etc. They have tried to sketch the process of global integration and its impact on economic growth. It has been pointed out that between 1970 and 1989, on average, per capita income in the open economies grew at a higher rate of over two percentage points relative to the closed ones, and therefore lends support to the assumption that economic reforms trigger higher economic growth. Edwards (1998) has analyzed the relationship between openness and total factor productivity growth. He has pointed out that relatively more open economies have experienced rapid productivity growth and ‘the results are forceful and persuasive’. Similarly, Waugh (2010) has argued how trade frictions are quantitatively significant in comprehending the glaring gap between standards of living and total factor productivity between rich and poor economies. Casacuberta et al. (2004) have analysed the Uruguayan manufacturing sector in terms of the impact of liberalized trade on productivity, labour and capital flows. Though openness has resulted in job creation and increase in productivity, this effect is offset by the unions. It has been suggested that the sectors, having higher tariff reductions and absence of unions, experience multiplying total factor productivity with the changing use of capital and labour. In line with this, Frankel and Romer (1999) have analyzed the correlation between trade and income but could not identify the direction of causation. They have suggested that trade has ‘quantitatively large and robust’ positive impact on income. According to them, changes in trade as a result of policy may not influence income the way it does in case of difference resulting from ‘geography based differences’. Kali et al. (2005) have analyzed the relationship between the trade structure and rate of economic growth. They have pointed out that the structure of trade (in terms of number of trading partners of a country and the trade dispersion among the partners) has significant impact on the economic growth. It has been suggested in the study that there is a positive correlation between the number of trading partners and growth; this effect is greater for ‘rich countries’. On the other hand, dispersion of trade has a negative correlation with growth for the sample of ‘poor countries’. Besides, Shirazi and Manap (2005) have examined export-led hypothesis for five South Asian countries (Bangladesh, India, Nepal, Pakistan and Sri Lanka). They have shown long-run relationship among real output, exports and imports for all the sample countries, except Sri Lanka. In stark contrast, Tang (2006) has tried to work out causality among increasing exports, imports and economic growth for China and concluded that there is no long-run relationship among the three variables.

1.3  Recent Literature on International Trade

7

Sakyi et al. (2015) have investigated the degree of impact of trade openness on the income levels and growth rates in the sample of 115 countries from the developing world, from 1970 to 2009. They have divided the countries into three income groups, namely low-income, lower middle-income and upper middle-income countries, and used non-stationary heterogeneous panel co-integration techniques. Their results indicated that in the long run there is a positive two way relationship between trade openness and income level, implying trade openness is either the cause or consequence of the income level; in the short run, similar results are found. Bhat et al. (2007) have worked out patterns of import intensity in the Indian economy and the manufacturing sector during the 1990s and beyond. According to them, relative to 1993–1994, in 1998–1999 import intensity increased in all broad sectors and also in branches of manufacturing sector. The change in import intensity resulted in a significant impact on the growth of output, employment and export. Liberalization of import is a key factor in India’s development strategy. They further discuss how import is important to produce for export, which is an essential feature of global integration and globalization of production process. However, in India, there are apprehensions about the desired results of the liberalized trade policies, that is whether these policies would result in an increase in demand for import with a corresponding increase in export. The growth of exports depends on several factors, for example capacity of domestic production, global demand, world trade environment, policy regime, along with its competitiveness vis-à-vis other economies. Harrison (1995) has identified few of the limitations of the empirical studies. According to him, many studies make use of different openness measures and methods as well as sample countries which do result in varying conclusions. In addition, there are at times problems with the interpretation of observed correlation between trade policies and growth. He has pointed out, policies, except trade policy, for example macroeconomic policies, have led to a spurt in export as well as in growth. Last but not the least, most of the analysis makes use of ‘cross- sectional averages or starting values for time series data’. Harrison has cited the example of Barro (1995), who has analyzed the effect of price distortions in 1960 on post-1960 gross domestic product (GDP) growth. According to him, such an approach is inappropriate in case of developing countries, because it is not possible to control for unobserved country-specific variations and it also ignores the changes undertaken over the time period for the specific country. Despite the voluminous empirical literature on the positive impact of trade on growth, debate seems far from resolved. According to Rodrik (1995), most of the work on trade and growth has measurement shortcomings. Harrison (1995) has highlighted that methodological problems make it inconvenient to link outcomes with that of the policies; causality tests as well as micro-level analyses led to mixed conclusions. Similarly, among others, studies by Musila and Yiheyis (2015) and Ulasan (2015) do not lend support to the existence of a positive relationship between trade and growth. Another line of argument highlights that there are factors other than trade which accentuate growth. There is voluminous empirical literature on cross-country regressions to explore relationship between growth rates and institutional and political indicators as well as economic policies. Memis and Montes (2006) have pointed

8

1 Introduction

out ‘that the conclusion that increased trade is sufficient for development is controversial. The best that can be said is that there is no consensus on a direct correlation between external integration and development. Rather, development is associated with several different strategies and policies depending on the political, social and economic structure of the country or the region’. In this backdrop, Neuhaus (2005) has observed that besides international trade, capital flow does have a significant effect on growth. Kneller (2002) has tried to analyze whether changes in the policy variables, for example fiscal policy, can nullify the effects of trade liberalization on GDP growth. According to a study conducted for a sample of developing economies, ‘Countries that liberalize their trade regimes do increase their spending on welfare, but once we control for fiscal policy, trade liberalization still has no effect on the rate of growth’. Gemmell (2001) has made an attempt to assess the theoretical and empirical evidence of the influence of fiscal policy (taxes, budget deficits etc.) on the long-run growth. A comparison was made between low-income countries and middle- and high-income (OECD) countries. He has also pointed out that though there was a large number of empirical evidence on the impact of fiscal policy on long run growth, a majority of this suffered from weak methodology, resulting in unreliable results. It has been noted that analysis conducted around late 1990s ignored the government budget constraints for testing the effects of fiscal policy, rendering ‘non-comparable or non-­robust results’. Evidence pertaining to the impact of redistribution (fiscal policy is a part of it) on growth is quite vague. There is less evidence available for low-income countries. Limited evidence available on the impact of fiscal policy on growth highlights that the impact varies in case of less developed countries vis-à-vis Organisation for Economic Co-operation and Development (OECD) countries. According to Gemmell (2001), though interpretations vary, ‘the robust evidence of a negative association between budget deficits and growth is beginning to emerge’. One facet of research relates to trade and diffusion of technology. Mayer (2001) has suggested that technology transfer to developing countries from advanced economies has multiplied significantly over the years. The growth-accounting results reflect that machinery imports and human capital stocks have a positive impact on economic growth. Through endogenous growth model, Guerrieri and Padoan (2006) have tried to show that multiplying technology transfer and business services are closely tied with trade openness. Furthermore, Saggi (2002) points out that trade enhances growth, only when the knowledge spillover is international in scope. Though the empirical evidence available is a bit ambiguous, there are evidences at the aggregate level, hinting at the positive impact of foreign direct investment on the economic growth of the host country. Schneider (2005) has empirically examined the role played by technology-intensive trade, intellectual property rights (IPR) and foreign direct investment in a country’s economic growth and innovation. He concludes that in developed and developing countries, high technology imports help in domestic innovation; and foreign technology has significant impact, vis-à-vis domestic technology, on per capita GDP growth; in developed countries, impact of IPRs on the rate of innovation is high; and last but not the least, results pertaining to foreign direct investment are inconclusive. Perkins and Neumayer (2005) have

1.3  Recent Literature on International Trade

9

examined two main factors causing diffusion of modern technology, viz., latecomer advantage of developing countries, which allows rapid spread of technology in comparison to developed economies, and the openness factor. They have analysed three different technologies across a panel of developed and developing countries. Interestingly, it was found that trade openness has a positive impact on the diffusion of all three technologies. There are several studies in line with such arguments, like Wheeler and Martin (1992), Coe and Helpman (1995), Coe et al. (1997), Gruber (1998), Caselli and Coleman (2001) etc. Similarly, Veermani (2014) has examined that the types of imported capital goods and the sources of their origin are important for growth. He has constructed an index denoted as IMKNOW that measures the knowledge level imbibed in a country’s import of capital goods. The results highlighted that the capital goods ‘exerts stronger impact on growth’ than intermediate manufactures. Furthermore, it was noted that it is significant to ‘look beyond the simple relationship between trade openness and growth.’ On the other hand, Andersson and Ejermo (2006) have traced the relationship between technology and export specialization. According to the study, there is a strong correlation between these two variables across regions. It has been highlighted that comparative advantage can be enhanced via knowledge-building capabilities. In line with this, size and structure of trade flows are largely affected by the technology specialization in origin and destination countries. Recent theories on the relationship between trade and growth have shown varying effects of trade (from none to beneficial and non-beneficial). Bidlingmaieri’s (2007) study found not-so-positive impact of trade on growth. He cited the case of cost advantage, where the ‘first mover’ offsets the gains for the new entrants due to increasing economies of scale; and also disadvantages attached to the multiplying specialization, especially for developing countries. In the latter case, if the country specializes in sectors with low productivity growth or has lower income elasticity of demand, then they will experience low productivity growth vis-à-vis developed economies. As per Redding (1997), it is referred as ‘specialization trap’. According to him, specialization of developing countries in accordance with the existing comparative advantage (in low technology intensive commodities) results in the reduction of welfare, whereas protectionist policies do increase it. The literature on increasing income inequality, unemployment and wage differential is quite vast. Egger et al. (2005) have studied Central and Eastern European countries and figured out the effects of trade openness on regional disparities. Those countries which have embarked on export openness in the period of 1991–1998 have experienced accentuation in regional disparities. As per the analysis, trade in intermediate goods is the major cause. They have estimated 23% rise in the average economy’s variance of real wages because of the rise in the intermediate goods export openness. Silva and Leichenko (2004) have worked out the impact of international trade on inter-regional and intra-regional income disparity prevalent in the United States. It has been suggested that income inequality accentuated from 1992 to 1994. Interestingly, the analysis points out that the regional income inequality cannot be regulated through trade policy. Though trade restrictions do result in higher import prices as well as regional inequality, there are evidences which have

10

1 Introduction

shown that free trade through cheaper exports has also led to higher inequality. The study also reveals that social policy is required to mitigate ‘the inequality enhancing effects of trade’. In a related work, Sjoberg and Sjoholm (2004) have analyzed the spatial concentration of Indonesia’s manufacturing industry over a period. They figured out that even after trade liberalization the spatial concentration has not declined, for instance industries indulging in international trade have experienced relative concentration. Other possible factors affecting the concentration of Indonesian manufacturing industries has also been discussed. Theron et al. (2007) have made an attempt to evaluate the effect of trade liberalization on employment and wages in South Africa. They found that unemployment, poverty and inequality have accentuated after trade liberalization. It has also been noted that the appraisal of the impact of trade liberalization on employment is quite complex, because there are other factors influencing the given sector. Similarly, Helpman et al. (2017) used a heterogeneous firm model of trade and inequality to derive estimates using employer-employee data for Brazil. The results suggested firm differences in wages, and these wage differentials are ‘systematically but imperfectly related to trade participation: exporters on average pay higher wages than non-exporters even after controlling for firm size’. On a similar line, Arbache (2001) said that the effects of trade liberalization on labour markets are quite inexplicable. Openness in Asian countries has resulted in reduced wage inequality, whereas Latin American and other economies have experienced an increase. There have been several attempts undertaken to elucidate this phenomenon, but none of them can be considered as a general theory. The study also highlights, ‘trade liberalization is a necessary, but not a sufficient condition to explain technological modernization and the increase in the stock of capital per capita, which are supposed to shift labour demand in favour of skilled workers thus causing wage inequality’. For example, in several developing economies, except openness, there are also other factors like ‘institutional framework’, human resource and political stability etc. which play an important role in adopting new technologies and attracting foreign capital. Similarly, Bowen Jr. (2006) has also pointed at the alarming rate of job loss due to trade-induced de-industrialization. Additionally, there is a related branch of the literature that has been explored for evidence regarding the differential impact of trade and economic growth on developing and developed countries. Khor (2000) has pointed out that globalization is an uneven process. Almost all countries are greatly affected, though it impacts different categories of countries differently. Even a country having a meagre share in global trade experiences significant effect. The benefits of trade liberalization accruing to developing countries is a highly controversial debate. On the one hand, it is perceived by developing countries as a means to attain growth. On the other, it results in eroding their economies and has ‘marginalized them’. Khor has written, ‘The notion that all are gainers and there are no losers in trade liberalization has proven to be overly simplistic’. Some countries have gained more in comparison to others, whereas there are many, especially developing countries, that have not gained at all ‘but have suffered severe loss to their economic standing’.

1.4  Geographic Perspectives on Globalization

11

In line with the above arguments, Shafaeddin (2005) has also analyzed the economic performance of a sample of developing countries that have embarked on the structural reforms and trade liberalization policies. Mostly East Asian countries have experienced rapid export growth along with the expansion and upgradation of industrial supply capacity. In contrast, growth in mostly African and Latin American countries was far from satisfactory. Wright and Rayment (2004) have argued that new policy orientation and rapid opening of developing countries to foreign trade and capital flows have failed to create an economic environment which supports robust economic growth. Kotilainen and Kaitila (2002) have also analyzed globalization in its various facets, with a major thrust on the economic effect on developing countries. They concluded that maximum gains can only be accrued by solving various development bottlenecks and further liberalizing imports in the developed economies. Nataraj (2007) has put forth that the regional trading agreements (RTAs) have emerged as a linchpin of the world trade. He has analyzed issues in the Doha round with reference to RTAs and emphasized on the rules governing RTAs in the World Trade Organisation (WTO) regime ‘as the key to growth for developing economies like India’. On the same note, MacPhee et al. (2014) have investigated the impact of about 12 RTAs on the intra- as well as extra-regional trade flows in the member developing economies. The results drawn from employing regression did not support the notion of ‘regional integration as a substitute for multilateral trade liberalization’; however, it is also noted that exceptions are always there. Several RTAs considered in the study did not succeed in intra-bloc trade creation. Although, three out of the five African RTAs considered in the analyses did boost intra-bloc trade. It is noted that the difference is attributed to the policy implementation. It is pertinent to point out that trade can be considered as a means to an end, rather an end itself. Despite explicit trends towards increasing trade openness over the last few decades, there are mixed evidences of positive and negative effects of trade and prevalence of methodological limitations to gauge it.

1.4  Geographic Perspectives on Globalization To understand geographical perspectives, it is important to first understand dimensions of space. According to Coe et al. (2007), the definition of space entered into ‘scholarly work’ during the 1950s and the 1960s in the field of economic geography when location theory was kind of dominant. They have discussed and highlighted three dimensions of space, namely territoriality and form, location and flows across space. It is worth understanding location within space and the multidimensional aspect inherent in the spaces of flows. It is important to capture how different countries are positioned in the global spaces of flows and also how spatial unevenness is a crucial component of the world economic system. Further, it was aptly stated that a significant part of a geographical perspective is to understand how economic processes are represented at multiple scales simultaneously, such as global, macro-­ regional (e.g. Association of Southeast Asian Nations (ASEAN), Southern Common

12

1 Introduction

Market (MERCOSUR for its Spanish initials)), national, regional, local and lived spaces. To understand different processes, for example globalization from a geographical perspective, it is important to capture the trends and patterns of the process embedded in the context of space and time. Globalization is a concept that has its roots in the nineteenth century, noticeable in the ideas of Karl Marx. Dicken (2015) said, ‘The “global” is claimed to be natural order, an inevitable state of affairs, in which time-space has been compressed, the “end of geography” has arrived and everywhere is becoming the same. In Friedman’s terms, the world is flat’. On the other hand, it is believed that geography still matters despite the shrinking of space due to technological innovation. In this reference, Harvey (1982) noted that globalization has triggered time-space compression, rendering the relative location within the world economy less important as compared to capital accumulation and place specificities far more vital. Even Massey argued on the similar lines. Likewise, Martin (1999) said, ‘Globalisation may well have eliminated space...... but it has by no means undermined the significance of location, of place’. Sheppard (2006) has very aptly debated the concept of positionality and globalization in the field of economic geography. He has applied the notion of positionality on international trade and staunchly supported and argued that positionality by no means be separated from ‘more territorial aspects of spatiality, such as place and scale’. He further added how interlinkages between places play a role in the emergence of spatial inequalities in the world economy. He has written, ‘positionality also implies that the conditions of possibility for a place to prosper depend not only on the local initiative, as suggested in the industrial districts literature and structural adjustment policy, but just as much on its interdependencies with and dependence on other places’. He tried arguing how positionality not only reflects but moulds the spatiotemporal networks of economic interdependence in the context of trade. In this backdrop of theoretical perspectives on globalization and geography, it is important to shed some light on the emergence of trade studies in geographical literature over the years. The intellectual history of the discipline and the influence of George G.  Chisholm’s contribution are worth noting in this regard. His pioneer work titled the ‘Handbook of commercial geography’ published in 1889 was a benchmark in laying the foundation of economic geography and the relevance of commercial geography. He has discussed the global production of commodities and the favourable geographical conditions for trade. Similarly, J. Russell Smith’s book titled ‘Industrial and commercial geography’ published in 1913 was also well received. He opined that trade and exchange are ‘irresistible impulses’. In these classical texts, as highlighted by Sheppard and Barnes (2005), ‘places are categorized, but only by the commodities they produce and trade, with the purpose of emphasizing geographical connectivity: the global system comes first, its constituent regions defined by commodity production came second...’. However, despite initial contributions in the trade research, which has been instrumental in evolving and institutionalizing economic geography, the issue was not taken up by geographers rigorously across the globe, and somewhere over the years, it got lost among other much researched geographical themes.

1.4  Geographic Perspectives on Globalization

13

On the similar lines, Dicken (2004) has raised the issue why geographers have ‘missed the boat’ by meagre contribution to globalization studies, and he has very explicitly noted the spatial perspectives on international trade as a ‘natural zone for geographers’. He strongly claimed that the absence of geographers in the globalization debate has distorted the geographical perspective on the issue. Andresen (2010) agreed with Dicken and examined the current research on ‘geography and international trade irrespective of its disciplinary affiliation’. In this context, he focussed on three research core areas that have directly or indirectly implied geography, namely geography and the pure theory of trade, the border effect and regional trading blocs, and stated, ‘The role of geography in the pure theory of trade relates the empirical work of economists to some theoretical work on the region by economic geographers, the border effect represents work performed originally by economists but significantly enhanced with a geographical perspective from economists and geographers, and regional trading blocs is work done by geographers’. It is worth noting that geography as a discipline as well as practice has a vital role in sustaining and resisting as well as comprehending the evolving dynamics of the global economy. Hence, it is becoming increasingly important and interesting to study the role of trade in shaping the space economy. However, the geographical literature is limited on international trade. By and large, research by geographers has confined to the case studies. For instance, Silva and Leichenko (2004) worked on the United States, Sjoberg and Sjoholm (2004) on Indonesia’ manufacturing industry, Hughes (2001) focused on Kenyan cut flower industry, Sparke et al. (2004) on Indonesia-­ Malaysia-­Singapore growth triangle, Bowen Jr. (2006) worked on New England, Breau (2007) on Canadian provinces and Horner and Murphy (2017) on Indian pharmaceutical firms. On a slightly different dimension, it is worth noting that there is growing literature on the global production networks and global value chains. In this context, Horner (2016) has worked on the evolving geographies of development, with the specific focus on the South-South trade. He has emphasized that global value chain and global production networks research can be extended to understand the dynamics of trade. There is also some geographical literature on regional trading blocs, as noted by Andresen (2010). For example, Poon and Pandit (1996) made use of principal component analysis to seek statistical evidence for the existence of regional trading blocs, although not establishing the global triad. It is noteworthy that they did find statistical support for six regional trading blocs, namely the United States, Japan, Germany, France, the United Kingdom and USSR (former). Subsequently, Poon has conducted similar studies (Poon, 1997; Poon et.al., 2000). Among others, noted geographical research was undertaken by O’Loughlin and Anselin (1996), Allen et al. (1998), Shin (2002) and Yadav (2012, 2017). Andresen (2010) has noted limitations in the current literature on regional trading blocs. He has written, ‘first is that most measures only incorporate import and export volumes, with little use of economy sizes….but no work so far has established if there is a geography of trade this is separate to that of the geography of production….. Second limitation, one of the premises of the exclusive nature of trading blocs is that certain countries are ‘left out’ of the world economy, decreasing global welfare. However, if every country is

14

1 Introduction

part of a regional trading bloc and trade within and between trading blocs has only increases, there are potentially no losers in the regionalization of the world economy! This is clearly a shortcoming with the methodology employed in the analyses……… Regardless, the explanation of the establishment of regional trading blocs needs to be theorized with geography at the core of that theory. There is little contention that space matters in international trade flows (from both economists and geographers). The primary contention is how space shapes the patterns of international trade’. In this connect, Michalak and Gibb (1997) have pointed out that although economic integration theory and preferential trade agreements have drawn attention from economist and political scientists, geographers have had little to comment on these issues. Commenting on the geographers’ contribution to trade studies, Angel (2002) observes that ‘Geographers have, for the most part, failed to synthesize the results of multiple case studies in ways that provide a broader assessment of global economic change and of the potential for positive engagement in these processes of change. Even when multiple studies are brought together in a single volume, differences in methodology and precise focus mean that the value of the work lies primarily in the individual case studies’. He further points out that one of the factors which has ‘marginalized’ geographers from policy analysis ‘is their failure to bring together case-study research as broader synthetic statements on global economic change.1 This claim raises a series of questions about theory and empirical research’. On the other hand, approach of economists is also limited in scope, missing out the spatial factor. In this context, Pose (2011) aptly highlights, ‘While geographers have been busy trying to detach themselves from geographical space, economists have stumbled upon space almost without realizing it. With very few exceptions— e.g. Hotelling 1929; Koopmans 1957; Myrdal, 1957—mainstream economists had hardly bothered with the spatial dimension of economic interaction. Geography, was reduced to a simple Euclidian space, modeled just as a set of points defined by a series of factor endowments. Endowments and exogenous technical differences determined what was produced where, and productivity in any given activity was assumed to be independent of the scale of the activity. Space was completely exogenous and, thus, inconsequential. Hence, space attracted nothing more than the occasional footnote in economic papers dealing with the location of economic activity. This all changed with the arrival of the ‘new economic geography’ (NEG) starting with Krugman. Under a NEG framework, transport and trade costs are key determinants of the location of activity…’. The World Bank Report (2009), also stressed on trade and its interplay with geography. The report presented a detailed discussion on key spatial concepts, such as density, distance and division, to understand economic geographies of development debate. Further Farole (2013) aptly pointed out that ‘we still lack a clear and detailed understanding of the interaction between trade and economic geography within countries, and in particular of the 1  He has acknowledged the contribution extended by Storper and Salais (1997) in this regard. They have ‘synthesized a broad body of case-study research on the territoriality of economic change through the identification of a set of ideal-type ‘regional worlds’ of production’.

1.5  Database and Analytical Framework

15

factors that determine the relative capacity of different locations to support the competitiveness of export-oriented firms’. On similar lines, Ottaviano (2011) has reflected on the firm heterogeneity with regard to the new, new trade theory. It was said that the majority of studies are centred on the issue of ‘macro-heterogeneity’ across locations and that of “micro-heterogeneity” was completely missed. In this backdrop, an attempt is made through this study to work out a spatial pattern and typology of the countries on the basis of trade pattern and growth. In the backdrop of the above discussion, focus of this book is: • To bring out the changing spatial pattern of the volume of trade over time with particular focus from 1990. The year 1990 is chosen with the Indian focus, since this was the time that India initiated the new economic policies. Several other countries had also initiated reforms during 1980s. • To note the change in the commodity composition of exports and imports in the international trade. • To compare volume of trade and commodity composition of exports and imports within trade bloc and across trade bloc. • To measure the degree of relationship between trade and other economic attributes. Furthermore, while attempting to achieve the above-mentioned objectives, it will in turn examine the following hypotheses: (1) countries have opened up by undertaking reforms in export and import policy in order to increase their trade; however, such expected increase is confined to a few countries; (2) South-South trade has significantly improved; (3) Commodity composition, particularly of the developing countries, has also changed from low value raw material to high value manufactured goods; (4) competitiveness leads to specialization in production. Consequently, one expects an increase in concentration in commodity composition of exports; (5) the process of commodity concentration of exports would lead to diversification of imports.

1.5  Database and Analytical Framework Database: Given the nature of objectives stated above, the analysis undertaken in the current book is based on secondary data. There are differences in definitions and other kinds of discrepancies between different sources of data, each with their own advantages and limitations. The data provided by different sources often do not tally with each other. There are data that are missing for specific time periods (Yadav 2012). Each country brings out trade-related data. For example, in India, the Reserve Bank of India (RBI) brings out an annual data series relating exports and imports and other economic indicators (the Hand Book of Statistics of the Indian Economy). Data provided by RBI basically emanates from the Directorate General of Commercial Intelligence and Statistics (DGCI & S), Ministry of Commerce and Industry, Government of India. The US Census Bureau (Foreign Trade Division,

16

1 Introduction

Data Dissemination Branch) is another such country specific source which provides trade-related data. There are also multilateral sources, like the World Bank, World Trade Organisation, UNCTAD, which disseminate data pertaining to trade and other economic indicators. The multilateral sources render the data more comparable than country-specific databases. To maintain the consistency, much of the data is extracted from UNCTAD. In regard to UNCTAD database, Yadav(2012) has written, “It provides statistics relevant for analyzing, international trade, foreign direct investment, indicators of development and more explicitly for comprehending the economic trends of developing countries over the past decades, particularly in the globalization context (UNCTAD Handbook of statistics, 2005).” The statistics are based on existing national and international data sources. Analysis of time series data span periods as long as available data permit. The export and import structure of individual countries by main regions of origin and destination are presented in percentages. Data are presented for as many individual countries as possible while trade partners are grouped in 13 major clusters according to the UNCTAD secretariat’s judgment as to their relevance for their analysis of the direction of international trade. The commodity groups are defined according to Revision 3 of the Standard International Trade Classification (SITC) of three digit standard’. Gross domestic product data from the UNCTADstat database is used to capture the economic size of various countries. Methodology: The Study Builds on the Existing Methodology Related to Trade Analysis  The first step required is to build data matrices reflecting structure of international trade across countries for which data can be collated for different time periods. In the present study, the time period from 1990 to 2015 is covered and data matrices have been created. This includes trade-related variables, such as value of exports and imports by region, structure of imports by origin and exports by destination, structure of international trade by product, export concentration and import diversification indices, and growth-related variables, such as gross domestic product. The time series data reveals comparability issues for countries (with changes in national boundaries), and therefore the analysis is limited to countries for which time-wise comparable data is available. Methods of analysis are discussed separately in detail in the respective chapters for the convenience of readers. In selected cases, a three-digit commodity classification is used particularly when dealing with concentration and diversification of commodity composition. Among other methods of measuring concentration, location quotient has been used. Data reduction techniques such as factor analysis and regression models for relating the economic structures and trade pattern have been employed. These models are used for both aggregate and disaggregated data. Data analysis is undertaken for different groups of countries over the time period of 1990 to 2015.

References

17

1.6  Presentation of the Book The present book takes the spatial view to understand international trade. The concepts and methods that have already proven important in the study of the systems at other spatial scales are applied in different chapters. The book is organized into six chapters. This introductory chapter includes the background, the theoretical construct of international trade, a brief review of literature, objectives, database and methodology and outline of the rest of the book. In the second chapter, an attempt is made to analyze volume and growth of trade in merchandise and services. Based on inter-regional flow of merchandise export and import, a typology of countries is developed in the third chapter. The changing composition of merchandise export and import is examined in the fourth chapter. In the fifth chapter, the relationship between economy (output) and trade patterns (merchandise trade) has been worked out. The sixth chapter deals with intra-regional trade, with reference to a trade block – MERCOSUR. Finally, in the backdrop of the study, the major observations and future research agenda have been sketched in seventh chapter.

References Allen J, Massey D, Cochrane A (1998) Rethinking the region. Routledge, London Andersson M, Ejermo O (2006) Technology and trade- an analysis of technology specialization and export flows. CIRCLE Electronic Working Paper Series 2006/05. http://www.circle.lu.se Andresen (2010) Geographies of international trade: theory, borders, and regions. Geogr Compass 4(2):94–105. https://doi.org/10.1111/j.1749- 8198.2009.00314.x Angel DP (2002) Studying global economic change. Econ Geogr 78(3):253–255. July Anonymous (2004) Sharpening your competitive edge. International Trade Forum Issue 3/2004. http://www.tradeforum.org. Accessed 1 Aug 2016 Aoyama Y, Murphy JT, Hanson S (2011) Key concepts in economic geography. Sage Publications, London Arbache JS (2001) Trade liberalization and labour markets in developing countries: theory and evidence. http://www.ipea.gov.br. Accessed 15 Sept 2016 Barro RJ (1995) Determinants of economic growth: a cross country empirical study. NBER working paper no 5698. http://www.nber.org. Accessed 25 Aug 2016 Bhat TP, Atulan G, Mahua P, Sahu Partha P (2007) Estimates of import intensity in India’s manufacturing sector: recent trends and dimensions. Working paper series 2007–08. isid.org.in. Accessed 15 Sept 2016 Bidlingmaieri T (2007) International trade and economic growth in developing countries. www. ifw-­kiel.de. Accessed 1 Sept 2016 Bowen JT Jr (2006) The geography of certified trade-induced manufacturing job loss in New England. Prof Geogr 58(3):249–265 Breau S (2007) Income inequality across Canadian provinces in an era of globalization: explaining recent trends. Can Geogr 51(1):72–90 Casacuberta C, Gabriela F, Nestor G (2004) The impact of trade liberalization on employment, capital and productivity dynamics: evidence from Uruguayan manufacturing sector. J Policy Reform 7(4):225–248 Caselli M (2006) On the nature of globalization and its measurement. UNO-CRIS Occasional Papers No. 1

18

1 Introduction

Caselli F, Coleman WJ (2001) Cross-country technology diffusion: the case of computers. Am Econ Rev 91(2):328–325 Chaudhary AM, Saleem M (2003) Comparative advantage of Pakistan’s exports, their complementarity, instability and market diversification. Asian Econ Rev 45(1):103–117 Chisholm M (1968) Geography and economics. G. Bell and Sons, Ltd, London Coe DT, Halpman E (1995) International R&D spillovers. Eur Econ Rev 39:859–887 Coe Neil M, Yeung H W-c (2001) Geographical perspectives on mapping globalisation. J Econ Geogr 1:367–380 Coe DT, Helpman E, Hoffmaister AW (1997) North-south R&D spillovers. Econ J 107:134–149 Coe NM, Kelly PF, Yeung HWC (2007) Economic geography: a contemporary introduction. Blackwell Publishing Ltd, Oxford Dicken P (2004) Geographers and ‘globalization’: (yet) another missed boat? Trans Inst Br Geogr 29:5–26. https://doi.org/10.1111/j.0020-­2754.2004.00111.x Dicken (2015) Peter Global Shift: Mapping the changing contours of the world economy, 7th edn. The Guilford Press, New York Dollar D, Kraay A (2004) Trade, growth and poverty. Econ J 114(493):22–49 Edwards S (1998) Openness, productivity and growth: what do we really know? Econ J 108:383–398 Egger P, Huber P, Pfaffermayr M (2005) A note on export openness and regional wage disparity in Central and Eastern Europe. Ann Reg Sci 39:63–71 Farole T (ed) (2013) The internal geography of trade. World Bank Publications, Washington, DC Fischer S (2003) Globalization and its challenges. Am Econ Rev 93(2):1–30 Frankel JA, Romer D (1999) Does trade cause growth? Am Rev 89(3):379–399 Gemmell N (2001) Fiscal policy in a growth framework. WIDER discussion paper no 2001/84. www.wider.unu.edu. Accessed 30 July 2016 Grant R (1994) The geography of trade. Prog Hum Geogr 18(3):298–312 Grossman GM, Helpman E (1989) Comparative advantage and long run growth. NBER working paper no 2809. http://www.nber.org. Accessed 30 July 2016 Gruber H (1998) The diffusion of innovations in protected industries: the textile industry. Appl Econ 30:77–83 Guerrieri P, Padoan PC (2006) What does the New Lisbon Strategy need? European policies and national policies. Rev Econ Conditions Italy 1:43–58 Harrison A (1995) Openness and growth: a time-series, cross-country analysis for developing countries. J Dev Econ 48:419–447 Harvey D (1982) Deconstructing the map. Basil Blackwell, Oxford Helpman E, Itskhoki O, Muendler M-A, Redding SJ (2017) Trade and inequality: from theory to estimation. Rev Econ Stud 84(1):357–405. https://doi.org/10.1093/restud/rdw025 Horner R (2016) A new economic geography of trade and development? Governing South–South Trade, Value Chains and Production Networks, Territory, Politics, Governance 4(4):400–420. https://doi.org/10.1080/21622671.2015.1073614 Horner R, Murphy JT (2017) South-­North and South-­South production networks: diverging socio-­spatial practices of Indian pharmaceutical firms. Global Netw. https://doi.org/10.1111/ glob.12159 Hughes A (2001) Global commodity networks, ethical trade and governmentality: organizing business responsibility in the Kenyan cut flower industry. Trans Inst Br Geogr 26:390–406 Kali R, Mendez, F, Reyes, J (2005) Trade structure and economic growth. http://editorialexpress. com. Accessed 2 Aug 2016 Kay J (1998) The adaptive capacity of the firm as a key to European competitiveness, November 30. www.johnkay.com. Accessed 2 Aug 2016 Khor M (2000) Globalization and the South: Some critical issues. UNCTAD Discussion Paper No147. http://www.unctad.org. Accessed 2 Aug 2016 Kneller R (2002) No miracles here: trade policy, fiscal policy and economic growth. GEP working paper no 02/26. papers.ssrn.com. Accessed 5 Sept 2016

References

19

Kotilainen, Kaitila (2002) Economic globalization in developing countries, www.elta.fi Krugman P (1995) Growing world trade: causes and consequences. Brookings Papers Econ Acitvity 1:327–377 MacPhee, Craig R, Sattayanuwat (2014) Consequences of regional trade agreements to developing countries. J Econ Integr 29(1):64–94 Martin R (ed) (1999) Money and the space economy. Wiley, New York Mayer J (2001) Technology diffusion, human capital and economic growth in developing countries. UNCTAD discussion paper no 154. http://www.unctad.org. Accessed 1 Nov 2016 Memis E, Montes Manuel F (2006) Assessing RTAs in the context of the flying geese framework. UNO-CRIS occasional paper no. 18. www.cris.unu.edu. Accessed 1 Nov 2016 Michalak W, Gibb R (1997) Trading blocs and multilateralism in the world economy. Annals of the Association of American Geographers 87(2):264–279 Musila JW, Yiheyis Z (2015) The impact of trade openness on growth: the case of Kenya. J Policy Model 37:342–354. https://doi.org/10.1016/j.jpolmod.2014.12.001 Nataraj G (2007) Issues for negotiation under the Doha Work Programme on RTAs: India’s negotiated stand, Economic and Political Weekly, pp. 1966–1974, May Neuhaus M (2005) Opening economies succeed: more trade boosts growth, November 11. www. dbreasearch.com. Accessed 11 Sept 2016 O’Loughlin J, Anselin L (1996) Geo-economic competition and trade bloc formation: United States, German, and Japanese exports, 1968–1992. Econ Geogr 72:131–160 Ottaviano G (2011) New economic geography: firm heterogeneity and agglomeration economies. J Econ Geogr 11(2):231–240 Panayotou T (2000) Globalization and environment. CID working paper no 53. www.cid.harvard. edu. Accessed 11 Sept 2016 Perkins R, Neumayer E (2005) The international diffusion of new technologies: a multitechnology analysis of latecomer advantage and global economic integration. Ann Assoc Am Geogr 95(4):789–808 Poon JP (1997) The cosmopolitanization of trade regions: global trends and implications, 1965–1990. Econ Geogr 73:390–404 Poon J, Pandit K (1996) The geographic structure of cross-national trade flows and regions states. Reg Stud 30:273–285 Poon J, Rigby DL (2017) International trade: the basics. Routledge, New York Poon JPH, Thompson ER, Kelly PF (2000) Myth of a triad? The geography of trade and investment ‘blocs. Trans Inst Br Geogr 25:427–444 Porter ME (1990) The comparative advantage of nations. Maclan, London Pose AR (2011) Economists as geographers and geographers as something else: on the changing conception of distance in geography and economics. J Econ Geogr 11:347–356. https://doi. org/10.1093/jeg/lbq034 Redding S (1997) Dynamic comparative advantage and the welfare effects of trade, November 24. http://www.nuff.ox.ac.uk. Accessed 7 Nov 2016 Rodrik D (1995) Trade policy and industrial policy reform. In: Behrman J, Srinivasan TN (eds) Handbook of development economics, vol Vol 9. Elsevier Science, Oxford/Amsterdam/ New York Rodrique R (2000) Trade policy and economic growth: a skeptic’s guide to cross- national evidence. NBER Working Paper No7081. www.nber.org. Accessed 1 Apr 2017 Sachs JD, Warner A (1995) Economic reform and the process of global integration. Brookings Papers Econ Acitv 1:1–118 Sachs JD, Yang X, Zhang D (1999) Trade pattern and economic development when endogenous and exogenous comparative advantage coexist. CID working paper no 3, March. http://www. cid.harvard.edu. Accessed 2 July 2016 Saggi K (2002) Trade, foreign direct investment, and international technology transfer: a survey. Policy Research working paper no 2349. http://www-­wds.worldbank.org. Accessed 30 July 2016

20

1 Introduction

Sakyi D, Villaverde J, Maza A (2015) Trade openness, income levels, and economic growth: the case of developing countries, 1970–2009. J Int Trade Econ Dev 24(6):860–882. https://doi. org/10.1080/09638199.2014.971422 Schneider PH (2005) International trade, economic growth and intellectual property rights: a panel data study of developed and developing countries. J Dev Econ 78(2):529–547 Shafaeddin SM (2005) Trade liberalization and economic reform in developing countries: structural change or de-industrialization? UNCTAD Discussion Paper No.179. http://www.unctad.org. Sheppard E (2006) Positionality and globalization in economic geography. In: Vertova G (ed) The changing economic geography of globalization. Routledge, London, pp. 45–72 Sheppard E, Barnes TJ (2005) A companion to economic geography. Blackwell Publishing Ltd., Malden Shin ME (2002) Measuring economic globalization: spatial hierarchies and market topologies. Environ Plan A 34:417–428 Shirazi NS, Manap TAA (2005) Export-led growth hypothesis: further econometric evidence from South Asia. Dev Econ XLIII(4):472–488 Silva J, Leichenko A (2004) Regional income inequality and international trade. Econ Geogr 80(3):261–286 Sjoberg O, Sjoholm F (2004) Trade liberalization and the geography of production: agglomeration, concentration, and dispersal in Indonesia’s manufacturing industry. Econ Geogr 80(3):287–310 Sparke M, Sidaway JD, Bunnell T, Grundy-Warr C (2004) Triangulating the borderless world: geographies of power in the Indonesia-Malaysia-Singapore growth triangle. Trans Inst Br Geogr 29:485–498 Stroper M, Salais R (1997) Worlds of production: the action frameworks of the economy. Harvard University Press, Cambridge MA Tang TC (2006) New evidence on export expansion, economic growth and causality in China. Appl Econ Lett 13(12):801–803 Theron J, Shane G, Margareet V (2007) Globalization, the impact of trade liberalization, and labour law: The case of South Africa. www.ilo.org. Accessed 4 Nov 2016 Thoman RS, Conkling EC (1967) Geography of international trade. Prentice-Hall, Inc, Englewood Cliffs Ulaşan B (2015) Trade openness and economic growth: panel evidence. Appl Econ Lett 22:163–167. https://doi.org/10.1080/13504851.2014.931914 Veermani C (2014) World’s knowledge spillovers: beyond openness and growth. J Econ Integr 29(2):298–328 Waugh ME (2010) International trade and income differences. Am Econ Rev 100(5):2093–2124 Wheeler D, Martin P (1992) Prices, policies and the international diffusion of clean technology: the case of wood pulp production. World Bank Discussion Papers World Bank (2009) World development report 2009: reshaping economic geography. World Bank, Washington, DC Wright R, Rayment P (2004) Globalisation reloaded: An UNCTAD perspective. UNCTAD Discussion Paper No.167. http://www.unctad.org, January Yadav P (2012) India’s changing trade pattern in the process of globalization. Procedia Soc Behav Sci 37:157–166 Yadav P (2017) Core and periphery: an analysis of the spatial patterns of international trade. Paper presented at the RSA annual conference 2017 in Dublin, 4–7 June, 2017

Chapter 2

Volume and Growth of Trade: Merchandise and Services

Abstract  International trade is an integral feature of the globalization process. Over the years, governments in most countries have increasingly opened their economies to foreign trade, either via the multilateral trading system or as part of their domestic reform programmes or increased regional cooperation and integration. The gaining grounds of free trade in proliferating economic growth and development across the globe is much talked about; however, empirical studies have noted the effects are heterogeneous across space and sectors. The present chapter analyses the trends and patterns of merchandise and service trade over the period 1990–2015 for the sample of 87 countries spread across different geographical regions. In addition, the chapter addresses the issue related to the conceptualization of service sector and also marginalization of the services in the globalization process despite being an important driving force in the changing structural landscape of the world economy. Keywords  Merchandise trade · Service trade · Developing countries · China

2.1  Introduction International trade is an integral part of the globalization process. Over the years, governments in most countries have increasingly opened their economies to international trade, whether through the multilateral trading system or as part of their domestic reform programmes or increased regional cooperation and integration. According to the WTO (2008), post World War II, international trade has experienced a long period of expansion, with global merchandise export increasing by more than 8% per annum in real terms over the period 1950–1973. Thereafter, trade growth slowed because of the two oil price shocks, inflation caused by monetary expansion and inadequate macroeconomic adjustment policies. In the 1990s, trade expanded again more rapidly, partly driven by innovations in the information technology (IT) sector. In 2001, despite a little contraction of trade due to the dot-com © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9_2

21

22

2  Volume and Growth of Trade: Merchandise and Services

crisis,1 the average expansion of world merchandise export continued to be high (according to the WTO, average of 6% over the period 2000–2007). The quantum of international trade, as per WTO press release (2016), was noted to have increased to 2.8% in 2016 and expected to surge to 3.6% in 2017, although still lower than the average of 5% since the 1990s. According to the WTO (2008), the most dynamic traders in the period 1950–1973 were the west European countries and Japan. Post–World War II reconstruction and the Korean War provided a major stimulus to Japanese and European exports in the early 1950s. From the early 1960s onwards, the six Asian newly industrialized economies (NIE) followed an outward-­oriented trade policy and succeeded in sharply increasing their merchandise exports. The dominant share of the United States in world trade in the early 1950s was eroded in the subsequent decades, and the share of regions (the United States and Canada) in world merchandise exports varied, largely due to the fluctuations of commodity prices and exchange rates. The oil-exporting developing countries (especially those in the Middle East) increased their share between 1973 and 1983 but lost almost all their gains when oil prices fell back thereafter. In 1993, after the disintegration of the Soviet Union and the fall of the Council of Mutual Economic Assistance (CMEA), industrial countries’ (i.e. Western Europe, North America and Japan) share of world merchandise exports reached a peak, in excess of 70%. Together with the six NIEs, they accounted for more than 80% of world trade in 1993. Over the 1990s, share of Japan in world’s export started to drop due to competition posed by the NIEs and China (WTO 2008). In 1994, creation of the North American Free Trade Agreement (NAFTA) was not sufficient enough to reverse the downward trend in the share of the United States and Canada in world trade. Similarly, the integration process of Europe continued to strengthen and expanded to cover the central European economies and the Baltic states but could not check the relative decline of European export (WTO 2008). The declining share of industrial economies can be because of the rise of China, the resurgence of the Commonwealth of Independent States (CIS) and recently the boom in commodity prices, which multiplied the shares of Africa, the Middle East and Central and South America, particularly regions exporting principally minerals and other primary products.

1  The dot-com bubble refers to the speculative bubble that took place in the ITC kind of industries during the late 1990s until its end in 2000/2001. Developed countries saw an increase in their total value and a very fast growth of GDP due to the mushrooming of technology companies at speculative rates. It was a speculative bubble broadly covering 1995–2000 with a huge increase on the Nasdaq stock market in industrialized developed countries. Nasdaq gives high valuations for many enterprises which have low or no profits, and fell nearly 50% from its peak in the second half of 2000. The bubble was collapsing rapidly by 2001. A majority of the dot-coms stopped trading and several have never made any net gain. Often investors called these disastrous dot-coms as ‘dotbombs’ (Maddison 2006)

2.3  Trends in Merchandise Trade

23

2.2  Database and Analytical Techniques The chapter attempts to look at the growth of merchandise trade of 87 countries (See Annexure 2.2). Initially data pertaining to the value of merchandise export and merchandise import for 208 countries was collated. And those countries having per cent share to total merchandise trade less than 0.01% in 1990 are dropped from the purview of the current book. In addition to this, only those economies having consistent data for 1990, 1995, 2000, 2005, 2010 and 2015 are retained. It is important to note at the outset that economies like Belgium-­Luxembourg, Czechoslovakia (former), the USSR (former) and Yugoslavia (former) are considered as undivided economies, therefore, post-break-up countries, for which data are available, are clubbed together for consistency through the years. For country groupings, see Annexure 2.1. The increasingly important area of trade in services is also discussed later in the chapter. Data pertaining to merchandise and service trade (measured in US dollars at current prices in millions) is taken from UNCTADstat. It is important to mention here that the analysis of international service trade is difficult because of lack of comprehensive and internationally comparable data (Ocampo and Vos 2008). Further, it is pertinent to note here that 15–20 years of cycle is not long enough to trace major changes in the sudden jump of the growth. Merchandise trade is calculated by adding merchandise export and import. Simple growth rate is calculated for 1990–1995 and 2010–2015. Compound annual growth rate (CAGR)2 for 1990–1995, 2000–2005 and 2010–2015 is calculated. Further, four classes are identified to compare simple growth rate during 1990–2015. For the ease of comparability, the classes were kept constant for 1990–1995 and 2001–2015, considering the former as a base period. Similarly, the results of CAGR are summarized by considering class groups for 1990–1995 as the base for subsequent time periods. Growth rates were calculated with both the methods of simple and compound annual growth rate for the mere purpose of checking the difference in results. Due to the absence of significant variation in results of the two methods, CAGR is considered for detailed analysis in this chapter. Furthermore, CAGR is also calculated for service trade (export of services and import of services) for three time periods for each country: 1990–1995, 2000–2005 and 2010–2013.

2.3  Trends in Merchandise Trade Merchandise trade grew rapidly over the past 50 years. During 1962–2000, the value of world merchandise trade grew at an annual average rate of about 10% and its volume by 6% (Ocampo and Vos 2008). According to Ocampo and Vos (2008), ‘Although, developed countries dominate all non-oil global markets, developing countries have 2  CAGR is an average growth rate over a period of several years. It is a geometric average of annual growth rates.

24

2  Volume and Growth of Trade: Merchandise and Services

rapidly expanded their participation, especially since the second half of the 1980s. More importantly, there was a significant shift in the structure of exports by developing countries (as a group) away from primary products towards manufactures’. Despite the ups and downs in the growth rate of world trade, the overall volume of trade continues to grow. Various factors in the global economy cause fluctuations in the growth rate of world trade, which span from world oil prices to changes in foreign exchange rates to geopolitical issues. Figure 2.1 displays value of merchandise trade, merchandise export and merchandise import for six points of time, namely 1990, 1995, 2000, 2005, 2010 and 2015. It is quite explicit from the figure that international trade in merchandise has certainly increased considerably since 1990. In 1990, the value of total merchandise trade was $7105 billion (in current US dollars) and $ 33,159 billion in 2015. The expansion of world trade is characterized by three significant trends. First, rise of emerging market economies as important trading partners; second, multiplying significance of regional trade; and third, the shift of higher technology exports toward dynamic emerging economies (IMF 2011). Similarly, trends in the value of merchandise export and import reflect consistent increase over the years (Fig. 2.1). In 1990, around 73% of world merchandise export was from developed economies3 (as a group). However, despite experiencing consistent decline in their share

Fig. 2.1  Trend in the value of world merchandise trade. (Source: Based on Annexure 2.3) 3  According to UNCTAD Handbook of Statistics 2006–07 countries are grouped into following major categories. Developed economies cover geographical regions like America, Asia, Europe and Oceania. Economies in transition comprises of countries forming the Commonwealth of Independent States. Developing economies include all countries and territories in America, Africa, Asia and Oceania not specified above (see Annexure 2.1 for detail country classification).

2.3  Trends in Merchandise Trade

25

Table 2.1  Share of merchandise export (shares in percent by country group)

World (in US billion dollars) Developing economies Economies in transition Developed economies World (in US billion dollars) Developing economies Economies in transition Developed economies

1980 1990 1995 Merchandise export 2050 3496 5169 29.67 24.12 27.58 4.17 3.39 2.62 66.16 72.50 69.80 Merchandise import 2091 3609 5226 24.03 22.16 28.60 4.00 3.88 2.35 71.97 73.96 69.05

2000

2005

2010

2015

6452 31.92 2.32 65.76

10,502 36.26 3.37 60.37

15,302 42.08 3.98 53.94

16,552 44.78 3.18 52.05

6655 28.82 1.38 69.80

10,778 31.77 2.23 66.00

15,421 39.04 2.94 58.02

16,607 41.98 2.30 55.72

Source: Calculated on the basis of data from the UNCTAD Handbook of Statistics, 2006–2007 and 2016

over the years, they still account for a dominant share in world merchandise trade (Table 2.1). Over the six points of time, participation of developing countries in the total world export has increased. A similar pattern could be seen in the case of merchandise import. An important feature of international trade over the years has been the growing participation of developing economies. Between 1990–1995 and 2010–2015, their merchandise export and import grew much more rapidly than the world average. (Figs. 2.2 and 2.3). However, it is interesting to note that the growth rate of trade in developing economies as compared to developed economies has been higher. As a result, their share has also increased over the years, as stated earlier. There have been sharp swings in the growth of international trade flows over the 1990s across country groups (Fig. 2.2 and 2.3). During the 1990s, world economy experienced a deep recession, and rapid recovery in emerging markets, together with the diverse movements of commodity prices, including oil, led to such fluctuations. However, the resilience of the US economy, a pick-up in economic activity in the EU and Japan, stronger than expected recovery in Latin America and the transition economies and sustained growth in Asia all helped to stimulate global trade. The prolonged period of boom in the United States has also left its mark on the global trading system (UNCTAD 2002). The sluggish growth of merchandise trade around 2015 was due to a combination of multiple factors, such as Chinese economic slowdown, recession in Brazil, falling prices for oil and other commodities and exchange rate volatility. It is possible to discern that over the past 25 years, global trade has been visibly affected by several factors, for example information technology advancement, financial crisis, multiplying number of the WTO members, natural disasters and, last but not the least, geopolitical relations. According to the WTO (2015), these factors have led to fluctuation in commodity prices, marked changes in the key traders and their partners and also growing significance of services in international exchange. Post 1995, trade has been considered as a key factor in helping to increase economic growth that was in line with the gains from free trade argument, which goes back to classical trade theories.

26

2  Volume and Growth of Trade: Merchandise and Services

Fig. 2.2  Growth of merchandise export. (Source: Based on Annexure 2.4 (UNCTAD Handbook of Statistics 2016))

Fig. 2.3  Growth of merchandise import. (Source: Based on Annexure 2.4)

2.3  Trends in Merchandise Trade

27

Table 2.2 shows top five players in international merchandise trade. The United States is the largest economy of the world, with a GDP of about 18 trillion US dollars in 2015. It is quite clear from the table that the United States is also the largest trading nation of the world over the period of 20 years; however, there is a decline in its share post 2005. As a major developed economy, its import has relied on raw materials and export on processed manufactured goods. The United States has also emerged as a key export market for a number of economies across the globe. In 2015, China with 12% share surpassed the United States and secured the top trading nation’s position. Table 2.2. captures the emergence of Chinese economy among the traditional top five traders of the world. Besides, Germany, Japan and the United Kingdom are other dominant traditional trading cores. The top five players reflect dominance of the developed North in the global commodity trade; however, there has been a decline in their relative share over the years. Also rise of developing countries such as China reflect upon the gaining significance of the global South and project upon the shifting power relations in the years to come. Hence, it will be interesting to capture the changing underlying dynamics of the evolving power relations in the global trade system. It will also help us in further comprehending the relative change in the relative locational utility of the new cores and the old ones in the emerging global economy. Figures 2.4 and 2.6 depict that over the period 1990–2015, economically diverse countries exhibit differential rate of growth in the merchandise trade. It is worth noting that the number of countries experiencing growth above global average has reduced from thirty one to twelve in the span of 25 years (1990–2015). Merchandise trade of the developing countries spanning geographically diverse regions have reported very high and high growth (see Figs. 2.4 and 2.5). It is also to be noted that the developing economies have grown faster than the developed ones between 1990–1995 and 2000–2005. Out of 12 countries registering fastest growth in trade, 6 were from East Asia and 5 from Central and South America; during 2000–2005, Table 2.2 Top five players in international merchandise trade (percent share to world merchandise trade) Country United States Germany Japan France United Kingdom

1990 1995 2000 Country 12.82 13.02 15.57 United States 10.93 9.48 7.98 Germany 7.36 6.45 5.76

7.48 5.75 4.85

6.55 China 5.07 Japan 4.82 France

Source: Calculation based on data UNCTADstat

2005 Country 12.38 United States 8.21 China 6.68 Germany 5.22 Japan 4.55 France

2010 Country 10.57 China 9.68 United States 7.53 Germany 4.76 Japan 3.69 United Kingdom

2015 11.93 11.50 7.18 3.84 3.28

28

2  Volume and Growth of Trade: Merchandise and Services

merchandise trade of about 29 countries grew at a very high rate. The reason for robust international trade in the latter case is primarily due to increased demand and rising prices of commodities. Besides developing America and Asia, African economies such as Angola, Nigeria and Zambia have shown impressive growth during

Fig. 2.4  Growth of merchandise trade, 1990–1995. (Source: Based on Annexure 2.5)

Fig. 2.5  Growth of merchandise trade, 2000–2005. (Source: Based on Annexure 2.5)

2.3  Trends in Merchandise Trade

29

Fig. 2.6 Figure 2.5: Growth of merchandise trade, 2010–2015. (Source: Based on Annexure 2.5)

2000–2005. This is attributed to the structural reforms and increase in the oil prices. Of the 29 fastest growing developing economies in terms of merchandise trade, 10 are from Asia. According to UNCTAD (2005), since 2002, strong demand from East and South Asia, particularly China and India, has been the major reason for pushing up the commodity prices. Demand from fast growing Asian economies for commodities such as fuels, minerals (iron ore, nickel and copper), natural rubber and soybeans is expected to remain strong, further benefitting their exporters. Although, future developments in the primary commodity market will also rely on additional supply capacity and how demand from the global North of the commodity will be influenced by the need to correct the exiting trade imbalances. It is worth noting that although rapidly growing developing countries are significant for global commodity demand, developed economies still account for ‘two-thirds of global non-fuel commodity imports, will continue to play an important role’. Among the developed economies, in 1990–1995, only Czechoslovakia (former) registered very high growth rate in trade. During 2000–2005, besides Czechoslovakia (former), Bulgaria, Hungary, Poland, Romania and Yugoslavia (former) also recorded very high growth of international trade. It is noteworthy that, between 2010 and 2015, Vietnam is the only economy during the three considered time period with very high growth in trade. According to the OECD, Vietnam’s economic performance in the last two decades is considered one of the most impressive in the global South, and this is largely due to successful implementation of the market reforms and global integration. Though the country did face trade-related bottlenecks. It is quite evident from Fig. 2.5 that during 2000–2005, international commodity trade registered a spectacular performance as compared to 1990–1995, although 2010–2015 trade growth declined across countries. During 2000–2005, the process

30

2  Volume and Growth of Trade: Merchandise and Services

of recovery strengthened, and this positive growth spread across the world, though at different rates (Fig. 2.5). Countries such as Angola, Algeria, Iran, Zambia and the USSR (former) experienced negative growth in the period 1990–1995; however, trade grew rapidly over 2000–2005. During the same period, it can also be seen that oil-producing economies experienced vigorous trade growth due to rising petroleum prices coupled with structural reforms. Major oil exporter countries like Saudi Arabia, the UAE, Kuwait, Nigeria, Angola, Algeria and Trinidad and Tobago (a large exporter of liquefied natural gas) experienced very high growth (Fig. 2.5). This increase was attributed to strong world demand, particularly on the part of the United States and other developed countries, and also the rise of emerging developing economies like China and India as key drivers of growth in the global economy, and their trade (particularly China) has exhibited robust growth post the 1990s. The growth of developing economies trade over 1990–2005 was intermittently halted by events such as the Asian financial crisis of 1997, the bursting of the IT bubble in late 2000 and 9/11 in 2001, but an overall consistent increasing trend has been observed (WTO 2007). Unlike 1990–1995, during 2000–2005, Southeastern and Western European economies like Bulgaria, Yugoslavia (former), Austria, Belgium-Luxembourg, Denmark, Germany, Netherlands, Norway and Spain registered a high growth rate in international merchandise trade from moderate and low growth (Refer Figs. 2.4 and 2.5). In 2000, the upturn in trade was expected to be particularly strong in Western Europe (UNCTAD 2000). Similarly, international trade of other western European economies like France, Italy, Portugal, Sweden and Switzerland grew moderately over 2000–2005. Bulgaria, Czechoslovakia (former), Hungary, Poland, Romania, the United Kingdom, Yugoslavia (former) and USSR (former) are other economies exhibiting robust growth over 2000–05. An underlying reason might be that among the transition economies, the strengthening import of Western Europe gave a major fillip to the export of Eastern and Central European economies. However, in the context of the USSR (former), trade got majorly affected by the mounting geopolitical tensions in the later years and also due to the global oil market. Besides, the United States’ trade continued to grow moderately from 1990–1995 to 2000–2005 and at a sluggish rate during 2010–2015. In this context, UNCTAD (2006) has pointed out, ‘growth in the volume of import decelerated in the USA as the economy slowed with the maturing of the economic cycle in 2005. Nonetheless, the merchandise trade balance of the United States recorded another record deficit despite faster export volume growth. The rising deficit is explained both by higher oil prices and an increase in the deficit in the non-petroleum trade balance’. Similar to the United States, Japan’s merchandise trade grew moderately from 1990–1995 to 2000–2005; thereafter, it recorded a negative growth rate. In this regard, it is pertinent to throw some light on the economic history of Japan, as it could partially explain such slackening in trade growth. UNCTAD (2006) has highlighted that Japan is the only large economy to experience three ‘fully-fledged recessions’ since the 1990s. The first big blow Japanese economy faced was due to restrictive monetary policy somewhere in 1991 and a huge real appreciation of the

2.3  Trends in Merchandise Trade

31

yen, and the Asian financial crisis in 1997/98 triggered the second long recession. In 1999 and the first half of 2000, there were signs of recovery, which was short-lived. Since the second quarter of 2001, Japanese economy was back in deep recession. During this period, problems cropped up at major firms which were facing recession and deflation, resulting in multiplying bankruptcies and record losses of companies, and also the banking system experienced further deterioration in the quality of its assets. Japan was further adversely hit by the sluggish growth in the US economy. As a consequence, not only overall growth fell, but also exports and private investment declined at a faster rate. International trade in commodities of East and Southeast Asian economies, such as, China, Vietnam, Malaysia, Philippines, Singapore, Thailand, South Korea and Indonesia, grew rapidly over 1990–1995. However, during 2000–2005, commodity trade of China, South Korea, Vietnam and Thailand grew rapidly, whereas Indonesia, Malaysia and Singapore registered moderate growth in their trade; Philippine trade grew sluggishly during the period. The country-specific challenges, policies and growth pattern differ significantly across East and South-east Asian countries in the considered time period (see Figs. 2.4, 2.5 and 2.6). In this context, it is worth noting that the Asian financial crisis (1997) adversely affected economies across the globe. The crisis was triggered by the depreciation of Thai currency, viz., baht, in July, 1997. It led to severe currency depreciations and an economic recession. It affected much of Asia. In this regard, Hunter et al. (1999) noted, ‘When crisis broke out with the collapse of the Thai baht in July 1997, many believed the damage would be confined to a handful of small Asian economies. Some fourteen months later, it’s clear how myopic that thinking was. The crisis has spread to other East Asian countries and to Russia. Just a few weeks ago, Latin America began to encounter serious troubles’. As it could be seen from Figs. 2.4 and 2.5, compared to developed economies, merchandise trade of developing economies grew rapidly over the years. Between 2000 and 2005, the global macroeconomic environment has been characterized by a slowdown followed by a strong recovery, and this is reflected in the trade performance of developing economy (WTO 2007). Such rapid growth across developing Asia, Latin America and Africa could be attributed to several factors, such as higher prices for primary commodities like fuels and mining products, the continuing export led growth in China and global economy rebounding from the IT crisis that began in 2000 (WTO 2007). Another WTO report (2006) puts to the fore, the sharp rise in prices and traded volumes of many primary commodities has often been considered a major factor explaining the relative strength of economies and commodity groups in international trade. A mark example of this is, of course, export growth of net oil exporters. The sharp rise in net oil imports of China, the United States and India since 2000 had been a major factor causing expansion of oil trade and also the increase in oil prices. In addition, UNCTAD (2007) noted, ‘Since 2002, there has been a “commodity boom.” International commodity prices showed a strong rising trend after their sharp fall in 1995–1997–2002. UNCTAD’s commodity price index (including fuels) in current US dollar terms has risen by 96%

32

2  Volume and Growth of Trade: Merchandise and Services

since 2002. The rise in prices has been driven by the boom in the prices of metals and minerals, and also of crude oil. The relative importance of factors behind the price rise differs from commodity to commodity. However, there are some common factors. These include the strong growth in import demand of developing countries that has given additional market opportunities, owing to the rapid pace of industrialization, especially in China, India and other emerging developing countries; the increased production of bio-fuels which mainly affect the markets for food products by pushing up the price of land and adding to effective demand for some products that are used both for food and for bio-fuel (for instance, sugar cane and maize); as well as emerging supply constraints in some commodity markets’. The multiplying consumer purchasing power in the developing Asia has also triggered demand for some agro-food commodities, for example coffee, tea and cocoa, which had been facing stagnant or falling demand in the traditional OECD highincome markets. This additional source of demand has contributed to a general recovery of commodity prices and improved prospects for commodity-producing countries. Sustained growth in the United States and economic recovery in Japan and Europe have also been contributing factors (UNCTAD 2007). Having discussed this, it is important to mention here that it is difficult to make distinct conclusions about the trade patterns of individual developing economies, and goes beyond the ambit of the current chapter. However, in subsequent chapters an attempt is made for better understanding of the dynamics of merchandise trade for individual economy considering the composition of merchandise trade in terms of major commodity groups. Broadly at the macroscale, global merchandise trade grew moderately between 1990 and 1995 and exhibited strong growth during 2000–2005, but it was rather sluggish between 2010 and 2015. The impact of the great recession in the global economy during the late 2000s and early 2010 has badly affected international market; however, the scale varies from country to country. In this context, the WTO reported (2014) that the period between 2009 and 2013 is referred to as ‘great trade collapse’. Fig. 2.6 depict dampened growth in trade across different countries from both the global South and global North (See Fig. 2.6). For example, oil exporters in Central and South America, Asia and Africa were adversely affected by an increased global supply of fuel, and the subsequent slash in the oil prices has further stalled growth in these regions. Also the degree of growth is uneven across the developing regions of Africa, developing Asia, Latin America and the Middle East primarily because of political instability, dependence on merchandise exports and also volatility of commodity prices in the global market.

2.4  Growth of Service Trade

33

2.4  Growth of Service Trade Services are ubiquitous in the world economy and are increasingly becoming a major driving force of the process. Coe and Yeung (2015) highlighted the crucial role services play as intermediaries in the world economy, and yet the empirical research has not sufficiently engaged with services and continues to be ‘production bias. On the similar lines, Beerepoot et al. (2017) has written that a new geography of services production has emerged, i.e. “a geography that is defined by new interregional and international divisions of labour and held together by increasingly complex global services production networks. The geographical changes are profound....” and it was further noted that the modern information and communication technologies (ICTs) and liberalisation of trade have boosted “especially producer oriented services to be traded across space and a much more varied location pattern.” Services are increasingly becoming vital for both developed and developing economies across globe. In this context, Grossman and Rossi-Hansberg (2008) has aptly added that the emerging trend in the global manufacturing, that are disintegrated functionally and geographically spread over distant places across developed and developing world, are tied via “trade in tasks”’. Recently, Kleibert (2017) said, ‘services were long believed to be untradeable across national borders, their production and consumption bound in a single place (and often in a single moment in time). Innovations in information and telecommunications technology (ICT) have disrupted this setting’. Therefore, an important question arises: what is trade in services? The typical examples may include communication technologies, transportation such as aviation and shipping and financial and banking services. However, service trade encompasses a heterogeneous category, for example engineering, construction, legal, healthcare, media, consulting and distribution, among others. In a comprehensive account of globalization, Friedman (2006) stated that technological advancement enabled the delivery of services from across what has according to him become a ‘flat’ world, and others referred this ICT-triggered globalization of services as a ‘tradability revolution’, ‘the next industrial revolution’ or a ‘next wave in globalisation’(UNCTAD 2004, Dossani and Kenney 2007; Blinder 2006). According to Cattaneo et al. (2010), ‘Technological progress has essentially transformed business practices, reducing cost, increasing the speed, improving the quality, and expanding the range of available services that now be traded across borders. Consequently, trade in services has now expanded both in length and breadth-to encompass more professions and in geographical reach, made possible through business process outsourcing and off-shoring practices’. Largely, potential of trade in services has remained untapped by developing economies, which might be due to the common belief that relative to commodities, services are non-tradable (Cattaneo et al. 2010), though in recent decades technological obstacles to service trade have been removed over the years. Although the existing technological gap between developed and developing economies did support the belief that as compared to relatively poor countries, service trade is more pronounced in advanced economies.

34

2  Volume and Growth of Trade: Merchandise and Services

However, recent research in this field negates such beliefs. In line with this, Mattoo and Payton (2007) argued that any economy, including least developed, could engage in service trade and gain from expanding open market. Moreover, potential vary across space and time. Also benefit might get overshadowed by associated costs. Often success relies on complementary reforms that could become too costly in countries with limited capital and human resources. Loungani et al. (2017) have discussed the instrumental role of technological advancement in boosting service trade, and also cheap and fast transportation add to the advantages of service export over commodity export. There is growing research on the impact of service-related policies, the role of service trade in the process of economic development and economic integration via trade and inclusiveness. Interestingly, the question coming to the fore is ‘Are the drivers of growth and development shifting away from manufacturing into services?’ However, it is too early to provide a definite answer to it. The percent share of service trade in global trade has experienced a visible spurt over the years. While merchandise trade still constitutes the bulk of total global trade, the share of services export has increased from about 3% in 1970 to around 23% in 2015. Similarly, the share of service trade to GDP hovered around 7% in 1980 and increased to 13% in 2015. It is worth mentioning that the services present a growing opportunity for diversifying export. Service export can be linked with growth; however, it is more pronounced in the case of developed economies, mainly due to their preponderance in the most dynamic sectors of the global export market for services. According to UNCTAD, international trade in services4 expanded across the globe at a very rapid pace in the late twentieth century, growing on average much faster than both the world GDP and world merchandise trade. However, this general picture covers a large heterogeneity at the country level. It is noteworthy that the interest regarding services is largely tied around its impact on growth, productivity and employment, although as far as trade in services in the world economy is concerned there is very little known to us (Loungani et al. 2017). Figure 2.7 shows positive trend in world’s export, import and trade in services. UNCTAD (2018) has provided a comprehensive discussion on the factors triggering the increasing share of services in the global economy. Major among them include: (i) rise in the consumer demand; (ii) liberalization of trade and technological advancement particularly in ICT led to geographical fragmentation of manufacturing processes, outsourcing and offshoring, which has resulted in global value chains(GVCs). GVC is quite visible across industries and also includes a 4  Services are defined as the economic output of intangible commodities that may be produced, transferred and consumed at the same time. However, services cover a heterogeneous range of intangible products and activities that are difficult to capture within a single definition and are sometimes hard to separate from goods. Services are outputs produced to order, and they typically include changes in the condition of the consumers realized through the activities of the producers at the demand of customers. Ownership rights over services cannot be established. By the time production of a service is completed, it must have been provided to a consumer. Given the general difficulties in statistically capturing certain aspects of trade in services, the figures presented here may be downward biased with regard to the actual flows of exports and imports of services (UNCTADSTAT).

2.4  Growth of Service Trade

35

Fig. 2.7  Trade in services. (Source: Based on Annexure 2.6) Table 2.3  Share of service trade (World export and import of service in billions of dollars and share in percent) Year Export World Developing economies Transition economies Developed economies Import World Developing economies Transition economies Developed economies

1990

1995

2000

2005

2010

2015

831.35 18.09 2.01 79.90

1222.22 22.36 1.35 76.29

1521.98 23.13 1.33 75.54

2573.22 24.51 1.85 73.65

3896.26 28.49 2.33 69.18

4826.03 31.04 2.22 66.74

875.19 22.14 3.49 74.37

1240.94 27.03 2.09 70.88

1519.39 27.41 1.75 70.84

2472.36 28.37 2.63 69.00

3739.25 34.86 3.20 61.94

4729.46 39.20 3.07 57.73

Note: ‘Percentage of the world total’ represents a country’s (group’s) trade in services as a percentage of the world total trade in services (calculated from Source: UNCTAD Handbook of Statistics 2016)

multiplying number of developing countries; (iii) growing significant role of services as providers of intermediate inputs; (iv) demographic changes, for example ageing has also affected the demand for services across the globe; (v) and slower productivity growth in services as compared to the manufacturing sector, as put

36

2  Volume and Growth of Trade: Merchandise and Services

forth by Baumol(1967), is the key reason for the increasing employment in the former. Geographical distribution of service trade is shown in Table  2.3. International trade in services, as could be seen in Table 2.3, is dominated by developed economies which account for the bulk share of export and import. For example, in 1990 and 2015, developed economies accounted for about 80% and 67% of the world export of services, whereas the proportional share is mere 18% and 31%, respectively, in the case of developing economies. The corresponding figures for service import of developed and developing countries in 1990 are 74% and 22%, whereas in 2015 their share stood at 58% and 39%, respectively. Similar to commodity trade, the participation of developing economies in service export and import has grown, and this is primarily driven by Asia. Unlike developing countries, developed economies constantly run a surplus in their balance of trade in services. In this context, it is interesting to note that in contrast to developing countries, several developed nations have compensated their trade deficits in merchandise trade with surpluses in service trade. The trend in trade deficit could be explained with the help of the product cycle theory.5 Although there are examples from the developing economies that have managed to improve their trade balance in services (UNCTAD 2018). Figures 2.9 and 2.10 show growth rate of the world service trade and also across aggregate country categories. It is quite explicit that the global export and import grew at a fluctuating rate over the years, which is closely tied with the ups and downs in the global economy as well as country-specific crisis. However, it is noteworthy that the service export and import of transition and developing economies grew faster as compared to the developed countries. The disaggregate picture, that is the country-level data given in Table 2.4, shows the top five players in the service trade over the years. It is clearly seen that over the period of past two decades, the United States and Germany retained their first and second position, respectively, in terms of the percent share of service trade to world service trade. Similar to merchandise trade, China has emerged among the top five cores in the service trade from 2010 onwards; Japan has been replaced by China again in the service trade post 2010. Figures 2.11, 2.12 and 2.13 illustrate spatio-temporal variation in the growth of service trade across 87 economies. It is noteworthy that the service trade of a large group of countries spanning both developed and developing parts of the world has grown moderately over the period from 1990 to 2013. Furthermore, the countries recording negative growth were about 12 during 1990–1995, mere three in 2000–2005 and during 2010–2013 the numbers swelled to 17. It is worth

5  The product cycle theory was given by Raymond Vernon in response to the Hecksher-Ohlin model to exlain international trade. The theory would predict, partly because of the differences in the factor intensity of trade products and countries factor endowments, developed countries are specializing in complex production processes and trade in services; however, most developing countries are intensifying their export primarily with less complex processes and merchandise products.

2.4  Growth of Service Trade

37

Fig. 2.8  Major categories of services. (Source: Based on UNCTAD Handbook of statistics, 2006–07)

Fig. 2.9  Growth rate of service export (values in percent). (Source: Calculation based on Annexure 2.6)

38

2  Volume and Growth of Trade: Merchandise and Services

Fig. 2.10  Growth rate of service import. (Source: Calculation based on Annexure 2.6)

Table 2.4  Top five players in international service trade (percent share to world service trade) Country United States Germany Japan France United Kingdom

1990 1995 Country 15.56 14.64 United States 8.65 8.54 Germany 7.36 7.63 United Kingdom 7.02 6.09 Japan 6.16

5.76 France

2000 2005 Country 16.73 13.45 United States 7.28 7.46 Germany 7.22 7.35 United Kingdom 6.06 4.85 France 4.65

4.55 China

2010 Country 12.63 United States 6.76 Germany 5.74 China

2013 12.37

4.77 United Kingdom 4.66 France

5.17

6.61 5.82

4.62

Source: Calculation based on data UNCTADstat

mentioning that all these countries were from the global South, for example during 1990–1995, about half were African countries, and so in the period 2010–2013. There has been a visible rise in the participation of developing economies in world service trade over the past decades. During 1990–1995, only China and Vietnam, out of 87 countries, experienced a very high growth in their service trade. Thereafter, no country has fallen in this group of very high growth. During the period 1990–1995, service trade grew at a high rate largely in the Southeast and East Asia, such as Indonesia, Malaysia, Thailand, Singapore, South Korea and Uruguay (only Latin American economy). Furthermore, Poland and Romania are the only developed countries with high service trade growth. During 1990–1995, countries experiencing moderate growth were mainly Latin American economies, such as Argentina, Brazil, Chile, Honduras, Peru and Venezuela; Asian countries, such as Bahrain, India and Sri Lanka; and also African

2.4  Growth of Service Trade

39

Fig. 2.11  Growth of service trade, 1990–1995. (Source: Based on Annexure 2.7)

Fig. 2.12  Growth of service trade, 2000–2005. (Source: Based on Annexure 2.7)

economies, such as Mauritius, Nigeria and Tunisia. Moderate compound annual growth in service trade was also experienced by European countries, such as Bulgaria, Hungary, Ireland, Netherlands and Portugal. During 1990–1995, none of the countries registered low growth in the service trade; although in 2000–2005, service trade of Argentina, Uruguay and Guinea grew sluggishly. In the subsequent

40

2  Volume and Growth of Trade: Merchandise and Services

Fig. 2.13  Growth of Service trade, 2010–2013. (Source: Based on Annexure 2.7)

period, 2010–2013, Uruguay managed to bounce back to high growth category. Argentina’s trade grew moderately and that of Guinea slipped back to its 1990–1995 level. It is pertinent to note the pattern exhibited by India, considering its growing significance in the global economy. Over the period, India did exhibit a promising pattern in the service trade. Ocampo and Vos (2008), has attributed this recent growth to be driven by its fast-growing service sector. Given the relatively lowincome level, the pattern of structural change reflects a premature shift in services. Based on advanced communications technology, services have become increasingly tradable. In this context, according to Bhat (2011), ‘The rapid growth of service sector observed in the domestic economy has thus been associated with an increased competitiveness in world markets. Since 1999, India is the second largest exporter of business services among the emerging Asian economies’. On the other hand, growth of service sector in Latin American economies has been attributed not so much with the dynamic transition as with the deindustrialization process, which to a certain extent pushed surplus labour into the low productivity tertiary sector (Ocampo and Vos 2008). Broadly, developed countries experienced moderate growth in service trade across time. However, the share of developed economies, as a group, is relatively higher than developing countries in terms of export and import of services, as stated earlier, but the rate of increase in the service trade of the developed world is much slower than that of the developing world as a whole. For example, as noted earlier in the chapter, the top five individual players in the global service trade in terms of percent share, which are the United States, the United Kingdom, Japan, Germany and France, have recorded growth below global average over the years; and it is

2.5  Concluding Remarks

41

interesting to note that the emergence of China is not only reflected in terms of its proportional share but also its robust growth (Figs. 2.11, 2.12 and 2.13). Among the developing economies, world trade in services has grown rapidly in Asia. This generalization covers a heterogeneous group of countries which have experienced different levels of growth over the years. It could be further added that similar to merchandise trade, economies growing relatively faster have specialized in service exports with stronger growth dynamics and with greater spillover effects. However, for some countries, particularly the developing regions, it may slowly shift into more dynamic service export by utilizing their endowments and appropriate policies, but it could be relatively difficult to embark on the similar path. Though these economies are still in the initial stages of building the necessary capabilities, ‘they can still effectively participate in other service sectors and thus promote the diversification of their economies’ (Ocampo and Vos 2008). It is important to mention that a further in-depth analysis is needed to capture the emerging dynamics of the geographies of service trade.

2.5  Concluding Remarks The brief glance at the period since the 1990s, it is possible to discern three major patterns that have moulded and restructured the global trading landscape in the globalized era: the persistent slowdown of developed economies, the consistently strong performance of East Asia and the uneven performance of other developing countries, both over spatial and temporal scale. All the geographic regions were affected to varying degrees by the trade slowdown, both merchandise and services. It could be said that the aggregate trade performance of developing economies has been promising. However, it has been neither a continuous process nor uniformly spread across the globe. Broadly, in the context of commodity trade, it can also be inferred that countries (the Middle East, Africa, the Commonwealth of Independent States and South and Central America) with the highest share of manufactured goods, fuels and ores and metals in their merchandise export have recorded the strongest compound annual export growth over the years. It is also pertinent to mention in this backdrop that the impacts of the region- and country specific-factors are no less important as compared to the influence of fluctuations in global commodity and financial markets. Whether a country would experience higher growth or not during the periods of economic expansion in the developing regions, or whether it would check growth failure during the downturn, largely depends on the interaction between domestic situations and also the way in which any economy is influenced by the global or regional market dynamics (Ocampo and Vos 2008). However, diverse experience of countries requires explanation, but this is beyond the gamut of the current chapter. Also no single explanation could suffice for the variety of historical experiences.

42

2  Volume and Growth of Trade: Merchandise and Services

References Baumol WJ (1967) Macroeconomics of unbalanced growth: the anatomy of urban crisis. Am Econ Rev 57:415–426 Beerepoot N, Lambregts B, Kleibert JM (eds) (2017) Globalisation and services-driven growth: perspectives from the Global North and South. Routledge, London/New York Bhat TP (2011) Structural changes in India’s foreign trade. isid.org.in. Accessed 25 Dec 2016 Blinder AS (2006) Offshoring: the next industrial revolution? Foreign Aff 85(2):113–128 Cattaneo O, Linda S, Michael E, Stern Robert M (2010) International trade in services: New trends and opportunities for developing countries. World Bank, Washington, DC Coe NM, Yeung HW-C (2015) Global production networks. In: theorizing economic development in an interconnected world. Oxford University Press, Oxford Dossani R, Kenney M (2007) The next wave of globalization: Relocating service provision to India. World Dev 35(5):772–791 Friedman TL (2006) The world is flat: A brief history of the twenty-first century. Penguin Books, New York Grossman GM, Rossi-Hansberg G (2008) Trading tasks: a simple theory of offshoring. Am Econ Rev 98(5):1978–1997 Hunter WC, Kaufman G, Krueger T (eds) (1999) The Asian financial crisis: origins, implications and solutions. Springer, Boston IMF (2011) Changing patterns of global trade. http://www.imf.org Kleibert JM (2017) Revisiting the role of services in global production networks and regional economic development. Paper presented at the RSA Annual Conference 2017 in Dublin, 4–7 June, 2017 Loungani P, Mishra S, Papageorgiou C, Wang K (2017). World trade in services: evidence from a new dataset, IMF working paper No.17/77 Maddison A (2006) The world economy: a millennial perspective and historical statistics. OECD, Paris Mattoo A, Payton L (2007) Services trade and development: the experience of Zambia. Palgave Macmillon, Washington, DC Ocampo JA, Vos R (eds) (2008) Uneven economic development. Orient Longman Pvt. Ltd, Hyderabad UNCTAD (2000) World economic situation and prospects. United Nations, New York UNCTAD (2002) Trade and development report. United Nations, New York/Geneva UNCTAD (2004) World investment report 2004: The shift towards services. United Nations, New York/Geneva UNCTAD (2005) Trade and development report. United Nations, New York/Geneva UNCTAD (2006) World economic situation and prospects. United Nations, Geneva UNCTAD (2007) Trade and development report. United Nations, New York/Geneva UNCTAD (2016) Trade and development report. United Nations, New York/Geneva UNCTAD (2018) Trade in services and employment. United Nations, New York WTO (2006) International trade statistics. www.wto.org. Accessed 20 Oct 2016 WTO (2007) Participation of developing economies in the global trading system. http://www.mdg-­ trade.org/WTCOMTDW162. Accessed 2 Nov 2016 WTO (2008) World trade report: trade in a globalizing world. www.wto.org. Accessed 15 Oct 2016 WTO (2015) World trade report

Chapter 3

Spatial Structure of Trade Flows

Abstract  Today’s trade is radically more complex. The global economy is shaped by two powerful factors. First is by the widespread of multinational corporations (MNCs). Second is through the resurgence of region formation. Regionalization has emerged as a well-known feature of international trade. This process is inextricably connected with political interest, cultural affinities and also historical perspectives. In this backdrop, the current chapter attempts regionalization of trade flows for 1990 and 2015. Factor analysis has been used to understand the dominant origin and destination of export and import flow. The functional regionalization of international trade flows reveals primacy of the north in the global trading system. However, developing economies such as China and India have emerged in the recent years along with the traditional cores. Furthermore, globalization has not diminished the economic significance of location. Keywords  Regionalization · Natural blocs · South-South trade · Q and R mode analysis

3.1  Introduction In the past 30 years, international trade flows have expanded noticeably and, generally, at a rate faster than global output, with a doubling of the value of trade in a 10-year period since the mid-1990s. A combination of multiple factors have played key role in the recent growth of trade, the growing integration of countries and also the increasing contribution of trade to development. These include liberalization of tariff rates and other trade barriers, preferential trade access, foreign direct investment via trade, autonomous unilateral structural reforms, technological advancements in transport and communications, strategic trade policies, etc. International trade of the present time rests on centuries of experience and rises from definite motivations. While deeply rooted in history, modern international trade is subject to such rapid and thorough change that we look to the distant past © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9_3

43

44

3  Spatial Structure of Trade Flows

primarily for certain reasons which facilitate or restrict commerce. Countries exchange goods for different reasons. The size, shape, relative position, natural and human endowment, types of organization, degree of political independence and political ideology or ideologies of specific countries are the key factors to be considered in assessing global trading behaviour (Thoman and Conkling 1967). According to Deichmann and Gill (2008), in 1910, British exports were spread almost evenly among Europe, Asia and other regions. However, by the 1990s, about 60% of British exports went to Europe and a mere 11% to Asia. A standard economic theory would envisage that with improved and cheaper transportation, trade with distant places would increase. Instead, trade multiplied between neighbours. The relative degree of trade integration seems to strongly differentiate economies with high growth from those with relatively slow growth. And also, the strategies for effective regional integration are not uniform across regions of the world. Globalization led by trade is also manifested in the changing geography of the present global economy, such as, besides the North, the emergence of a dynamic South as a key driver of global trade and also an expansion in South-South trade in goods and services (UNCTAD 2008). A significant contributing factor has been the impressive growth in the proportion of international merchandise and services trade of several dynamic developing countries such as Brazil, India, China, Mexico, South Africa and South Korea among others. This growth has resulted in new and better opportunities for both trade and development. Another related feature has been a significant rise in trade between the countries of the South, as stated earlier

South-South trade account for about one quarter of global trade in 2015 ASIA Intra Asia 3272.76 billion

$225.67 billion $ 234.08 billion

$ 111.65 billion

$ 170.92 billion

AMERICA Intra America 160.55 billion $

$ 15.83billion

AFRICA Intra Africa 68.53 billion $

$ 12.64 billion

Fig. 3.1 Intra- and inter-regional trade among developing countries (2015). (Source: UNCTAD 2016)

3.1 Introduction

45

(Fig.  3.1). According to United Nations Conference on Trade and Conference (UNCTAD), their share in merchandise global exports increased from around 20% in 1970 to an all-time record of about 36% in 2006. Since 2008, developing economies as a group have been exporting relatively more to the South vis-a-vis global North. In 1990, South-South trade accounted for about 8% of the total world trade and increased to 25% in 2015.The trade-GDP ratio for the world has also increased from 31% in 1990 to 45% in 2015; figures for developed economies are 29% and 41% and for developing economies as a group are 41% and 50% in 1990 and 2015, respectively. The pattern reflects a greater openness on the one hand and trade dependence on the other. The significance of export earnings as a source for development finance also increased. UNDP (2013) has also indicated that the multiplying South-South trade and investment can lay the base for moving the industrial capacity to other less-developed regions. Since the mid-1980s, developing economies have also considerably enhanced their presence in developed economies, as discussed in the previous chapter. In the 2000s, export from the South (particularly China) accounted for 32% of total imports by developed economies, as compared to 25% in the 1980s. As regards export from developed countries to the South, it remained at about 23% during the same period, suggesting that the South is capturing a greater market both in the North and in the South. Japan, among developed economies, shows the strongest trade linkage with the South. More than 60% of Japan’s import flows from the South, and more than half of its export flows to the South. In the 2000s, about 52% of US import came from the South, having consistently increased from 35% in the 1980s. As for the EEC-15, the share of the South in its import slightly fell between the 1980s and 1990s, reflecting at the reduction in import from Africa. It then increased to 20% in the period between the 1990s and 2000s, which is about the same level as that of the 1980s, owing primarily to a substantial increase in imports from China. The aggregate trade performance of developing countries has been spectacular (UNCTAD 2008). However, it has been neither a continuous process nor uniformly spread across developing regions. It is also required to further add that North-South trade remains important, with the North providing the main markets for developing countries as whole. However, in parallel, South-South trade has emerged from the fringe of world trade to becoming more and more central, both in terms of quantity and quality. In this reference, UNCTAD report (2016) highlights that the global trade between 2004 and 2014 was primarily driven by the growing South-South trade. By 2015, the value of trade between developed countries touched about 4600 billion dollars, which was similar to that of North-North trade. However, in 2015, trade was slashed for both developed and developing countries. The quantum of South-South trade varies across developing regions, such as 70% in South and East Asia and 40% in South America. Since 2005, China has emerged as key trading partner for developing as well as developed countries and has a pivot role in intra-regional trade. So much so, it is speculated that the rise of China may lead to unification of Asia. It is to be noted that trade with China was more resilient in the year 2015, while a large part of the trade downturn was related to other South-South flows (UNCTAD 2016). The emerging developing countries, as put by Nadvi

46

3  Spatial Structure of Trade Flows

(2014), emphasize scale, that is large size in the context of population, and physical and economic geography. Due to this, these countries have rich natural resources, industrial base, large domestic market and expanding middle class population. Furthermore these ‘Rising Powers’ have not only been bourgeoning their participation in the international trade of commodity and services, ‘but also countries where there is increasing competition by national and global players for market shares within their growing domestic markets…Thus, unlike the Asian Tiger economies (most notably South Korea and Taiwan) which were feted two decades ago, the Rising Powers are of a size where they can potentially tip the balance of global economic power from the North’. It is interesting to note that today’s trade is radically more complex. The global economy is shaped by two ‘powerful forces’. First is by internalization of capital through the widespread of multinational corporations (MNCs). As noted by Gereffi (1989), spread of MNCs has drawn attention to the role of economic networks relative to nation-states as the relevant unit of analysis in the study of global production systems. Second, is through the resurgence of region formation. In this regard, Poon (1997) stated that such emergence of regions as a medium for organizing trade has generated ‘fears that global trade is becoming less free, resulting in macroeconomic disequilibria’. Hence, it could be said that regionalization has emerged as a widespread feature of international trade. Bonapace (2005) has added in this context that regionalization has suggested a new complementary strategy to multilateralism for developed and developing economies. In this regard, Gaulier et al. (2004) argued that out of eighty sample countries more than half of their international trade is concentrated within a single Triad region, namely America, Asia-Oceania or Eurafrica. The regional concentration is particularly strong in Eurafrica, with region accounting for more than 75% of foreign trade for most countries therein. Such a pattern of regional concentration is more limited in America, however, still important for US neighbours like Mexico and Canada. On the other hand, Asia-Oceania appears as the region having the weakest regional polarization, excluding large economies like Japan, China and South Korea, and intra-regional trade accounts for about 55–60% of total trade. Poon and Pandit (1996) investigated the importance of spatial structure in the process of regionalization. They find, rather than a triad structure of global trade, the clusters are not geographically contiguous regions but ‘functional units’ defined by the quantum of trade flows and understood by the bilateral trade intensity between the members. They argue in accordance with the new trade theory that ‘scale economies and large efficient markets are instrumental in shaping regional configuration’. Similarly, Kohl and Brouwer (2014) have written that trade clusters emerge not only due to proximity but also on account of other geographic, political, historical and cultural characteristics. Furthermore, they noted that over the decades there is popular debate on the ‘irrelevance of geography in the face of economic globalisation’. Friedman (2005) has captured this phenomena as the ‘world is flat’. On the other hand, there is also a growing literature advocating the ‘relevance of geography’, for example McCann (2011) highlighted that the world has become ‘steeper’ reflecting at the growing importance of geography in the

3.1 Introduction

47

global economy. Geography moulds development prospects and also to a certain extent suggests the type of instruments needed. According to John Kay, ‘Geography is still important. Globalization has not diminished the economic significance of location’ (Yeung 2002). In light of the preceding discussion, the general pattern of trade reveals that relatively more intense trade takes place within geographical regions. In the case of intra-regional flow analysis, Berry in the year 1966 used this set of techniques for identifying nodes in India (Box 3.1). In his study, these nodes were the 36 major trade blocks of India constituting combinations of states and major cities. According to Raza and Agarwal (1986), ‘the analysis of dyadic interactions becomes all the more important if it is postulated that close neighbours will tend to have greater volume of interactions than pairs occupying distant locations in space’. In this backdrop, a similar approach is adopted in this chapter for international trade to find out nodal shipper/receiver and dominant destination/origin. Hence, an attempt is made in this chapter to work out functional regionalization of international trade flows.1 Functional regions have played a significant role in the geographical research for decades. The concept of functional region was introduced in geography in the second half of the twentieth century (Haggett 1965). It also relates to the classical locational models given by Thunen (1826), Christaller (1933) and Losch (1940). However, the concept has evolved over the years. Broadly functional regions are a result of the interactions between places. In this context, Harrison (2008) has written, ‘ We are living in a regional world and single essential definitions of the region have been put firmly into the shadows by the recognition that spatial configurations are not necessarily or purposively territorial or scalar, but constituted through spatiality of flow, porosity, and relational connectivity associated with globalization…It can be seen that regions are already proving to be an important object of inquiry in the development of the theoretical debates’. This chapter has drawn upon the regionalization analysis from the paper presented by Yadav (2017) and is updated accordingly. The paper examined the structural interdependency via global trade in the context of the core-periphery debate. Factor analysis on flow matrices is used for analyzing the pattern of inter-regional and intra-regional interactions of 87 countries2 at two points of time, 1990 and 2015. Factor analysis is not the only technique by which such spatial linkages can be identified, graph theory has also been frequently used. But a major advantage of factor 1  According to Berry (1968), traditionally geographers used three classificatory approaches to conceptualize regions, namely homogeneity, nodality or polarization and programming or policy orientation. The first emphasizes on the homogeneity of places located within the regions with respect to a given set of properties; the second stresses on nodality, usually of areas around some central urban place; and the third is concerned primarily with ‘either administrative coherence and the identity between the area being studied and available political institutions for effectuating policy decisions, or else with the appropriate regional framework to insure achievement of a set of goals’. 2  Countries accounting for more than equal to 0.01% of the world merchandise trade were selected for the analysis. Eighty-seven countries were picked fitting this criterion. Countries which are excluded from the purview of the analysis together accounted for about 10% of the world merchandise trade.

48

3  Spatial Structure of Trade Flows

procedures in the analysis of flow patterns is that they avoid several arbitrary decisions, use all of the data in the matrix to produce regional description and can also be applied to ordinal and even nominal data (Clark et al. 1974). Box 3.1: Early Example of Factor Analysis in the Literature ‘The first examples of experimentation with such procedures were Kendall’s pioneering work on the distribution of crop productivity in Great Britain and Hagood’s studies of agricultural regions of the United States, which may be contrasted with Odum’s masterful exposition of traditional cartographic methods of regionalization using variables and the classic statement of the National Resources Planning Board. Attempts to develop alternative simpler statistical procedures immediately postwar were made by Weaver and Zobler, but with the development of modern large-scale computers in the United States after 1958 and their rapid adoption for scientific purposes the much more satisfactory but computationally more demanding methods proposed by Kendall could be used and extended. Berry applied factor and discriminatory analysis to the regionalization of economic development and later added improved dimensional analysis and grouping methods to the process. In these studies he showed how more than 50 variables used to characterize relative economic development based upon the scores of countries infact displayed only four basic patterns of spatial variation, and he then proceeded to regionalize economic development based upon the scores of countries with respect to these four patterns. The methods have also been used to derive indices of economic health of areas from a variety of original variables, to classify cities and towns, and to analyze differences in regional development and welfare in Canada. All of the above studies used scalar data and produced uniform regions. Stone, however, showed that vector data could be used in a similar analytic process. Russett has extended the work in his examination of international trading regions, and Berry showed that identical procedures of taxonomy were relevant in two cases. Harman remains the basic source on factor analysis, but see also Hotelling, Kendall, and Anderson. Sokal and Sneath, Rubin, and Ward, are useful sources on grouping methodology. As an exploratory method, factor analysis has been used by Adelman and Morris (1973) and Roberts and McBee (1968) to identify the patterns of development and modernization in developing countries’ (Source: Berry 1966).

3.2  Database and Methodology Johnston (1976) has remarked that global trade system can be contextualized as an ecosystem, as a spatial system. Keeping this in view, the present chapter, with the geographical view of world trade system, enquires whether there is any order in the spatial location and interaction via flows, whether there is spatial and functional

3.2 Database and Methodology

49

organization in the pattern of interaction over the years. Furthermore, in the backdrop of discussion done earlier in the chapter, it is quite clear that the continued interest in international merchandise trade is due to its crucial role in economic development as it binds producers and consumers located in different countries into a global economic system.3 In this chapter, the regional pattern of the value of merchandise export and import flows across the world is analyzed. The data matrix with the dimension of 87*87in, which is measurement in terms of value of merchandise export and import for each of the 87 countries, is considered for two points of time, 1990 and 2015, with an aim to shed some light on the underlying regional pattern of trade flows. Factor analysis identified functional areas using flow data matrix. The origins (exporting countries) are in the rows and the destinations (importing countries) in columns. There are several attempts made to utilize this statistical technique with varying objective (Refer Box 3.1). Factor analysis, a branch of multivariate analysis, identifies the underlying structure of a set of variables by selecting a smaller number of hypothetical variables called factors (Bumb 1982). Factor analysis collapses huge data sets into manageable summary. This summary helps in understanding the dominant origin and destination of export and import flow pattern. This matrix can be examined by analyzing similarities among rows (R mode analysis) and columns (Q mode analysis). One way of regionalization refers to the catchment area of imports and the other hinterland of exports. The result of analyzing such matrix is to study spatial pattern of trade flow, showing normal varimax rotated factor loadings for all factors with eigen values exceeding unity, and factor scores were computed by Bartlett’s method.4 The scores in conjunction with the factor loadings help in interpreting what characteristics of the trade flows are being represented by the factor. The factor loadings highlight groups of countries with similar patterns of trade flow, and the scores revealed the significance of 87 shippers/receivers to the components (extracted by varimax rotation). One origin/destination dominated each set of scores, linking the countries with high loadings on any component with the primary shippers/receivers as represented by component scores. As discussed in chapter one, there are various sources of data, each having its advantage as well as disadvantage. Data for the present chapter has been extracted from Direction of trade statistics (DOTS Trade matrix) and International Monetary Fund (IMF). Data consists of values of merchandise export and import flows of 87 countries by their partners, measured in million US dollars.  Draft IMTS 2010, see http://unstats.un.org/unsd/statcom/doc10/BG-IMTS2010.pdf  There are two main classes of factor scores computation methods, viz. non-refined and refined methods. Non-refined methods are relatively simple, whereas refined methods use more sophisticated, technical approaches and provide estimates that are standardized scores. Bartlett scores approach is one of the refined methods. With this approach only the common factors have an impact on factor scores. The sum of squared components for the ‘error’ factors (i.e. unique factors) across the set of variables is minimized, and resulting factor scores are highly correlated to their corresponding factor and not with other factors. However, the estimated factor scores between different factors may still correlate. One advantage this method has is that it produces unbiased estimates of the true factor scores (Distefan et al. 2009). 3 4

50

3  Spatial Structure of Trade Flows

3.3  The Factor Analysis of Flow Patterns (Q Mode Analysis) The flow maps (Figs. 3.1 and 3.2) illustrate factor analytic functional regionalization of the merchandise trade flows for 1990 and 2015. These figures show groups of receivers (hinterlands) with common patterns of assembly (dominant shippers). The results of factor analysis (Q mode) reveal underlying ten factor structures in 1990, which explains 91% of the variance. However, factors explaining up to 4% of the variance are retained for analysis, and the rest are dropped for both Q mode and R mode analysis. In 2015, there were eleven factors, and taken together explains around 93% of the total variance. The results further show that each region is functionally organized via trade flow around core economies. The five dominant shippers in 1990 were the United States, Germany, France, Japan and the United Kingdom5; however, in 2015, there are three, namely China, Germany and the United States6 (Figs. 3.1 and 3.2). Figure 3.2 reveals that the US-oriented region predominantly consists of developing Central and South America. Japan is the only Asian developed economy in this group. This pattern is driven by multiple factors which underline dynamics of regional economic integration in the Americas. The dynamics of such integration

Fig. 3.2 Dominant shippers with their Hinterland (1990). (Source: Based on author’s calculations)

5  The United States explains about 47% of the variance, Germany 16%, France 8%, Japan 7% and the United Kingdom explains about 4% of the variance. 6  China as the dominant shipper explains about 47% of the variance, followed by Germany and the United States explaining the maximum variance of about 16% and 12%, respectively

3.3 The Factor Analysis of Flow Patterns (Q Mode Analysis)

51

could be explained in light of US economic strategy, which offers enhanced market access to countries willing to adopt extensive economic reforms (Phillips 2005; Shadlen 2008). Further, Shadlen (2008) added that such preferences were institutionalized in the form of regional and bilateral trade agreements (RBTA’s) which were modelled on the North American free trade agreement (NAFTA). Further, Oman (1994) has observed that with NAFTA, the United States has renewed its focus on Latin America, shifting back plants to Mexico and initiating free trade arrangements with the aim of a hemisphere-free trade area. However, response to US strategy varied across the region and over time. For example, Brazil showed interest in free trade area of the Americas only in 2003. However, according to Shadlen (2008), ‘standard economic explanations based on country-size or industrial structure or export profile cannot sufficiently account for this variation, given that these potential explanatory factors are either fixed or change very slowly’. Figure 3.3 shows in 2015, besides developing America, Western Asian economies such as Bahrain and Saudi Arabia have emerged as markets for US export. Over a period of 15 years, spatial pattern of US export flow indicated persistence of intra-­ regional trade. Factor two identifies commodity movements predominantly within the European region with Germany as a dominant shipper (Fig. 3.2). The USSR (former), Turkey and Iran are the only non-European countries importing German commodities. In 2015, Germany maintained its position at the second factor. However, the market of German merchandise export got further geographically concentrated with the major focus on EU members, reflecting on the intensive regional orientation.

Fig. 3.3 Dominant shippers with their Hinterland (2015). (Source: Based on author’s calculations)

52

3  Spatial Structure of Trade Flows

Factor three identifies France as a dominant exporter to African countries, such as Algeria, Gabon, Guinea, Morocco, Senegal and Tunisia, among others (Fig. 3.2). The reason for the same could be that France still continues to play an important role in Africa, especially in its former colonies, through aid programs, commercial activities, military agreements and cultural impact. Germany, Spain and Jordan are other key export markets. Besides African economies, European economies, namely Germany and Spain, also constitute hinterland of the French export. Japan’s export flow (factor four) reflects on the predominance of the intra-­ regional movement of commodities. Besides developing Asian countries like China, Indonesia, the Philippines, Singapore, South Korea, Thailand and Sri Lanka, developed economies like the United States and New Zealand are also receiving Japanese export. Poon (1997) has pointed out in this context that Japanese firms are actively pursuing a pan-Asian strategy through industrial integration in the region. She further stated that such global management of economic activities by corporations within coherent spatio-economic units is likely to increasingly shape regionalization patterns. New Zealand is the only non-Asian economy among key Japanese markets. In this context, Poon and Pandit (1996) added that the presence of New Zealand in the Japan-oriented region is indicative of the shift of this economy’s integration more with the Asian countries than European. Figure 3.2 also depicts the United Kingdom (factor five) as a dominant shipper with export flowing towards European, African and Asian economies. Ireland and Norway are the major European markets; African countries like Kenya, Ghana, Nigeria, South Africa, Zambia and Zimbabwe and India and Saudi Arabia are significant Asian destinations. Similar to France, the United Kingdom has trade relations with its former colonies. Figure 3.3 shows the emergence of China as the dominant export core (first factor) in 2015. The China-oriented region primarily includes developing Asian and African economies, such as India, South Korea, Malaysia, Thailand, the Philippines, Vietnam, Indonesia, Bangladesh, Pakistan, Iran, Jordon, Kenya, Nigeria, Guinea, Cameroon and Senegal. Developed economies, namely Sweden and Japan, are also part of the market hinterland for Chinese commodities. It reveals the integration of China chiefly with the developing ‘South’ economies, indicating at the mounting significance of the South-South trade and also the emergence of China as an alternative to the traditional trading cores for the commerce of developing economies. Figures 3.2 and 3.3 clearly bring out the existence of intra-regional trade flows in either years of reference. From the pattern analyzed in 1990 and 2015, roughly three trading hubs could be seen, viz., Pan-America, Pan-Europe and Pan-Asia. In this background, Loughlin and Anselin (1996) have put forth that in view of the German school of Geopolitik, dominated by Karl Haushofer, a stable equilibrium could be produced and maintained by the division of the globe into three zones  – Pan-­ America (North and South America), Pan-Europe (Europe, the Middle East and Africa) and Pan-Asia (Asia and Australasia). Each would comprise of a core and a periphery, and this complimentary trade relationship would reduce the necessity to trade outside the blocs. Bonapace (2005) has also noted that regionalism has become a key component of the new international order. According to him, regionalism is a

3.4 The Factor Analysis of Flow Patterns (R Mode Analysis)

53

complex process which is inextricably linked to political objectives, cultural affinities and historical perspectives. MacLeod and Jones (2007) have noted that distance is the predominant factor in creating economic regions with the underlying impact of politics, history and culture on the relative distance (spatial flows) as discontinuous or supporting factors. Such regions, as generated by these spatial flows, have strong still fragile links due to the process of ‘territorial restructuring in a world of political and economic turbulence’. Furthermore Poon et al. (2000) have added that the changes in trade clusters is primarily driven by a set of centrifugal and centripetal forces that are working simultaneously and result in the patterns of relationships, such as trade clusters, where ‘space is both sticky (important) and fluid (flexible) at the same time’.

3.4  The Factor Analysis of Flow Patterns (R Mode Analysis) In 1990, there were ten significant factors, which taken together account for about 92% of the variance in the spatial pattern of import flow. In 2015, there were seventeen factors, cumulatively accounting for about 93% of the variance. Figures  3.3 and 3.4 show selected five factors, each for 1990 and 2015. Dominant receivers were the United States, Germany, Japan, France and USSR (former) in 19907 and

Fig. 3.4  Dominant receivers with their major Shippers/Origins(1990). (Source: Based on author’s calculations)

7  The United States accounts for 45% of variance, Germany 17%, Japan 8%, France 6% and USSR (former) 5%.

54

3  Spatial Structure of Trade Flows

the United States, Germany, China, Japan and India in 2015.8 The rise of developing emerging economies, such as China and India, indicates new trading cores in the changing global economic system. According to Subacchi (2008), the debates regarding the shift of economic power have gained popularity to the ‘point of cliche’ mainly due to the China influence. Subacchi further highlighted that besides traditional nodes like the United States, Europe and Japan, the reason behind the rise of countries like China and India, from periphery to the core of the global economy, is not only their large sizes but also the potential to affect international geopolitical relations. It is worth noting that the degree and nature of global integration varies significantly between the two emerging Asian economies. The share of China’s merchandise export to world export has increased from 2% in 1990 to 14% in 2015; however, India’s share to world merchandise export inched up from 0.5% in 1990 to 2% in 2015. Figure 3.4 shows that US import from diverse geographical sources spread across America (Canada, Brazil, Colombia, Mexico, Panama, Honduras, Venezuela, etc.), Asia (Japan, Bangladesh and Sri Lanka) and Africa (Angola, Nigeria and Guatemala). In 2015, the United States remained at the position of first factor (Fig.  3.5). The pattern more or less remained same, though China has replaced Japan over the period of 15 years as the source of US merchandise import.

Fig. 3.5  Dominant receivers with their major shippers/origins (2015). (Source: Based on author’s calculations)

8  The United States explains 39% of the variance, Germany16%, China11%, Japan 5% and India 4%.

3.4 The Factor Analysis of Flow Patterns (R Mode Analysis)

55

Germany at the second factor is the dominant market particularly for the European economies, such as Austria, Belgium-Luxembourg, Bulgaria, Denmark, France, Greece, Italy, Malta, the Netherlands, Portugal, Switzerland, Poland and Sweden. Other countries, like the USSR (former), European-Asian economy Turkey and African countries, viz., Ghana and Zimbabwe, are also exporting to Germany. It is noticeable that there is a preponderance of EU members among the group of shippers, and Turkey is benefitting mainly because of EU-Turkey customs union (1995) and free trade agreement with the EU (2004), however, without full membership. In 2015, Fig.  3.4 reveals that Germany’s trading pattern became more Eurocentric. Bangladesh is the only developing Asian economy that has emerged in the group. The regionalization pattern is similar to the spatial structure of its export. Japan (third factor), as seen in Fig. 3.4, is the dominant receiver of the merchandise commodities predominantly from Asian economies, spread across southeast (Indonesia, Thailand and Malaysia), east (Mongolia and South Korea) and oil-based west Asia (Saudi Arabia, UAE and Kuwait). Japan has emerged as an important core in the Pacific, with its influence extended to the Middle East, Africa and, of course, Asia; Chile and New Zealand are also visible in the Japanese import hinterland. In 2015, as Fig.  3.5 shows, Japan (slipped to the fourth factor)-oriented region got more Asia centric, with the rise of China also in its trade hinterland. Further, Fig. 3.4 shows in 1990, France (fourth factor) was also the dominant receiver of shipments from European countries, such as Belgium-Luxembourg, Germany, Italy, Portugal, Spain and the United Kingdom, and African economies, such as Algeria, Cameroon, Cote d’Ivoire, Gabon, Mauritius, Morocco, Senegal and Tunisia. This reflects upon the fact that France not only retains strong regional trade ties but also has political and economic influence in its former African colonies. Also, in 1990, USSR (former) was among the major import cores with the hinterland comprising of eastern European economies and India, Syria and Egypt. China and India have evolved among traditional dominant importers in 2015 (Fig. 3.5). It is worth to note that over the years Chinese economy has not only risen as a dominant player in the global trade but also as a crucial trading partner for both developed and developing countries. China’s spatial flow structure highlights that the economy largely imports from the developing world, constituting diverse geographical regions, namely Asia (Iran, Malaysia, Singapore and Thailand), Africa (Angola and South Africa) and Latin America (Brazil, Chile and Peru). This is reflective of proliferating South-South trade and also gaining significance of emerging economies in the global market and in the process reshaping the spatial imprint of evolving interdependencies via international trade flows. Furthermore, Fig. 3.5 also shows emergence of India as the new import core in 2015. Switzerland is the only developed economy in the India-oriented group. The hinterland of Indian import primarily consists of peripheral economies from the developing Asia, Africa and Latin America. China and India hinterland exhibit significant intra-­regional trade structure, and also put these countries in a competitive footing in the global economy for not only strengthening trade ties with the geographically proximate economies or natural trade partners, but also poses strong geopolitical challenge before them both regionally as well as globally.

56

3  Spatial Structure of Trade Flows

Broadly, it could be seen that the analysis captures regionalization tendencies in 1990 for some nodal receivers, and the pattern gets further intensified in the last 15 years. The core import regions also reflect strong historical tendencies, for example the relationship between European countries like the United Kingdom and its ex-colonies, the common cultural traits of Latin America (LAIA) and the former communist bloc of the USSR (former)/Eastern Europe (Poon 1997). As far as developing economies are concerned, intra-regional integration has deepened through regional trading agreements. In most of the instances, such agreements go beyond trade issues and encompass common policies on resources, infrastructure growth, food security as well as trade facilitation and promotion of technological innovation, cultural exchanges and cooperation across other sectors.

3.5  Conclusion Regionalization is a process which is visible as the key feature of the international trade order. This process is inextricably connected with political interest, cultural affinities and also historical perspectives. The functional regionalization of international trade flows reveals primacy of North in the global economy. The analysis reinforces the significance of traditional traders, such as the United States and Germany, in the global trading system; however, the emergence of Asian economies, such as China and India, also highlights the growing role of the developing South in changing the structure of the traditional global interdependencies and hence changing the underlying regional dynamics of trade network. Over the years, a tectonic shift in the global trading system has been quite visible. In line with this, analysis of the regional pattern of merchandise trade revealed that North-South trade remains essential, because the North is yet the main market for developing countries. In parallel, South-South trade has also expanded significantly. Furthermore it is noted that stronger trade nodes have increasingly entrenched geographically (confirming Krugman and others observations), bringing out the creation of natural blocks on the basis of spatial contiguity. Such natural trading units are further intensified economically and politically via free trade deals by the EU to North Africa and Eastern Europe, NAFTA’s extension to the rest of the Latin America and also formation of the South Asian Preferential Trade Area and the Association of Southeast Asian Nations (ASEAN) free trade area. So developing economies are also bonding together into the Preferential Trade Area (PTA), which improves their negotiating position in trade and other international deals. For example, MERCOSUR, which is a custom union among Argentina, Brazil, Paraguay and Uruguay. The inter-regional integration in the South has also intensified, for example IBSA (India-Brazil-South Africa) Cooperative Initiative and trade agreements between Asian and Latin American countries and also bilateral agreements between Asian and African economies.

References

57

References Adelman I, Morris CT (1973) Economic growth and social equity in developing countries. Stanford University Press, Stanford Berry BJL (1966) Essays on commodity flows and the spatial structure of the Indian economy. University of Chicago, Chicago Berry BJL (1968) A synthesis of formal and functional regions using a general field theory of spatial behaviour. In: Berry BJL, Duane FM (eds) Spatial analysis: a reader in statistical geography. Prentice-Hall Inc, Englewood Cliffs, pp 419–428 Bonapace T (2005) Regional trade and investment architecture in Asia-Pacific: emerging trends and imperatives, RIS Discussion Paper No. 92/2005, www.ris.org Bumb B (1982) Factor analysis and development. J Dev Econ 11:109–112 Clark D, Davies WKD (1974) The application of factor analysis in human geography. The Statistician 23(3/4):71–84 Deichmann U, Gill I (2008) The economic geography of regional integration. Financ Dev 45(4):45–47 DiStefano C, Zhu C, Mindrila MD (2009) Understanding and using factor scores: considerations for the applied researcher. Pract Assess Res Eval 14(20):1–11 Friedman TL (2005) The world is flat: a brief history of the twenty-first century. Farra, Straus and Giroux, New York Gaulier G, Jean S, Ünal-Kesenci D (2004) Regionalism and the regionalisation of international trade, CEPII Working Paper No 2004-16 Gereffi G (1989) Development strategies and the global factory. Ann Am Acad Pol Soc Sci 505:92–104 Harrison J (2008) The region in political economy. Geogr Compass 2(3):814–830 Haggett P (1965) Locational network analysis in human geography. Arnold, London Johnston RJ (1976) The world trade system: some enquiries into it spatial structure. G. Bell and Sons Ltd., London Knox P, John A, Linda MC (2014) The geography of the world economy. Routledge, New York Kohl T, Brouwer AE (2014) The development of trade blocs in an era of globalization. Environ Plan A 46:1535–1553 Loughlin J, Anselin L (1996) Geo-economic competition and trade bloc formation: United States, German, and Japanese exports, 1968–1992. Econ Geogr 72:131–160 MacLeod G, Jones M (2007) Territorial, scalar, networked, connected: in what sense a ‘regional world’? Reg Stud 41:1177–1191 McCann P (2011) Globalization and economic geography: the world is curved, not flat. Camb J Reg Econ Soc 1:351–370 Nadvi K (2014) “Rising powers” and labour and environmental standards. Oxf Dev Stud 42(2):137–150. https://doi.org/10.1080/13600818.2014.909400 Oman C (1994) Globalization and regionalisation: the challenge for developing countries. OECD, Paris Phillips N (2005) US power and the politics of economic governance in the Americas. Latin Am Polit Soc 47(4):1–25 Poon J (1997) The cosmopolitanization of trade regions: global trends and implications, 1965–1990. Econ Geogr 73:390–404 Poon J, Pandit K (1996) The geographic structure of cross-national trade flows and region states. Reg Stud 30(3):273–285 Poon J, Thompson ER, Kelly PF (2000) Myth of the triad? The geography of trade and investment blocs. Trans Inst Brit Geogr 25:427–444

58

3  Spatial Structure of Trade Flows

Raza M, Agarwal Y (1986) Transport geography of India: commodity flows and the regional structure of Indian economy. Concept Publishing Company, New Delhi Roberts RE, McBee GW (1968) Modernization and economic development in Mexico: a factor analytic approach. Econ Dev Cult Chang 16(4):603–612 Shadlen K (January 2008) Globalisation, power and integration: the political economy of regional and bilateral trade agreements in the Americas. J Dev Stud 44(1):1–20 Subacchi P (2008) New power centres and new power brokers: are they shaping a new economic order? Int Aff 84(3):485–498 Thoman R, Conkling E (1967) Geography of international trade. Prentice-Hall Inc., Englewood Cliffs UNCTAD (2016) Trade and development report. United Nations, New York/Geneva UNCTAD (2008) Globalization for development: the international trade perspective. United Nations, New York/Geneva UNDP (2013) The rise of south: human progress in a diverse world. United Nations Development Programme, New York Yadav P (2017) Core and periphery: an analysis of the spatial patterns of international trade. Paper presented at the RSA Annual Conference 2017 in Dublin, 4–7 June, 2017 Yeung HWC (2002) The limits to globalization theory: a geographic perspective on global economic change. Econ Geogr 78(3):285–305. July

Chapter 4

An Analysis of the Commodity Composition of International Trade

Abstract  The role played by global trade in growth disparity across countries is deeply rooted in the kind of commodities and services these countries produce and in the potential for these products in the international market. Hence, different economies depending on their comparative advantage in terms of economic and resource endowments tend to participate at varying rate not only in international trade but also in the commodities they exchange. The chapter attempts to study the changing structure of commodity trade in the globalized era across aggregate country groups, and also country wise for major commodity groups. Change in commodity composition of developing economies shows mixed results. It is interesting to note that developing economies are slowly moving towards trade in manufactures, and the pattern is more pronounced in the case of East Asian economies. However, developing African and American countries have been the slowest regions to diversity away from the primary commodities. Furthermore, South-South trade is also gaining significance in the changing geography of commodity trade. Keywords  Centre-periphery · Manufactures · Primary commodities · Location quotient

4.1  Introduction The role played by international trade in growth disparities across countries originates in differences in the kind of commodities and services these countries produce and in the potential for these products (UNCTAD 2005). In this regard, Raza and Agarwal (1986) point out that similarities and dissimilarities in spatial patterns of inter-regional trade between developed and developing economies depend on two things – first, organization of their space economy and second, on the nature of the commodity. Different economies across the globe specialize at varying levels and intensity in different economic activities and trade that is broadly regulated by their resource endowments and the relationship between exchange, cost as well as © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9_4

59

60

4  An Analysis of the Commodity Composition of International Trade

distance. Hence, different countries depending on their economic base, tend to participate at varying rate not only in international trade but also in commodities they trade. It is quite obvious by now that there exist huge intra- and inter-regional disparities in import and export flows. Possible reasons causing such regional variations at different levels could be ecological, socio-economic conditions and political structure (Raza and Agarwal 1986). Grant (1994) has noted that changes in the sectoral composition of global trade have changed the static notions pertaining to conventional comparative advantage. It was further highlighted that the basis of comparative advantage is not wholly natural but ‘products of the structure of the world economy (the hierarchical core, periphery and semi-periphery structure) and human creations that reflect political ideologies and resultant government-policy orientations’. Besides change in the quantum and product composition of global trade, price volatility of internationally traded goods affects nature of impact of global trade on individual country. The effect of price fluctuations in international market for primary and manufacturing goods gauged via terms of trade and purchasing power of its export1 is in the short run determined by the composition of the trade (export and import) basket and in the medium term by its flexibility in being able to adapt the composition of its export and import to the dynamics of global demand and supply, respectively (UNCTAD 2005). On similar lines, there have been numerous studies undertaken to establish relationship between commodity prices of the traded commodities and the economic growth. For example, Gruss (2014) analyzed the impact of increasing commodity prices on the Latin American and Caribbean countries (LAC) and noted the positive relationship between increase in the price and rising output of these economies, although there was no relationship established between the price level and the growth of output. Countries with poor endowments tend to have relatively greater disadvantage with respect to gains of international trade. As noted by Ocampo and Vos (2008), countries still relying on a few primary export commodities, for example cotton, coffee or minerals, have experienced great volatility in the global market prices of their export, and also over the years they have experienced fall in prices of their exporting commodities as compared to manufactured goods. This has resulted in constraining the available resources, and weak institutions failed to implement strong policies and mobilize the domestic and external capital resources for private investment and infrastructure, which are required to diversify. In this backdrop, it is worth mentioning that the nature of growth of any economy to a certain extent is better captured by the commodity composition of its export and import than mere quantum of total shipments and receipts. Also commodity-wise analysis is also needed because of the wide variations in the commodity composition of trade for different countries cutting across diverse geographical regions. Kristjanpoller et al. (2016), in a recent study, examined the relationship between 1  Terms of trade refers to the evolution of a country’s export prices relative to its import prices and the purchasing power of its exports defined as the export value deflated by import prices (UNCTAD 2005).

4.2  Methodology and Database

61

commodities’ boom and growth and also which of the different commodities’ export account for the higher level of growth. They concluded ‘ore and mineral exports, fuel exports and food exports generally had a negative effect on GDP per capita growth but ore and mineral exports had a positive effect on LAC countries during the boom. During the boom period, agricultural exports had negative effects, especially for LAC countries. Fuel exports had a positive effect on LAC and non-LAC countries during the boom. Manufacturing exports in general had a positive effect on economic growth, but in the boom period this effect almost disappeared for the LAC countries’. In this context, the present chapter attempts to analyze structure of the international trade reflected in terms of levels and trends of inter-regional commodity flows. However, it is worth mentioning that the scope of the present chapter is limited because of the broad category of commodities considered.

4.2  Methodology and Database For the purpose of the current chapter, major commodity flow data have been collated for three major regions, namely developed, developing and economies in transition, and also for individual 87 countries. Three major country groups are classified by UNCTAD according to the three categories of development, and each category is further divided into geographical regions (refer Annexure 4.1). Five major commodity groups, viz., all food items, raw materials from agricultural sources, ores and metals, fuels and manufactured goods, are used in the study. Figure 4.1 shows a brief description of categorization of major commodity groups given by

Fig. 4.1  Commodities, at the 3-digit level or by broad product group, classified in accordance with the United Nations Standard International Trade Classification (SITC). (Source: UNCTAD Handbook of Statistics, 2006–07; see Annexure 4.2 for detailed commodity composition)

62

4  An Analysis of the Commodity Composition of International Trade

UNCTAD. Analysis and interpretation cover the period from 1995 to 2015 for comparison of five major commodity groups across three major geographical regions. For analyzing regional variation (here 87 countries) in commodity composition of export and import, data for each country both on the volume of export and import of the commodity groups stated earlier, for two points of time, namely 1995 and 2015, have been used. Due to lack of availability of consistent and appropriate data for commodity groups across 87 countries for 1990, these 2 years are selected, particularly for analysing and grouping countries on the basis of the relative concentration of commodity groups (see Fig. 4.1. for method of classifying commodity groups) in their merchandise export and import. Box 4.1 summarizes data discrepancies as noted by UNCTAD. Simple graphs and tabulation are used to present the data. Location quotient (LQ) is used to measure the degree of concentration of commodity groups in export and import basket of 87 countries for two points of time, as stated earlier. Location quotient tells us about the relative concentration or dispersion of an attribute. Export and import flow data of 87 countries for five major commodity groups for 1995 and 2015 are used to calculate LQ. Concentration Index (LQ) for merchandise export and import is calculated by using the following formulae:   Export / import of Commodity X in country i / Total merchandise trade of Country j      World Export / import of commodity X / Total world merchandise trade   

LQ  



Value could be more or less than 1. Value of LQ above unity reflects the degree of concentration. Countries having LQ 1 or more than one in any one of the five commodity groups are classified under single commodity group. Countries

Box 4.1: Brief Note on Data Merchandise export by destination and import by origin by six major commodity groups is collated from UNCTAD and presented in Tables 4.7, 4.8, 4.9, 4.10 and 4.11. It is important to know that export from destination differs considerably in some cases from data on import as reported by countries of destination for a variety of reasons, among which the following may be of some relevance: • Most import data are reported on a c.i.f. rather than an f.o.b. basis. • Imports arrive at the destination and are registered with some time lag from the date they were recorded as export. • There may be considerable differences between the recorded destination of export and the actual destination as shown in import statistics. Similarly, the classification used by the exporting country may differ from the one assigned by the importing country. (Source: UNCTAD Handbook of Statistics, 2006–07)

4.3  Composition of Merchandise Export

63

recording LQ value of more than one in five major commodity groups are classified as countries that specialize in multiple commodities. The following are the derived commodity groups: (1) Single commodity group • • • • •

All Food items Agricultural raw materials Ores and metals Fuels Manufactured goods

(2) Multi-commodity group • Multiple commodities with dominance of primary commodities • Multiple commodities with dominance of manufactured goods

4.3  Composition of Merchandise Export Figure 4.2 shows aggregate export values of total world merchandise primary commodities (excluding fuels) and manufactured goods. It is quite clear from the figure that value of export of all products has consistently risen from 1995 to 2015. Similarly, the product composition of world merchandise export reveals that the value of manufactured goods has multiplied over the period of 20 years, and also its value is significantly higher than that of primary commodities (excluding fuels).

Fig. 4.2  World merchandise export of major commodity groups. (Source: Based on UNCTADSTAT database)

64

4  An Analysis of the Commodity Composition of International Trade

The primary products have low elasticity of income, high protection and subsidization for agricultural commodities, technology change and price developments (Ocampo and Vos 2008). The market for manufacturing products is relatively more dynamic. The composition of the world’s merchandise trade and that of the three major country groups in the analysis are summarized below in Tables 4.1 and 4.2. During the period of 20 years, the share of manufactured goods is dominant in the world export basket (Table  4.1). It is noteworthy, despite a higher percent share, manufactures export from developed country group has marginally dipped and that of developing country group has inched up relatively. And economies in transition have experienced a decline in share of primary commodities (excluding fuels) from about 21% in 1995 to 17% in 2015 and also dip in the case of manufactured goods from around 32% in 1995 to 25% in 2015. Spatial distribution of world merchandise export of five major commodity groups is summarised in Tables 4.2, 4.3, 4.4, 4.5 and 4.6. Major share of world export of all food items is destined to developed economies with a share of around 68% over the period of 10 years, although in 2015 the share dipped to 57%. However, between 1995 and 2005, economies in transition account for a meagre share of about 4%. Table 4.1  Distribution of export of primary commodities and manufactured goods (percent to total merchandise export) Primary commodities, excluding fuels 1995 2000 2005 2010 2015 World 14.98 11.33 11.49 13.39 13.08 Developed economies 14.17 11.29 11.70 14.01 13.80 Developing economies 16.50 11.09 10.93 12.58 11.97 Economies in transition 21.01 15.87 13.73 13.41 17.14

Manufactured goods 1995 2000 2005 72.71 73.23 70.43 76.81 77.82 76.50 65.73 66.99 64.48 31.60 29.53 25.40

2010 65.46 70.90 62.66 20.99

2015 75.21 71.59 69.99 24.67

Source: Calculations based on UNCTADSTAT database Table 4.2  Distribution of world merchandise export of all food items by destination (values in percent) Year 1995 2005 2015

Developed economies 67.9 68.7 56.5

Economies in transition 4.2 4.2 4.2

Developing economies 25.4 26.3 38.9

Source: UNCTAD Handbook of Statistics, 2016 Table 4.3  Distribution of world merchandise export of raw materials from agricultural sources by destination (values in percent) Year 1995 2005 2015

Developed economies 67.9 61.9 48.5

Economies in transition 0.7 1.6 2.0

Source: UNCTAD handbook of statistics, 2016

Developing economies 30.4 35.9 49.0

4.3  Composition of Merchandise Export

65

Table 4.4  Distribution of world merchandise export of ores and metals by destination (values in percent) Year 1995 2005 2015

Developed economies 69.3 60.5 44.6

Economies in transition 1.1 1.3 1.1′

Developing economies 27.5 36.8 53.3

Source: UNCTAD handbook of statistics, 2016 Table 4.5  Distribution of world merchandise export of fuels by destination (values in percent) Year 1995 2005 2015

Developed economies 68.6 66.0 49.8

Economies in transition 2.7 1.2 1.2

Developing economies 25.8 29.3 45.3

Source: UNCTAD handbook of statistics, 2016 Table 4.6  Distribution of world merchandise export of manufactured goods by destination (values in percent) Year 1995 2005 2015

Developed economies 67.9 66.1 56.3

Economies in transition 1.4 2.4 3.3

Developing economies 29.4 31.2 40.2

Source: UNCTAD handbook of statistics, 2016

Developing economies’ share hovered around 25% from 1995 to 2005 but increased to 39% in 2015. Share of raw materials from agricultural sources in world export reflects more or less a similar pattern. However, share of developed economies has experienced a decline from 68% in 1995 to about 49% in 2015, and on the other hand, share of developing economies has grown from 30% in 1995 to 49% in 2005. Developed economies account for the largest share in world export of ores and metals for a decade; however, its share has persistently declined from about 69% in 1995 to 60% in 2005 and 45% in 2015. On the other hand, the proportion of ores and metals in export of developing economies has increased from about 28% in 1995 to 36% 2005, and in 2015 they accounted for more than half of the global export. Similar to other commodities, share of export destined to economies in transition remains more or less stable over the same time period. In the case of fuels and manufactured goods, similar pattern is observed. There is a visible increase in the share of developing economies over the years. Due to higher prices, discovery of new natural resources and increased efficiency in production, developing countries registered increase over the years. It could be broadly inferred from the foregoing analysis, in case of raw materials from agricultural sources the share of developed economies in export has declined over the period of 10  years, but in case of developing economies their share has increased. It is important to note, developed economies still account for larger share in food commodities. This is mainly because of heavy subsidy enjoyed by their

66

4  An Analysis of the Commodity Composition of International Trade

agriculture. In fact, developing economies are unable to compete with the subsidized agriculture of developed economies, and it has resulted in their import of food produced by developed economies (UNCTAD 2008). This could be also observed in analysis done further in the chapter. More or less similar pattern could be seen in case of manufactures. UNCTAD has noted in this regard that developing countries as a group have made progress towards one of their major economic development goals, that is to multiply the degree of processing of their commodities. However, it is worth noting that the level of impact varied across countries and commodities. Tables 4.7, 4.8, 4.9, 4.10 and 4.11 summarize share, direction and composition of international trade, viewed at major geographical regions and also by commodity Table 4.7  Composition of export of all food items by destination (Share in percent) 1995

2005

2015

1995

2005

2015

1995

2005

2015

Export To

From Developed economies Economies in transition Developing economies

Economies in Developing Developed economies transition economies 76.05 79.12 70.92 3.67 2.84 1.88 18.79 17.09 26.80 30.23 23.89 23.07 60.61 53.33 33.87 6.83 22.06 42.80 56.74 51.45 37.90 2.81 3.91 2.87 40.10 44.25 58.74

Source: Based on UNCTADstat database Table 4.8  Composition of export of raw materials from agricultural sources by destination (Share in percent) 1995

2005

2015

1995

2005

2015 1995

2005

2015

Export To

From Developed economies Economies in transition Developing economies

Developed economies 76.20 70.33 57.54 63.44 47.42 32.18 50.97 46.90 34.46

Economies in transition 0.41 1.15 1.50 7.19 11.34 9.99 0.25 0.70 0.88

Developing economies 22.93 27.72 40.92 24.95 41.22 57.74 48.56 52.14 63.93

Source: Based on UNCTADstat database Table 4.9  Composition of export of ores and metals by destination (share in percent) 1995

2005

2015

1995

2005

2015

1995

2005

2015

Export To

From Developed economies Economies in transition Developing economies

Developed economies 70.20 59.07 44.38 55.69 38.88 30.61 63.37 46.15 31.68

Source: Based on UNCTADstat database

Economies in transition 0.82 0.69 0.40 31.92 29.49 27.84 2.37 1.31 1.05

Developing economies 21.89 34.71 51.43 10.51 31.63 41.54 34.07 52.47 66.90

4.3  Composition of Merchandise Export

67

Table 4.10  Composition of export of fuels by destination (share in percent) 1995

2005

2015

1995

2005 2015 1995

2005

2015

Export To

From Developed economies Economies in transition Developing economies

Economies in Developed economies transition 81.25 84.17 68.45 1.11 0.57 62.17 67.09 57.37 22.60 7.42 62.92 56.96 34.51 0.13 0.05

1.12 6.94 0.11

Developing economies 12.92 11.02 23.92 9.83 9.43 20.88 35.93 42.50 64.62

Source: Based on UNCTADstat database Table 4.11  Composition of export of manufactured goods by destination (Share in percent) 1995

2005

2015

1995

2005

2015

1995

2005

2015

Export To

From Developed economies Economies in transition Developing economies

Developed economies 73.59 74.01 68.90 35.37 36.07 39.82 55.07 51.55 42.07

Economies in transition 1.12 2.27 2.31 32.25 31.35 32.30 0.85 1.46 1.70

Developing economies 24.27 23.28 28.57 29.80 32.56 27.82 43.81 46.79 55.94

Source: Based on UNCTATstat database

groups. Table 4.7 reveals export pattern of major country groups for all food items. It could be seen that it is more of intra-regional export of all food items from developed economies, for example in 1995 about 76% of its export of all food items went to developed economies and this share has further risen to 79% in 2005, though dipped to 71% in 2015. Among developed economies, Europe receives a major share. Of the total export flowing to developed regions, in 1995 about 58% was destined to Europe which increased to 62% in 2005. Despite the decline to 55% in 2015, Europe continues to account for about half of the export destined for developed markets. The United States received 5% in 1995, and over the period of 2005–2015 its share pegged at 7%. Canada accounted for around 3% in 1995–2015. Other developed economies accounted for a relatively meagre share of 1% in 1995–2005 and about 2% in 2015. The proportional share of economies in transition accounts for about 4% in 1995 and 3% in the later years. Developing economies received for about 19% in 1995, and its share has slipped to 17% in 2005, however, inched up to 27% in 2015. Among developing economies group, Asia bags a relatively larger share as compared to Africa and America. Asia accounts for 11%, 10% and 17% in 1995, 2005 and 2015, respectively. Broadly it could be seen that export flow of all food items is more lateral than vertical, that is more among developed economies and much less among developing economies. Major share of all food items from economies in transition is traded within the region in 1995. Over the decade, more than about 50% is exported to economies in transition. However, share of the region has declined from 61% in 1995 to 53% in

68

4  An Analysis of the Commodity Composition of International Trade

2005 and 34% in 2015. On the other hand, Table 4.7 reveals that the developing economies have experienced an impressive gain in their share from mere 7% in 1995 to 22% in 2005 and 43% in 2015. Of the total export directed to the developing economies, in 1995 Asia received the largest share of 5%, and it increased to 22% in 2015; whereas Africa and America got roughly around 1% each over the years. It is to be noted that among the Asian regions, eastern, southern and south eastern regions received higher share in 1995; however, post 2005, China emerged as an important market among them. Developed economies received 30% of all food items export from economies in transition in 1995, and since then over the decade its share pegged around 23%. Among the developed economies, Europe is the major market for all food items with the share of about 25% in 1995 and 32% in 2015. As compared to developed economies and economies in transition, developing economies reflect relatively a more diversified market for their export of all food items (see Table 4.7). In 1995, about 57% of all food items export was received by developed economies, as compared to 40% share of developing economies. However, the share of developed economies has shown some signs of decline over the years but yet accounted for about half of the all food items exported from developing economies in 2005 and mere 38% in 2015. Among developed economies, Europe (28 % in 1995, 25% in 2005 and 18% in 2015), the United States (13% in 1995, 15% in 2005 and 12% in 2015) and Japan (14% in 1995, 9% in 2005 and 5% in 2015) are key markets; although each of them has experienced relative decline in their respective shares. Intra-regional export flow has increased significantly over the period of 20 years. Major share of all food items is received by Asia, around 29–30% over 1995–2005 and around 43% in 2015, predominantly Eastern, Southern and Southeastern Asia with about 23% in 1995–2005, and an increase to 34% in 2015. Export share flowing to economies in transition hovered around 3–4% over the period of 20 years. It can be concluded that South-­South trade in all food items has strengthened over the years; for instance, presently it accounts for more than half of the region’s export of the commodity. Table 4.8 reveals significant North-North trade in raw materials from agricultural sources, accounting for more than 70% in1995 and 2005, although it declined to 58% in 2015. Similar to spatial pattern of all food items export, Europe is the major destination among developed economies for raw materials from agricultural sources, with 48% in 1995–2005 and 45% in 2015. Distantly followed by the United States with 12% in 1995, 13% in 2005 and 17% in 2015. Japan’s share in the total export from developed economies has consistently declined from 11% in 1995, 5% in 2005 and 3% in 2015. On the other hand, economies in transition receive a meagre share. Developing economies are emerging as a potential market for agricultural export from developed economies; their share grew to 41% in 2015 from 23% in 1995. Among developing economies, Asia is the major recipient (accounting for 18% in 1995, 22% in 2005 and 33% in 2015) of raw materials from agricultural sources. Table 4.8 reveals that unlike, all food items, export of raw materials from agricultural sources from economies in transition is relatively less concentrated intra regionally. About 63% in 1995, 48% in 2005 and 32% in 2015 directed to the

4.3  Composition of Merchandise Export

69

developed economies; however, as noted, their share has almost halved. Among developed economies, Europe is a major market though with fluctuating share from 54% to 40% and 45% in 1995, 2005 and 2015, respectively. As compared to this, developing economies have increased their share from 25% in 1995 to 58% in 2015. Among developing economies, Asia (predominantly Eastern, Southern and Southeastern Asian economies, and most interestingly China in the recent decades) is an important destination, because this region received about 22% in 1995, 39% in 2005 and 37% in 2015 of the total share destined to developing economies in the aforesaid years. The intra-group trade has increased marginally from 7% in 1995 to about 12% in 2015. Developing economies have increased trade in raw materials from agricultural sources among themselves from 49% in 1995 to 52% in 2005 and 64% in 2015(Table 4.8). Asia, again, emerged as a major destination among developing economies, with a consistent rise in its share from 40% in 1995 to 44% in 2005 and further to 59% in 2015. Share of raw materials from agricultural source received by developed economies has shown signs of relative decline from 51% in 1995 to 47% in 2005 and further slashed to 35% in 2015. Among developed economies, Europe (23% in 1995, 21% in 2005 and 16% in 2015), the United States (12% in 1995,14% in 2005 and 11% in 2015) and Japan (15% in 1995, 9% in 2005 and 6% in 2015) are dominant markets. Share of economies in transition remained below 1% over the years (1995–2015). Table 4.9 reveals that there is a strong tendency for developed economies to receive a major share of ores and metals exported from developed economies, although the group has registered a consistent decline from 70% in 1995 to 59% in 2005 and further to 44% in 2015. The leading market of ores and metals is Europe among developed economies, with a share of nearly 49%, 46% and 39% in 1995, 2005 and 2015, respectively. Europe was followed by the United States with export share of about 11% in 1995–2005 and 7% in 2015. In the case of developing economies, their share has increased from 22% in 1995 to 35% in 2005 and 51% in 2015. Like other commodity groups, Asia is the dominant market among developing economies for export of ores and metals, receiving 20% in 1995, 26% in 2005 and 43% in 2015 (predominantly Eastern, Southern and Southeastern Asia, and in recent years China also). According to UNCTAD report, ‘Significantly the market share of developed countries has fallen much more for metals, where there are few barriers to trade, than for agricultural commodities, where tariffs are high (especially of processed items) and where developed countries subsidies to domestic producers make it difficult for developing countries to compete’. Table 4.9 shows that economies of transition export more than half of ores and metals to developed economies, for example 56% in 1995; subsequently the relative share has declined. Europe, again, is the leading market among other developed economies, accounting for about 58% in 1995–2005 and increasing to 62% in 2015. On the other hand, share of ores and metals received by developing economies registered a spectacular increase from 11% in 1995 to 32% in 2005 and 42% in 2015. Asia is the main developing market with a consistent rise in share from 7% in 1995 to 19% in 2005 and about 21% in 2015. Major Asian markets are located in the

70

4  An Analysis of the Commodity Composition of International Trade

Southern, Eastern and Southeastern Asia, with China emerging in recent decades. Trade in ores and metals among economies in transition more or less hovered around 30% over the period of 20 years. Similar to the geographical composition of other commodity groups, developing economies’ trade in ores and metals further reiterates the increasing significance of the global South as key markets for export flowing from either developed or developing part of the world. The tectonic shift in the global trading system is largely tied with the remarkable rise of emerging developing countries that has created an alternative demand for commodities of developing South, and hence demand increased noticeably, for example China, India, Brazil and South Africa. In this context, Table 4.9 shows that the developing economies export to their developed counterpart account for about 63% in 1995, which slipped to 46% in 2005 and 32% in 2015. Like other product groups, Europe is the nodal recipient of ores and metals, precious stones and non-monetary gold of developing countries, taken as a group. In this context, it is worth noting that the share of Europe has declined over the years, for example in 1995 it was 28% whereas 16% in 2015. Despite of this declining trend for 20 years, Europe has managed to secure its top market position among other developed markets such as the United States, Canada or Japan. In 1995, the United States received 12% and Japan 16%; although in 2015 their share dipped to 8% and 6%, respectively. Furthermore, economies in transition account for a meagre share of about 1% in 2015 relative to 2% in 1995. It is interesting to note that the quantum of export from economies in transition, as noted earlier in the chapter, to developing countries has significantly multiplied in recent decades. On the contrary, trade among developing economies has almost doubled in the period of 20 years, and intra-group trade has increased from 34% in 1995 to 52% in 2005 and 67% in 2015. Similar to spatial pattern of commodity groups analysed earlier, Asia is again an important market with a share of 36% in 1995 and 44% in 2005; and a major proportion is again received by Southern, Eastern and Southeastern Asia. These regions together accounted for 34%, 45% and 60% in 1995, 2005 and 2015, respectively. The geographical composition of intra-­Asian trade reflects the importance of Eastern, Southern and Southeastern Asia, and also in particular China, for the increasingly rising regionalism in the multilateral trading system. Table 4.10 shows that more than 80% of fuel is traded among developed economies in the years 1995 and 2005; however, their share dipped to 68% in 2015. Europe accounts for more than half of the export destined for developed economies over the years but marginally slipped to 47% in 2015. The United States, on the other hand, receives about 18% in 1995, inched up to 23% in 2005, declining thereafter to 16% in 2015. It could further be seen from Table 4.10 that only around 11% of fuel export from developed economies went to developing economies over the period of 10 years. In 2015, the share increased to 24%. Among developing economies, Asia (predominantly Southern, Eastern and Southeastern Asia) received about 6% per in 1995 and 2005; however, it increased to 9% in 2015. Share of developing America pegged at 4% over the decade, although in 2015 it raised to 10%. As compared to other commodity groups, it is worth noting in the case of fuel

4.3  Composition of Merchandise Export

71

export that the market in the global South has geographically diversified in the recent decades. A significant proportion of fuel export from economies in transition is directed to developed economies, for example 62% in 1995 and 67% in 2005, but dropped to 57% in 2015. Of the total share exported to the global North, Europe accounted for 60% in 1995 and 2015. The share of developing economies hovered around 9 per over 10 years but increased to 21% in 2015. Among developing markets, in 1995, America received 6% and Asia 4%. It is interesting to note that in the period of 20 years, Asia has shown a marked increase in its share to 14% in 2015, and developing America share hovered around 4%. Eastern, Southern and Southeastern Asia (2% in 1995, 4% in 2005 and 13% in 2015) and Western Asia (2% in 1995, 3% in 2005 and 0.7% in 2015) are major Asian destinations. Also to be noted, China has increased its share in the fuel export from economies in transition over the years, from 0.1% in 1995 to 7% in 2015. Intra-group trade has registered a steep decline to 7% in the 2000s from 23% in 1995 (Table 4.10). Developing economies sends more than half of their fuel export to developed economies over the period of 10 years. It can be seen from Table 4.10 that the share of developed economies has halved from 1995 to 2015. Unlike other commodities, major markets of developing economies are relatively less geographically concentrated, for instance Europe (20% in 1995, 17% in 2005 and 12% in 2015), the United States (19% in 1995, 22% in 2005 and 10% in 2015) and Japan (21% in 1995, 14% in 2005 and 11% in 2015) are the key recipients. South-South fuel trade has almost doubled over the years. In 1995, developing economies received 36% of fuel export, which increased to 43% in 2005 and further to 65% in 2015. It is worth stating here that fuel is heavily localized, such as in the Middle East and within a few other areas, such as Indonesia and Malaysia, and also in some African countries. Among developing economies, Asia is the key destination with a share of 28% in 1995, 35% in 2005 and 55% in 2015. Southern, Eastern, Southeastern Asia is the dominant Asian market for fuel export of developing economies. Fuel export to economies in transition is insignificant (see Table 4.10). Table 4.11 summarizes spatial structure of trade in manufactures. Developed economies predominantly traded manufactured goods among themselves, for example out of the total export, 74% goes to their developed counterparts in 1995 and 2005. In 2015, despite marginal decline to 69%, North-North trade in manufactures is significant. Out of this proportion, about 50% is destined to Europe in the reference years. Developing economies, on the other hand, received around 24% of manufactured goods from developed economies in 1995 and 2005; however, the share increased to 29% in 2015. Asia accounts for around 17% of manufactures received by developing economies in 1995 and about 19% in 2015. Economies in transition received only about 2% of manufactured goods exported from developed economies. Table 4.11 shows that economies in transition export to developed economies hovered around 35% in 1995 and 2005 and increased to 40% in 2015,to developed economies. Among developed economies, Europe is the major recipient with about 27% in 1995, 31% in 2005 and 44% in 2015. The share of manufacture trade among

72

4  An Analysis of the Commodity Composition of International Trade

economies in transition hovered around 32% in the period of 20 years. However, the share of developing economies was 30%, 33% and 28% in 1995, 2005 and 2015, respectively. Asia (largely directed to the Southern, Eastern and Southeastern regions) received about 26% of the total manufactures destined to developing economies in 1995 and 2005, and its share decreased to 14% in 2015. Table 4.11 shows that developing economies export relatively larger proportion of manufactured goods to the developed part of the world. For example, developed economies received about 55% of manufactures in 1995, 52% in 2005 and 42% in 2015. It is interesting to note that unlike the regional structure of commodity groups analyzed earlier in the chapter, among developed economies, the United States receives about 26% in 1995, 23% in 2005 and 18% in 2015. The other market is Europe with around 18% over 1995 and 2005, though registering a slight dip to 15% in 2015. Japan received about 9% in 1995 and 7% in 2005 and mere 5% in 2015. Developing economies received 44% of manufactured goods in 1995; its share increased to about 47% in 2005 and 56% in 2015. Asia is the major recipient among other developing countries, with about 36% in 1995 and 46% in 2015. Similar to other commodity groups, Southern, Eastern and Southeastern economies received major share with about 36% in 1995 and 46% in 2015, among other Asian economies. Hence, it is quite clear that South-South trade in manufactures is also expanding rapidly. Economies in transition accounted for a meagre share of 1% in 1995 and 2% in 2005 and 2015. On a similar note, it is worth to highlight that the old geography of international trade has been moulded by colonialism. The Industrial Revolution aided colonial powers to achieve technological superiority and enabled them to occupy key positions in global economic relations with respect to developing economies. This unbalanced pattern persisted even in the post-colonial era. Such a situation triggered varying rates of development in different economies across the globe. Such a situation infuses inter-regional disparities in the global economy, which tend to be high, and poor economies would continue to remain depressed. In addition to what was said earlier, Raza and Agarwal (1986) discussed, ‘In the absence of a balancing mechanism in inter-regional exchange, it is possible for a region to draw on the resources of another without a corresponding inflow into the latter from the former’. They analysed inter-regional commodity flows to probe into such ‘suction mechanisms’ within the economy. However, quite a transformation is reshaping the global economy, and how trade is happening in present times. The centuries-­old international trade where developing economies served as hinterlands of resources and markets for processed products of the traditional nodes, that is developed economies, also is changing. The study of inter-regional commodity flows could also be seen from the analysis stated earlier. However, the present analysis of suction mechanism would be confined to two major commodity groups, viz., raw materials from agricultural sources and manufactured goods. According to the foregoing analysis, high proportion of raw materials from agricultural sources is traded among developed economies vis-a-vis developing economies. However, Table 4.8 shows that from 1995 to 2015, this proportion has consistently declined and that of developing economies

4.3  Composition of Merchandise Export

73

has almost doubled; nevertheless, the North-North trade accounts for more than half of the North’s export of raw materials from agricultural sources even in the recent years. On the other hand, developing economies export a relatively higher share of raw materials from agricultural sources to developed economies vis-a-vis developing economies in 1995. It is also important to note in this context, in 20 years the share received by developing economies has markedly surpassed the share going to developed economies from them. If inter-regional trade of manufactures is looked into (Table  4.11), a major share from developed economies is exported to developed countries as compared to developing economies. Unlike raw materials from agricultural sources, the relative share of these two major regional groups in export of developed economies broadly remains unchanged through the considered period. In 1995 and 2005, more than 74% of manufactured goods went to developed economies, but in 2015 their share slipped to about 69%. In case of export from developing economies to developing countries of manufactured goods,, a reverse pattern is observed. In 1995, about 44% went to developing economies, 47% in 2005; thereafter, their share increased to around 56%. It is worth noting that the stark difference in the share of manufactures marketed in developed and developing world is maintained in the context of developed economies export over the period of 20 years; however, in the case of developing economies, gap in the export market share is not that acute, and also trade among developing economies has gained momentum in the recent decades. This broadly reflects the changing global trading landscape and the emergence of new spaces of flows. However, the dominance of developed economies in international trade is still evident, but developing economies are steadily moving away from the hinterlands of the world economic system driven by trade. One noticeable aspect of this trend is the swelling South-South trade and a ‘new geography of trade’ is emerging. The traditional global trade pattern, that is primary commodity exports from developing economies and manufactures export from developed economies, is being replaced by a more complex pattern. Outward-­ oriented strategies, liberalization, regional trade agreements and the growing role of manufacturing networks covering many countries have given a strong boost to this process of development. These networks are particularly well developed in East Asia. Countries such as China, Korea, Malaysia, Singapore, Thailand, other East Asian economies and lately India have emerged as new growth poles in the global industrial economy, of course with differential rate and nature of integration. However, it is important to note here that not all developing economies are dynamic players at the international economic platform, though there are signs of improved performance over the years. In line with this, UNCTAD (2004) points out, ‘Strategic policies and actions by more successful developing countries and their firms, TNC strategies and the globalization of production systems, mobility of factors of production and business, changes in demand patterns and market access conditions, shifts in factor intensity of tradables, cost competitiveness and technological changes (including ICTs) are among the factors that contribute to the performance of the South’. Furthermore it is noted that the ‘primacy’ of the developed economies (popularly called the North) in international economic relations will persist. Also

74

4  An Analysis of the Commodity Composition of International Trade

the gap between the ‘North’ and the ‘South’ is quite stark, and would take a very long period to fade. Further trade expansion within the developing country group has reflected especially their emerging status as both exporter and importer. In this context, Athurkorala (2011) pointed out the multiplying trade complementarity among countries from the global South. Their production structure has diversified over the years, and the regional division of labour within production networks has further intensified the trade complementarities among these economies. A broad overview of the trading pattern in terms of its composition of the three major country groups, viz., developed, developing and economies in transition, sheds some light on the interesting similarities and differences in the preceding paragraphs. An attempt is also made, subsequently, to shed some light on the variations in the concentration of commodities in the export and import basket of various countries constituting these broad country groups. Tables 4.12, 4.13, 4.14 and 4.15 show country classification on the basis of concentration of commodities in export and import based on the value of location quotient. Table 4.12 shows that in 1995, commodity composition of export of Ireland, El Salvador, Panama and Mauritius has highly concentrated in a single commodity, viz., all food items. However, in 2005 and 2015, Ireland has diversified its export base from only all food items to multi-commodity with manufactured goods (along with all food items). Electrical machinery and apparatus, processed foods, chemical products, clothing and textiles and beverages are important export commodities. Also Panama’s export diversified to multi-commodity with primary commodities. In 2005, in addition to El Salvador and Mauritius, Iceland export concentrated on a single commodity from multiple primary commodities (Table  4.13). In 2015, Iceland and El Salvador diversified their export to multi-commodity with primary Table 4.12  Commodity composition of global export basket (1995) Single commodity group All food items – El Salvador, Ireland, Mauritius, Panama Raw materials from agricultural source Ores and metals – Guinea, Zambia Fuels – Algeria, Angola, Nigeria, Brunei Darussalam, Iran, Kuwait, Saudi Arabia, Trinidad and Tobago Manufactured goods – Germany, Italy, Malta, Sweden, Switzerland, UK, China, Japan, South Korea, Singapore Multi-commodity group Primary commodities – Canada, Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, Bahrain, Bulgaria, Honduras, Paraguay, Peru, Uruguay, Venezuela, Cameroon, Côte d’Ivoire, Egypt, Gabon, Ghana, Guatemala, Kenya, Morocco, South Africa, Zimbabwe Denmark, Greece, Hungary, Iceland, Netherlands, Norway, Poland, India, Indonesia, Jordan, Malaysia, Mongolia, Philippines, Sri Lanka, Syrian Arab Republic, Thailand, Turkey, UAE, Vietnam, USSR (former), Fiji, New Zealand Combination with manufactured goods – Austria, Belgium-Luxembourg, Czechoslovakia (former), Finland, France, Portugal, Romania, Spain, Yugoslavia SFR (former), USA, Mexico, Dominican Republic, Bangladesh, Pakistan, Senegal, Tunisia Source: Based on Annexure 4.1

4.3  Composition of Merchandise Export

75

Table 4.13  Commodity composition of global export basket (2005) Single commodity group All food items – El Salvador, Iceland, Mauritius Raw materials from agricultural source Ores and metals Fuels – Algeria, Angola, Nigeria, Trinidad & Tobago, Venezuela, Brunei Darussalam, Kuwait, Saudi Arabia, Iran, UAE Manufactured goods – Czechoslovakia (former), Germany, UK, Hungary, Italy, Malta, Switzerland, Bangladesh, China, Japan, Philippines, Singapore, South Korea Multi-commodity group Primary commodities – Bulgaria, Denmark, Netherlands, Norway, Guatemala, Canada, Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, Honduras, Panama, Paraguay, Peru, Uruguay, Bahrain, Indonesia, India, Jordan, Mongolia, Sri Lanka, Syrian Arab Republic, Vietnam, Cameroon, Cote d’Ivoire, Egypt, Fiji, Gabon, Ghana, Guinea, Kenya, Morocco, Zambia, Zimbabwe, Senegal, South Africa, New Zealand, USSR (former) Combination with manufactured goods – Austria, Belgium-Luxembourg, Finland, France, Greece, Ireland, Poland, Portugal, Romania, Spain, Sweden, Turkey, Yugoslavia SFR (former), USA, Mexico, Malaysia, Pakistan, Thailand, Dominican Republic, Tunisia Source: Based Annexure 4.2 Table 4.14  Commodity composition of global export basket (2015) Single commodity group All food items – Argentina, Mauritius Raw materials from agricultural source Ores and metals Fuels – Algeria, Angola, Brunei Darussalam, Kuwait, Malta, Nigeria, Saudi Arabia, Trinidad and Tobago, Venezuela Manufactured goods – Bangladesh, China, Czechoslovakia (former), Germany, Hungary, Italy, Japan, South Korea, Mexico, Switzerland Multi-commodity group Primary commodities – Bahrain, Bolivia, Brazil, Bulgaria, Cameroon, Canada, Chile, Colombia, Costa Rica, Cote d’Ivoire, Denmark, Dominican Republic, Ecuador, Egypt, Fiji, Gabon, Ghana, Greece, Guatemala, Guinea, Honduras, Iceland, India, Indonesia, Iran, Kenya, Malaysia, Mongolia, Morocco, Netherlands, New Zealand, Norway, Panama, Paraguay, Peru, Senegal, South Africa, Spain, Sri Lanka, Syrian Arab Republic, UAE, Uruguay, USSR (former), Zambia, Zimbabwe Combination with manufactured goods – Austria, Belgium-Luxembourg, El Salvador, Finland, France, Ireland, Jordan, Pakistan, Philippines, Poland, Portugal, Romania, Singapore, Thailand, Tunisia, Turkey, UK, USA, Vietnam, Yugoslavia SFR (former) Source: Based on Annexure 4.3

commodities and manufactures, respectively. Furthermore, over the period of 20 years, Argentina (see Tables 4.12, 4.13 and 4.14) has increasingly specialized in export of all food items in 2015, as compared to multiple primary commodities during previous years. It is worth noting here that these economies have experienced growth in their merchandise trade at varying rates through the period under consideration (See Chap. 2).

76

4  An Analysis of the Commodity Composition of International Trade

Tables 4.12, 4.13 and 4.14 show that in 1995, 2005 and 2015, none of the country’s export is concentrated in raw materials from agricultural sources; however, a large number of them fall in the group of multi-commodity with primary commodities, having raw materials from agricultural sources as one of their export item. As far as ores and metals are considered, it is quite clear that in 1995 only two African economies, viz., Guinea and Zambia, showed heavy export reliance on ores and metals. Africa is known for its abundance in ores and metals. It is interesting to note that 2005 onwards, these economies shifted their export concentration to more diversified commodity composition with primary commodities; however, ores and metals remain an important export item. During 1990–1995, merchandise trade recorded negative growth rate in Zambia and low in Guinea. On the other hand, in 2000–2005 and 2010–2015, trade grew moderately for Guinea, though Zambia’s trade grew at a very high rate during 2000–2005 and sluggishly in the period 2010–2015. These countries are heavily dependent on export of only oil and oilrelated products. Economies heavily dependent on export of only fuel and related products, in 1995, were mainly Asian and African economies, such as Brunei Darussalam, Iran, Kuwait, Saudi Arabia, Algeria, Angola and Nigeria. In 2005, Venezuela and the UAE joined this group of fuel exporters. Besides the existing countries, in 2015, Malta has shifted from manufactures export to fuels. Unlike in 2005, UAE export diverse primary commodities in 2015. Tables 4.12, 4.13 and 4.14 further reveal that developed European economies like the United Kingdom, Germany, Switzerland, Italy and Malta and Asian economies like Japan, China, South Korea and Singapore continued specialization in manufactured goods in 1995 and 2005. Sweden and the United Kingdom diversified its export base from single commodity to multi commodities with manufactures in 2005 and 2015, United Kingdom joined this group in the latter year. Merchandise trade expanded in all these economies at differential rates in 1995–2000 and 2000–2005 (see Figs. 2.4, 2.5, and 2.6). Switzerland registered an increase to 8% from 0.4%, Germany 11% from mere 1%, Italy 10% from 2%, the United Kingdom 7% from 5%, Sweden to 9% from 2%, China to 25% from 11%, Singapore to 10% from 2%, Japan to 5% from 2%, Korea to 10% from 5% in 1995–2000 and 2000–05, respectively. Malta, on the other hand, registered a decline to 0.64% from 4%. There are also new entrants like Czechoslovakia (former), Hungary, Philippines and Bangladesh in 2005; excluding the Philippines, these countries persisted with specialization in manufactures. In 1995, countries such as Czechoslovakia (former) and Bangladesh had diversified export base with manufacturing goods; on the other hand, Hungary and the Philippines exported primary products. In 2015, the Philippines has diversified its export to multiple commodities with manufactures. This change is reflective of the changing economic structure of the Philippines from one based on agriculture to manufacturing, and its drive towards becoming a newly industrialized country among other Southeast Asian economies is discussed later in the chapter. It is important to note that the nature of manufacturing goods exported by developed and developing countries varies considerably in terms of technology and capital intensity; moreover it is largely labour-intensive manufactures dominating the export basket of the global South relative to developed trading cores.

4.3  Composition of Merchandise Export

77

The concentration in export of primary commodities (includes raw materials from agricultural sources, all food items, fuels, and ores and metals) is spread across developing America, Asia, Africa, Oceania and a few developed countries. Subsequent trade and financial reforms, particularly during the 1990s, triggered export-led growth. However, export growth was not built on dynamic industrialization in many cases. It was either based on a continued or in some cases deepening reliance on export of primary commodities, particularly in South America, or on assembling manufacturing processes, for example in case of Mexico and Central America (Ocampo and Vos 2008). This could also be seen from Tables 4.12, 4.13 and 4.14, for example South American economies like Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Peru and Uruguay, and also Central American economies like Costa Rica, Guatemala and Honduras are mainly exporting primary commodities over the period from 1995 to 2015. Panama also joined this group in 2005 and persisted in 2015; however, in 1995 its concentration of exports was on single commodity group  – all food items. The only difference observed is it is either increasing concentration or diversification, but the nature of commodity composition shows visible tilt in favour of primary commodities, for example Argentina and Panama. The export base of Latin American economies indicates their fundamental comparative advantage in primary commodities. The region-­specific characteristics with the example of MERCOSUR are later dealt with in chapter six of the book. However, the shift towards manufactures in South America mainly reflects the rapid growth of manufactured exports from Mexico (UNCTAD 2005). Further, it is added by the UNCTAD report, ‘Several smaller Central American and Caribbean countries, which formerly specialized in food and beverages, have also become exporters of manufactures, owing largely to the expansion of their maquila plants’. Similarly, in Asia, Bahrain, India, Indonesia, Jordan, Malaysia, Mongolia, the Philippines, Sri Lanka, Thailand, Turkey, the UAE and Vietnam are primary commodity exporters in 1995. However, it is worth noting that with the expansion of trade in 2005, there is a mark shift in the export base from primary commodities to manufactures, for example Malaysia, Thailand, the Philippines and Vietnam over the years (see Tables 4.12, 4.13 and 4.14). African economies such as Cameroon, Cote d’Ivoire, Egypt, Gabon, Ghana, Kenya, Morocco, South Africa, Syria Arab Republic and Zimbabwe (in 2005, new entrants in this group are Gabon, Senegal, Zambia and Guinea) show a clear predominance of primary commodities in their export base over the period of 20 years. This is mainly because of the abundance of natural resources and cheap labour; they have comparative advantage in primary commodities, like all food items, raw materials from agricultural sources and, more importantly, ores and metals and fuels. According to UNCTAD (2006), the comparative advantage of sub-Saharan Africa ‘is gradually shifting towards countries with similar resource endowments that have additional capabilities, ranging from quality control systems to the widespread availability of ICT, that play a role in securing market shares. In absolute terms, the region is also the smallest exporter’. Poon and Rigby (2017) have noted that the concentration of exports in primary commodities create problems for developing

78

4  An Analysis of the Commodity Composition of International Trade

countries ‘if the exports are subjected to unstable markets through volatility in prices and deteriorating terms of trade.’ There are also a set of developed economies exporting largely primary commodities over the years, such as Canada, Bulgaria, Denmark, Iceland, the Netherlands, Norway, the USSR (former) and New Zealand. Canada is unusual among developed economies, in terms of importance of primary sector. It is not only one among the few developed economies that are key exporters of energy but also of agricultural products. The Netherlands also holds key position worldwide in value of agricultural export. There are also examples of developing countries exporting diversified commodities such as raw materials from agricultural source, all food items, ores and metals and fuels along with manufactured goods. In 1995, industrialized economies such as Austria, Belgium-Luxembourg, Czechoslovakia (former), Finland, France, Portugal, Spain, Yugoslavia (former), Romania and the United States exported multi commodities with manufactured goods. In 2005, more or less these economies persisted with their commodity composition, except Czechoslovakia (former), which has specialised its export in single commodity, that is manufactured goods. In addition to developed countries, developing economies like Bangladesh, Mexico, Pakistan, the Dominican Republic, Senegal and Tunisia are also in this group (Table 4.12). However, over the years, there is a visible change in export basket of these economies (Tables 4.13 and 4.14), for example Bangladesh and Mexico have in the later years concentrated in exporting manufactures only; the Dominican Republic and Senegal tilted towards primary goods export. Also in the 2000s Poland and Ireland moved to the group of diversified export base with manufactures as compared to their 1990s pattern. Turkey also joins this group in latter years. Tables 4.15, 4.16, 4.17 show relative concentration/diversification of merchandise import of the selected countries for 1995, 2005 and 2015, respectively. It is interesting to draw comparison of import composition of developed and developing countries. For example, developing American countries such as Guatemala, and Uruguay have diversified import and export with primary commodities over the years. On the other hand, Brazil, Colombia, Costa Rica, Paraguay and Venezuela have a diverse import basket with manufactures; however, export is diverse with a predominance of primary goods. Similarly, a majority of African countries, in the selected sample, export is diversified with primary commodities and import it too. Algeria, Angola and Nigeria export is highly specialized with a single commodity, that is fuels. However, their import composition varies. Algeria imported diversified primary commodities in 1990s and shifted to manufactures in 2000s. Angola has changed from importing primarily all food items to diversified commodities with manufactures. On the other hand, Nigeria shifted from diversified manufacturing in 1995 to import of all food items in 2015 (refer Tables 4.15, 4.16 and 4.17). Countries such as South Africa, Zambia and Zimbabwe have witnessed increasing concentration of their import in the latter year. In 1995 and 2005 South Africa and Zimbabwe import was diversified with manufactures whereas in 2015 the former imported fuels; and latter all food items. There are also countries, namely Cameroon, Cote de I’voire, Egypt, Guinea,

4.3  Composition of Merchandise Export

79

Table 4.15  Commodity composition of global import basket (1995) Single commodity group All food items – Angola, Iran, Paraguay, Greece, Netherlands Raw materials from agricultural source Ores and metals Fuels – Romania Manufactured goods – Canada, Argentina, Ecuador, Mexico, Ghana, Ireland, Sweden, Switzerland, Malaysia, New Zealand Multi-commodity group Primary products – Algeria, Egypt, Cameroon, Cote d’Ivoire, Guatemala, Guinea, Kenya, Mauritius, Morocco, Senegal, Tunisia, Zambia, Bahrain, Bangladesh, China, Jordan, Indonesia, Mongolia, Japan, Pakistan, Philippines, India, South Korea, USSR (former), Turkey, Belgium-­ Luxembourg, Bulgaria, Iceland, Italy, Poland, Portugal, Spain,, Brazil, El Salvador, Fiji, Honduras, Panama, Uruguay Manufactured goods – Bolivia, Chile, Venezuela, Colombia, Costa Rica, Dominican Republican, Peru, Trinidad and Tobago, Gabon, Brunei Darussalam, Kuwait, Saudi Arabia, Singapore, Sri Lanka, Syria, UAE, Vietnam, Nigeria, France, UK, Thailand, Norway, USA, Zimbabwe, Denmark, South Africa, Malta, Germany, Finland, Yugoslavia SFR (former), Austria, Czechoslovakia (former), Hungary Source: Based on Annexure 4.4 Table 4.16  Commodity composition of global import basket (2005) Single commodity group All food items – Dominican Republic, Nigeria, Spain, USSR (former) Raw materials from Agricultural source Ores and metals Fuels – Vietnam Manufactured goods – Canada, Costa Rica, Mexico, Philippines, Hungary, Romania, Zambia Multi-Commodity Group Primary Products – Bahrain, Bangladesh, India, Indonesia, Japan, Pakistan, South Korea, Sri Lanka, Thailand, Cameroon, Egypt, Guinea, Kenya, Mauritius, Morocco, Cote d’Ivoire, Senegal, Tunisia, Zimbabwe, Brazil, Chile, El Salvador, Guatemala, Honduras, Panama, Peru, Uruguay, Trinidad and Tobago, Finland, Greece, Italy, Netherlands, Portugal, Yugoslavia SFR (former) Manufactured goods – Algeria, Angola, Gabon, Ghana, South Africa, Turkey, Bolivia, Brunei Darussalam, China, Iran, Kuwait, Malaysia, Saudi Arabia, Singapore, Syria, UAE, Argentina, Colombia, Ecuador, Paraguay, Venezuela, Austria, Belgium-Luxembourg, Bulgaria, Czechoslovakia (former), Denmark, France, Germany, Iceland, Ireland, Malta, Norway, Poland, Sweden, Switzerland, UK, New Zealand, USA Source: Based on Annexure 4.5

Kenya, Morocco and Senegal, that have been importing and exporting diversified primary commodities over the two decades. Mixed pattern is visible for Asia. There are a number of countries consistently importing diverse primary commodities in the period of 20  years, for example Bahrain, Jordon, Bangladesh, Pakistan, India, Mongolia, Indonesia and South Korea. Malaysia, Thailand and Sri Lanka have come to import diverse primary

80

4  An Analysis of the Commodity Composition of International Trade

Table 4.17  Commodity composition of global import basket (2015) Single commodity group All food items – Nigeria, USSR (former),Zimbabwe Raw materials from agricultural source Ores and metals – Zambia Fuels – Malta, Panama, Singapore, South Africa, Manufactured goods – Argentina, Bolivia, Canada, Czechoslovakia (former), Hungary, Mexico, Switzerland, USA Multi-commodity group Primary products – Bahrain, Bangladesh, Belgium-Luxembourg, Bulgaria, Cameroon, China, Cote d’Ivoire, Dominican Republic, Ecuador, Egypt, El Salvador, Fiji, Finland, Ghana, Greece, Guatemala, Guinea, Honduras, Iceland, India, Indonesia, Italy, Japan, Jordan, Kenya, South Korea, Malaysia, Mauritius, Mongolia, Morocco, Netherlands, Pakistan, Portugal, Senegal, Spain, Sri Lanka, Syria, Thailand, Trinidad and Tobago, Tunisia, Uruguay, Yugoslavia (former) Manufactured goods – Algeria, Angola, Austria, Brazil, Brunei Darussalam, Chile, Colombia, Costa Rica, Denmark, France, Gabon, Germany, Iran, Ireland, Kuwait, New Zealand, Norway, Paraguay, Peru, Philippines, Poland, Romania, Saudi Arabia, Sweden, Turkey, UAE, UK, Venezuela, Vietnam Source: Based on Annexure 4.6

commodities by 2015. West Asian economies such as Brunei Darussalam, the UAE, Kuwait and Saudi Arabia recorded to have diverse import with manufactures and export heavily concentrated on fuels. Southeast Asian countries, namely Philippines, Singapore and Vietnam, have witnessed changing relative concentration of different commodities. The Philippines import basket has changed from diversified primary products to diverse import with manufactures; on the other hand, Singapore primarily imports fuels in 2015 as compared to its diversified import basket with manufactures till 2005. China imports diverse primary commodities, and its export is highly specialised in manufactures, as noted earlier in the chapter. The pattern of commodity concentration in the import basket of most developed countries reveals significance of manufactures in case of single commodity concentration or multi-commodity structure. For example, the United States, Canada, Hungry, Czechoslovakia (former) and Switzerland import heavily concentrate on manufactures by 2015. Countries such as Romania, Ireland, Sweden, Switzerland, Germany, France, the United Kingdom, Norway, Denmark, Germany, and Austria have a multi-commodity import basket with manufactured goods. Then there are countries with a preponderance of primary commodities as compared to manufactures, namely Greece, the Netherlands, Japan, Italy, Portugal, Spain and Finland. The USSR (former) import has become highly concentrated with all food items since 2005. As noted in the preceding discussions, most of these countries also heavily export manufactured goods, however, with relative variability.

4.4  Concluding Remarks

81

4.4  Concluding Remarks It is important to note that it is difficult to draw meaningful conclusions about the collective trade composition and behaviour of these 87 economies, on the basis of aggregate figures, as they make up such a large and heterogeneous group. However, it does reflect on the compositional structure of the export and import basket and the degree of concentration/diversification that has taken place over the years. The present analysis provides a broad understanding of the trade structure of different countries and in that context the emerging nature of interdependency in the multilateral trade system. In the light of this statement, the following are the major conclusions drawn: (i) From the foregoing analysis, it is quite clear that manufactured goods have gained vis-a-vis primary commodities. Also developed economies continue to dominate the world merchandise export; however, developing economies have shown signs of improvement in terms of their share in the global export of the selected commodities over the years. (ii) Over the period of 20 years, patterns of international trade have been changing in favour of trade between developed and developing countries. Developed countries still trade mostly among themselves, but the share of their exports going to developing countries has also risen. At the same time, developing countries have increased trade among themselves, reflecting at the increasing South-South trade. Although, developed countries remain their main trading partners, the key markets for their export and the main source of their import. (iii) A brief analysis of the centre-periphery relationship in international trade flows shows a slight deviation from the suction model. It has come to the fore from the pattern of trade in raw materials from agricultural sources and manufactured goods that the traditional global trade pattern – primary commodity exports from developing countries, and manufactures exports from developed countries  – is being replaced by a more complex pattern. The dominance of developed economies in global trade is still evident, but developing economies are moving steadily away from the hinterlands of the world trading economy. It could be observed that a new geography of trade is emerging, particularly with the coming up of Asian tigers. (iv) industrially advanced economies increasingly export and import manufactures and they also account for a higher share of export of food products vis-a-vis developing economies. (v)The structure of export and import of economies has not shown rapid change over the period of 20 years (1995–2015). Industrialized economies such as the United States, Austria, Belgium-Luxembourg, France, Finland and Poland have multi-commodity export and import basket with manufactured goods. On the other hand, for European countries like Germany, Italy and the United Kingdom, export relies heavily on single commodity, viz., manufactured goods. As for their import basket, the United Kingdom and Germany have multi-commodity with a combination of manufactures and that of Italy is multi-commodity with a tilt towards primary commodities. (vi) The structure of exports of developing economies has not shown any marked changes. Developing Africa and America have been the slowest regions to diversify away from exports of primary commodities, and the picture slightly varies across Asia (vii) As compared to other developing economies, export

82

4  An Analysis of the Commodity Composition of International Trade

of first-tier industrialising economies like China, Singapore and South Korea heavily concentrates on a single commodity, that is manufactures. All regions lost market share to the first-tier newly industrialized economies, that is, Southeast Asia and China. This contributed to the growth divergence among developing economies. It will be interesting to further study the impact of changing power dynamics based on the compositional structure of the trade basket of developed and developing countries in the global trading system. In light of the discussion, it is worth raising/reiterating the question why do a large number of developing countries still heavily concentrate on primary commodities as their main source of export earnings. Is it sustainable for less developed and developing countries? The answer, partly, lies in their level of development and nature of global interdependency since the colonial times. It is widely accepted that the structure of export and import does get reflected in the pattern of spatial dependency. In this context, it is pertinent to note that the comparative advantage in primary commodities that is primarily characterized by price volatility in the global market and adverse terms of trade might keep these less developed countries trapped in the ‘cycle of underdevelopment’. Having said this, it is worth to mention that in order to comprehend the emerging complex structure and underlying dynamics of the global trade, it is necessary to undertake a detailed study commodity wise as well as country wise, and that is beyond the scope of the present study.

References Athurkorala P-C (2011) South-South trade: an Asian perspective. ADB Economics Working Paper Series no 265 Grant R (1994) The geography of international trade. Prog Hum Geogr 18(3):298–312 Gruss B (2014) After the boom—commodity prices and economic growth in Latin America and the Caribbean. IMF Working Paper WP/14/154 Kristjanpoller W, Olson Josephine E, Salazar Rodolfo I (2016) Does the commodities boom support the export led growth hypothesis? Evidence from Latin American countries. Latin Am Econ Rev 25(6). https://doi.org/10.1007/s40503-­016-­0036-­z Ocampo JA, Rob V (eds) (2008) Uneven economic development. Orient Longman pvt. ltd., Hyderabad Poon J, Rigby DL (2017) International trade: the basics. Routledge, London/New York Raza M, Agarwal Y (1986) Transport geography of India: commodity flows and the regional structure of Indian economy. Concept Publishing Company, New Delhi UNCTAD (2004) New geography of international trade: South-South cooperation in an increasingly interdependent world. http://unctad.org. Accessed 17 Jan 2017 UNCTAD (2005) Trade and development report. United Nations, New York/Geneva UNCTAD (2006) World economic situation and prospects. United Nations, Geneva UNCTAD (2008) Globalization for development: the international trade perspective. United Nations, New York/Geneva

Chapter 5

Output and Trade Relation

Abstract  The impact of openness captured via trade has become a serious issue for several economies. Over the years, there have been an increasing concern and empirical evidence challenging the very export-led growth strategies and also differential and its limited positive impact on economic growth. A number of models exist in the literature to study the causality between foreign sector on the domestic economy and vice-versa. This chapter briefly covers the literature about the changing perspective on the trade and output relationship. Furthermore, an empirical analysis is done to capture causality from the gross domestic product (GDP) to trade with the particular focus on 1990 and 2015. The results revealed that the variation in GDP explained a significant variation in the export values. However, residual mapping did show mixed results across countries. Hence, this chapter also reflects on the fact that besides GDP there are other factors that could affect trade, for example geopolitical environment, labor market conditions, capital inflow, institutions and also economic reforms. Keywords  Trade · Economic growth · Export led growth strategy · Regression and residual mapping

5.1  Introduction The growth process is highly uneven across space and time. To capture this process from the geographical perspective, it is important to understand the patterns and processes embedded in space. Johnston (1976) has presented a comprehensive discussion on trade and development in the spatial context. He pointed out that with regard to this, development relates to the improvement of the standard of living in a number of countries occupying a particular position in the global trade system. He further adds, ‘It is assumed that alteration of the trade system or of the trading patterns of countries, will assist in the development process, thus inferring a temporal and causal relationship from the cross-sectional associations between trade parameters and development levels’. Broadly, such spatio-temporal patterns reflect differences in economic © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9_5

83

84

5  Output and Trade Relation

organization, resource endowment, demographic characteristics, political system, relative roles in the global economic system and the degree of global integration. Among other factors, trade has often been considered to be an essential driver of growth for countries, particularly the developing part of the world. Maddison (2006) has discussed about three interactive processes which could have possibly sustained economic performance. These processes encompass: conquest of relatively empty regions which had productive land, new resources and also possessed capability to support transfer of population, crops etc.; international trade and capital flows; technological and institutional innovation. It has also been widely accepted that economic growth is an extremely complex process, which depends on a number of factors like trade, fluctuations in price, capital accumulation including both physical and human, distribution of income, political situations and even more on geographical characteristics (Medina-Smith 2001). Of all factors, export has come out as one of the key means to an end called economic growth and development. Farole (2013) has noted that trade has a major role to play in the interaction between location, growth and inequality. Trade has increasingly become a key driver of growth, both in the short and long run. The mercantilists are known to be the first to consider export promotion as a key part of their economic policy and were also criticized for their erroneous reasoning by Adam Smith (Afxentiou and Serletis 1991). However, in modern times, exports are examined in the light of general equilibrium analysis, and further to this approach empirical international relationships between exports and other variables, particularly GNP, have also seen. Several scholars, like Keesing (1967), Bhagwati(1978), Krueger (1978), Chenery and Strout (1966), Voivodas(1973) and Scitovsky (1954), have discussed various gains of export growth, such as opening of economies to global communication regarding new ideas and methods of production which strengthen competitive behaviour and boost efficiency; also the ability of small countries to overcome market constraints and benefit from cost advantages of increasing returns to scale; relaxing foreign exchange limitations that normally interrupt development initiatives; and also positive externalities which affect the entire economic structure of an economy. The gains from export expansion, as stated in the preceding paragraph, according to Afxentiou and Serletis (1991), lose some of their significance in industrialized economies. In line with this, Afxentiou and Serletis (1991) put to the fore, ‘Generally any increase in exports led to growth of income in either developed and developing economies, this is mainly due to the fact that exports are one of the components of national income. However, positive externalities from exports which are generally expected in developing economies, are significantly reduced by the advance infrastructure of developed economies. Further, industrialized countries which are already competitive, and due to their industrialized nature they have continuously adopted advanced production techniques, therefore, neither the enhanced competitive behaviour nor diffusion of advanced production techniques due to export, reap much of the expected benefits’. In light of the arguments, the present chapter would deal with one of the key process, that is international trade, in relation to economic growth, a relationship subject to considerable controversy. There are several attempts undertaken over the years to work out the relationship between the two, but yet there is no common consensus.

5.2 Relation Between Economic Growth and Export: The Background

85

5.2  R  elation Between Economic Growth and Export: The Background Of all factors, export has emerged as one of the key driving factors for economic growth and development. Export-led growth (ELG) is an economic development strategy in which international trade, and more specifically export, is considered to be an engine of a country’s growth and development. There has been a general global shift towards the ELG strategy in recent years. This change has been found to be alike because of the actual and potential economic benefits this strategy accorded to both developing and developed countries (Sentsho 2000). Rapid growth of export is often linked with economic growth in several ways (UNCTAD 1992). Export expansion positively affects growth; it encourages the optimum allocation of resources, enhances capacity utilization, which allows taking advantage of scale economies, and promotes technical change (Balassa 1978, 1985). According to Jung and Marshall (1985), export growth has led to rising output, employment generation and consumption, which boosts demand for a country’s output. Bhat (2011) also sheds some light on the issue. According to him, export leads to growth via export-encouraging policies, for example export subsidies or exchange rate depreciation, which will in turn boost growth. The positive externalities encourage economic growth. The reverse side of this argument puts forth that the economic growth promotes exports which depend on the gains in productivity, which will lead to comparative advantages in certain sectors, which in turn result in the growth of export. Several studies were undertaken to explore the relationship between export and economic growth (Refer Annexure 5.1 for summary). In line with this approach, extensive studies were done by Greenaway and Sapsford (1994, Shan and Sun (1998), Marin (1992), Serletis (1992), Hodne (1994), Henriques and Sadorsky (1996), Islam (1998), Ghatak et  al. (1997), Al-Yousif (1997), Ghatak and Price (1997), Sharma and Dhakal (1994), Ukpolo (1994), Balassa (1978), Krueger (1980), Feder (1982), Bhagwati and Shrinivasan (1978), Kavoussi (1984), Ram (1985), Tyler (1981), Fosu (1990) and Kugler and Dridi (1993). However, whether the export-led growth strategy will also be beneficial to the small resource-based economies across the globe, for instance like that of sub-Saharan Africa, is still a debatable issue (Olorunfemi and Olowofeso 2006). Though, the cause and effect link between export and growth of an economy, both from the theoretical and the empirical angles lack universal consensus and turn out to be one of the much debated issues. Some authors, as already mentioned earlier, find causation from exports to GNP, while others have concluded that exports retard GNP growth (Holman and Graves 1995). In contrast, some believe that it is economic growth which multiplies export potential. Among others, Ocampo and Vos (2008) argue that despite sound theoretical arguments, the statistical substantiation of the causal relationship between export and growth is somewhat mixed, and it seems to vary over geographical and temporal scale. They further state that the growth performance is expected to be correlated with the

86

5  Output and Trade Relation

particular composition of the exports that a country chooses over a period of time. On similar lines of argument, Dimkpah (2002) has pointed out that export growth and economic development are positively correlated; however, the impact is stronger in middle-income countries than in low-income countries. According to him, in less developed economies, policies favouring exports would ‘facilitate economic growth’. Colombatto (1990) has pointed out that higher growth rates in export do not necessarily benefit economic growth. Similarly, Banga (2014) stated that tradeled growth is far more complex than it is perceived. She has tried to capture the link via global value chains1 (GVCs) in the context of forward and backward linkages, and also policy should aim at augmenting export of domestic value added contents. In this backdrop, she concludes, ‘linking into GVCs is increasingly being considered as a new development challenge in developing countries. GVCs are expected to bring gains to the linked countries via improved competitiveness, better access to global markets, and expansion of production and jobs. However, whether countries realize these gains or not is still not clear mainly because the tools to measure a country’s extent of participation in GVCs and distribution of incomes generated from GVCs across countries are limited. The rising share of intermediate products in total trade has challenged the use of traditional tools like export shares in assessing countries’ competitiveness. Higher export shares may not necessarily imply higher competitiveness if exports contain a large share of imported intermediate products. In a similar fashion, higher exports may not guarantee more domestic production and jobs if the domestic value-added content of exports does not rise’. In analyzing the link between trade liberalization and growth, both theoretical and empirical literature have been prolific. Models of ‘endogenous growth’ have been constructed, under the theoretical approach that suggests openness should be positively associated with growth (Grossman and Helpman 1989; Barro and Salai-­Marti 1995; and Obstfeld and Rogoff 1996). Such models support the view that openness stimulates growth through numerous channels such as access to technology, access to intermediate and capital goods or increased competition (Grossman and Helpman 1989; Lucas 1988). However, other models showed the not-so-­ optimistic picture, like how openness can also push countries into less dynamic sectors and harm growth (Rodriguez and Rodrick 2000). As dealt by Winters (2004), an alternative approach to linking trade liberalization and growth is through productivity. Coe et al. (1997) show a positive effect of a country’s openness on total factor productivity. However, in the short run some factor owners could suffer if productivity increases faster than output (Parker et al. 1995). Veeramani (2014) has highlighted that it is important to go beyond the simple relationship between the trade openness and growth. He hypothesized that the kinds of imported capital goods and their origin impact growth; he further noted that there are higher levels of knowledge embedded in the capital goods relative to the intermediaries, and also 1  Global value chains, with the help of concepts like ‘governance’ and ‘upgrading’, explain the ways in which new patterns of global trade, production and employment mould avenues for development and competitiveness (Gereffi 2018).

5.2 Relation Between Economic Growth and Export: The Background

87

trade positively influences growth. On the other hand, there are studies looking at the other side of the story, that is openness could also harm growth through increased macroeconomic volatility (Easterly and Kraay 2000). Caselli (2015) has reflected on this view by drawing examples from new studies to as old as of Newbery and Stiglitz (1984). He stated, ‘Openness to international trade leads to higher GDP volatility. The origins of this view are rooted in a large class of theories of international trade predicting that openness to trade increases specialisation’. As a result of specialization in production, the country becomes more sensitive to any shocks specific to the sector/commodity of specialization and exposed to higher macro-volatility. Numerous other studies have appeared on the issue. Some of these concentrate on a single country and some on multi-country analysis. Hence, it can be seen that it is very difficult to bring out a distinct relationship between export expansion and the process of economic growth. In this regard, Goncalves and Richtering (1987) said, the reasons behind this relationship that varies across economies might depend on their respective historical experiences of development and domestic as well as international environment. They further said, ‘These differences are often not taken into account the indiscriminate use of bivariate tests on the statistical relationship between export performance and output growth. Making broad generalizations and drawing strong conclusions on the basis of such tests, leaving out other important explanatory variables, should be discouraged’. In the twentieth century, Nurkse (1959) argued, unlike the nineteenth century, trade is not an engine of growth. Kravis (1970) pointed out that trade and capital flows were ‘handmaidens’ but not engines of growth. He further noted that the key factors affecting growth were ‘internal’, like land, human resource, economic and social institutions. Reynolds (1983), for the post–Second World War era, has put forth the existence of examples of non-export-induced growth, though he has also pointed out on the basis of few newly industrializing economies ‘a tendency for a high growth rate to be associated with export success’. Goncalves and Richtering (1987) highlighted that a notable feature of inter-country comparisons of the relation between export increase and economic growth is their restrictive nature. Nurkse (1961) supported the ‘balanced growth theory’, whereas Prebisch (1962) advocated the import substitution policy and opposed the export-led growth approach. In this backdrop, Zestos and Tao (2002) highlighted that trade and development theories have had significant impact on the long-run economic policies adopted by countries across the world. Export-led growth approach could be seen in the case of Southeast Asian economies, on the other hand, Latin American economies represented import substitution strategy. In addition to views stated at the outset, there are efforts undertaken to work out the reverse causality from GNP (gross national product) to export on the basis of theoretical grounds supporting plausibility of such causality. In the light of this, Afxentiou and Serletis (1991) stated that such causality is plausible. They further added that such causality could be envisaged in cases where the long run accumulation of physical and human capital, in combination with new technology, augments overall productivity of an economy, and in this very process the country gets

88

5  Output and Trade Relation

benefited with favouring comparative advantage in international trade. In this context, Kindleberger (1973) further adds that in other cases like high growth economies with relatively high propensities to save, due to their low absorption look for foreign markets to utilize sizeable parts of their increased output. In line with this viewpoint, Kaldor’s (1964, 1967, 1968) contribution is worth citing, which was aptly dealt by Stavrinos (1987). According to Kaldor (1967), ‘the very fact of a faster growth of output could be expected to act as a stimulus to exports when output and capacity are both enlarged, productivity is increased and unit costs are reduced. It is then easier to sell abroad’. However, a major weakness such reasoning faces is that it is based on the implicit assumption that while progress is made in one industrial country facilitating growth of export, there is total passivity in other industrial countries. This assumption is a misfit in the dynamic industrial world (Afxentiou and Serletis 1991). In this backdrop, an exploratory attempt is made to work out such causality from GDP to merchandise export. As compared with previous studies, the present analysis is not country specific or region specific, rather it is attempted at a broader scale and covers 87 countries with varying levels of development and openness.

5.3  Sources of Data and Database There are differences in definitions and other kinds of discrepancies between different sources of data, each with their own advantages and limitations (refer to section 1.6.12). The data set for merchandise export and import as well as GDP are taken from UNCTADstat. Data are reported in current US million dollars.

5.4  Methodology The major objective of the present chapter is to bring out the relationship between international trade (measured here in terms of merchandise export) and the size of the economy (Gross Domestic Product). According to Mahmood (1977), in case of social sciences, it is difficult to find out exact relationships, and he further adds, ‘we have to tolerate a certain amount of error while approximating them into any exact form of relationships’. Hence, knowing the limitation of the statistical tools and analysis, an effort is being made to explore the relationship between the two variables. Most of the previous studies relate economic growth to export or openness. In this chapter, an attempt is made to explore economic growth and trade nexus from an alternative perspective, as mentioned in the preceding paragraphs. The relationship being worked out is between the proportion of merchandise export to world merchandise export as a dependent variable and gross domestic product as an independent variable.

5.4 Methodology

89

For the analysis, statistical, graphical and cartographic techniques like regression,2 scatter plots,3 residual mapping etc. have been used. The simplest method for examining the relationship between two different characteristics is a scatter diagram. Most often, a scatter diagram is used to provide basic evidence of cause-and-effect relationships. It shows relationships between two variables; however, correlation does not necessarily mean a direct cause and effect relationship. Further, a single analysis using a bivariate model should not be expected to be sufficiently accurate because it is difficult to scale down the dynamic world process to a few or couple of indicators. Pal (1998) further argued that in the real world any dependent phenomenon is a function of several independent variables, yet largely used in identifying relationships and is also relied upon in the present study. On similar lines, Raza and Agarwal (1986) have noted that despite the limitations of unilateral models, many such models have been helpful in understanding the ‘factors generating the commodity flows’. Regression analysis is primarily used to estimate the unknown parameters. It is worthy to note that one has to be cautious as it is the simplest version, having said that yet regression models are quite useful for examining the aggregate pattern. For the present chapter, scatter plots have been attempted on different scales – arithmetic, logarithmic and semi-logarithm. Though for the final analysis, results pertaining to arithmetic and logarithmic scales are retained. In line with this, Mukherjee et al. (1998) have pointed out that in econometric practice, logarithmic transformation is very popular. One of the reasons is that functions which can be linearized with the help of logarithms have coefficients ‘which lend themselves to meaningful economic interpretations, like growth rate or elasticity’; second is that the logarithmic transformation frequently (but not always) does the “trick” with socioeconomic data which are most often skewed to the right. It has been noted that logarithms have the property of shrinking the distance between two or more values which are greater than 1. Data used for scatter plots are used in the regression analysis with GDP as independent variable and proportion of merchandise export to world merchandise export as a dependent variable. Further, standardized residual4 values for each country are plotted on the map. Residuals, as discussed by Pal (1998), provide an insight into how well the regression equation predicts the dependent variable for a particular observation. 2  Regression analysis helps in analyzing the statistical association between two (or more) variables. Mukherjee et al. (1998) also pointed out that a regression model only depicts statistical association between two variables, but in itself it cannot establish the direction of causality between them. Whether a causal link exists between two variables and which way the causality runs is a matter which can only be settled by sound theoretical reflection. 3  In bivariate analysis (as to graphics), a simple but powerful tool is the scatter plot (Mukherjee et al. 1998). It is the simplest method of seeing the relationship between two variables by plotting the two variables against each other on a graph. 4  Pal (1998) has pointed out that for mapping residuals should be standardized; this is done by dividing the absolute residuals by their standard deviation, which in this case is the standard error of the estimate. By this the magnitude of the residuals is affected but the relative pattern remains the same.

90

5  Output and Trade Relation

5.5  An Analysis of the Relation Between Growth and Export As stated earlier in the chapter, an initial step was undertaken to explore the relationship existing between economic growth or gross domestic product and the share of merchandise export for the entire spectrum of countries with varying levels of development and openness. It is quite explicit from the scatter plots that the two variables are not linearly related (Fig. 5.1). Further, it can be broadly inferred that there exists a positive correlation between gross domestic product (GDP) and merchandise export (share of merchandise export to world merchandise export). Figures 5.3 and 5.4 show scatter diagrams plotted on log scale, clearly reflecting a log linear relationship pattern in 1990 and 2015. It can be said that with the change of scale, the nature of correlation varied. It can also be seen from scatter plots that relatively low exporting countries or small players in international merchandise export are relatively less affected by their gross domestic product/economic size. In case of higher income economies, for example the United States, the United Kingdom, Germany, France and Japan, with China recently added to this group, the share of merchandise export is also high, reflecting that gross domestic product does influence the same.

Fig. 5.1  Association between merchandise export and GDP (1990) (arithmetic scale). (Source: Based on Annexure 5.1)

5.5 An Analysis of the Relation Between Growth and Export

91

Fig. 5.2  Association between merchandise export and GDP (2015) (arithmetic scale). (Source: Based on Annexure 5.2)

Fig. 5.3  Association between merchandise export and GDP (1990) (logarithmic scale). (Source: Based on Annexure 5.1)

92

5  Output and Trade Relation

Fig. 5.4  Association between merchandise export and GDP (2015) (logarithmic scale). (Source: Based on Annexure 5.2)

Regression analysis5 is also attempted for the same data set in this chapter. Most importantly, the regression confirms that GDP is a major determinant of international trade. The R2 is 0.73 and 0.71 for 1990 and 2015, respectively. The positive relationship may not reflect an effect of GDP on export. Further, in order to capture the relationship between GDP and export across different set of countries, and to check the predictability of the model at that scale, residual mapping is done. For this purpose, standard residuals6 are mapped and analyzed. It will show the spatial variation in the degree of fit of the regression equation (Pal 1998). Geographers often plot these residuals, as it plays a scientific role in geographical research. In regions where residuals show that regression is a poor indicator, there are factors other than the independent variable which affect the dependent variable (Mahmood 1977; Pal 1998). Through residual mapping, it is convenient to identify and regionalize the countries having high negative and high positive. To maintain comparability, the values of class divide are kept constant for 1990 and 2015. The relationship between trade and gross domestic product may differ significantly according to circumstances and nature of economic structure of countries.

5  Regression analysis provides a functional relationship by which the value of one variable can be estimated from the value of another variable, and one variable is thus considered as dependent upon the other (Pal 1998). 6  The absolute residual of a particular observation of X or Y does not have much use in research because of the problem of units of measurement when comparing two or more sets of values. In this context, the standardized residual is preferred which expresses the value of the absolute residual in terms of a normal distribution of residuals. Standardized residual bands, like standard error bands, run parallel to the regression lines. Therefore, they do not give undue emphasis to the residuals in Y related to either the large or small values of X (as do absolute residuals) (Pal 1998).

5.5 An Analysis of the Relation Between Growth and Export

93

The residual values reflect how individual values differ from their predicted values, that is it captures variations in the levels of merchandise export not explained by variations in gross domestic product. In 1990, a group of countries exporting more than expected from their size of the economy includes large industrialized countries like Germany, France, Netherlands, Belgium-Luxembourg, the United Kingdom, Canada and Italy as well as smaller ones like Singapore. Singapore has been one of the most dynamic elements in the world as well as one of the resurgent Asian economies. It is a strategically placed city state ‘with a vocation for entrepot trade’, and the growth received impetus from the government via promotion of high savings, development in education, boosting export and acquiring foreign technology. Its own manufacturing production grew more sophisticated, labour costs shot up and it became a major capital exporter. Despite being a small country, Singapore, as stated earlier, has significant entrepot trade as compared to the national production. It basically constitutes re-exports without any extra processing or value addition. The rest of the Western European countries and Canada (Western offshoots) falling in this group have quantitative historical evidence better than most other parts of the world (Maddison 2006). In line with this, all these countries are industrially advanced rich nations with a preponderance of manufacturing goods in export basket. Broadly it can be said that in case of higher exporting countries, GDP does make a difference, and there seems to exist a relation between economic size and export in the group of economies stated in the preceding paragraphs. Germany accounts for 11% of world trade, France around 7%, the United Kingdom about 6% and the Netherlands, Belgium-Luxembourg and Canada account for approximately 4%, whereas Singapore’s share pegged at 2 % of the world trade. The merchandise export of Singapore grew rapidly during 1990–1995 at the rate of about 16%; the export of Belgium-Luxembourg, the Netherlands and Canada experienced moderate growth rate with around 8% and French export grew at the pace of 6%, whereas economies like Germany, the United Kingdom and Italy registered sluggish growth in their export (about 4%) during that period. Countries constituting high negative residual group depict that these economies export less than expected from their economic size. This group consists of 48 countries spread largely across the global South, that is Africa, Asia and Latin America; it also consists of a few Oceania and developed economies (Fig.  5.5). It can be deducted in the case of these countries, there are factors other than GDP that influence the magnitude of the merchandise export. This group comprises of the largest, viz., the United States, and smallest country, viz., Fiji (a small island developing economy), in terms of gross domestic product, among the 87 countries across the globe. It is worth noting here that in this group only the United States accounts for around 13% of the world merchandise export, while rest of the countries have less than 1% of share in world merchandise export. The composition of the export basket varies across the group, measured broadly in terms of the largest export share of individual commodity (SITC Rev. 3 classification, 1digit tier) group in total product export. It ranges from major exporters of petroleum (e.g. Bahrain, the Syrian Arab Republic, Brunei Darussalam, Gabon) to manufactured goods (e.g. the United States, the Dominican Republic, Malta), as well as food items (Panama, Iceland,

94

5  Output and Trade Relation

Fig. 5.5  Residuals from regression of merchandise export on GDP, 1990. (Source: Based on Annexure 5.3)

Honduras) etc. This diverse set of economies has registered varying growth rates for the period of 1990–1995, ranging from negative to very high. Zestos and Tao (2002) have broadly explained the weaker causality in the case of the United States. They have pointed out that being a large industrial economy with a huge domestic market has been instrumental ‘in some ways in the past economically isolated from the rest of the world.’ Further, dollar being an important and tradable currency for a long time, the United States was able to import and invest in other countries despite the level of its export. The large capital surplus has also most likely to impact the causality. In 2015, there is a visible change in the country composition of high positive, medium and high negative residual group (Fig. 5.6). In the period of twenty-five years, countries from the global South has risen in the global trading system and has intensified their trade. For example, China, South Korea, the UAE and Mexico have emerged in the high positive residual category, reflecting upon the fact that these economies export above the predicted levels, and did so virtually the entire time period. Besides, the USSR (former) has also exhibited signs of improvement in its trading capacity over the years. The most striking result is the remarkable expansion of the Chinese economy (by GDP) and its export potential. China’s economic size has swelled significantly over the years from being the eleventh largest to the second largest economy in the world from 1990 to 2015. Its merchandise export share to world export has also mounted up from mere 2% in 1990 to 14% in 2015. China is predominantly an exporter of manufactured goods with a share of 92% in its export basket. On the other hand, the USSR (former) has not shown remarkable growth in terms of its GDP size; it has slipped from the seventh to the eighth

5.6 Concluding Remarks

95

Fig. 5.6  Residuals from regression of merchandise export on GDP, 2015. (Source: Based on Annexure 5.4)

position among the 87 countries under analysis from 1990 to 2015, respectively. Similarly, its merchandise export share to world export has also not shown any remarkable change, for example, its share has hovered around 3% during 1990–2015. The USSR (former) is predominantly exporting fuels over the years and has shown signs of increasing relative concentration of fuels vis-a-vis other commodities in its export basket from 1990 to 2015. The countries constituting high negative group broadly remain the same, however, with a few exceptions such as Algeria, Nigeria, Venezuela and New Zealand that revealed to export less than their economic size in 2015. Hence, it is quite explicit that the results are mixed and the predictability of the model varies across countries over time.

5.6  Concluding Remarks A variety of models exist in the literature to study the causality between foreign sector and national economy and vice-versa. The current chapter has made an attempt to study the relations between GDP and export with the help of unilateral models such as simple linear regression. The preceding analysis shows that economic size does influence the levels of merchandise export. It is noted that correlation between the two variables is positive. There is a log linear relationship between GDP and merchandise export over the period of 25 years. Furthermore, the residual mapping highlights how distinct individual cases are from their predicted values. Clearly the relationship between GDP and export levels is not exactly the same

96

5  Output and Trade Relation

everywhere. It is to be noted that a variety of factors might have led to such variations. In particular, the country-specific dynamics play a major role in explaining variation in export levels across GDP size. There is a close relationship between the country- and region-specific factors and elements that are particular to a specific conjuncture. One of the important determinants would be the nature and structure of the policy towards openness. It is also worth noting that the policy is not the same across countries, and that does get reflected in the differential impact of GDP on export. Whether a country experiences growth-led export or export-led growth during different periods of economic history and global integration also depends on the interaction between the domestic conditions and the way in which a country is influenced by the dynamics of the global or regional market. Hence, it could be said that it is more country specific and needs to be explained through other variables and at different spatial scales, which of course is beyond the gamut of the current chapter.

References Afxentiou PC, Serletis A (1991) Exports and GNP causality in the industrial countries: 1950–1985. Kyklos 44:167–179 Al-Yousif YK (1997) Exports and economic growth: some empirical evidence from the Arab gulf countries. Appl Econ 29(6):693–697 Balassa B (1978) Exports and growth: further evidence. J Dev Econ 5(2):181–189 Balassa B (1985) Export, policy choices, and economic growth in developing countries after the 1973 oil shocks. J Dev Econ 18:23–35 Banga R (2014) Linking into global value chains is not sufficient: do you export domestic value added contents? J Econ Integr 29(2):267–297 Barro RJ, Sala-i-Martin X (1995) Economic growth. McGraw-Hill, Cambridge, MA Bhagwati J (1978) Foreign trade regimes and economic development: anatomy and consequences of exchange control regimes. Ballinger, Cambridge, MA Bhagwati JN, Shrinivasan TN (1978) Trade policy and development. In: Dornbusch R, Frenkel JA (eds) International economic policy: theory and evidence. The Johns Hopkins University Press, Baltimore, pp 1–38 Bhat TP (2011) Structural changes in India’s foreign trade. isid.org.in. Accessed 15 Dec 2017 Caselli F (2015) what effect does trade openness have on GDP? https://www.weforum.org. Accessed 22 May 2019 Chenery HS, Strout AM (1966) Foreign assistance and economic development. Am Econ Rev 56:679–733 Coe DT, Helpman E, Hoffmaister AW (1997) North-South R & D spillovers. Econ J 107(440):134–149 Colombatto E (1990) An analysis of exports and growth in LDCs. Kyklos 43(4):579–597 Dimkpah YO (2002) The stage of economic development, exports, and economic growth: an empirical investigation. Afr Econ Bus Rev:3 Easterly W, Kraay A (2000) Small states, small problems? Income, growth and volatility in small states. World Dev 28(11):2013–2027 Farole T (ed) (2013) Trade, location and growth. In: The internal geography of trade. World Bank Publications, Washington, DC Feder G (1982) On export and economic growth. J Dev Econ 12:59–73 Fosu AK (1990) Export composition and the impact of export on economic growth of developing economies. Econ Lett 34:67–71

References

97

Gereffi G (2018) Global value chains and development: redefining the contours of 21st century capitalism. Cambridge University Press, Cambridge Ghatak S, Price SW (1997) Export composition and economic growth: Cointegration and causality evidence for India. Weltwirtschaftliches Archiv: Rev World Econ 133(3):538–553 Ghatak S, Milner C, Utkulu U (1997) Exports, export composition and growth: cointegration and causality evidence for Malaysia. Appl Econ 29(2):213–223 Gonçalves R, Richtering J (1987) Intercountry comparison of export performance and output growth. Dev Econ 25(1):3–18 Greenaway D, Sapsford D (1994) What does liberalisation do for exports and growth? Weltwirtschaftliches Arch 130:152–174 Grossman Gene M, Helpman E (1989) Comparative advantage and long run growth. NBER Working Paper No 2809. http://www.nber.org. Accessed 25 Dec 2016 Henriques I, Sadorsky P (1996) Export-led growth or growth-driven exports? The Canadian case. Can J Econ 29(3):540–555 Hodne F (1994) Export-led growth or export specialization? Scand Econ Hist Rev 42(3):296–310 Holman JA, Graves PE (1995) Korean exports economic growth: an econometric reassessment. J Econ Dev 20(2):45–56 Islam MN (1998) Exports expansion and economic growth: testing for cointegration and causality. Appl Econ 30(3):415–425 Johnston RJ (1976) The world trade system: some enquiries into its spatial structure. G.Bell and Sons Ltd, London Kaldor N (1964) Essays on economic policy. G. Duckworth, London Kaldor N (1967) Strategic factors in economic development. Cornell University Press, New York Kaldor N (1968) Productivity and growth in manufacturing industry: a reply. Econ New Ser 35(140):385–391 Kavoussi RM (1984) Export expansion and economic growth: Further empirical evidence. J Dev Econ 14(1/2):241–250, January/February Keesing DB (1967) Outward looking policies and economic development. Econ J 77:303–320 Kindleberger CP (1973) International economics. D.B. Taraporevala sons & Co. Pvt. Ltd, Bombay Kravis I (1970) Trade as a handmaiden of growth: similarities between the nineteenth and twentieth centuries. Econ J 80:850–872 Krueger AO (1978) Foreign trade regimes and economic development: liberalization attempts and consequences. Ballinger, Cambridge, MA Krueger AO (1980) Trade policy as an input to development. Am Econ Rev 70:288–292 Kugler P, Dridi J (1993) Growth and exports in LDCs: a multivariate time series study. Int Rev Econ Bus 40(9):759–767 Lucas RE Jr (1988) On the mechanisms of economic development. J Monet Econ 22:3–42 Maddison A (2006) The world economy: a millennial perspective and historical statistics. OECD, Paris Mahmood A (1977) Statistical methods in geographical studies. Rajesh publications, New Delhi Marin D (1992) Is the export-led growth hypothesis valid for industrialized countries? Rev Econ Stat 74(4):678–688 Marshall (1985) Exports, growth and causality in developing countries. J Dev Econ 18:1–12 Medina-Smith EJ (2001) Is export – led growth hypothesis valid for developing countries? a case study of Costa Rica. Policy Issues in International Trade and Commodities Study Series, No 7 United Nations Conference on Trade and Development Mukherjee C, White H, Wuyts M (1998) Econometrics and data analysis for developing countries. Routledge, New York/London Newbery D, Stiglitz J (1984) Pareto inferior trade. Rev Econ Stud 51(1):1–12 Nurkse R (1959) Patterns of trade and development. The Wicksell Lectures, Stockholm Nurkse R (1961) Trade theory and development policy. In: Ellis HS (ed) Economic development of Latin America. St. Martin, New Yok

98

5  Output and Trade Relation

Obstfeld M, Rogoff K (1996) Foundations of international macroeconomics. MIT Press, Cambridge, MA Ocampo JA, Vos R (eds) (2008) Uneven economic development. Orient Longman Pvt. Ltd, Hyderabad Olorunfemi S, Olowofeso OE (2006) Time series analysis of export and economic growth in the Ecowas countries. Soc Sci 1(4):344–350 Pal SK (1998) Statistics for geoscientists: techniques and application. Concept, New Delhi Parker Ronald L, Riopelle R, William FS (1995) Small enterprises adjusting to liberalization in five african countries. The World Bank, Washington, DC Prebisch R (1962) Towards a dynamic development policy for Latin America, United Nations Ram R (1985) Exports and economic growth: some additional evidence. Econ Dev Cult Chang 33:415–423 Raza M, Agarwal Y (1986) Transport geography of India: commodity flos and the regional structure of Indian economy. Concept publishing company, New Delhi Reynolds LG (1983) The spread of economic growth to the third world. J Econ Lit 21(3):941–980 Rodriguez, Rodrik (2000) Trade policy and economic growth: a skeptic’s guide to cross- national evidence. NBER Working Paper No7081, April. www.nber.org. Accessed 10 Jan 2017 Scitovsky T (1954) Two concepts of external economies. J Polit Econ 62:143–151 Sentsho J (2000) Exports revenue as determinant of economic growth: evidence from Botswana. University of Bostwana, Gaborane Serletis A (1992) Export growth and Canadian economic development. J Dev Econ 38:133–145 Shan J, Sun F (1998) On the export-led growth hypothesis: the econometric evidence from China. Appl Econ 30:1055–1065 Sharma SC, Dhakal D (1994) Causal analysis between exports and economic growth in developing countries. Appl Econ 26:1145–1158 Stavrinos VG (1987) The inter-temporal stability of Kaldor’s first and second laws in the UK. Appl Econ 19:1201–1209 Tyler WG (1981) Growth and export expansion in developing countries: some empirical evidence. J Dev Econ 9(1):121–130 Ukpolo V (1994) Export composition and growth of selected low-income African countries: evidence from time-series data. Appl Econ 26:445–449 UNCTAD (1992, 2007) Trade and development report, United Nations, New York/Geneva Veeramani C (2014) World’s knowledge spillovers: beyond openness and growth. J Econ Integr 29(2):298–328 Voivodas C (1973) Exports, foreign capital inflow and economic growth. J Int Econ 3:337–349 Winters LA (2004) Trade liberalization and economic performance: an overview. Econ J 111(493):F4–F21 Zestos GK, Tao X (2002) Trade and GDP growth: causal relations in the United States and Canada. South Econ J 68(4):859–874

Chapter 6

Regional Trade Blocks: A Case Study of Mercosur

Abstract  Regionalization is a widespread feature of international trade. Lately, developing countries have joined developed ones under new varieties of regional arrangements such as North-South RTAs and North-South-South RTAs, highlighting increasing North-South interdependence. Deepening integration among developing economies can provide additional opportunities for improved South-South trade. Similarly, in Latin America, several efforts were made to create free trade areas at the sub-regional level. One of the most significant initiative is the formation of MERCOSUR by Argentina, Brazil, Paraguay and Uruguay. The origins of MERCOSUR can be traced to the July 1986 agreement between Argentina and Brazil. The MERCOSUR trade bloc’s purpose is to allow for free trade between member states, with the ultimate goal of full South American economic integration. MERCOSUR’s original full members include Argentina, Brazil, Paraguay and Uruguay, and it is the largest trading bloc in South America. Yet experts say MERCOSUR has become somewhat paralyzed in recent years. The MERCOSUR economies are different in terms of their economic size, level of development and nature of their markets. At the same time, there are other structural and political differences which led to differential pattern of international trade of individual MERCOSUR member economies intra-regionally and inter-regionally. Hence, an attempt is made through this chapter to compare volume of trade and commodity composition of export and import among MERCOSUR member countries, and with the rest of the world, from 1990 to 2015. Interestingly, the trade bloc partners over the period of past two decades did not fully channelize the optimum potential intrinsic in the regional trade, although the effect of regional integration is positive, but as compared to other similar examples it is not that impressive. Keywords  Regionalisation · MERCOSUR · Regional trade agreements (RTAs) · South-South Trade

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9_6

99

100

6  Regional Trade Blocks: A Case Study of Mercosur

6.1  Introduction The past 30 years have seen a substantial increase in the number of regional and sub-regional trade agreements signed across the globe. While this particular pattern in some regions has led to intensification of mutual trade and also helped many to gain from multiplying export, it has not systematically resulted in increased intra-­ trade within the trade groupings that have been created (UNCTAD 2006). Regional economic and trade cooperation, including bilateral and regional trade agreements (RTAs), has intensified in a number of developing economies. Regional arrangements offer important possibilities to enlarge economic space, attract the foreign direct investment (FDI) to the region on better terms, and pool economic, human, institutional, technological and infrastructural resources and networks of participating countries. The regional arrangements were slow to take off in the beginning, which is still the case with many. However, as complementarities among economies emerged, confidence in opening up to one another also developed. More recently, developing countries have joined their developed partners under new varieties of regional arrangements, such as North-South RTAs and North-South-South RTAs, reflecting growing North-­South interdependence. Deepening integration among developing countries can provide further opportunities for enhanced South-South trade. Some of these arrangements, such as MERCOSUR, have had a substantial impact on the expansion of trade in specific sectors among participating countries, as well as between these countries and the rest of the world (UNCTAD 2006). Second Wave of Regionalism: Regionalisation is a widespread feature of international trade (Gaulier et  al. 2004). According to Estevadeordal et  al. (2000), the world is undergoing a second wave of regionalism. The first wave of regionalism in the 1950s and 1960s was mostly short-lived, except in the case of Western Europe; however, the second wave witnessed many successful attempts to form integrated trading areas all over the world since the mid-1980s. Those recent attempts are often referred to as the ‘New Regionalism’. They further stated that in Europe, a successful attempt to create a single market by 1992 led to a deeper economic integration that involved monetary unification. In Asia and the Pacific, many economies united into an economic union called the Asian Pacific Economic Cooperation (APEC), and by 2010 they aimed to achieve free trade and investment in the region for industrialized economies and by 2020 for developing economies. Similarly, in North America, the United States, Canada and Mexico formed the North American Free Trade Agreement (NAFTA) in 1994 and agreed to eliminate tariff and non-tariff barriers in the region by 2009. In Latin America, several efforts were made to create free trade areas at the sub-­ regional level. One of the most significant initiatives is the formation of MERCOSUR by Argentina, Brazil, Paraguay and Uruguay. The origins of MERCOSUR can be traced to the July 1986 agreement between Argentina and Brazil that established the Economic Integration and Cooperation Program (PICE). The MERCOSUR trade bloc’s purpose, as stated in the 1991 Treaty of Asunción, is to allow for free trade between member states, with the ultimate goal of full South American economic

6.1 Introduction

101

integration. MERCOSUR’s full members include Argentina, Brazil, Paraguay, and Uruguay. Venezuela’s entry as a full member is still pending ratification by Brazil and Paraguay. The treaty considered complete liberalization of trade in goods and the creation of a common market resulting in free circulation of goods and services. Since the end of the 1980s, the tariff rates, both most favoured nation (MFN) tariffs and preferential tariffs, have been drastically reduced and the quantum of international trade, both intra-regional and extra-regional, has dramatically increased. Other bilateral and sub-regional groupings have also made substantial progress in the same direction. The launching of the Free Trade Area of the Americas (FTAA) in 1995 is now in full negotiations with the potential of becoming the largest, in a geographical sense, experiment in the New Regionalism approach to economic integration. Further, relations with the rest of the world would be conducted through a Common External Tariff1 (CET). In 1996, association agreements were signed with Chile and Bolivia, establishing free trade areas with these countries. These countries have geographical advantage in terms of the proximity and hence can emerge as natural partners. In this reference, Knapp (2004) has highlighted that geographically proximate countries have close trade networks. This is largely attributable to the relative shorter distance, spatial contiguity, common cultural traits and broadly exhibiting relative similarity in the level of development. These factors are applicable in the context of MERCOSUR members. The common market of the South ‘MERCOSUR’ is the largest trading bloc in South America. MERCOSUR’S primary interest has been in eliminating obstacles to regional trade, like high tariffs, income inequalities, or conflicting technical requirements, for bringing products to market. Yet experts say MERCOSUR has become somewhat paralyzed in recent years, with its members divided over the future of the organization. Some countries, like Brazil, want to keep MERCOSUR focused on regional trade. Other countries, like Venezuela, which has yet to attain full membership in the bloc, would like to expand the group’s mandate to political affairs. The creation of a new regional organization in 2008, the Union of South American Nations (Unasur), has raised further questions about MERCOSUR’s utility. It is now the fourth largest trading bloc in the world, after the European Union (EU), North American Free Trade Agreement (NAFTA), and the Association of South East Asian Nations (ASEAN). It is worth noting that currently, in the developing South, MERCOSUR and ASEAN are the two largest PTAs (Mansfield and Pevenhouse 2013). MERCOSUR offers a unique case of regional economic integration because its member countries unilaterally reduced their MFN tariffs against non-members at the same time they were reducing intra-area preferential tariffs. This constitutes a striking difference from other regional free trade arrangements where the formation of a trading bloc is rarely accompanied by a parallel lowering 1  The common external tariff (CET) adopted in 1995 by MERCOSUR implies substantial overall tariff reduction compared to those existing in the member countries in the 1980s (Estevadeordal et al. 2000)

102

6  Regional Trade Blocks: A Case Study of Mercosur

of external tariffs, for example in the case of the EU or NAFTA (Goto and Hamada 1999). Note that, as Ethier (1998) pointed out, in this particular case, the magnitude of additional preferential tariff reductions is smaller than the MFN tariff reductions. In the context of the discussion, a major objective of the chapter is to compare volume of trade and commodity composition of export and import among MERCOSUR member countries, and MERCOSUR with the rest of the world. Data for the current chapter are mainly extracted from UNCTADstat and the World Bank (WITS database); distance data for gravity model is taken from the following web link: http://www.chemical-­ecology.net/java/capitals.htm. Furthermore, an attempt is made in the chapter to bring out the relation between trade ratio and gross domestic product (GDP) ratio and also between actual trade flows (trade ratio) and expected trade (ratio of gravity model estimates) flows between individual MERCOSUR partners with the rest 86 economies. Regression analysis is attempted in both the cases with the mere purpose of checking out the difference in results. The regression results of GDP ratio on trade ratio is retained in this chapter. However, in the interest of the reader, excerpts from my doctoral research pertaining to gravity model estimates is annexed for reference (refer Annexure 6.21). Ratios are calculated by the following formulae:

GDP ratio  PiPj /  PiPj,

where Pi and Pj are the GDP of the origin and destination;



Gravity model ratio  PiPj /  dij 

2

 /   PiPj /  dij   , 2

where Pi and Pj are the GDP of the origin and destination and dij2 is the square of the distance between the capital cities of origin and destination;

Trade ratio  CiCj /  CiCj,

where CiCj is the trade between origin and destination. Standardized residuals are accordingly grouped in high positive, medium positive, medium negative and high negative group. As medium positive and medium negative are very close to estimate, so the analysis is confined to high positive and high negative residuals.

6.2  A  Brief Analysis of the Trading Pattern of MERCOSUR Economies Trade among the MERCOSUR members, rest of the region i.e. South America, and also with the rest of the world has visibly expanded over the years (Figs. 6.1 and 6.2). The export figures depicted in Fig. 6.1 highlight that the increase in the export to the

6.2  A Brief Analysis of the Trading Pattern of MERCOSUR Economies

103

rest of the world is steeper than the intra-trade group. The value of total export of MERCOSUR in aggregate has multiplied from 90,686 million US dollars in 1995 to 300,210 million US dollars in 2015. Figure 6.1 further shows that intra-group value of export has increased from 17,308 million US dollars in 1995 to 40,118 million US dollars in 2015. Export of MERCOSUR with the rest of the region has increased from about 33,750 million US dollars in 1995 to 73,931 million US dollars in 2015. MERCOSUR import shows a similar pattern of increasing trend over the period of 20 years. Total import of MERCOSUR in aggregate has increased from 92,506 million US dollars in 1995 to 281,188 million US dollars in 2015. Intra-group import grew to 38,711 million US dollars in 2015 from 16,356 million US dollars in 1995. Import from the rest of the region has increased from 31,796 million US dollars in 1995 to 74,021 million US dollars in 2015. It is quite clear from Fig. 6.2 that the value of import from the rest of the world has also followed an increasing pattern. It is worth noting that the quantum of intra-group trade is significantly less as compared to its trade with the rest of the world. Among MERCOSUR members, Argentine and Brazilian economies are much larger than Paraguay and Uruguay. Therefore, as put forth by Estevadeordal et al. (2000), changes in aggregate MERCOSUR trade flows is mostly driven by changes taking place in export and import of these two larger economies. The degree of trade openness varies among MERCOSUR members. For example, in 1990, the openness index for Argentina and Brazil was 11% and 13%, respectively; however, in the case of smaller economies, such as Paraguay, it was as high as 50% and 33% in the case of Uruguay. Over the period of 20 years, in 2015, Argentina, Brazil and Paraguay

Fig. 6.1  MERCOSUR: Merchandise export. (Source: UNCTADstat)

104

6  Regional Trade Blocks: A Case Study of Mercosur

Fig. 6.2  MERCOSUR: Merchandise import. (Source: UNCTADstat) Table 6.1  Share of merchandise export from MERCOSUR partners Argentina 1990 2015 Argentina Brazil Paraguay Uruguay

48.8 14.6 13.8

70.74 21.75 23.42

Brazil 1990 77.6

2015 79.98

82.3 85.1

72.61 69.28

Paraguay 1990 2015 8.0 10.43 28.8 14.02 1.1

Uruguay 1990 2015 14.3 9.59 22.3 15.24 3.0 5.64

7.30

Source: Calculation based on DOTS Trade Matrix database, IMF

further opened their economies with the index of 18%, 21% and 67%, respectively; in the case of Uruguay, it hovered around 32%. It is quite explicit that the degree of openness in the context of the larger countries, like Brazil and Argentina, is lower as compared to that of the smaller MERCOSUR partners, Paraguay and Uruguay. In this reference, Garcia et al. (2013) aptly noted ‘countries that have a smaller domestic market, require foreign trade to stock the products they need. This is the case of Paraguay and Uruguay.’ Tables 6.1 and 6.2 summarize intra-trade bloc pattern of export and import flows, respectively, in 1990 and 2015. In the case of Argentina, Brazil is the major trading partner over the years. In 1990, Brazil accounted for about 78% of Argentina’s merchandise export, whereas the share of Paraguay and Uruguay was a mere 8% and 14%, respectively. In 2015, the share of Brazil further multiplied to 80%, on the hand other, Paraguay and Uruguay’s share hovered around 10%. A similar pattern is

6.2  A Brief Analysis of the Trading Pattern of MERCOSUR Economies

105

Table 6.2  Share of merchandise import from MERCOSUR partners Argentina 1990 2015 Argentina Brazil Paraguay Uruguay

63.4 27.5 46.1

83.04 37.70 41.81

Brazil 1990 82.4

2015 92.89

71.2 51.8

59.94 54.18

Paraguay 1990 2015 7.0 4.05 13.9 7.14 2.0

Uruguay 1990 10.5 22.5 1.2

2015 3.06 9.82 2.37

4.01

Source: Calculation based on DOTS Trade Matrix database, IMF

observed in the case of Brazilian export, with Argentina receiving a significant share over the years and the smaller players losing along the way. Furthermore, it is explicitly illustrated in the Table 6.2 and 6.3 that the smaller economies Paraguay and Uruguay mainly export to Brazil, followed by Argentina. Also the latter has improved its share visibly in 2015. Besides, Paraguay and Uruguay’s export to each other is significantly low. Hence, it could be broadly inferred from the analysis that there is change in the relative significance of MERCOSUR member economies in individual partners export market over the period of 20 years. Secondly, Argentina and Brazil, which are the dominant economies in MERCOSUR, also have larger accessibility in each other’s market for their export, relative to Paraguay and Uruguay. In the case of Paraguay and Uruguay, Brazil is a key market for their export but has experienced decline in its share recently. Table 6.2 shows import partners of MERCOSUR economies for 1990 and 2015. Like export, Argentina and Brazil show heavy dependence on each other for their import, and the relative position of Paraguay and Uruguay is highly skewed in their markets. For example, Argentina imports 82% in 1990 from Brazil and about 92% in 2015. Paraguay and Uruguay accounts for 7% and 11%, respectively, in 1990, but their share got further eroded in 2015. On the contrary, Paraguay in 1990 imported about 71% from Brazil, followed by Argentina with 28% and Uruguay mere 1%. In 2015, Brazil remains the dominant shipper with a slight dip in the share and yet accounts for more than half of the Paraguayan import. Argentina’s share inched up to 38% in 2015. On the contrary, Uruguay over the period of 20 years still accounts for an insignificant share as compared to other two MERCOSUR members. Hence, it could be inferred from the foregoing observations that Paraguay imports more from Argentina than its export to Argentina. Brazil comes out to be the dominant receiver as well as shipper. Unlike Argentina and Brazil, Uruguay’s import partners are more diverse. Also its import partners are far less concentrated than its export markets. Like export, Argentina and Brazil are the dominant sources of Uruguay’s import in the considered years.

106

6  Regional Trade Blocks: A Case Study of Mercosur

6.3  An Analysis of the Relation Between Trade and GDP Figures 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9 and 6.10 show spatial distribution of standardized residual from regression of GDP ratio on trade ratio for MERCOSUR members for 1990 and 2015.2 Countries constituting medium positive and medium negative are close to estimates; hence, analysis is focussed on the categories with high positive and high negative residual values that reflect on the deviation from the estimate. Figures 6.3 and 6.4 reflect the relation between GDP ratio and trade ratio in 1990 and 2015, respectively. In 1990, countries trading more than their size warrant with Argentina are largely developed and highly industrialized, such as, the United States, Canada, the United Kingdom, Austria, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Norway, Sweden, the USSR (former) and Japan, including also a set of developing economies, for example, China, India, Pakistan, the Philippines, Saudi Arabia, South Korea, the UAE, and also Cameroon, Nigeria, and Zimbabwe. On the other hand, countries trading less than their size warrants cover predominantly developing America, including MERCOSUR members, and other countries falling in this group are Belgium-Luxembourg, the Netherlands, Portugal, Spain, Egypt, Iran and Malaysia. In 2015, more or less similar spatial pattern of trading is visible. However, it is to be noted here the number of developing

Fig. 6.3  Argentina: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.1)

2  It is to be noted here the correlation between trade ratio and GDP ratio is significant for all MERCOSUR members for both the years. For further analysis, regression is worked out with GDP ratio as dependent variable and trade ratio as independent variable.

6.3  An Analysis of the Relation Between Trade and GDP

107

Fig. 6.4  Argentina: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.1)

Fig. 6.5  Brazil: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.2)

economies got skewed with only China, India and South Korea remaining in the group of high positive group. Group of countries trading less with Argentina, than what their economic size warrant, has increased in number, and also MERCOSUR partners continue to trade less than the expected levels over the period of 25 years.

108

6  Regional Trade Blocks: A Case Study of Mercosur

Fig. 6.6  Brazil: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.2)

Fig. 6.7  Paraguay: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.3)

6.3  An Analysis of the Relation Between Trade and GDP

109

Fig. 6.8  Paraguay: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.3)

Fig. 6.9  Uruguay: Residuals from regression of GDP ratio on trade ratio, 1990. (Source: Based on Annexure 6.4)

110

6  Regional Trade Blocks: A Case Study of Mercosur

Fig. 6.10  Uruguay: Residuals from regression of GDP ratio on trade ratio, 2015. (Source: Based on Annexure 6.4)

Figures 6.5 and 6.6 illustrate geographical distribution of residuals from regression of GDP ratios on trade ratio with reference to Brazil for 1990 and 2015 respectively. In 1990, Brazil’s actual trade is recorded to be more than the estimated figures for a geographically diversified group of countries spread across the global North and South (See Fig.  6.5). Developing trading partners are from Asia, Africa and America. In 2015, as shown in Fig.  6.6, countries trading more with Brazil are developed industrialized economies, and also newly industrialized Asian economies. Export basket of these economies is either dominated by manufactured goods or diversified with a combination of manufactured goods. Brazil’s international strategy included commercial ties with other developing economies, which could also be a possible reason for the observed pattern. It is to be noted here, in 2015, the United States and Paraguay seem to trade more with Brazil. Countries trading less with this large MERCOSUR economy, than what is estimated based on the GDP, are geographically concentrated in south and central America, African economies (Algeria, Morocco, Nigeria and South Africa) and Asian countries, viz., Saudi Arabia, Iran, the UAE, Malaysia, Singapore Thailand, Vietnam and South Korea). Similar to Argentina, Brazil is trading less with the MERCOSUR partner countries; even with the United States. Although in 2015, Brazil’s trade improved with the United States and Paraguay. Figures 6.7 and 6.8 reveal the spatial pattern of the GDP and trade relation for Paraguay. It is quite clear that with Paraguay, economies trading more than the expected levels are mainly with the developed countries, such as, the United States, Canada, Germany, the United Kingdom, France, Italy, the USSR (former) and Japan. Also China and India are seen in this group. On the other hand, over the period of 25 years, countries trading less than their size are largely

6.4  Trade Structure by Major Commodity Groups: MERCOSUR

111

concentrated in South America and other developing part of the world. Like large MERCOSUR member economies, Paraguay’s trade with China and India is higher than what is estimated. Furthermore, in 2015 (see Fig. 6.7), Paraguay has improved trade with Indonesia and Saudi Arabia. Also the USSR (former) moved from high positive to high negative group over the past two decades. Furthermore, trade with Latin American countries, particularly, Argentina, Brazil and Uruguay, is far less than expected. The residual maps did not reveal any change in the intra-trade bloc scenario. Similar to other MERCOSUR countries, even Uruguay over the years seems to be trading more than what its size warrants with the group of developed industrialized economies constituting the traditional ‘Core’ of the world economy. In 2015, emergence of southeast and west Asian economies, such as Indonesia, Saudi Arabia and Iran is evident from Figure 6.10. Uruguay’s trade with China has also significantly improved in 2015. It is worth mentioning here, the intra-regional trade has continued to be less as compared to inter-regional trade.

6.4  T  rade Structure by Major Commodity Groups: MERCOSUR Given that not major shifts have occurred in the direction of members’ trade toward MERCOSUR, a related question is what products are most important in this exchange and how has the composition of exports changed? This aspect is brought out in Tables 6.3 and 6.4, which reflect on commodity composition of MERCOSUR economies in terms of major commodity groups. In the foregoing analysis on commodity composition of merchandise trade, it has been noted that largely merchandise export of Latin American economies specialize in primary commodities. MERCOSUR economies follow similar patterns of commodity composition at country level. In 1995, per cent share of all food items in Argentina’s merchandise export was around half of its total merchandise export, followed by manufactured goods with 34%. Furthermore, the relative share of other commodity groups was quite meagre; the pattern persisted even after two decades. In 2015, the preponderance of all food items further intensified with the increased share of about 64%, whereas, manufactures share marginally dipped to 30%. In the case of Brazil, the commodity composition of export over the years got relatively diversified with manufactured goods and all food items as key commodities. In 1995, manufactures accounted for about half of the Brazilian export, although dipped to 37% in 2015; all food items over the years gained in share from 29% in 1995 to 38% in 2015. An increase in the per cent share of fuels from 0.9% to 7% is also noticeable in the considered years.

112

6  Regional Trade Blocks: A Case Study of Mercosur

Table 6.3  Merchandise export of MERCOSUR economies by major commodity groups (percent share in total export) All food items Countries 1995 2015 Argentina 49.8 63.53 Brazil 29.2 38.24 Paraguay 43.9 67.13 Uruguay 44.4 61.64

Raw materials from agricultural sources 1995 2015 4.4 1.05 5.3 4.85 36.4 1.86 14.9 13.61

Ores and Manufactured metals Fuels goods 1995 2015 1995 2015 1995 2015 1.6 2.94 10.3 2.67 33.9 29.81 10.5 12.23 0.9 7.41 54.0 37.27 0.3 0.75 0.2 20.82 19.3 9.44 0.7 0.24 1.1 0.29 38.9 24.22

Source: Based on UNCTAD handbook of statistics, 2006–07 Table 6.4  Merchandise import of MERCOSUR economies by major commodity groups (per cent share in total export) All food items Countries 1995 2015 Argentina 5.5 2.7 Brazil 10.7 5.1 Paraguay 18.5 8.7 Uruguay 10.4 12.8

Raw materials from agricultural sources 1995 2015 2.0 1.0 2.7 1.1 0.2 0.91 4.0 2.10

Ores and metals 1995 2015 2.7 2.8 3.4 3.3 0.7 0.9 1.2 1.0

Fuels 1995 4.2 12.1 6.5 10.1

2015 11.1 14.5 13.7 11.8

Manufactured goods 1995 2015 85.6 82.4 71.1 75.9 74.0 75.7 74.4 72.2

Source: Based on UNCTADstat database

The export basket of Paraguay shows remarkable transformation over the years. Table 6.3 shows heavy reliance of Paraguayan export on all food items with 44% in 1995 and further concentration with 67% in 2015. Over the period of 20 years, the share of raw material from agricultural sources in the Paraguayan export has significantly declined from 36% in 1995 to 2% and also manufactures share dipped considerably to 9% in 2015; on the other hand, fuels have experienced a significant increase from mere 0.2% in 1995 to 21% in 2015. Similarly, Uruguay exports about 44% of all food items in 1995, and with an increase in its share it accounts for more than half of Uruguayan export in 2015. Unlike Paraguay, manufactured goods also account for about 39% in 1995; however, its proportion has declined in 2015 to 24%. This broadly refers to gaining significance of all food items with respect to other commodity groups in Uruguay’s export; in the case of Paraguay all food items and fuels have emerged as key export commodities. Table 6.4 indicates relative concentration of manufactured goods in the import basket of MERCOSUR economies in 1995 and 2015. Similar to their export pattern, fuel import has also gained relatively in its share in 2015, for example, Argentina and Paraguay. Hence, it is quite clear from the analysis of commodity composition of MERCOSUR economies that the composition of trade basket has shown some signs of relative change in the considered years. First, there has been a visible reshuffling in the proportional share of dominant traditional commodity groups. Second, due to their rich natural resources, these economies export primary commodities, particularly, all food items, and import largely manufactured goods.

6.4  Trade Structure by Major Commodity Groups: MERCOSUR

113

It is to be noted that the aforesaid analysis is based on the aggregate commodity group. In the subsequent paragraphs, an attempt is made to throw some light on the spatial pattern trade in the commodities at the disaggregate level. However, this study is limited to only four major commodities, hence restricting the scope of the current chapter. Figures 6.11, 6.12, 6.13, 6.14, 6.15, 6.16, 6.17, 6.18 illustrate the top ten trading partners of the individual MERCOSUR countries in four commodities, viz., food and live animals, beverages and tobacco, mineral fuel/lubricants and manufactured goods, for two points of time, 1995 and 2015. Figure 6.11 shows, of the total trade of Argentina in food and live animals, the top ten players account for 69% in 1995 and 49% in 2015. Predominance of the developed North is quite evident, for example, Netherlands (8%), the United States, P.Rico (7%), Germany (5%), Spain (5%), Italy (5%), Japan (5%) and the United Kingdom (3%) are among the top rankers. Furthermore, 28% of Argentina’s trade in food and live animals is with Brazil and Paraguay, taken together. In this context, it is to be noted that Brazil has the topmost rank among others with a lion’s share of 26%. In 2015, Brazil’s share dips considerably to 10%, yet it retains its top position. Besides, Southeast Asian countries, namely, Indonesia, Malaysia and Vietnam, have emerged in the period of 20 years. Broadly, Argentina has indicated a visible tilt towards its developing counterparts vis-à-vis traditional core economies. The top ten trading partners of Argentina in beverages and tobacco account for 85% in 1995 and 72% in 2015 (See Figs. 6.11 and 6.12). MERCOSUR members,

Fig. 6.11  Top Ten Trading partners of Argentina, 1995. (Source: Based on Annexure 6.5, 6.9, 6.13, 6.17)

114

6  Regional Trade Blocks: A Case Study of Mercosur

Fig. 6.12  Top Ten Trading partners or Argentina, 2015. (Source: Based on Annexure 6.5, 6.9, 6.13, 6.17)

Fig. 6.13  Top Ten Trading partners of Brazil, 1995. (Source: Based on Annexure 6.6, 6.10, 6.14, 6.18)

6.4  Trade Structure by Major Commodity Groups: MERCOSUR

115

Fig. 6.14  Top Ten Trading partners of Brazil, 2015. (Source: Based on Annexure 6.6, 6.10, 6.14, 6.18)

Fig. 6.15  Top Ten Trading partners of Paraguay, 1995. (Source: Based on Annexure 6.7, 6.11,6.15, 6.19)

116

6  Regional Trade Blocks: A Case Study of Mercosur

Fig. 6.16  Top Ten Trading partners of Paraguay, 2015. (Source: Based on Annexure 6.7, 6.11,6.15, 6.19)

Fig. 6.17  Top Ten Trading partners of Uruguay, 1995. (Source: Based on Annexure 6.8, 6.12, 6.16, 6.20)

6.4  Trade Structure by Major Commodity Groups: MERCOSUR

117

Fig. 6.18  Top Ten Trading partners of Uruguay, 2015. (Source: Based on Annexure 6.8, 6.12, 6.16, 6.20)

taken together, account for 39% in 1995. Brazil and Paraguay are the top rankers. However, their share registers a sharp decline in 2015, with Brazil’s proportion lowering to mere 5% and Paraguay’s to 4%. In 1995, developed countries, such as, Spain, the United Kingdom, Belgium-Luxembourg, France, Germany, the United States and Japan, are also seen among the top ten players. In 2015, there is an increase in the relative share of developed economies in the trade of beverages and tobacco with Argentina. For example, the United States surpasses Brazil and, with the mounting share of 28%, emerges at the top. Further, the emergence of new trading partners could also be seen, for example, the Canada, China and the Netherlands. In this context, from the geographical distribution of Argentina’s trade in beverages and tobacco, it could be inferred that there has been a significant tilt towards the North at the cost of the MERCOSUR members. Taken together, the top ten trading partners of Argentina accounts for 98% in 1995 and 96% in 2015 (see Figs. 6.11 and 6.12). There is marked reshuffling in the relative position of the top three key players in the Argentinean trade of mineral fuels. For example, in 1995, Brazil (34%), Chile (28%) and the United States (15%) hold the top three position; however, in 2015, the United States improves its share to 26%, followed by Brazil (19%) and China (12%). Broadly, it could be inferred that there is a marked predominance of intra-­regional trade, and also significance of

118

6  Regional Trade Blocks: A Case Study of Mercosur

European economies is far more limited relative to Argentina’s trade in other commodities analyzed earlier. The top ten trading partners of Brazil account for 60% of the total trade in food and live animals in 1995 (Fig. 6.13), although their share slashes to 45% in 2015. Similar to Argentina, since 1995–2015, as illustrated in Figs.  6.13 and 6.14, the geographical structure of Brazilian trade in food and live animals has diversified. For example, Asian economies, viz., China, Saudi Arabia and Hong Kong, have come up as important markets. Furthermore, Argentina is the only MERCOSUR economy to appear among the top ten traders of Brazil in 1995, though replaced by Venezuela in 2015, the per cent share of European economies taken together record a decline in 2015. Furthermore, the top ten trading partners of Brazil in beverages and tobacco are shown in the Figs. 6.13 and 6.14. These economies, considered together, account for 77% in 1995 and 69% in 2015 of the Brazilian trade. Paraguay is the only MERCOSUR economy among the top ten trading countries with a share of 16 % in 1995 and mere 6% in 2015. Developed economies, such as Belgium-Luxembourg (top most), the United States, Germany, the United Kingdom and the Netherlands are other trading partners of Brazil. Further, it could be inferred from Fig. 6.14 that the geographical distribution of Brazilian trade in beverages and tobacco become more diverse. China emerges as an important trading partner at the second position, with a share of 11%. Other new entrants among the top ten players are the Russian federation, Indonesia, Vietnam and Turkey. Figures 6.13 and 6.14 show that in addition to traditional markets of the United States and Europe, Brazilian mineral fuel trade has diversified, with a remarkable emergence of developing economies, particularly Asia, for example, China and India. Figures 6.15 and 6.16 portray the top ten trading partners of Paraguay in 1995 and 2015. These economies, taken together, account for more than 90% of trade in food and live animals over the years. Out of this 90%, 60% in 1995 and mere 16% in 2015 happened between MERCOSUR members. Paraguay’s trade in this commodity is spatially entrenched with a significant proportion of trade with Brazil over the years. However, in 2015, trade becomes more geographically diverse with the coming up of new trading economies, such as Poland, the United Kingdom, the Russian Federation and Israel. Also it is to be noted that there is a marked reshuffling of the top five trading partners of Paraguay in food and live animals. Brazil, the Russian Federation, Uruguay, Chile and Peru secured the top five positions in the recent year. Like other MERCOSUR members, even in the case of Paraguay’s trade, the United States did seem to lose its relative importance over the decades. Figure 6.17 shows in 1995 about 35% of the Paraguayan trade in beverages and tobacco is with MERCOSUR economies. Of this, Argentina accounts for a share of 24% and has come out to be the dominant partner in 1995. France with 18%, South Africa custom union (13%), Belgium-Luxembourg and Brazil with 11% each, are among the top five rankers. In 2015, Paraguay experienced several changes in the geographical distribution of its trade in beverages and tobacco. First, none of the

6.4  Trade Structure by Major Commodity Groups: MERCOSUR

119

African economies are among the key players; second, Argentina and Brazil are replaced by other South, Central American and the Caribbean island nations, in 2015; third, the share of European economies, taken together, has also experienced a decline. Figures 6.17 and 6.18 make it quite explicit that Uruguayan top ten trading partners in food and live animals in 1995 and 2015 account for about 82% and 72% of its trade, respectively. In 1995, MERCOSUR economies, taken together, account for about 49% of the food and live animals trade, and it is pertinent to mention that Brazil holds the top rank with a share of 45%. Other trading countries cover largely Europe and America, and Israel was the only Asian economy to be seen among others. In 2015, the share of MERCOSUR economies dips to 15%, with only one MERCOSUR member, that is Brazil, as the second most important partner. The rise of China at the top position with the share of 17% in 2015 is quite interesting to note, and also the United States registers an improvement in its share with 10%. Figure 6.18 reveals a very high spatial concentration of Uruguayan trade in beverages and tobacco reveals a very high spatial concentration. As compared to other MERCOSUR economies, top ten trading partners of Uruguay accounted for a share of 99% in 1995 and about 90% in 2015. It is pertinent to note in 1995, 87% of Uruguay’s trade is with MERCOSUR economies, and of which Brazil accounts for 52%. However, in 2015, there is a slight decline in relative share of these member economies, and also Paraguay emerges as the dominant player. New countries join the top ten trading partners in place of traditional ones, such as Aruba (2%), the United States(3%) and Panama (3%). Figures 6.15, 6.16, 6.17 and 6.18 show similar information for smaller MERCOSUR economies, viz., Paraguay and Uruguay. It could be inferred from the spatial distribution of trade in mineral fuel that it is highly skewed in favour of Latin American economies. In case of Paraguay, Argentina is the only trading partner with a share of 100% of the trade in mineral fuel; however, in 2015, there is a drastic decline in the share of Argentina, and Brazil emerges as the topmost trading partner with a share of 79%. In the context of Uruguay, Argentina is the nodal trading partner with a remarkable share of 84% in 1995, but in 2015, Free Zones emerges at the top position with 99% of Uruguay’s trade in mineral fuel. Figures 6.11, 6.12, 6.13, 6.14, 6.15, 6.16 and 6.17 show top ten trading partners of Argentina, Brazil, Paraguay and Uruguay for manufactured goods as well. In case of all these economies, spatial structure of trade in manufactures broadly exhibits the following pattern. First, the emergence of Asian economies, particularly China, among the top ten trading partners of individual MERCOSUR economies, and it also reflects on the growing South-South trade; second, relatively less dependence on European economies, mainly in case of larger MERCOSUR economies, viz., Argentina and Brazil; third, trade in manufactured goods with the United States is higher in the case of Argentina and Brazil, unlike Paraguay and Uruguay; fourth, these two smaller countries record relatively higher proportion of trade with MERCOSUR partners, as compared to Argentina and Brazil.

120

6  Regional Trade Blocks: A Case Study of Mercosur

6.5  Concluding Remarks The MERCOSUR economies are different in terms of their economic size, level of development and nature of their markets. Considering GDP, population and territory, Paraguay and Uruguay are smaller MERCOSUR members. Argentina and Brazil are the key players in the region. At the same time, there are other structural and political differences which led to differential pattern of international trade of individual MERCOSUR member economies intra-regionally as well as inter-­ regionally. There are multiple factors shaping the performance of MERCOSUR. In this backdrop, it has emerged in the current chapter that the trade has increased over the years; however, the degree of integration varies across countries; in case of intra-­ MERCOSUR export and import flows, Paraguay and Uruguay have limited accessibility in Argentina and Brazil. On the other hand, Paraguay and Uruguay experience dominance of these two big economies, particularly Brazil. On the contrary, Argentina and Brazil relatively trade more with each other than with the smaller MERCOSUR economies; residuals from regression of GDP ratio on trade ratio for 1990 and 2015 across MERCOSUR members broadly reflect that developed countries seem to trade more than their size with individual MERCOSUR economies, and the actual trade is more than the expected levels. On the other hand, developing countries trade less than their size with Argentina, Brazil, Paraguay and Uruguay over the years. However, there are a few exceptions to this finding. Furthermore, it is worth noting that the trade between the MERCOSUR countries is much less than the potential trade over the past two decades; The commodity composition of the trade basket reveal that these economies largely export primary commodities and import manufactures. However, Brazil is an exception with export and import heavily relying on the manufactured goods.

References Estevadeordal A, Goto J, Saez R (2000) The new regionalism in the Americas: the case of MERCOSUR. Working Paper 5, April, INTAL ITD Ethier WJ (July 1998) The new regionalism. Econ J 108(449):1149–1161. https://doi. org/10.1111/1468-­0297.00335 Garcia E, Navarro PM, Gomez HE (2013) The gravity model analysis: an application on MERCOSUR trade flows. J Econ Policy Reform 16(4):336–348. https://doi.org/10.108 0/17487870.2013.846857 Gaulier G, Jean S, Ünal-Kesenci D (2004) Regionalism and the regionalisation of international trade. Working Paper No 2004-16, CEPII Goto J, Hamada K (1999) Regional economic integration and article XXIV of the GATT. Rev Int Econ 7(4):555–570. https://doi.org/10.1111/1467-­9396.00181 Knapp T (2004) Models of economic geography: dynamics, estimation and policy evaluation. Labyrint Publications, Ridderkerk Mansfield E, Pevehouse J (2013) The expansion of preferential trading arrangements. Int Stud Q 57(3):592–604 UNCTAD (2006) World economic situation and prospects. United Nations, Geneva

Chapter 7

Concluding Observations

7.1  Introduction World trade has grown rapidly during the last two decades of the twentieth century. Increasing trade is a consequence of a paradigm shift in economic policies of many countries. Growth in international trade and flow of foreign capital have had infused efficiency in some economies and may have also adversely affected others. Both developed and developing countries are affected by integrating market economy, but in varying degrees. International trade is an important area of geographical enquiry. However, geographers in general, and more so in India, have not addressed this theme for many years. By and large, research by geographers is confined to case studies but fail to generalize. An attempt is made through this book to capture the changing dynamics of global trade from the geographical perspective. This approach broadly entails to study the changing nature, pattern and structure of the trade network embedded in space and time. In this background, the study has thrown light on how different countries responded differently to the globalization process via trade flows. This chapter not merely summarizes what has already been brought out in the preceding chapters but also reflects on the analytical results in an integrated manner.

7.2  Major Issues Countries Are Affected by Globalization, But in Varying Extent  International trade is one of the key drivers of globalization, which itself is an uneven process across time and space. Almost all countries are affected by it; however, its impact varies. Similarly, countries in order to increase their trade have opened up by undertaking reforms in export and import policy; although, such expected increase is confined to few countries (See Sects. 2.3. and 2.4). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9_7

121

122

7  Concluding Observations

Sectoral change in the trade basket at different scales has been noted. The traditional pattern of global trade characterized by export of primary commodities from developing countries and manufactured exports from developed countries is being replaced by a more complex pattern. Analysis of world trade in manufactures reflects the significance of the developed countries and also the extent to which the manufacturing sector has emerged in the global trading system. (See Sect. 4.3). It could be observed that a new geography of trade is emerging, particularly with the coming up of Asian tigers. For example, Asian economies such as Malaysia, the Philippines and Thailand experienced a shift in the structure of their export from largely primary commodities to manufactures. Emerging economies, for example China, Singapore and South Korea, continued to have predominance of manufactured goods in their export basket, although India remained dependent on primary commodities. It is also seen that African and South American economies are the slowest regions to diversify. The commodity concentration of export and import vary across developing economies. It reflects at the differential pattern of development and structural change these countries are undergoing that has affected their pattern of spatial dependency and commodity exchange. How sustainable is this exchange particularly for the less developed countries is the question raised at various global platforms; however, answer to this is yet to be received. Similarly, the regionalization process also captures the underlying structural spatial dependency via trade flows in the globalized era. Regionalization of trade reflects dependence of less developed countries on trade with developed economies that are either geographically or geopolitically close. For example, the United States has emerged as the nodal receiver as well as shipper of merchandise trade of South American economies; France is the focal trading partner of its former African colonies. Furthermore, pattern of origin and destination of import and export flow reflect significant regional polarization. For example, the United States, Germany, Japan, China and India are nodal suppliers and receivers largely for economies that were part of the same geographical region (See Sects. 3.3 and 3.4). In this context, Michalak and Gibb (1997), drawing examples from Krugman(1991), Summers(1991) and Hufbauer (1990), said, ‘trading blocs merely formalize the already existing practice of geographically proximate countries or natural partners, who are bound to trade with each other more than with distant or unnatural partners’. According to Yadav (2012), ‘Despite the globalization trends, countries still prefer to trade with countries that are proximate which underlines the fact that more or less significant volume of the trade is intra-regional. Such a pattern perhaps is also influenced by the need to maintain positive international relations’. Nevertheless, it is worthy to note that the degree and nature of intra-regional trade does vary from one trade cluster to the other. Furthermore, predominance of the traditional cores persists in the global trade network; however, the rise of emerging countries as new cores, such as China (in particular) and India, will surely impact the transformation of the global trade system.

7.2  Major Issues

123

Continued Growth in the South-South Trade  South-South integration is key to rebalancing the world economic system. Developing countries play a substantial role in international trade, and their importance is only expected to grow. The pattern suggests that emerging countries will come to occupy a predominant position in the global trading system. Furthermore, the rise of the South as market presents developing countries with greater opportunities to transform their productive structures and develop more sophisticated export sectors. It can be conjectured from the analysis that trade among South economies has increased remarkably over the years; however, this increase is confined to a few. Regional economic and trade cooperation, including bilateral and regional trade agreements (RTAs), has intensified in a number of developing economies. These initiatives resulted in deepening integration among developing countries, which gave impetus to South-­South trade. Rise of China in the commodity trade is remarkable and has to a greater extent crowded out traditional and regional partners in the recent years. It is further noted that there is a visible shift of China and India from the periphery to the centre of the world trade system, consequently reshaping the spatial interdependency. In this context, Bonapace (2005) put forth, ‘Currently, there is great interest in Asia in the potential of regional economic integration for exploiting their synergies for mutual benefit as evident from proliferation of initiatives at sub-regional and bilateral FTAs in the region’. These attempts have led to the emergence of Asian economic integration, which is reflected in the distinct tilt in favour of these countries. Therefore, the role of China and India is seen through this notion of Asian unification in the foreseeable future. Competitiveness Leads to Specialization in Production and Diversification of Consumption  Competitiveness leads to specialization in production due to comparative advantage and diversification of consumption. Consequently, one expects an increase in concentration in commodity composition of exports. The process of commodity concentration of exports would lead to diversification of import. The index of commodity concentration of export and import shows emerging Asian economies such as China, Singapore and South Korea, recorded to have specialized export base in manufactured goods; on the other hand, their import basket is diversified with multi commodity structure. This could also be seen in case of developed economies, for example Germany, Italy, Switzerland and Japan. Regionalism in International Trade  The growing presence of regionalism will influence the nature and structure of the global economy. It is seen through the present study that not only globalization but also regionalization and regionalism are at the centre of the economic transformation driving the contemporary world economy and have an impact on the evolving spatial dependency via trade flows. Hence, the emerging forms of trade relations based on regionalism and blocs across developed and developing economies will determine the nature of ‘geoeconomic’ terms in the years to come, and the question of welfare gains stemming from it remains a debatable issue. In this context, Yadav (2012) puts forth that increasing competition in the world economy ‘necessitates further collaboration among countries in the region to

124

7  Concluding Observations

multiply the advantages offered by economic integration’; also it has been pointed out that there is heavy reliance of the emerging economies on the European and US market. The example of MERCOSUR broadly hints at the fact that once perceived as most promising bloc it has failed to meet the expectations. Role of big countries like Brazil, differences in political and economic ambitions, similarity in resource endowments and demand structure, higher dependence on non-members (e.g. the United States, the EU and China), trade diversion, geopolitics, lack of infrastructure and size asymmetries are possible inhibiting factors. Hence, the expected welfare gains from such regional blocs are limited and also question its very sustenance in the multilateral trading system. Last but not the least, this book is an attempt to provide a holistic picture of the changing trends and pattern of international trade in the spatial context. However, as mentioned at the very outset, it does not give a complete and detail picture of the dynamism of trading system. Therefore, the hitherto unresolved and unanswered dimensions  – both theoretical and methodological  – provide agenda for future research, as it is beyond the scope of the current book. Future research needs to stress and study these issues particularly from the geographical perspectives. Furthermore, it will be equally interesting to conduct intensive research on the regional and local economies in the context of interlinkages via trade as it will give a complete understanding of the welfare gains from trade across spatial scales and also how globalization is linking the local spaces with the global. Such studies are very significant in comprehending the structural connects and disconnects between the cores and peripheries found at different geographical scales in the globalized era. In the context of sectoral composition of trade, it is important to note that despite the increasing role of services in the globalization process, it is largely ignored by the researchers, and more so by geographers. Broadly, an effort is made via this book to throw some light on this sector, but there is immense scope in the future to undertake a comprehensive study.

References Bonapace T (2005) Regional trade and investment architecture in Asia-Pacific: Emerging trends and Imperatives, RIS Discussion Paper No. 92/2005. www.ris.org Hufbauer G (1990) Europe 1992: an American perspective. The Brookings Institution, Washington, DC Krugman P (1991) The move toward free trade zones. Econ Rev 76:5–25 Michalak W, Gibb R (1997) Trading blocs and multilateralism in the world economy. Ann Assoc Am Geogr 87(2):264–279 Summers L (1991) Regionalism and the world trading system. The WTO and Reciprocal Preferential Trading Agreements. Ed. Caroline Freund. Edward Elgar Publishing, 2007, pp. 421–427 Yadav P (2012) Impact of globalization on India: an analysis of international trade and capital flows. LAP Lambert Academic Publishing, Germany

Annexures

Annexure 2.1 The country distribution presented emanates from the UNCTAD Handbook of Statistics (2006–07), which follows those used by the Statistics Division, Department of Economic and Social Affairs (DESA) of United Nations. The term ‘economies’ covers regions, countries and territories. Three main groups of countries/territories are as follows: • Developed economies • Economies in transition • Developing economies The exact composition of each group is shown below: Developed economies In Europe: Andorra, Austria, Belgium, Cyprus, Czech Republic, Denmark, Estonia, Faero Islands, Finland France, Germany, Gibraltar, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Luxemburg, Malta, Monaco, Netherlands, Norway, Poland, Portugal, San Marino, Slovakia, Slovenia, Spain, Sweden, Switzerland, the United Kingdom of Great Britain and Northern Ireland In America: Bermuda, Canada, Greenland, Saint Pierre and Miquelon, the United States of America, Puerto Rico In Asia: Israel, Japan In Oceania: Australia, New Zealand

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 P. Yadav, Geographical Perspectives on International Trade, SpringerBriefs in Geography, https://doi.org/10.1007/978-3-319-71731-9

125

126

Annexures

Economies in Transition Europe: Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Republic of Moldova, Romania, Russian Federation, Serbia and Montenegro, the former Yugoslav Republic of Macedonia, Ukraine Asia: Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyzstan, Tazikistan, Turkmenistan, Uzbekistan Developing economies America South America: Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Falkland Islands (Malvinas), Guyana, Paraguay, Peru, Suriname, Uruguay, Venezuela Central America: Belize, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama Caribbean islands: Anguilla, Antigua and Barbuda, Aruba, Bahamas, Barbados, British Virgin Islands, Cayman Islands, Cuba, Dominica, Dominican Republic, Grenada, Haiti, Jamaica Montserrat, Netherlands Antilles, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Trinidad and Tobago, Turks and Caicos Islands, the United States Virgin Islands Africa North Africa: Algeria, Egypt, Libyan Arab Jamahiriya, Morocco, Sudan, Tunisia, Western Sahara Southern Africa: Botswana, Lesotho, Namibia, South Africa, Swaziland Middle Africa: Angola, Cameroon, Central African Republic, Chad, Congo, Democratic Republic of the Congo, Equatorial Guinea, Gabon, Sao Tome and Principe Eastern Africa: British Indian Ocean territory, Burundi, Comoros, Djibouti, Eritrea, Ethiopia, Kenya, Madagascar, Malawi, Mauritius, Mayotte, Mozambique, Rwanda, Seychelles, Somalia, Uganda, the United Republic of Tanzania, Zambia, Zimbabwe Western Africa: Benin, Burkina Faso, Cape Verde, Cote d’Ivoire, Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali, Mauritania, Niger, Nigeria, Saint Helena, Senegal, Sierra Leone, Togo

Annexures

127

Asia Eastern Asia: China, South Korea, Hong Kong SAR, China Macao SAR, China, Mongolia, the Democratic People’s Republic of Korea, Taiwan, Province of China Southern Asia: Afghanistan, Bangladesh, Bhutan, India, Iran, Maldives, Nepal, Pakistan, Sri Lanka South-Eastern Asia: Brunei Darussalam, Cambodia, Indonesia, the Lao People’s Democratic Republic, Malaysia, Myanmar, the Philippines, Singapore, Thailand, Timor-Leste, Vietnam Western Asia: Bahrain, Iraq, Jordan, Kuwait, Lebanon, Oman, Occupied Palestinian territory, Qatar, Saudi Arabia, Syrian Arab Republic, Turkey, the United Arab Emirates, Yemen

Annexure 2.2

Annexure 2.3 Value of Merchandise Trade (US Million $)

Merchandise Trade Import Export Trade

1990 3609.25 3495.68 7104.93

1995 5225.56 5168.92 10394.48

2000 6654.56 6452.32 13106.89

Source: UNCTAD Handbook of Statistics, 2016

2005 10777.64 10502.49 21280.13

2010 15420.51 15302.14 30722.65

2015 16607.24 16551.59 33158.83

Annexures

128

 nnexure 2.4 Annual Average Growth Rate of Merchandise A Export and Merchandise Import (Values in Per Cent)

Export World Developing economies Transition economies Developed economies Import World Developing economies Transition economies Developed economies

1990–1995 7.76 10.94 4.75 6.77

1995–2000 3.6 5.8 1.8 2.8

2000–2005 11.4 14.4 20 9.5

2005–2010 6.3 9.2 9.4 4.1

2010–2015 1.5 2.7 −2.6 0.8

1990–1995 7.25 12.87 1.18 5.6

1995–2000 4 3.1 −4.9 4.6

2000–2005 11.3 13.5 21.8 10

2005–2010 5.9 10.4 12.1 3.2

2010–2015 1.4 3 −2.9 0.5

Source: UNCTADstat (http://unctadstat.unctad.org)

 nnexure 2.5 Compound Annual Growth A in Merchandise Trade

S.NO. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Year Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire

CAGR 1990–1995 −2.11 −1.42 20.12 6.56 0.93 14.02 7.50 9.37 13.29 6.93 1.69 −3.62 7.50 14.37 19.46 14.27 17.09 5.26

CAGR 2000–2005 16.29 24.25 6.04 12.56 12.64 8.71 12.52 11.44 11.51 9.09 21.28 11.03 5.54 14.43 24.56 11.52 6.61 16.32

CAGR 2010–2015 −1.75 −4.01 −1.39 −0.26 −4.88 8.85 −0.71 8.30 −1.23 −3.49 3.57 2.40 1.35 −0.61 5.88 2.28 1.77 1.82 (continued)

Annexures S.NO. 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62

Year Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland

129 CAGR 1990–1995 21.38

CAGR 2000–2005 20.89

6.65 11.57 13.18 5.16 21.98 3.85 5.49 5.50 2.86 4.92 11.57 5.80 13.44 1.75 14.28 6.85 1.70 9.48 12.65 −4.04 11.64 4.54 8.30 8.33 8.39 14.05 13.31 20.89 9.46 4.56 12.81 −10.93 8.61 8.51 7.87 1.36 4.11 8.45 10.79 17.44 17.20 16.76 14.94

10.71 1.03 18.71 12.24 5.08 10.28 9.11 7.79 12.99 10.81 11.89 9.71 15.04 5.53 9.60 16.49 12.44 20.89 8.34 17.70 6.82 9.60 5.28 17.90 13.87 10.40 17.93 7.27 0.72 7.08 5.05 14.35 11.02 11.33 12.00 19.12 11.01 15.79 31.48 9.02 15.76 3.85 18.79

CAGR 2010–2015 2.84 0.15 3.48 0.91 0.31 4.25 5.19 −2.88 −1.01 −1.50 0.56 4.76 −4.08 4.95 6.30 4.75 0.81 3.33 2.71 −0.03 −8.85 1.64 −1.46 −2.75 4.51 4.97 1.56 −1.27 0.68 −0.70 1.75 5.25 6.51 2.25 −0.33 2.72 −4.99 −2.68 2.32 5.33 2.44 1.81 3.21 2.96 (continued)

Annexures

130 S.NO. 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87

Year Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe World

CAGR 1990–1995 5.85 7.69 2.67 3.95 16.40 6.84 8.12 14.34 5.41 3.92 4.58 17.79 6.32 8.16 10.20 8.46 4.29 8.28 10.37 −0.50 4.61 21.41 2.77

CAGR 2000–2005 9.14 23.69 17.37 15.47 9.55 13.81 12.20 5.33 8.74 9.57 18.29 11.85 15.58 10.42 18.25 18.94 7.28 5.23 4.85 19.83 9.90 18.10 17.28

−7.64 6.13 7.94

19.66 2.09 10.18

CAGR 2010–2015 −0.83 3.15 0.78 3.40 −0.47 −1.78 0.33 5.89 −2.03 7.81 −27.41 2.08 1.70 −2.36 3.23 6.05 1.53 3.26 2.27 −2.98 −7.91 15.88 1.46 4.24 −0.48 1.54

Source: UNCTADstat

Annexure 2.6 Value of Service Trade (US dollars at current prices and current exchange rates in millions) 1990 1995 2000 World 1,706,540 2,463,160 3,041,370 Developing economies 344,150 608,690 768,530 Transition economies 46,760 Developed economies 1,315,120 1,812,060 2,226,080 Source: UNCATDstat

2005 5,045,580 1,331,990 112,600 3,600,990

2010 7,635,510 2,413,560 210,590 5,011,370

2013 9,219,370 3,119,210 314,650 5,785,510

Annexures

131

 nnexure 2.7 Compound Annual Growth Rate A of Service Trade

Year Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg  Bolivia  Brazil  Brunei Darussalam  Bulgaria  Cameroon  Canada  Chile  China  Colombia  Costa Rica  Côte d’Ivoire  Czechoslovakia (former)  Denmark  Dominican Republic  Ecuador  Egypt  El Salvador  Fiji  Finland  France  Gabon  Germany  Greece  Ghana  Guatemala  Guinea  Honduras  Hungary  Iceland  India  Indonesia  Indonesia

1990–1995 8.16 2.47 14.78 3.10 9.60 15.34 4.50 3.52 11.86 – 13.52 −10.87 4.64 12.24 34.15 6.43 10.19 −2.96 3.31 13.67 7.21 6.65 6.89 7.38 5.84 4.63 −3.24 7.36 11.91 8.52 12.96 −0.69 10.61 10.66 3.63 9.73 – 17.34

1995–2000 3.98 6.51 5.01 −1.99 5.11 1.78 5.97 4.97 5.77 – 7.25 16.17 7.19 4.91 8.42 3.21 11.27 −2.16 0.75 10.68 9.54 2.19 5.16 12.68 −4.59 0.95 −1.19 −0.44 1.01 12.79 13.32 3.32 −6.98 15.22 4.71 10.60 16.05 – 1.87

2000–2005 17.39 18.62 0.15 13.15 21.44 7.25 13.44 11.11 9.08 12.31 15.31 7.37 7.59 10.88 18.97 6.79 5.16 11.58 13.30 12.31 3.31 8.28 7.75 4.37 12.25 15.25 10.16 2.78 11.22 9.86 16.94 11.47 0.47 6.28 17.03 15.74 22.71 – −100.00

2005–2010 16.04 22.99 14.82 4.56 5.92 14.50 9.88 7.75 18.41 9.08 6.61 3.68 7.02 10.18 17.55 10.85 8.14 5.54 10.52 7.13 6.28 7.23 8.90 0.25 1.01 9.07 9.69 −100.00 6.52 3.38 13.49 10.62 4.86 5.66 7.54 0.49 18.31 4.15 -

2010–2013 −1.48 7.32 6.04 6.28 −7.89 15.23 6.14 24.81 10.09 −100.00 4.89 −100.00 2.86 5.61 14.71 10.94 9.14 −100.00 3.50 4.60 4.72 7.85 −3.46 11.33 8.32 3.20 5.35 5.67 −3.72 17.95 3.75 −100.00 8.98 3.03 10.41 6.19 10.10 – (continued)

Annexures

132 Year  Iran  Ireland  Italy  Japan  Jordan  Kenya South Korea  Kuwait  Malaysia  Malta  Mauritius  Mexico  Mongolia  Morocco  Netherlands  New Zealand  Nigeria  Norway  Pakistan  Panama  Paraguay  Peru  Philippines  Poland  Portugal  Romania  Saudi Arabia  Senegal  Singapore  South Africa  Spain  Sri Lanka  Sweden  Switzerland  Syria  Thailand Trinidad and Tobago  Tunisia  Turkey Union of Soviet Socialist Republics United Arab Emirates United Kingdom United States

1990–1995 −7.79 13.59 4.10 8.38 4.13 7.98 19.30 7.77 23.26 7.24 9.42 1.14 −5.50 3.30 9.80 9.47 12.19 1.32 6.48 7.92 8.72 8.83 −100.00 24.11 10.27 18.87 −2.20 −1.73 17.74 8.19 6.95 13.35 1.30 6.02 14.24 21.48 −6.26 8.79 12.10 –

1995–2000 10.93 22.95 −1.05 −0.38 0.23 −8.70 5.54 0.00 2.91 0.62 5.25 9.27 9.51 3.93 2.32 0.78 −0.36 4.05 −5.40 3.76 −4.74 5.13 – 1.77 1.62 2.42 5.71 −6.20 3.86 0.52 6.86 4.87 6.79 2.46 −0.41 −2.73 10.00 0.63 8.48 14.32

2000–2005 26.32 23.48 9.92 5.82 7.72 12.00 10.94 14.87 6.24 11.67 8.97 3.74 29.92 19.38 10.81 13.86 10.39 12.48 25.23 9.97 −0.32 7.07 3.80 10.39 9.70 23.20 8.21 14.86 13.60 16.60 13.56 7.23 11.44 10.82 9.38 9.73 8.83 9.29 6.24 20.11

2005–2010 11.87 9.35 3.01 4.08 15.49 13.87 10.85 12.87 9.07 17.30 10.67 2.29 7.30 10.84 4.90 3.31 23.83 7.01 4.12 12.87 7.11 12.45 23.18 14.39 8.12 9.93 14.45 6.47 11.94 6.74 5.55 9.01 6.61 10.51 15.45 11.21 −2.59 8.05 7.17 12.33

2010–2013 −100.00 5.76 1.95 1.47 2.95 8.26 6.29 3.47 9.91 4.45 9.39 7.81 29.25 1.99 7.33 6.12 0.14 2.94 −2.44 16.17 10.71 12.29 7.21 5.50 3.15 13.74 0.31 −100.00 8.78 −1.96 4.00 13.60 7.20 7.39 −100.00 12.92 −100.00 −3.13 7.97 15.19

13.44 6.17 6.31

8.30 9.13 7.12

17.58 11.04 5.94

17.39 3.42 7.28

17.90 2.85 5.74 (continued)

Annexures

133

Year Uruguay Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe World

1990–1995 20.90 11.85 68.06 – −100.00 −100.00 7.62

1995–2000 −0.54 −2.90 7.60 9.33 – – 4.31

2000–2005 0.85 3.53 7.92 17.93 10.60 −4.70 10.65

2005–2010 13.39 17.42 14.81 8.26 9.82 3.71 8.64

2010–2013 15.00 11.18 10.89 4.03 −100.00 −100.00 6.48

Note: Countries not classified due to unavailability of required data: Brunei Darussalam, the United Arab Emirates, the USSR (former), Vietnam, Yugoslavia, SFR (former), Zambia, Zimbabwe

 nnexure 3.1 Results of Q Mode Analysis, A Communalities (1990)

Initial Extraction

Initial Extraction

Initial Extraction

Algeria Angola Argentina Austria Bahrain Bangladesh BelgiumLuxembourg Bolivia

1 1 1 1 1 1 1

0.981 0.356 0.924 0.975 0.799 0.81 0.878

Ghana Greece Guatemala Guinea Honduras Hungary Iceland

1 1 1 1 1 1 1

0.697 0.987 0.944 0.916 0.991 0.956 0.863

Panama Paraguay Peru Philippines Poland Portugal Romania

1 1 1 1 1 1 1

0.959 0.954 0.947 0.985 0.985 0.811 0.97

1

0.893

India

1

0.912

1

0.967

Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile Colombia

1 1

0.917 0.895

Indonesia Iran

1 1

0.981 0.916

Saudi Arabia Senegal Singapore

1 1

0.983 0.849

1 1 1 1 1

0.95 0.978 0.987 0.966 0.979

Ireland Italy Japan Jordan Kenya

1 1 1 1 1

0.929 0.968 0.889 0.942 0.934

1 1 1 1 1

0.971 0.867 0.835 0.976 0.847

Costa Rica

1

0.991

1

0.977

1

0.972

Côte d’Ivoire

1

0.92

South Korea Kuwait

Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand

1

0.963

0.956

0.959

Malaysia

1

0.909

Trinidad 1 and Tobago Tunisia 1

0.946 0.987

Malta Mauritius

1 1

0.933 0.965

Turkey USSR (former)

0.988 0.868

Czechoslovakia 1 (former) Denmark 1 Dominican 1 Republic

1 1

0.979

(continued)

Annexures

134 Initial Extraction

Initial Extraction

Ecuador Egypt

1 1

0.983 0.983

Mexico Mongolia

1 1

0.988 0.753

El Salvador

1

0.93

Morocco

1

0.931

Fiji Finland

1 1

0.375 0.838

1 1

0.899 0.986

France Gabon

1 1

0.915 0.973

Netherlands New Zealand Nigeria Norway

1 1

0.976 0.666

Germany

1

0.818

Pakistan

1

0.944

Initial Extraction UAE United Kingdom United States Uruguay Venezuela

1 1

0.959 0.922

1

0.602

1 1

0.94 0.985

Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1 1

0.958 0.974

1 1

0.896 0.966

Extraction Method: Principal Component Analysis

 nnexure 3.2 Results of Q Mode Analysis, Factor A Loadings (1990)

Receiver Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic

Rotated component matrixa 1 2 3 0.238 0.33 0.88 0.268 0.174 0.373 0.75 0.391 0.192 0.033 0.978 0.08 0.347 0.04 0.045 0.153 0.106 0.072 0.202 0.712 0.49 0.515 0.136 0.048 0.826 0.277 0.146 0.107 0.01 0.015 0.068 0.958 0.117 0.015 0.182 0.968 0.977 0.011 0.063 0.836 0.221 0.153 0.934 0.15 0.122 0.977 0.036 0.052 0.045 0.112 0.944 −0.013 0.688 0.031 0.114 0.885 0.213 0.983 −0.002 0.052

4 0.095 0.024 0.132 0.064 0.144 0.639 0.083 0.14 0.237 0.182 0.045 0.053 0.124 0.262 0.235 0.16 0.054 −0.028 0.099 0.106

5 −0.003 0.144 0.039 0.034 0.155 0.119 0.274 0.069 0.118 0.498 0.01 0.041 0.095 0.133 0.086 0.069 0.076 0 0.299 0.065 (continued)

Annexures

Receiver Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania

135 Rotated component matrixa 1 2 3 0.92 0.167 0.108 0.661 0.414 0.429 0.954 0.046 0.05 0.104 −0.012 0.03 0.125 0.676 0.143 0.169 0.839 0.087 0.062 0.058 0.977 0.167 0.144 0.587 0.232 0.265 0.182 0.097 0.808 0.366 0.958 0.074 0.049 0.159 0.142 0.921 0.981 0.017 0.046 0.019 0.881 0.06 0.502 0.465 0.122 0.423 0.346 0.242 0.274 0.243 0.101 −0.026 0.771 0.188 0.167 0.173 0.105 0.119 0.773 0.51 0.875 0.149 0.095 0.479 0.268 0.682 0.11 0.353 0.26 0.575 0.095 0.074 0.426 0.432 0.203 0.284 0.066 0.028 0.016 0.433 0.277 −0.062 0.241 0.8 0.979 0.038 0.065 −0.053 0.762 0.008 0.158 0.275 0.896 0.255 0.778 0.305 0.53 0.151 0.095 0.296 0.517 0.406 0.155 0.536 0.127 0.561 0.272 0.16 0.222 0.000 0.033 0.554 0.064 0.088 0.913 0.141 0.079 0.607 0.111 0.072 0.064 0.926 0.076 0.097 0.633 0.514 0.102 0.451 0.061

4

5 0.222 0.193 0.094 0.348 0.125 0.072 0.085 0.206 0.143 0.131 0.106 0.079 0.14 0.013 0.125 0.401 0.907 0.44 0.083 0.057 0.104 0.141 0.381 0.79 0.564 0.646 0.083 0.255 0.124 0.403 0.04 0.139 0.681 0.194 0.11 0.621 0.946 0.216 0.104 0.744 0.051 0.058 0.003

0.095 0.156 0.076 −0.017 0.367 0.293 0.05 0.369 0.709 0.186 0.07 0.11 0.068 0.009 0.525 0.463 0.076 0.162 0.906 0.217 0.106 0.331 0.726 0.078 0.366 0.075 0.212 0.243 0.071 0.002 0.068 0.313 0.439 0.587 0.488 0.252 0.015 0.08 0.081 0.093 0.054 0.258 0.01 (continued)

Annexures

136

Receiver Saudi Arabia Senegal Singapore Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey USSR (former) United Arab Emirates United Kingdom United States Uruguay Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

Rotated component matrixa 1 2 3 0.529 0.263 0.233 0.038 0.114 0.982 0.456 0.08 0.082 0.168 0.666 0.62 0.162 0.145 0.067 0.197 0.774 0.178 0.152 0.896 0.344 0.239 0.557 0.624 0.213 0.102 0.076 0.936 0.02 0.06 0.044 0.397 0.872 0.376 0.791 0.269 0.177 0.888 0.075 0.268 0.38 0.267 0.427 0.702 0.431 −0.057 0.182 0.083 0.311 0.151 0.118 0.935 0.194 0.139 −0.045 0.433 0.24 0.061 0.877 0.186 0.261 0.447 0.05 0.478 0.478 0.199

4

5 0.527 0.055 0.766 0.086 0.724 0.12 0.103 0.128 0.919 0.114 0.005 0.194 0.178 0.622 0.214 0.726 0.056 0.127 0.607 0.015 0.32 0.283

0.506 0.029 0.134 0.264 0.252 0.381 0.123 0.09 0.061 0.238 −0.018 0.179 0.01 0.393 0.066 0.143 0.08 0.142 −0.009 0.001 0.716 0.612

Extraction Method: Principal Component Analysis; Rotation Method: Varimax with Kaiser Normalization a Rotation converged in 8 iterations

Annexure 3.3 Results of Q Mode, Factor Scores (1990)

Shipper Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam

1 −0.15 −0.11 −0.05 −0.20 −0.13 −0.15 −0.18 −0.04 0.03 −0.12

2 −0.10 −0.22 −0.24 0.58 −0.25 −0.21 0.44 −0.21 −0.38 −0.25

3 −0.11 −0.19 −0.24 −0.24 −0.20 −0.20 0.82 −0.21 −0.12 −0.20

4 −0.25 −0.23 −0.33 −0.17 −0.19 −0.18 −0.14 −0.27 −0.37 −0.16

5 −0.26 −0.17 −0.41 −0.34 −0.13 −0.17 0.69 −0.25 −0.28 −0.18 (continued)

Annexures Shipper Bulgaria Cameroon Canada Chile Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria

137 1 −0.14 −0.14 −0.09 0.00 0.07 −0.01 −0.21 −0.20 0.01 −0.13 −0.01 −0.13 0.01 −0.13 −0.11 −0.45 −0.11 0.12 −0.14 −0.13 0.15 −0.14 −0.10 −0.19 −0.13 −0.33 −0.06 −0.06 −0.09 0.03 −0.48 −0.14 −0.13 −0.03 −0.13 −0.12 −0.13 −0.13 0.41 −0.13 −0.14 −0.08 0.01 −0.32

2 −0.23 −0.26 −0.10 −0.26 −0.25 −0.25 −0.41 0.76 0.12 −0.24 −0.25 −0.20 −0.25 −0.25 0.42 −0.22 −0.25 8.64 −0.24 −0.14 −0.26 −0.26 −0.24 0.20 −0.22 0.00 −0.27 0.03 −0.04 1.77 −0.11 −0.25 −0.24 −0.28 −0.25 −0.37 −0.24 −0.24 −0.26 −0.24 −0.23 0.63 −0.18 −0.49

3 −0.16 −0.07 −0.03 −0.25 −0.23 −0.21 0.35 −0.46 −0.25 −0.20 −0.22 −0.17 −0.21 −0.20 −0.26 8.84 −0.09 −0.16 −0.19 −0.11 −0.22 −0.10 −0.20 −0.31 −0.20 −0.15 −0.18 −0.22 −0.11 0.98 −0.05 −0.19 −0.19 −0.10 −0.10 −0.08 −0.19 −0.20 −0.22 −0.19 −0.11 0.63 −0.09 0.43

4 −0.25 −0.22 0.92 −0.25 −0.14 −0.23 −0.27 0.00 −0.20 −0.22 −0.20 −0.21 −0.26 −0.20 −0.14 −0.10 −0.22 0.05 −0.20 −0.24 −0.28 −0.22 −0.23 −0.07 −0.23 0.21 0.19 −0.04 −0.20 −0.12 8.78 −0.17 −0.21 1.15 −0.03 0.71 −0.22 −0.22 −0.24 −0.22 −0.21 −0.13 −0.09 −0.29

5 −0.19 −0.15 0.02 −0.24 −0.25 −0.19 0.24 −0.79 0.91 −0.15 −0.22 −0.17 −0.20 −0.13 0.32 −0.35 −0.18 −0.14 −0.16 −0.21 −0.19 −0.15 −0.16 −0.47 −0.14 −0.17 −0.21 −0.37 0.01 −0.85 0.05 −0.13 −0.10 −0.33 0.03 −0.16 −0.15 −0.14 −0.31 −0.16 −0.16 1.51 −0.76 0.72 (continued)

Annexures

138 Shipper Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey USSR (former) United Arab Emirates United Kingdom United States Uruguay Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1 −0.02 −0.22 −0.10 −0.10 −0.04 −0.13 −0.13 −0.04 −0.14 −0.10 −0.15 −0.40 0.04 −0.13 0.13 0.00 −0.12 −0.20 −0.11 −0.15 −0.15 −0.27 −0.24 −0.70 8.99 −0.07 0.65 −0.13 −0.18 −0.13 −0.18

2 0.23 −0.16 −0.25 −0.23 −0.22 −0.25 0.20 −0.21 0.18 −0.45 −0.28 −0.29 0.13 −0.23 0.12 0.37 −0.09 −0.25 −0.24 −0.22 −0.08 0.17 −0.28 0.03 −0.22 −0.21 −0.28 −0.14 0.41 −0.25 −0.24

3 −0.27 −0.12 −0.20 −0.21 −0.21 −0.19 −0.25 0.10 −0.26 −0.02 −0.06 −0.14 0.77 −0.16 −0.23 −0.10 −0.19 0.01 −0.20 −0.11 0.08 −0.28 −0.14 −0.01 0.35 −0.20 −0.23 −0.23 −0.32 −0.21 −0.25

4 −0.24 −0.04 −0.23 −0.22 −0.23 −0.09 −0.23 −0.29 −0.16 0.49 −0.22 0.66 −0.25 −0.18 −0.06 −0.06 −0.21 0.31 −0.22 −0.19 −0.10 −0.33 0.22 −0.16 0.41 −0.23 −0.29 −0.18 −0.05 −0.21 −0.20

5 0.39 −0.22 −0.16 −0.19 −0.21 −0.17 −0.36 0.13 −0.43 0.23 −0.15 −0.64 −0.13 −0.19 1.67 0.23 −0.22 −0.13 −0.16 −0.15 −0.33 −0.30 0.12 8.40 0.53 −0.21 −0.29 −0.24 −0.51 −0.06 0.23

Annexure 3.4 Results of Q Mode, Communalities (2015)

Country Algeria Angola Argentina Austria

Initial 1 1 1 1

Extraction 0.967 0.749 0.917 0.973

Country South Korea Kuwait Malaysia Malta

Initial 1 1 1 1

Extraction 0.944 0.935 0.867 0.933 (continued)

Annexures

139

Country Bahrain Bangladesh Belgium Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy

Initial 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Extraction 0.891 0.993 0.813 0.93 0.944 0.824 0.893 0.98 0.993 0.995 0.913 0.994 0.998 0.963 0.918

Country Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal

Initial 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Extraction 0.977 0.994 0.94 0.971 0.841 0.989 0.984 0.761 0.991 0.919 0.974 0.991 0.981 0.957 0.942

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0.942 0.993 0.96 0.975 0.987 0.808 0.924 0.858 0.955 0.697 0.989 0.95 0.996 0.973 0.993 0.923 0.848 0.898 0.902 0.96 0.728 0.866

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0.941 0.945 0.994 0.807 0.834 0.988 0.939 0.974 0.871 0.914 0.891 0.983 0.883 0.953 0.982 0.933 0.658 0.989 0.946 0.987 0.969 0.852

Japan Country Jordan Kenya

1 Initial 1 1

0.962 Extraction 0.913 0.984

Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Vietnam Yugoslavia, SFR (former) Zambia Country Zimbabwe

1 Initial 1

0.99 Extraction 0.991

Extraction Method: Principal Component Analysis

Receiver Algeria Angola Argentina Austria Bahrain, Kingdom of Bangladesh Belgium Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador

0.379 0.028 0.125 0.953 0.233 0.063 0.641 −0.014 0.256 −0.002 0.821 0.151 0.036 0.112 0.243 0.109 0.03 0.185 0.942

0.854 0.031 0.053

2

0.545 0.699 0.38 −0.022 0.527 0.97 0.118 0.153 0.555 0.749 0.024 0.875 0.039 0.552 0.101 0.335 0.141 0.723 0.11

0.172 0.13 0.346

1

0.025 0.979 0.9

0.116 0.194 0.492 0.02 0.667 0.085 0.297 0.407 0.732 0.003 −0.03 0.12 0.992 0.767 0.457 0.915 0.984 0.108 −0.01

3

0.019 0.101 0.147

0.139 0.169 0.711 0.029 0.166 0.131 −0.01 0.858 0.134 −0.04 −0.03 0.098 0.03 0.289 0.072 0.166 0.068 0.046 0.007

4

6 −0.052 0.274 0.015 0.103 0.165 −0.065 −0.011 0.045 0.015 0.287 −0.11 −0.182 0.032 −0.028 0.777 0.005 0.026 −0.1 0.076 0.023 0.008 0.002

0.577 0.251 0.058 −0.12 −0.01 −0 0.308 −0.01 0.109 −0.15 0.08 0.338 0.023 0.043 −0.05 0.055 0.022 0.576 −0.04 −0.04 0.044 0.042

5

0.415 0.028 0.046

7 −0.056 0.127 0.047 0.001 −0.012 0.003 0.314 0.006 0.048 −0.094 −0.113 0.032 0.024 0.033 0.096 0.037 0.023 0.021 0.051

Annexure 3.5: Results of Q Mode Analysis, Factor Loadings (2015)

0.003 0.004 0.004

0.003 0.172 0.014 0.013 −0.007 0.02 −0.01 −0.026 0.023 −0.041 −0.044 0.049 0.003 0.017 −0.021 0.01 0.002 0.045 0

8

0.043 0.045 0.032

0.356 −0.04 0.058 −0 0.054 0.006 0.114 0.04 0.051 0.04 0.133 0.113 −0.01 0.041 −0.04 0.033 0.011 0.211 −0.02

9

−0.077 0.014 0.025

10 0.163 0.035 −0.006 −0.187 −0.042 0.132 −0.043 −0.013 −0.004 0.01 0.406 −0.038 −0.006 −0.01 0.096 −0.007 0.001 −0.031 −0.06

0.026 0.037 −0.011

11 −0.023 −0.081 0.015 −0.059 0.234 −0.043 0.303 0.007 −0.005 0.382 −0.028 −0.043 0.054 −0.024 0.06 0.016 0.036 0.046 −0.069

140 Annexures

Receiver Receiver Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya Kuwait Malaysia Malta Mauritius

2

2 0.363 0.046 −0.076 0.665 0.849 0.103 0.283 0.113 0.763 0.038 0.144 0.026 0.951 0.439 0.155 −0.008 0.172 0.288 0.819 0.125 0.282 0.096 0.315 0.019 0.317 0.132

1

1 0.72 0.145 0.658 0.107 0.087 0.656 0.226 0.919 0.305 0.258 0.903 0.091 0.106 −0.009 0.869 0.853 0.903 0.063 0.224 0.862 0.814 0.943 0.703 0.836 0.554 0.85

3 3 0.294 0.979 0.061 0.013 0.155 0.173 0.153 0.186 0.008 0.959 0.12 0.99 0.006 0.273 0.318 0.124 0.035 0.263 0.115 0.423 0.349 0.183 0.523 0.18 0.071 0.048

4 0.189 0.059 −0.04 −0.03 0.02 0.031 −0.09 0.104 −0 0.074 0.106 0.048 0.005 0.224 0.082 0.021 0.174 0.028 0.005 0.086 0.13 0.076 0.065 0.016 −0.01 0.066

4 5 0.242 0.017 0.247 −0.03 −0.04 0.625 0.478 0.143 0.253 0.035 0.175 0.022 −0.03 0.036 0.007 −0.13 0.107 0.082 0.17 0.047 0.143 0.036 0.177 −0.16 0.181 0.319

5 6 −0.034 0.016 0.533 −0.016 −0.047 0.246 −0.28 −0.213 0.018 0.012 −0.252 0.016 0.034 −0.075 0.044 0.31 −0.075 0.021 −0.012 −0.016 −0.013 −0.143 0.114 0.247 0.281 −0.065

6 7 0.056 0.025 0.054 0.582 0.109 0.122 0.239 0.098 0.048 0.033 0.121 0.024 0.006 0.626 0.03 −0.091 0.043 0.177 0.176 0.05 0.008 0.025 0.025 −0.093 0.043 −0.019

7 8 0.007 0.003 −0.041 −0.007 −0.001 0.042 −0.065 0.094 −0.019 0.008 0.036 0.005 −0.007 −0.035 0.05 −0.047 0.03 0.04 0.004 0.045 0.013 0.146 0.016 −0.037 −0.014 0.302

8 9 0.082 0.003 −0.02 0.06 0.252 0.082 0.112 0.035 0.209 0.013 0.042 0.008 −0.01 −0.03 0.006 0.065 −0.03 −0.01 0.257 −0.01 0.032 0.003 0.07 0.066 0.023 0.167

9

10 10 0.362 −0.012 −0.02 0.354 −0.075 −0.044 0.117 0.03 0.402 0.007 0.004 0 −0.072 −0.048 0.066 0.048 0.195 −0.041 0.073 0.076 0.101 0.038 0.013 0.025 0.637 −0.06

11 11 −0.029 0.014 −0.115 0.009 0.145 −0.083 0.375 0.011 0.065 0.001 −0.012 0.029 −0.032 0.357 0.009 0.171 −0.159 0.727 0.024 −0.063 −0.031 0.005 0.11 0.185 0.005 0.038 (continued)

Annexures 141

Receiver Mexico Mongolia Morocco Receiver Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa South Korea Spain Sri Lanka Sweden Switzerland Syria Thailand

2

0.043 0.157 0.325 2 0.741 0.159 0.151 0.474 0.083 −0.029 −0.024 0.078 0.038 0.959 0.429 0.936 0.358 0.14 0.767 0.073 0.175 0.549 0.768 0.063 0.847 0.836 0.031

1

0.08 0.751 0.236 1 0.404 0.778 0.887 0.073 0.967 0.688 0.254 0.497 0.928 0.141 0.042 0.06 0.625 0.866 −0.043 0.795 0.862 0.755 0.25 0.967 0.12 −0.029 0.863

0.99 0.055 0.14 3 0.28 0.537 0.252 0.099 0.139 0.585 0.405 0.822 0.274 0.005 0.007 −0.03 0.597 0.085 −0.05 0.418 0.389 0.271 0.078 0.077 0.041 0.25 0.227

3 0.037 0.031 0.082 4 0.091 0.019 0.111 −0 0.088 0.017 0.861 0.23 0.048 0.025 0.043 0.004 0.121 0.095 −0.05 0.029 0.065 0.143 0.044 0.038 0.001 0.046 0.014

4 0.022 −0 0.368 5 −0.04 0.003 0.161 −0.02 0.023 −0.1 0.031 0.053 −0.03 −0.02 0.088 0.149 0.128 0.37 0.083 0.038 −0.07 0.012 0.481 0.071 0.04 0.109 −0.14

5 0.055 −0.046 −0.016 6 0.038 0.215 −0.214 0.149 −0.135 0.261 0.022 −0.001 0.198 0.046 −0.015 −0.018 0.164 −0.182 −0.114 0.127 0 −0.014 0.028 0.062 −0.054 0.138 0.234

6 0.025 0.109 −0.023 7 0.125 0.006 0.132 0.703 0.01 −0.066 0.023 0.044 −0.002 0.107 0.034 −0.142 0.064 0.068 −0.224 0.044 −0.012 0.072 0.059 −0.049 0.313 −0.005 −0.066

7 0.001 0.028 0.013 8 0.042 −0.007 0.079 −0.016 0.048 −0.039 −0.004 0.012 −0.008 0.01 −0.009 −0.025 0.013 0.061 −0.056 −0.001 0.015 0.027 0.004 −0.007 −0.013 0.039 −0.027

8 0.002 −0.01 0.803 9 0.082 0.043 0.022 −0.01 0.004 0.054 0.005 0.025 −0 0.051 0.863 0.05 0.044 0.174 −0.12 −0.03 0.032 0.053 0.042 0.059 0.037 0.021 0.032

9

10 −0.008 0.561 0.036 10 0.109 0.017 0.016 0.013 0.02 0.018 0.023 0.053 0.012 0.02 0.011 0.119 0.015 0.016 0.352 −0.029 0.063 −0.1 −0.126 0.005 −0.07 −0.087 −0.017

11 0.034 −0.131 0.003 11 0.05 0.144 0.057 0.067 −0.036 0.125 −0.006 −0.031 −0.016 −0.022 0.014 −0.004 0.075 −0.004 −0.037 0.011 0.068 0.06 0.158 0.112 0.173 0.33 0.122

142 Annexures

1 0.959 0.034 0.183 0.484 0.353 3 0.004 0.311 0.128 0.803 0.118 −0.01 0.009 −0

−0.031 −0.032

0.152 0.189

3

0.041 0.467 0.742 0.364 0.766 2 0.207 0.071 0.591 0.037 0.014 0.851

2

0.131 0.142 0.358 0.76 0.332 1 0.778 0.396 0.737 0.399 0.948 0.176

0.111 0.046 0.034 0.102 0.019 4 0.072 0.851 0.097 0.416 0.075 0.052 −0.01 −0

4

0 0

5 −0.01 0.72 0.108 0.05 0.145 5 −0.01 0.036 0.087 0.056 0.044 0.164 −0.007 −0.014

0.093 0.061 0.067 0.057 −0.035 6 0.03 −0.015 −0.101 −0.049 0.183 0.016

6

−0.009 −0.016

0.031 −0.186 0.119 0.012 0.215 7 0.05 0.033 0.022 0.033 0.038 −0.193

7

a

Extraction Method: Principal Component Analysis; Rotation Method: Varimax with Kaiser Normalization Rotation converged in 10 iterations

Receiver Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom Receiver United States Uruguay USSR (former) Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe 0.982 0.977

8 −0.012 −0.026 0.023 0.04 0.005 8 0.011 0.005 0.015 0.004 0.007 −0.021 0 −0

0.001 0.221 0.102 0.016 0.158 9 −0.01 0.051 −0.01 0.01 −0.03 −0.07

9

−0.003 −0.001

10 0.08 0.173 0.446 −0.007 −0.134 10 −0.035 0.006 −0.082 0.036 0.061 0.154 0.015 0.002

11 0.126 0.068 −0.029 0.138 0.012 11 −0.008 0.028 −0.044 0.005 −0.095 −0.007

Annexures 143

Shipper Algeria Angola Argentina Austria Bahrain Bangladesh Belgium Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Shipper Czechoslovakia (former) Denmark Dominican Republic Ecuador −0.156 −0.088

−0.502

−0.223

−0.413

0.271 −0.511

2 0.648 −0.274

−0.519

−0.291

0.051

6.489 −0.267

1 −0.331 −0.283

3 −0.160 −0.242

0.344 0.207

−0.333

4 −0.314 −0.037

0.682 −0.130

−0.229

−0.230

6.198

−0.374

−0.149

0.507

0.004

2.297 −0.214

4

−0.188 −0.151

−0.479 0.400

−0.332 −0.295

3

2

1

5 −0.025 −0.290

0.315 −0.264

−0.287

−0.131

−0.141

0.264

−0.023 0.049

5

Annexure 3.6. Results of Q Mode, Factor Scores, (2015)

7 −0.583 1.927

0.091 −0.239

−1.647 −0.271

6 −0.699 −0.343

0.311

−0.518

0.167

0.564

−0.414 −0.932

7

0.029

−0.334

0.432

−0.914

−0.136 −0.503

6

8 −0.247 −0.215

0.324 −0.188

−0.205

−0.160

−0.102

−0.035

−0.167 −0.205

8

9 −0.603 −0.328

−0.060 −0.031

−0.145

−0.259

−0.101

0.626

−0.013 −0.792

9

10 −0.130 −0.160

0.183 −0.245

−0.127

0.206

0.141

−0.420

−0.137 −0.047

10

11 0.018 0.269

−0.681 −0.394

−0.367

−0.257

0.317

0.166

−0.352 −0.032

11

144 Annexures

Shipper Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta −0.029 −0.138 −0.146 −0.152 −0.117 −0.110 0.153 −0.128 −0.082 −0.253 −0.144

6.041

−0.239

−0.485 0.411 −0.471

−0.456

−0.246 1.519 −0.303 −0.464

−0.445

−0.518 −0.480

−0.168

−0.288

−0.267 −0.360 −0.244

0.270

−0.234 −0.396 0.979 −0.233

0.607

0.507 −0.243

−0.132

−0.132

−0.166 −0.368

−0.295 0.391

−0.211 0.071

3 −0.122

2 −0.440

1 −0.180

−0.517 −0.238

−0.147

−0.262 0.201 −0.505 −0.240

−0.380

−0.303 −0.231 −0.241

−0.218

0.390

−0.209 −0.336

4 −0.269

−0.542 −0.232

1.423

−0.058 2.202 −1.448 −0.227

−0.463

−0.244 −0.035 −0.244

−0.179

−1.540

−0.251 5.351

5 −0.044

0.842 −0.288

4.713

−0.341 −0.041 2.126 −0.261

−0.034

−0.325 −0.662 −0.289

−0.395

1.141

−0.327 0.852

6 −0.231

−0.323 −0.172

1.257

0.176 −2.301 −0.668 −0.231

−0.271

−0.183 −1.304 −0.136

−0.555

0.131

0.258 −0.110

7 −0.281

−0.208 −0.150

−0.117

−0.146 −0.171 −0.370 −0.153

−0.235

−0.149 −0.228 −0.149

−0.162

0.272

−0.135 −0.145

8 −0.092

−0.135 −0.137

−0.957

−0.161 −0.419 0.278 −0.142

−0.113

−0.135 −0.834 −0.127

−0.202

−0.019

−0.179 1.345

9 −0.039

−0.504 −0.287

0.532

−0.459 1.799 −0.368 −0.278

−0.381

−0.284 0.331 −0.298

0.417

−1.750

−0.397 −1.069

10 −0.117

0.289 −0.330 (continued)

−1.921

0.034 0.929 1.098 −0.285

−0.211

−0.379 −0.195 −0.335

−0.277

−0.869

−0.207 0.322

11 −0.306

Annexures 145

Shipper Mauritius Mexico Mongolia Shipper Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand

2

2 −0.446 0.797 −0.556

−0.284

−0.475 −0.457 −0.462 0.525 −0.397 0.019

−0.495 −0.733 −0.448 0.014 −0.469 −0.192 0.024

1

1 −0.183 −0.100 −0.110

−0.261

−0.252 −0.328 −0.170 −0.259 −0.239 −0.310

−0.211 1.131 −0.310 −0.371 −0.226 −0.274 −0.009

−0.156 −0.571 −0.183 0.014 −0.140 −0.286 −0.034

−0.136 −0.048 −0.132 −0.194 −0.209 −0.124

−0.210

3 −0.166 −0.185 −0.203

3

−0.254 −0.716 −0.214 0.080 −0.242 −0.248 −0.201

0.012 0.435 −0.275 −0.356 −0.074 −0.277

−0.087

4 −0.261 −0.403 −0.323

4

−0.123 −1.863 −0.110 −0.595 −0.239 −0.570 −0.194

−0.215 −0.285 −0.323 −0.131 0.555 −0.204

−0.171

5 0.135 1.045 0.285

5

−0.356 2.736 0.205 −0.253 −0.288 0.156 −0.212

−0.285 −0.163 −0.153 −0.661 0.226 −0.550

−0.450

6 −0.347 −1.429 0.750

6

−0.146 −1.258 −0.202 −0.266 −0.170 3.974 −0.169

−0.193 −0.220 −0.189 0.040 0.092 −0.709

1.299

7 −0.162 2.969 0.060

7

−0.166 −0.669 6.768 −0.073 −0.152 −0.093 −0.222

−0.147 −0.153 −0.173 −0.269 0.058 −0.199

−0.174

8 −0.168 −0.368 −0.218

8

−0.148 0.813 −0.006 6.365 −0.148 0.083 −0.050

−0.114 −0.111 −0.134 −0.518 −0.219 −0.132

−0.225

9 −0.171 0.094 −0.246

9

−0.352 0.279 −0.066 0.291 −0.297 0.209 −0.250

−0.297 −0.255 −0.301 −0.206 −0.542 0.483

−0.496

10 −0.426 −0.284 −0.549

10

−0.348 2.557 0.042 −0.485 −0.336 −0.734 0.060

−0.362 −0.307 −0.289 −0.263 −0.508 −0.332

0.362

11 −0.361 1.943 −0.791

11

146 Annexures

Shipper Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Shipper Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

2

0.177

0.299 −0.037

0.972 2

0.157

−0.476

1

−0.135

0.004 −0.424

−0.222 1

−0.329

−0.249

−0.145

−0.117 −0.234

−0.276

−0.354 4

0.132 0.090

−0.292 6.794 −0.149 3

−0.097

4

−0.146

3

−0.251

−0.186

−0.377 5

0.156 0.105

0.623

5

−0.276

−0.650

0.057 6

0.110 0.226

−0.485

6

−0.181

−1.046

1.036 7

0.624 0.122

−0.969

7

−0.023

−0.213

0.040 8

0.352 −0.003

−0.199

8

−0.133

−0.649

0.089 9

−0.549 −0.069

−0.143

9

−0.283

0.457

5.958 10

0.155 −0.023

0.644

10

−0.324

−0.304

−0.528 11

5.083 0.338

−0.162

11 Annexures 147

148

Annexures

 nnexure 3.7 Results of Factor Analysis (R Mode), A Communalities (1990)

Country

Initial Extraction Country

Initial Extraction Country

Initial Extraction

Algeria Angola Argentina Austria Bahrain Bangladesh BelgiumLuxembourg Bolivia

1 1 1 1 1 1 1

0.986 0.978 0.92 0.969 0.795 0.982 0.947

Ghana Greece Guatemala Guinea Honduras Hungary Iceland

1 1 1 1 1 1 1

0.963 0.963 0.894 0.693 0.992 0.959 0.952

Panama Paraguay Peru Philippines Poland Portugal Romania

1 1 1 1 1 1 1

0.963 0.97 0.969 0.995 0.975 0.883 0.982

1

0.65

India

1

0.967

1

0.966

Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile Colombia

1 1

0.977 0.976

Indonesia Iran

1 1

0.988 0.906

Saudi Arabia Senegal Singapore

1 1

0.956 0.742

1 1 1 1 1

0.659 0.963 0.994 0.978 0.983

Ireland Italy Japan Jordan Kenya

1 1 1 1 1

0.979 0.972 0.95 0.911 0.899

1 1 1 1 1

0.936 0.954 0.818 0.967 0.963

Costa Rica

1

0.984

1

0.985

1

0.973

Côte d’Ivoire

1

0.902

South Korea Kuwait

Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand

1

0.924

0.993

Czechoslovakia (former) Denmark Dominican Republic Ecuador

1

0.945

Malaysia

1

0.642

Trinidad 1 and Tobago Tunisia 1

1 1

0.842 0.984

Malta Mauritius

1 1

0.904 0.971

1 1

0.959 0.854

1

0.978

Mexico

1

0.992

1

0.975

Egypt

1

0.973

Mongolia

1

0.606

1

0.945

El Salvador

1

0.811

Morocco

1

0.973

1

0.394

Fiji Finland

1 1

0.857 0.813

1 1

0.884 0.969

1 1

0.948 0.992

France Gabon

1 1

0.913 0.958

Netherlands New Zealand Nigeria Norway

Turkey USSR (former) United Arab Emirates United Kingdom United States Uruguay Venezuela,

1 1

0.969 0.921

1 1

0.973 0.976

Germany

1

0.846

Pakistan

1

0.94

Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1 1

0.907 0.938

0.955

Annexures

149

Annexure 3.8 Results of R-Mode, Factor Loadings (1990)

Shipper Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland

1

2

3

4

0.513 0.941 0.574 0.079 0.389 0.902 0.095 0.515 0.875 0.05 0.066 0.336 0.99 0.596 0.964 0.96 0.248 −0.015 0.145 0.973 0.985 0.313 0.824 0.25 0.159 0.22 0.615 0.238 0.345 0.153 0.94 0.673 0.98 0.083 0.244 0.554 0.282 0.013 0.157

0.235 0.117 0.212 0.967 −0.028 0.224 0.731 0.107 0.229 −0.036 0.709 0.211 −0.008 0.424 0.177 0.23 0.447 0.479 0.805 −0.017 0.024 0.296 0.326 0.065 0.534 0.809 0.076 0.223 0.842 0.85 0.083 0.289 0.139 0.629 0.503 0.247 0.036 0.506 0.449

0.041 0.019 0.144 0.031 0.472 0.118 0.069 −0.028 0.271 0.981 0.038 0.049 0.065 0.577 0.068 0.011 0.106 −0.03 0.108 0.04 0.018 0.122 0.001 0.234 0.025 0.077 0.098 0.129 0.129 0.027 0.042 0.008 0.074 −0.013 0.185 0.333 0.951 0.74 0.069

0.639 0.261 0.138 0.01 0.029 0.116 0.549 0.028 0.147 −0.048 0.085 0.874 0.037 0.13 0.066 0.031 0.457 −0.046 0.112 0.064 0.042 0.195 −0.008 −0.078 0.126 0.02 0.732 0.64 −0.028 0.281 0.017 0.376 0.034 −0.033 0.181 0.016 −0.014 0.052 0.199

5 0.153 0.03 0.185 0.147 0.004 0.152 0.09 −0.021 0.071 0.021 0.09 0.072 0.038 0.053 0.034 0.039 0.149 0.827 0.101 0.032 0.024 0.767 0.013 −0.006 0.514 0.103 0.05 0.2 0.153 0.196 0.032 0.008 0.037 0.737 0.101 0.657 0.043 0.099 0.041 (continued)

Annexures

150 Shipper Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey USSR (former) United Arab Emirates United Kingdom United States Uruguay Venezuela Vietnam

1

2

3

4

0.241 0.945 −0.024 0.099 0.824 0.252 0.509 0.063 0.214 0.989 0.017 −0.002 0.105 0.568 0.962 0.138 0.595 0.931 0.075 0.763 0.869 0.059 0.119 0.149 0.728 −0.06 0.769 0.146 0.896 0.351 0.261 −0.039 0.745 0.989 −0.015 0.251 0.063 0.086 0.525 −0.027 0.234 0.988 −0.041

0.744 0.153 −0.013 0.604 0.091 0.055 0.064 0.654 0.33 −0.006 0.175 0.286 0.879 0.106 0.1 0.472 0.478 0.305 0.094 0.298 0.09 0.787 0.688 0.337 0.004 0.134 0.113 0.6 0.23 0.714 0.862 0.085 0.153 −0.006 0.56 0.896 0.802 0.003 0.619 0.194 0.18 0.04 −0.075

0.086 0.039 0.094 0.077 0.535 0.828 0.586 0.015 0.027 0.067 0.556 0.125 0.02 0.73 −0.002 0.068 0.435 −0.01 0.026 0.456 0.464 −0.004 0.042 0.005 0.621 0.076 0.337 0.051 0.187 0.09 0.171 −0.032 0.603 0.006 0.015 0.055 0.278 0.98 0.116 0.525 0.036 0.047 0.322

0.473 0.055 0.069 0.04 0.031 0.08 −0.005 0.212 0.431 0.065 0.054 0.923 0.219 −0.004 0.11 0.153 0.139 0.002 0.064 0.06 0.036 −0.05 0.476 0.052 0.164 0.943 0.043 0.678 0.063 0.169 0.267 0.326 0.06 0.051 0.731 0.114 0.236 0.011 0.464 0.062 0.055 0.046 0.024

5 0.098 0.057 0.04 0.068 0.04 0.029 0.072 0.25 0.036 0.024 0.061 0.054 0.073 0.11 0.027 0.038 0.141 0.043 −0.033 0.125 0.038 0.545 0.046 0.916 0.045 0.042 0.041 0.126 0.111 0.069 0.154 0.879 0.066 0.027 0.163 0.248 0.107 0.028 0.089 0.019 0.153 0.026 0.912 (continued)

Annexures Shipper Yugoslavia, SFR (former) Zambia Zimbabwe

151 1

2 0.108 0.011 0.363

0.53 0.059 0.692

3 −0.03 0.86 0.282

4

5 0.195 0.354 0.022

0.752 0.036 0.056

Extraction Method: Principal Component Analysis; Rotation Method: Varimax with Kaiser Normalization a Rotation converged in 8 iterations

 nnexure 3.9 Results of R-Mode Analysis, Factor A Scores (1990)

Receiver Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana

1 −0.152 −0.153 −0.065 −0.193 −0.111 −0.098 0.269 −0.124 −0.397 −0.120 −0.126 −0.076 −0.020 0.000 −0.057 0.124 −0.121 −0.180 −0.190 −0.087 −0.083 −0.093 0.064 −0.127 −0.168 −0.432 −0.138 0.273 −0.133

2 −0.132 −0.223 −0.506 0.284 −0.299 −0.244 0.816 −0.277 −0.567 −0.272 −0.263 −0.299 −0.108 −0.397 −0.261 −0.150 −0.285 0.395 −0.061 −0.276 −0.271 −0.081 −0.269 −0.298 0.317 0.653 −0.290 8.424 −0.268

3 −0.211 −0.218 −0.409 −0.253 −0.150 −0.175 −0.061 −0.200 0.094 −0.209 −0.224 −0.273 0.443 −0.246 −0.193 −0.326 −0.242 −0.069 −0.127 −0.230 −0.217 −0.256 −0.304 −0.173 −0.148 0.169 −0.224 −0.068 −0.220

4 −0.017 −0.127 −0.189 −0.078 −0.130 −0.171 0.559 −0.158 −0.189 −0.149 −0.171 0.213 −0.231 0.047 −0.182 −0.259 0.013 −0.380 −0.125 −0.158 −0.164 −0.240 −0.198 −0.162 −0.220 8.554 −0.069 −1.338 −0.100

5 −0.144 −0.237 −0.270 0.530 −0.186 −0.137 −0.421 −0.204 −0.324 −0.201 0.092 −0.246 −0.226 −0.188 −0.194 −0.267 −0.207 0.280 −0.052 −0.203 −0.201 −0.011 −0.211 −0.186 −0.341 0.311 −0.200 0.502 −0.210 (continued)

Annexures

152 Receiver Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago

1 −0.167 0.103 −0.147 −0.039 −0.131 −0.139 −0.255 −0.099 −0.089 −0.005 −0.235 −0.038 −0.121 −0.126 −0.021 −0.120 0.048 −0.135 −0.135 −0.053 −0.135 −0.133 0.096 −0.133 −0.132 −0.103 −0.161 0.053 −0.116 −0.036 −0.144 −0.167 −0.134 −0.148 −0.105 −0.134 0.027 0.009 −0.110 −0.229 −0.279 −0.110 −0.050 −0.122

2 0.056 −0.106 −0.278 −0.243 0.057 −0.252 0.007 −0.210 −0.034 −0.010 1.777 −0.138 −0.233 −0.254 −0.306 −0.252 −0.263 −0.246 −0.272 −0.232 −0.280 −0.219 0.484 −0.498 −0.222 0.232 −0.199 −0.255 −0.293 −0.339 −0.316 0.073 −0.111 0.353 −0.187 −0.280 −0.532 0.564 −0.228 0.466 0.531 −0.122 −0.234 −0.283

3 −0.144 −0.317 −0.218 −0.262 −0.179 −0.225 0.077 −0.098 −0.099 −0.264 0.010 8.761 −0.206 −0.160 1.136 −0.235 0.152 −0.227 −0.202 0.018 −0.227 −0.154 0.535 −0.190 −0.204 −0.263 −0.028 −0.266 −0.217 −0.273 0.195 −0.169 −0.198 −0.126 −0.024 −0.214 1.404 −0.059 −0.169 −0.256 0.021 −0.266 0.433 −0.225

4 0.013 −0.275 −0.096 −0.184 −0.256 −0.151 0.559 −0.071 −0.111 0.122 1.507 −0.348 −0.179 −0.159 −0.231 −0.174 −0.166 −0.140 −0.162 −0.155 −0.150 −0.041 1.262 −0.292 −0.062 −0.095 −0.345 −0.190 −0.147 −0.173 −0.094 −0.218 0.185 −0.307 −0.082 −0.069 −0.379 1.069 −0.198 −0.246 −0.155 −0.198 −0.114 −0.148

5 −0.033 −0.289 −0.205 −0.218 0.111 −0.202 −0.295 −0.175 −0.001 −0.325 1.634 0.187 −0.128 −0.210 −0.188 −0.127 −0.191 −0.200 −0.194 −0.210 −0.180 −0.177 −0.123 −0.159 −0.168 −0.300 −0.236 −0.236 −0.194 −0.209 −0.117 0.174 −0.313 −0.202 0.130 −0.208 0.478 −0.518 −0.175 0.123 −0.015 −0.173 −0.088 −0.196 (continued)

Annexures

153

Receiver Tunisia Turkey USSR (former) United Arab Emirates United Kingdom United States Uruguay Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1 −0.204 −0.156 −0.307 −0.058 −0.507 9.049 −0.122 −0.055 −0.130 −0.091 −0.145 −0.161

2 −0.215 −0.040 −1.030 −0.217 1.095 −0.291 −0.291 −0.265 −0.246 0.511 −0.204 −0.273

3 0.144 −0.122 −0.437 −0.109 0.171 −0.087 −0.214 −0.193 −0.216 −0.249 −0.190 −0.156

4 −0.061 −0.019 −0.529 −0.228 −0.702 0.292 −0.160 −0.153 −0.158 −0.260 −0.245 −0.112

5 −0.202 0.123 8.774 −0.043 −0.167 0.233 −0.198 −0.216 −0.191 0.031 −0.244 −0.205

 nnexure 3.10 Results of R-Mode Analysis, A Communalities (2015)

Country Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil BruneiDarussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador

Initial 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Extraction 0.946 0.974 0.842 0.947 0.916 0.954 0.908 0.808 0.911 0.881 0.806 0.881 0.987 0.964 0.973 0.926 0.965 0.737 0.935 0.818 0.971 0.947

Country Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa

Initial 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Extraction 0.833 0.903 0.81 0.988 0.98 0.931 0.922 0.985 0.843 0.877 0.908 0.724 0.951 0.89 0.95 0.946 0.949 0.94 0.937 0.633 0.654 0.901 (continued)

Annexures

154 Country Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan

Initial 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Extraction 0.865 0.884 0.917 0.846 0.856 0.832 0.819 0.941 0.822 0.911 0.884 0.959 0.938 0.853 0.85 0.933 0.89 0.847 0.882 0.93

Jordan Kenya South Korea Kuwait

1 1 1 1

0.934 0.858 0.955 0.853

Country Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay U.S.S.R. Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe Extraction Method: Principal Component Analysis.

Extraction Method: Principal Component Analysis

Initial 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Extraction 0.864 0.956 0.66 0.957 0.869 0.892 0.891 0.871 0.84 0.598 0.811 0.23 0.975 0.835 0.967 0.981 0.725 0.92 0.924

Shipper Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark

Rotated component matrix 1 2 3 0.254 0.346 0.094 0.075 0.037 0.953 0.196 0.052 0.396 0.116 0.954 0.048 0.355 0.001 0.089 0.563 0.718 0.129 0.132 0.796 0.057 0.223 −0.03 0.085 0.426 0.126 0.823 −0.02 −0.05 0.108 −0.03 0.756 0.07 0.027 0.12 0.646 0.982 0.062 0.117 0.322 0.052 0.878 0.857 0.197 0.083 0.907 0.081 0.231 0.941 0.104 0.237 0.452 0.452 0.088 −0.01 0.961 0 0.253 0.809 0.156 4 0.04 −0.05 0.03 0 0.24 0.07 0.03 0.03 0.03 0.93 −0.04 0.06 0.02 0.21 0.39 0 0.01 0 −0.01 0.03

5 0.112 0.204 0.12 0.042 0.046 0.089 0.157 −0.01 0.035 0.012 0.004 0.606 −0 0.031 0.029 0.037 0.001 0.436 0.012 0.003

6 0.818 0.117 0.078 −0.041 0.019 0.216 0.252 0.035 0.023 0.031 0.09 0.246 0.019 0.066 0.033 0.142 0.048 0.151 −0.006 0.012

7 0.046 0.002 0.039 −0.01 0.845 0.089 −0.01 −0.03 0.04 −0.03 0.041 0.018 0.048 −0.01 0.1 0.007 0.004 −0.05 −0.03 0.001

Annexure 3.11 Results of R-Mode Analysis, Factor Loadings (2015)

8 0.15 0.02 0.79 0.02 −0 0.03 0.02 0.86 0.15 −0 −0 −0 0.02 0.19 0.08 0.13 0.02 0.02 0 0.02

9 0.115 0.007 0.032 −0.11 0.062 0.137 0.249 −0.06 0.099 −0 −0.06 0.082 0.007 0.024 0.128 0.064 0.084 0.227 −0.04 0.221

10 0.166 0.012 −0.02 −0.06 0.053 0.099 0.314 0.019 0.004 −0.01 0.083 0.088 0.025 −0.01 0.052 -0.03 0.044 0.233 −0.09 −0.09

11 0.077 0.005 −0.04 −0.01 −0.04 −0.09 −0.06 −0.09 0.012 −0.02 0.454 0.046 0.021 0.016 0.037 0.02 0 0.009 −0.05 −0.13 (continued)

Annexures 155

Shipper Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea

Rotated component matrix 1 2 3 0.969 0.07 0.113 0.944 0.119 0.151 0.216 0.308 0.115 0.936 0.023 0.06 0.912 0.09 0.164 0.21 0.738 0.184 0.247 0.822 0.185 0.125 0.089 0.773 0.408 0.331 0.325 0.037 0.111 0.358 0.177 0.618 0.035 0.947 0.048 0.099 0.102 0.233 0.049 0.954 0.205 0.062 0.005 0.959 0.006 0.136 0.373 0.063 0.697 0.175 0.239 0.399 0.065 0.532 −0.09 0.003 0.846 0.723 0.395 0.097 0.34 0.778 0.119 0.653 0.112 0.689 0.683 0.006 0.163 0.523 0.281 0.128 0.341 0.055 0.906 4 0.02 0.02 −0.01 −0.02 0.18 0.04 0 0.16 −0.02 0.15 −0.03 0.02 0.15 −0 −0.01 0.01 0.09 0.67 0.18 0.03 0.01 0.1 0.07 0.08 0.09

5 0.098 0 0.22 −0.02 −0.06 0.059 0.083 0.098 0.128 0.867 −0.02 −0.03 0.856 −0 0.003 0.103 −0.06 0.161 0.29 0.089 0.135 −0.01 0.255 0.122 0.001

6 0.02 0.053 0.143 0.023 0.037 −0.06 0.208 0.359 0.266 −0.049 0.183 0.03 0.175 0.03 0.019 0.385 0.007 0.004 −0.057 0.116 0.252 −0.009 −0.019 0.01 −0.035

7 0.045 0.021 0.725 0.002 0.049 −0.01 0.075 −0.02 0.01 0.027 0.213 0.017 0.05 0.01 −0.03 −0.04 0.437 0.081 0.129 0.053 0.04 0.066 0.601 0.218 0.049

8 0.03 0.09 0.02 −0 0.01 0.09 0.05 −0 0.02 0.01 0 0.02 0.01 0.01 0.01 0.01 0.03 0.03 0.02 −0 0.03 0.04 0 0.02 0.06

9 0.016 0.042 0.04 −0.04 0.071 0.422 0.151 0.1 0.361 0.014 −0.02 0.015 −0.01 0.008 −0.05 0.728 0.13 0.011 −0.04 0.276 0.088 0.02 −0.06 0.621 −0

10 0.015 −0.01 0.179 −0.02 0.083 −0.11 −0.07 0.045 0.435 0.139 0.158 0.001 −0.17 −0.01 −0.11 0.018 0.228 0.018 −0.07 0.211 0.221 0.01 −0.07 0.157 0.006

11 −0.01 0.082 0.286 0.014 0.016 0.124 −0 0.186 0.174 0.056 0.55 0.001 0.006 −0.02 0.046 −0.03 −0.05 0 0.157 −0.12 −0.08 −0.01 0.028 0.095 0.005

156 Annexures

Shipper Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka

Rotated component matrix 1 2 3 0.216 −0.04 0.534 0.349 0.091 0.699 0.197 0.783 0.06 0.366 0.234 0.068 0.987 0.056 0.082 −0.08 0.008 0.98 0.06 0.241 0.038 0.053 0.931 0.038 0.44 0.093 0.841 0.057 0.121 0.091 0.087 0.732 0.071 0.586 0.326 0.493 0.549 0.23 0.515 0.004 0.091 −0.043 0.45 0.122 0.796 0.431 0.148 0.607 −0.01 0.963 −0.007 0.098 0.496 0.049 −0.02 0.916 0.002 0.322 −0 0.634 0.009 0.077 0.342 0.245 0.032 0.704 0.34 0.396 0.715 0.102 0.698 0.051 0.858 0.269 0.17 4 0.65 0.45 0.27 −0.01 0.01 −0.1 −0 −0.01 0.24 0.28 0.01 0.01 0.27 0.01 −0.01 0.59 −0.01 −0.06 −0.01 0.59 −0.03 0.3 0.22 −0.01 0.11

5 0.293 0.003 0.001 0.078 −0 0.013 0.26 0.066 −0.04 0.722 0.036 0.046 0.084 0.052 0.073 −0.13 0.024 0.116 0.022 0.249 0.578 0.046 0.196 0.089 0.212

6 0.004 −0.029 0.198 0.244 0.027 −0.046 0.869 0.089 0.007 0.31 0.058 0.069 0.036 0.039 0.056 0.019 −0.004 0.805 0.126 0.078 0.13 −0.033 −0.007 0.237 −0.009

7 0.082 0.035 0.075 0.323 0.042 −0.03 0.024 0.014 0.082 0.026 0.012 0.397 −0.06 −0.01 −0.03 −0.02 −0.01 0.023 0.028 0.094 0.088 0.008 0.076 0.008 0.144

8 0.02 0.02 −0 −0 0.03 0.02 0.07 0 0.02 0.29 0.01 0.02 0.02 0.95 0.16 0.03 0.03 0.01 0.02 0.02 −0 −0 0.04 0.04 0.04

9 0.073 0.018 −0.03 0.27 −0.01 −0.02 0.048 0.022 0.078 0.181 0.485 0.172 0.084 0.065 0.026 0.037 0.097 0.127 −0.06 −0.01 0.044 0.019 0.129 0.038 0.151

10 −0.01 0.041 0.346 0.602 0.012 0.001 0.189 0.088 0.058 −0.01 0.167 0.138 −0.11 −0.01 0.025 0.014 −0.08 −0.02 0.094 0.042 0.383 0.027 0.02 0.544 0.122

11 0.024 −0.06 −0.09 −0.1 0.023 0.008 0.04 −0.17 −0.03 −0.08 −0.25 −0.04 −0 0.153 −0.05 −0.05 0.031 −0.09 0.266 0.03 −0 −0.07 −0.08 0.029 0.044 (continued)

Annexures 157

Rotated component matrix 1 2 3 0.278 0.632 0.171 0.388 0.737 0.335 −0.06 −0.03 −0.021 0.493 0.102 0.638 0.853 0.02 0.037 0.021 0.529 0.021 0.16 0.75 0.061 −0.02 −0.04 0.236 0.586 0.551 0.305 −0.02 0.139 0.318 0.198 0.118 0.65 0.001 0.429 0.308 0.792 0.023 0.42 0.707 0.205 0.566 −0.01 0.719 −0.012 −0.07 −0.01 0.944 −0.02 0.069 0.945 4 −0.02 0.17 −0.05 0.46 −0.02 0 −0.05 0.65 0.01 0.11 −0.09 0.08 0.08 0.31 −0.04 0.01 −0.09

5 0.054 0.283 −0.06 −0.01 −0.01 0.114 0.094 0.32 0.181 −0.01 −0 0.077 0.379 −0.03 −0.03 0.122 0.048

6 0.062 0.094 −0.033 −0.015 0.097 0.438 0.072 −0.073 0.118 0.001 0.044 −0.081 −0.003 0.047 0.002 −0.035 0.017

7 −0 0.133 0.923 0.064 −0 −0.02 0.217 0.096 0.05 0.013 −0.02 −0.01 0.057 0.078 −0.02 0.073 0

Extraction Method: Principal Component Analysis; Rotation Method: Varimax with Kaiser Normalization a Rotation converged in 7 iterations

Shipper Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

8 0 0.05 −0 0.04 0.39 0 0.05 −0 0.02 0.16 0.7 0.13 0.05 0.06 −0 0.01 0.04

9 0.377 −0.05 −0.02 0.048 −0.02 −0.17 0.312 −0.01 0.087 0.066 0.022 0.446 −0.06 0.069 −0.18 −0.02 0.076

10 0.042 0.096 −0.04 0.071 0.015 0.573 0.171 −0.02 0.112 0.135 −0.01 −0.05 −0.04 0.066 0.006 0.032 0.063

11 −0.06 −0.02 −0.04 −0.06 −0.05 0.167 0.235 0.014 −0.03 −0.22 −0.02 0.566 0.013 0.001 0.414 0.032 0.055

158 Annexures

Annexures

159

 nnexure 3.12 Results of R-Mode Analysis, Factor A Scores (2015)

Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran

1 −0.17646 −0.15096 0.02350 −0.14792 −0.13083 −0.11529 0.13746 −0.12366 −0.34449 −0.12370 −0.13220 −0.12756 0.23260 0.10631 −0.73919 0.05420 0.11940 −0.17709 −0.23002 −0.16657 0.16965 0.06942 −0.17621 0.23281 −0.12461 −0.13951 −0.45425 −0.13649 −0.41563 −0.07784 −0.16431 0.26906 −0.13791 0.27906 −0.17047 −0.12712 −0.01034 −0.10288 −0.24829

2 −0.19055 −0.31513 −0.38306 0.35739 −0.36153 −0.37501 0.88680 −0.36699 −0.38868 −0.35913 −0.08496 −0.38401 −0.06622 −0.37292 −0.00985 −0.33115 −0.36114 −0.45235 1.13341 0.05228 −0.39335 −0.33386 −0.21906 −0.36761 −0.35931 −0.05894 1.62409 −0.36994 8.03457 −0.35595 −0.06612 −0.39212 −0.40913 −0.40759 0.25218 −0.32912 −0.46824 −0.33242 −0.28133

3 −0.19466 −0.15059 −0.27425 −0.19379 −0.18919 −0.12041 0.04354 −0.11849 −0.32837 −0.20248 −0.23387 −0.24521 0.12087 −0.19039 8.98886 −0.12077 −0.20666 −0.23666 −0.20797 −0.18870 −0.26283 −0.08732 −0.19416 −0.24661 −0.19235 −0.17551 −0.23514 −0.19347 −0.20405 −0.33725 −0.20401 −0.25766 −0.18291 −0.26083 −0.21532 −0.21173 0.08241 0.17762 −0.37160

4 −0.28530 −0.27106 −0.36279 −0.22932 −0.11532 −0.10964 −0.71364 −0.31734 −0.08723 −0.23373 −0.24359 −0.31157 −0.33926 −0.37060 −0.92893 −0.30700 −0.28388 −0.37786 −0.25993 −0.32782 −0.36202 −0.20467 0.08493 −0.33631 −0.24971 −0.33618 0.04335 −0.30300 0.26244 −0.25527 −0.23506 −0.41684 −0.39256 −0.38520 −0.22994 −0.27128 2.11249 0.50845 0.86748

5 −0.18335 −0.16678 −0.37075 −0.31040 −0.16371 −0.17419 0.98954 −0.22211 0.03057 −0.19122 −0.29890 0.24336 −0.04662 −0.27949 −0.10298 −0.26139 −0.20508 0.69387 −0.33798 −0.19535 −0.15949 −0.23887 −0.35465 −0.19754 −0.20688 −0.26479 1.01340 −0.05023 −0.07360 0.46775 −0.22543 −0.19085 0.19330 −0.20338 −0.29724 −0.17668 8.18733 −0.27733 0.18755 (continued)

Annexures

160

Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey UAE UK US Uruguay USSR (former) Venezuela

1 −0.02361 −0.27236 −0.14383 −0.17815 −0.16936 −0.14862 −0.29220 −0.01420 −0.13541 −0.12156 0.16267 −0.12470 −0.18064 −0.00222 −0.02868 −0.06134 −0.16293 −0.18130 0.24827 −0.14702 0.09673 −0.09318 −0.14868 −0.15821 −0.14723 −0.35608 −0.11897 −0.06589 −0.27584 −0.14438 −0.10109 −0.18502 0.01673 −0.14421 −0.12139 −0.13449 −0.13400 −0.18758 −0.19495 0.02462 9.04924 −0.14168

2 −0.09011 1.58578 −0.06268 −0.32709 −0.36393 −0.35345 −0.30727 −0.34852 −0.32713 −0.35950 −0.12481 −0.35815 −0.26240 0.36709 −0.35928 −0.32737 0.09707 −0.47727 −0.40214 −0.36048 −0.38873 −0.31568 0.63117 −0.21988 0.28212 −0.06014 −0.35089 0.03959 −0.41986 0.12305 −0.35108 0.61923 0.20855 −0.37149 −0.27872 −0.45454 −0.32238 0.32965 −0.30396 1.74966 0.42105 −0.36022

3 −0.29738 0.21409 0.47677 −0.19185 −0.19219 0.57981 −0.20159 0.28280 −0.20963 −0.20856 0.12800 −0.21578 −0.21358 0.20509 −0.26307 −0.22529 −0.06566 −0.24364 −0.16106 −0.12070 −0.20117 −0.02462 −0.16930 −0.06062 −0.22771 −0.19067 −0.23964 0.16232 0.19746 0.12025 −0.18070 −0.19892 0.12184 −0.18888 0.08854 0.01533 −0.23460 −0.00768 0.19900 −0.11087 0.61262 −0.20431

4 −0.37072 −0.19052 7.16085 −0.18827 −0.13405 3.22720 −0.31753 0.95062 −0.27499 −0.25070 −0.20301 −0.25492 −0.21519 0.12036 0.25694 −0.30659 −0.43129 0.64947 −0.40172 −0.32603 −0.31847 0.31675 −0.29360 −0.41601 −0.23972 −0.19187 −0.25478 2.31414 −0.38972 −0.58928 −0.17880 −0.30733 −1.02525 −0.05945 1.64611 −0.25101 −0.22270 −0.17423 −0.37791 −0.09476 0.04756 −0.26669

−0.24899

0.99168

−0.30881

−0.05700

5 0.59659 −0.12376 −1.57996 −0.16539 −0.25592 −0.49656 −0.32066 −0.15824 −0.18587 −0.18761 −0.53340 −0.18436 −0.17466 0.72494 −0.34792 0.37955 −0.24151 0.01592 −0.26084 −0.22381 −0.27479 −0.41584 −0.14877 0.15363 −0.39798 −0.41540 −0.01228 −0.77536 0.61706 0.76476 −0.25953 −0.20489 2.29325 0.01628 −0.60925 −0.34001 −0.27994 −0.02309 −0.45097 −0.25407 −0.06810 −0.13829 0.34065 (continued)

Annexures

161 1

Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

2 −0.38838 0.64108

3

4

0.02252 −0.11876

0.12387 −0.24951

0.32886 −0.21701

5 −0.21244 −0.45500

−0.16520 −0.16853

−0.34884 −0.33435

−0.03938 −0.05850

−0.34371 −0.30928

−0.15901 −0.14653

 nnexure 4.1 Values of Location Quotient for Export of Major A Commodity Groups (1995)

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany

All food items 0.12 0.11 5.27 0.41 0.29 1.12 1.24 2.29 3.09 0.01 1.97 2.86 0.85 2.60 0.88 3.48 7.05 6.29 0.65 2.81 1.59 5.61 1.05 6.05 5.91 0.26 1.53 0.02 0.57

Raw materials from agricultural sources 0.02 0.02 1.58 1.22 0.02 0.98 0.52 3.55 1.93 0.00 1.13 10.00 3.55 5.13 0.66 2.08 1.89 5.89 1.36 1.16 0.12 1.12 2.23 0.42 2.78 3.07 0.54 4.76 0.42

Ores and metals 0.16 0.05 0.50 1.02 7.92 NA 1.25 10.85 3.22 0.01 3.03 2.56 2.16 14.95 0.64 0.20 0.34 0.05 0.93 0.42 0.40 0.10 1.96 0.87 0.07 0.95 0.79 0.62 0.84

Fuels 12.43 12.85 1.35 0.13 6.83 0.06 0.39 1.91 0.12 11.96 0.87 3.81 1.26 0.03 0.47 3.79 0.11 1.29 0.56 0.37 0.00 4.68 4.86 0.01 0.06 0.25 0.31 10.80 0.13

Manufactured goods 0.04 0.00 0.44 1.15 0.25 1.12 1.04 0.24 0.70 0.11 0.80 0.10 0.85 0.16 1.09 0.41 0.34 0.19 1.08 0.86 1.08 0.10 0.52 0.50 0.47 1.09 1.03 0.03 1.17 (continued)

Annexures

162

Countries Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland

All food items 6.19 3.19 6.90 0.83 9.22 2.27 8.02 2.37 1.21 0.38 2.21 0.70 0.05 2.39 5.95 0.24 0.03 1.02 0.23 3.12 0.82 0.23 3.33 2.19 4.61 0.29 0.95 1.25 7.93 4.65 3.34 2.24 1.11 0.76 0.70 0.10 1.64 0.43 1.32 1.65 2.13 0.25 0.32

Raw materials from agricultural sources 5.37 1.63 1.51 0.43 1.22 0.85 0.20 0.56 2.42 0.35 0.44 0.24 0.21 0.65 2.69 0.48 0.02 2.28 0.02 0.25 0.49 10.06 1.24 1.48 6.74 0.87 0.60 1.40 0.17 13.22 0.98 0.75 1.01 1.69 1.21 0.05 3.12 0.40 2.48 0.60 1.67 2.57 0.27

Ores and metals 2.66 2.46 0.14 20.85 0.14 1.53 3.69 1.31 1.83 0.18 0.36 0.44 0.33 6.12 0.86 0.32 0.09 0.42 0.16 0.06 0.88 18.31 3.53 0.87 1.51 0.07 2.88 0.05 0.35 0.08 13.94 2.15 2.22 0.56 1.06 0.17 3.26 0.62 3.83 0.73 0.23 0.97 0.79

Fuels 0.63 0.87 0.26 0.08 0.00 0.40 0.01 0.26 3.31 11.20 0.06 0.16 0.08 0.00 0.80 0.26 12.36 0.93 0.20 0.00 1.34 0.00 0.29 0.96 0.21 12.09 6.66 0.13 0.41 0.03 0.70 0.32 1.07 0.42 1.04 11.33 1.98 0.92 1.83 0.23 0.06 0.27 0.01

Manufactured goods 0.17 0.66 0.37 0.29 0.12 0.89 0.15 0.91 0.66 0.12 0.99 1.17 1.27 0.72 0.36 1.21 0.06 0.98 1.25 0.91 1.01 0.13 0.67 0.85 0.41 0.03 0.38 1.08 0.26 0.25 0.19 0.87 0.93 1.08 1.02 0.15 0.65 1.12 0.71 1.02 0.96 1.11 1.22 (continued)

Annexures

Countries Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe Total

163 All food items 1.30 2.11 0.89 1.04 2.07 0.37 0.83 1.12 4.71 0.69 0.30 3.35 1.00 0.29 4.62 1.00

Raw materials from Ores and agricultural sources metals 2.54 0.24 2.02 0.19 0.06 0.07 0.24 0.53 0.53 1.01 0.10 1.25 0.28 0.85 1.42 0.80 5.44 0.20 2.14 3.25 0.05 2.02 1.19 0.15 1.16 1.63 0.20 26.59 2.47 3.56 1.00 1.00

Fuels 8.17 0.10 6.25 1.11 0.17 9.53 0.83 0.24 0.14 4.98 10.02 2.46 0.44 0.43 0.17 1.00

Manufactured goods 0.23 0.95 0.56 1.03 0.97 0.25 1.07 1.05 0.51 0.51 0.18 0.60 1.02 0.08 0.48 1.00

Source: Based on UNCTAD Handbook of Statistics, 2006–2007

 nnexure 4.2 Values of Location Quotient for Export of Major A Commodity Groups (2005)

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia

All food items 0.02 0.02 6.85 0.93 0.08 0.76 1.25 3.01 3.80 0.01 1.58 2.80 1.04 2.87 0.47 2.57

Raw materials from agricultural sources 0.00 0.01 0.86 1.20 0.02 0.82 0.82 1.12 2.48 0.00 1.15 9.15 3.11 4.42 0.33 2.93

Ores and metals 0.15 0.04 1.00 0.86 4.54 0.06 0.90 5.16 3.13 0.03 4.52 1.93 1.79 17.44 0.55 0.42

Fuels 7.06 7.12 1.20 0.34 5.46 0.02 0.51 3.54 0.44 6.69 0.77 4.03 1.54 0.16 0.17 2.91

Manufactured goods 0.01 0.01 0.42 1.13 0.11 1.25 1.08 0.15 0.72 0.09 0.83 0.05 0.81 0.18 1.24 0.48 (continued)

Annexures

164

Countries Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay

All food items 4.31 5.56 0.60

Raw materials from agricultural sources 1.87 5.17 0.94

Ores and metals 0.36 0.07 0.60

Fuels 0.04 2.00 0.31

2.65 2.10 4.09 1.43 5.15 6.73 0.27 1.58 0.16 0.68 9.67 3.26 4.89 1.26 9.39 0.94 8.53 1.47 1.69 0.53 1.27 0.97 0.08 2.18 5.31 0.15 0.05 1.02 0.90 4.30 0.78 0.26 3.07 2.02 7.49 0.23 0.78 1.74 12.35 10.94

1.65 0.23 2.72 1.64 0.45 2.93 3.28 0.60 4.69 0.55 1.95 1.54 2.04 1.46 1.86 0.40 0.53 1.10 3.16 0.18 0.29 0.37 0.34 0.17 5.64 0.50 0.02 1.61 0.05 0.30 0.31 8.52 1.14 1.94 6.66 0.23 0.32 0.93 0.47 5.71

0.42 0.46 0.14 0.97 0.91 0.19 1.07 0.66 1.10 0.81 0.38 2.60 0.15 24.64 2.30 0.59 5.88 2.53 2.59 0.37 0.27 0.47 0.56 3.77 0.54 0.52 0.16 0.36 0.09 0.16 0.56 17.69 2.67 0.86 1.31 0.08 1.90 0.12 1.21 0.41

0.70 0.00 4.28 4.15 0.22 1.80 0.32 0.30 6.03 0.16 0.41 0.69 0.41 0.50 0.05 0.20 0.10 0.95 1.99 6.19 0.05 0.25 0.06 0.01 1.21 0.40 6.73 0.97 0.08 0.01 1.07 0.39 0.36 0.85 0.11 6.96 5.03 0.30 0.05 0.00

Manufactured goods 0.88 0.34 1.19 0.91 1.13 0.10 0.36 0.78 0.31 1.14 1.10 0.05 1.21 0.31 0.76 0.76 0.02 0.32 1.19 0.26 0.90 0.63 0.12 1.20 1.18 1.30 0.97 0.48 1.22 0.07 1.02 1.24 0.93 1.04 0.28 0.85 0.92 0.43 0.01 0.24 1.10 0.12 0.19 (continued)

Annexures

Countries Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe

165 All food items 3.00 0.88 1.39 1.25 0.43 0.10 4.18 0.25 1.31 2.07 3.45 0.53 0.39 1.95 1.73 0.45 1.50 1.54 0.44 0.81 1.04 7.99 0.59 0.07 2.94 1.13 1.94 4.45

Raw materials from agricultural sources 1.11 0.34 0.76 1.26 1.44 0.05 1.32 0.21 1.77 0.73 1.37 2.65 0.27 2.13 2.90 0.02 0.40 0.33 0.17 0.39 1.52 5.80 1.87 0.03 1.95 1.56

Ores and metals 14.88 0.70 1.26 0.90 1.30 0.06 0.86 0.34 7.26 0.77 1.19 1.00 0.84 0.39 0.39 0.13 0.35 0.77 0.81 0.94 0.85 0.16 2.17 0.68 0.18 1.89

Fuels 0.82 0.14 0.38 0.34 0.77 6.43 1.52 0.92 0.79 0.32 0.00 0.38 0.16 4.90 0.32 5.04 0.93 0.26 4.87 0.73 0.22 0.35 4.11 6.38 1.86 0.43

Manufactured goods 0.23 1.20 1.07 1.10 1.07 0.13 0.61 1.13 0.72 1.04 0.94 1.12 1.24 0.18 1.05 0.35 1.01 1.11 0.35 1.08 1.14 0.40 0.38 0.11 0.67 1.04

3.51 5.21

19.41 7.25

0.05 0.02

0.22 0.50

 nnexure 4.3 Values of Location Quotient for Export of Major A Commodity Groups (2015)

Countries Algeria Angola Argentina Austria

All food Raw materials from Ores and items agricultural sources metals 0.08 0.02 0.09 0.01 0.01 0.07 7.13 0.69 0.86 0.86 1.06 0.99

Fuels 8.26 8.51 0.23 0.18

Manufactured goods 0.04 0.01 0.40 1.14 (continued)

Annexures

166

Countries Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait

All food Raw materials from items agricultural sources 0.53 0.15 0.43 0.93 1.16 0.95 2.07 0.43 4.29 3.19 0.02 0.03 1.81 0.93 2.36 13.60 1.40 2.94 2.72 4.76 0.31 0.27 1.80 2.73 4.56 1.83 6.93 5.02 0.52 0.77 2.29 2.29 3.04 0.63 5.36 3.84 2.26 1.40 2.26 0.58 5.69 3.97 0.30 5.57 1.27 0.64 0.27 6.13 0.65 0.51 5.60 3.20 2.39 1.23 5.35 1.95 0.87 1.47 4.73 0.92 0.92 0.41 5.09 0.46 1.44 1.31 2.39 3.30 0.81 0.34 1.14 0.38 0.98 0.45 0.10 0.59 1.88 0.26 4.90 8.14 0.14 0.55 0.16 0.36

Ores and metals 6.47 0.11 0.88 6.90 3.58 0.04 4.27 0.93 2.11 15.79 0.34 0.38 0.31 0.07 0.51 0.42 1.02 0.30 1.43 0.36 0.68 1.47 0.54 1.63 0.76 1.64 2.65 1.82 14.60 1.07 0.39 11.86 1.12 1.59 2.05 0.32 0.64 0.78 2.48 0.78 0.59 0.18

Fuels 2.96 0.06 0.75 4.44 0.64 8.16 0.95 3.97 1.78 0.06 0.11 4.72 0.00 1.55 0.28 0.46 0.10 3.25 2.01 0.12 1.51 0.64 0.30 6.21 0.20 2.65 2.63 0.27 3.05 0.33 0.21 0.14 1.15 2.06 5.01 0.06 0.28 0.17 0.01 0.39 0.54 6.99

Manufactured goods 0.52 1.26 1.03 0.08 0.50 0.07 0.76 0.12 0.74 0.19 1.27 0.32 0.75 0.17 1.20 0.93 0.90 0.10 0.66 1.02 0.31 1.03 1.10 0.14 1.19 0.12 0.50 0.54 0.06 0.66 1.17 0.16 0.91 0.59 0.36 1.17 1.14 1.26 1.00 0.49 1.20 0.22 (continued)

Annexures

Countries Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia (former) Zambia Zimbabwe Total

167 All food Raw materials from items agricultural sources 1.22 1.22 0.86 0.14 3.53 0.68 0.79 0.21 0.21 3.84 2.34 0.60 1.81 1.98 6.76 7.78 0.46 0.84 1.11 0.56 2.22 0.90 2.26 1.13 7.54 1.22 2.94 0.99 0.88 0.63 1.45 0.85 1.40 1.62 1.21 1.17 0.27 0.06 4.48 1.07 0.35 0.31 1.37 1.60 1.88 0.67 2.88 1.65 0.74 2.82 0.46 0.15 5.39 1.23 1.57 2.55 0.36 0.02 1.60 0.41 1.38 0.30 0.64 0.23 0.79 0.44 1.15 1.53 6.92 8.94 1.02 1.76 0.04 0.03 1.67 1.79 1.22 1.97 1.14 1.29 5.85 4.22 1.00 1.00

Ores and metals 1.15 0.15 0.13 0.76 19.79 2.33 0.78 0.92 0.26 1.85 0.45 0.75 0.22 14.25 1.49 1.06 0.63 0.71 0.62 2.87 0.40 7.23 1.03 0.16 1.23 0.66 0.47 0.34 0.40 0.41 1.17 1.56 1.11 0.87 0.07 1.78 0.26 0.22 1.42 21.04 3.34 1.00

Fuels 1.43 2.26 0.02 0.53 1.89 0.19 1.28 0.16 7.80 5.19 0.10 2.46 1.79 0.75 0.11 0.29 0.66 0.40 5.96 1.62 1.16 1.07 0.44 0.16 0.56 0.11 0.72 0.35 4.94 0.62 0.28 3.85 0.68 0.70 0.02 4.75 7.85 0.39 0.51 0.13 1.23 1.00

Manufactured goods 0.90 0.88 0.90 1.13 0.04 0.91 0.85 0.31 0.04 0.31 1.02 0.63 0.13 0.20 1.14 1.06 1.01 1.08 0.35 0.40 1.09 0.65 0.99 0.93 1.05 1.24 0.54 1.03 0.51 1.03 1.07 0.59 1.08 1.02 0.32 0.36 0.10 1.04 1.01 0.19 0.21 1.00

Annexures

168

 nnexure 4.4 Values of Location Quotient for Import A of Major Commodity Groups (1995)

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia

All food items 3.10 2.72 0.58 0.60 1.30 1.82 1.33 1.01 1.13 1.43 0.83 1.83 0.62 0.72 0.74 1.04 1.08 2.22 0.55 1.36 1.43 0.80 2.99 1.55 1.52 0.64 1.13 2.01 1.11 0.90 1.68 1.25 3.29 1.33 0.60 1.24 0.52 0.93

Raw materials from agricultural sources 1.02 0.28 0.65 1.03 0.31 1.08 0.76 0.54 0.87 0.19 0.88 0.80 0.57 0.55 1.67 0.86 0.40 0.30 0.91 1.02 0.54 0.90 2.25 0.64 0.21 1.17 0.82 0.22 0.92 0.36 0.79 0.48 0.32 0.35 0.97 0.51 1.48 1.98

Ores and metals 0.40 0.10 0.69 1.00 1.20 0.58 1.39 0.66 0.87 0.84 1.17 1.44 0.87 0.57 1.14 0.67 0.50 0.38 1.21 0.56 0.16 0.47 0.69 0.41 0.21 1.47 0.89 0.29 1.10 0.72 0.76 0.30 0.20 0.31 1.09 1.17 2.04 1.17

Fuels 0.13 0.08 0.52 0.55 4.57 0.96 0.87 0.56 1.50 0.02 4.34 0.30 0.46 1.12 0.49 0.36 1.06 2.40 1.15 0.44 1.14 0.73 0.15 1.14 1.69 1.11 0.86 0.43 0.83 0.78 0.90 1.54 2.38 1.43 1.47 0.90 3.42 0.92

Manufactured goods 0.86 0.96 1.14 1.10 0.60 0.92 0.96 1.08 0.94 1.09 0.66 0.95 1.13 1.06 1.05 1.09 1.03 0.76 1.03 1.04 0.99 1.09 0.80 0.96 0.94 1.00 1.01 1.00 1.00 1.08 0.95 0.97 0.63 0.98 1.00 0.99 0.73 0.97 (continued)

Annexures

Countries Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States of America Uruguay

169 All food items 2.24 0.98 1.25 1.74 2.20 1.06 0.58 1.64 0.53 1.07 1.79 0.71 1.50 2.05 1.47 0.78 1.19 0.73 1.87 1.13 1.95 1.42 1.09 1.01 1.44 0.90 1.77 3.42 0.49 0.73 1.44 1.67 0.71 0.70 1.79 0.41 1.66 1.32 0.74 1.07 1.08 0.52 1.10

Raw materials from agricultural sources 0.78 0.44 1.88 2.04 0.68 0.60 1.79 0.35 0.42 0.26 1.02 0.80 0.21 2.03 0.76 0.38 0.29 0.86 1.77 0.26 0.06 0.61 0.85 1.04 1.16 0.75 0.39 0.70 0.29 0.76 0.95 0.57 0.73 0.66 1.07 1.36 0.22 1.34 1.79 0.29 0.80 0.70 1.28

Ores and metals 0.83 0.59 1.32 1.72 0.65 0.52 1.63 0.43 0.86 0.26 0.22 0.61 0.17 1.02 0.91 0.89 0.25 1.72 0.68 0.35 0.18 0.21 1.02 0.83 0.55 0.92 0.92 0.42 0.59 0.63 1.00 0.40 0.98 0.79 0.35 0.86 1.45 0.76 1.49 0.55 0.88 0.71 0.30

Fuels 0.23 0.45 0.94 2.05 1.62 1.82 1.78 0.07 0.29 0.49 0.88 0.28 2.40 1.70 0.96 0.66 0.64 0.36 2.03 1.70 0.81 1.09 1.41 1.14 1.02 2.67 0.03 1.24 1.02 1.07 1.04 0.30 0.73 0.37 0.14 0.87 0.06 0.90 1.60 0.20 0.45 1.06 1.25

Manufactured goods 0.94 1.11 0.92 0.71 0.81 0.95 0.90 1.08 1.17 1.11 0.95 1.14 0.86 0.75 0.96 1.09 1.08 1.07 0.76 0.97 0.98 0.99 0.95 0.99 0.96 0.85 1.04 0.71 1.11 1.06 0.94 1.04 1.08 1.13 1.02 1.08 1.03 0.97 0.91 1.13 1.06 1.08 0.99 (continued)

Annexures

170

Countries USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe

All food items 1.68 1.50 0.54 0.51 1.04 0.65

Raw materials from agricultural sources 0.61 1.43 0.80 1.23 0.74 0.62

Ores and metals 1.05 0.93 0.53 1.12 0.55 0.56

Fuels 2.40 0.14 1.34 1.34 1.64 1.15

Manufactured goods 0.78 1.01 1.05 1.01 0.96 1.07

 nnexure 4.5 Values of Location Quotient for Import A of Major Commodity Groups (2005)

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland

All food items 2.78 2.16 0.41 0.88 0.98 1.95 1.18 1.48 0.63 2.49 0.80 2.60 0.82 0.89 0.47 1.26 0.93 2.12 0.81 1.66 1.68 1.15 3.31 2.50 2.12 0.76

Raw materials from agricultural sources 0.96 0.40 0.86 1.25 0.28 2.44 0.70 0.81 0.87 0.14 0.91 0.99 0.71 0.58 2.03 0.92 0.60 0.26 0.85 1.27 0.65 0.75 2.65 0.94 0.21 1.67

Ores and metals 0.40 0.09 0.92 0.94 1.87 0.68 1.04 0.19 1.01 0.32 2.06 0.92 0.75 0.92 2.23 0.70 0.35 0.26 0.96 0.46 0.30 0.31 0.99 0.26 0.23 1.72

Fuels 0.07 0.08 0.34 0.84 3.20 0.94 0.89 0.71 1.25 0.09 0.44 1.81 0.64 1.49 0.67 0.18 0.74 1.94 0.63 0.47 0.94 0.82 1.06 1.11 1.98 0.96

Manufactured goods 1.05 1.13 1.20 1.04 0.53 0.90 1.01 1.06 0.99 1.10 1.08 0.69 1.11 0.93 1.03 1.16 1.10 0.76 1.10 1.07 0.99 1.06 0.73 0.87 0.76 0.98 (continued)

Annexures

Countries France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain

171 All food items 1.12 2.73 1.06 1.85 1.62 1.56 2.64 2.16 0.62 1.12 0.51 1.16 1.25 1.21 1.33 1.53 2.01 1.23 0.64 1.93 0.76 1.70 2.44 0.87 1.86 1.52 1.40 1.10 1.86 0.98 1.52 1.79 1.18 1.64 1.00 0.90 1.67 0.86 2.13 4.04 0.41 0.71 1.33

Raw materials from agricultural sources 0.82 0.22 0.89 0.76 0.70 0.68 0.65 0.43 0.63 0.76 1.22 1.95 1.19 0.61 1.60 1.37 0.68 0.93 1.12 0.32 0.69 0.45 1.10 0.80 0.25 1.57 0.93 0.45 0.26 1.05 2.41 0.29 0.66 1.00 0.49 1.07 0.91 0.59 0.40 0.92 0.22 0.68 0.77

Ores and metals 0.69 0.30 1.08 0.55 0.79 0.26 0.05 0.23 0.54 1.08 1.45 0.87 0.60 0.39 1.16 1.62 0.51 0.38 1.77 0.39 0.98 0.19 0.27 0.66 0.12 0.85 0.92 0.61 0.34 2.04 0.89 0.27 0.24 0.26 0.62 0.79 0.70 0.70 1.09 0.57 0.43 0.75 0.86

Fuels 0.92 0.25 0.82 0.54 1.23 1.07 1.18 1.49 0.73 0.64 2.70 2.08 0.73 0.50 0.87 1.80 1.62 1.51 1.78 0.09 0.57 0.73 1.15 0.38 1.82 1.47 1.11 0.83 0.89 0.29 1.48 1.23 1.00 1.35 0.91 0.80 1.06 0.96 0.02 1.57 1.24 1.09 0.97

Manufactured goods 1.03 1.04 1.03 1.04 0.91 0.98 0.87 0.84 1.12 1.06 0.68 0.75 1.05 1.12 0.97 0.75 0.81 0.91 0.84 1.14 1.12 1.04 0.87 1.16 0.82 0.85 0.95 1.06 0.99 1.09 0.83 0.93 1.03 0.91 1.05 1.06 0.94 1.05 1.10 0.62 1.06 1.03 0.99 (continued)

Annexures

172

Countries Sri Lanka Sweden Switzerland Syrian Arab Republic Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States of America Uruguay USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe

All food items 1.85 1.09 0.80 2.62 0.60 1.30 1.22 0.44 1.34 1.35 0.63 1.16 1.79 1.42 0.93 1.25 0.90 1.38

Raw materials from agricultural sources 0.71 0.93 0.63 2.30 1.20 0.32 1.49 1.69 0.42 0.83 0.78 1.74 0.80 0.62 2.13 1.00 0.74 1.35

Ores and metals 0.95 0.89 0.93 0.71 1.07 1.13 0.80 1.69 1.24 0.68 0.60 0.42 0.82 0.38 0.81 1.13 0.70 10.13

Fuels 0.95 0.82 0.50 0.14 1.27 2.39 0.94 1.01 0.06 0.62 1.23 1.66 0.89 0.06 1.03 0.98 0.73 1.04

Manufactured goods 0.94 1.03 1.13 1.00 0.98 0.70 0.99 1.00 1.16 1.06 1.01 0.86 0.96 1.19 0.98 0.97 1.09 0.47

 nnexure 4.6 Values of Location Quotient for Import A of Major Commodity Groups (2015)

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile

All food items 2.26 2.07 0.32 0.94 2.04 1.97 1.15 0.92 0.60 1.84 1.19 1.85 0.99 1.05

Raw materials from agricultural sources 0.88 0.39 0.62 1.38 0.48 4.17 0.80 0.36 0.64 0.10 0.70 1.05 0.61 0.45

Ores and metals 0.37 0.15 0.67 1.10 1.92 0.48 0.95 0.14 0.80 0.39 2.21 0.40 0.72 0.32

Fuels 0.37 0.37 0.92 0.66 0.17 0.60 1.06 0.68 1.20 0.50 1.35 1.54 0.60 1.17

Manufactured goods 1.00 1.04 1.12 1.05 0.98 0.91 0.98 1.13 1.03 1.04 0.86 0.85 1.09 1.02 (continued)

Annexures

Countries China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan

173 All food items 0.76 1.22 1.38 2.32 0.69 1.66 1.84 1.03 2.13 2.08 2.14 1.09 1.01 1.95 0.97 1.72 1.65 1.67 2.33 1.74 0.65 1.33 0.74 1.14 1.72 1.36 1.28 1.20 2.36 1.01 0.70 1.90 1.05 0.72 2.65 0.72 1.46 1.31 1.41 1.36 1.70 1.13 1.41

Raw materials from agricultural sources 2.17 0.51 0.71 0.34 0.73 1.75 0.78 0.54 1.64 1.12 0.27 1.47 0.65 0.42 0.88 0.68 0.64 0.85 0.52 0.57 0.73 0.75 1.28 1.79 1.04 0.40 1.43 1.12 0.76 0.76 0.99 0.33 1.07 0.14 1.46 0.65 0.23 1.12 0.98 0.52 0.51 0.77 2.59

Ores and metals 2.40 0.38 0.48 0.26 0.76 0.38 0.25 0.33 0.70 0.26 0.28 1.36 0.51 0.25 1.02 0.24 0.85 0.22 0.13 0.17 0.64 2.85 1.76 0.76 0.32 0.33 1.14 1.55 0.45 0.30 1.62 0.67 1.40 0.38 0.24 0.55 0.12 1.06 0.84 0.47 0.50 1.21 0.93

Fuels 1.03 0.79 0.82 1.84 0.59 0.54 1.48 1.63 1.07 1.08 1.90 1.17 0.87 0.61 0.81 1.01 2.21 1.18 2.03 1.01 0.71 1.03 2.54 1.29 0.23 0.63 1.08 1.73 1.52 1.34 1.96 0.05 1.04 2.40 1.29 0.57 1.90 1.50 1.45 0.83 0.96 0.36 1.89

Manufactured goods 0.92 1.05 1.02 0.77 1.12 1.02 0.87 0.94 0.86 0.90 0.78 0.93 1.06 1.01 1.04 0.96 0.74 0.94 0.74 0.97 1.11 0.86 0.73 0.93 1.08 1.07 0.94 0.82 0.79 0.99 0.84 1.09 0.96 0.86 0.79 1.14 0.87 0.88 0.89 1.03 0.96 1.08 0.77 (continued)

Annexures

174

Countries Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia (former) Zambia Zimbabwe Total

All food items 0.86 1.02 1.31 1.37 1.01 1.76 1.07 1.65 2.38 0.50 0.79 1.33 1.55 1.34 0.74 2.20 0.77 1.48 1.31 0.71 1.03 1.20 0.72 1.50 1.49 2.24 0.90 1.37 0.60 2.25 1

Raw materials from agricultural sources 0.11 0.55 0.73 0.38 1.02 1.11 0.99 0.47 0.89 0.28 0.59 0.73 1.34 0.77 0.51 1.38 1.02 0.51 1.18 1.75 0.36 0.80 0.60 1.27 0.83 0.52 1.62 1.14 0.32 0.27 1

Ores and metals 0.09 0.23 0.24 0.43 0.84 0.66 0.66 0.91 0.33 0.41 0.67 0.88 0.33 0.69 0.62 0.70 0.92 0.64 0.84 1.70 0.68 0.66 0.50 0.25 0.57 0.29 0.84 1.25 4.58 0.33 1

Fuels 1.84 1.13 0.85 0.98 0.63 1.09 0.55 0.06 1.57 1.86 1.59 1.16 1.16 0.85 0.42 1.02 1.29 1.66 1.18 0.67 0.45 0.70 0.76 0.98 0.87 0.39 0.48 1.01 0.79 0.73 1

Manufactured goods 0.95 1.03 1.04 1.01 1.07 0.91 1.09 1.10 0.79 0.97 0.95 0.95 0.94 1.01 1.16 0.87 0.98 0.87 0.94 1.03 1.12 1.05 1.11 0.98 0.99 1.01 1.09 0.94 0.89 0.95 1

 nnexure 5.1 Relation Between GDP and Merchandise A Export (1990)

S. no. 1 2

Countries Algeria Angola

GDP (current US million dollars) 61751.38 13661.85

Proportion of merchandise export to world merchandise export 0.37 0.11 (continued)

Annexures

S. no. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

Countries Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea

175 GDP (current US million dollars) 153185.7 166066.8 4909.13 28136.57 219620.8 4867.583 406,897 3900.702 20726.03 11845.71 593942.3 34481.45 398623.7 56924.87 6800.559 11892.83 57092.29 138248.2 9522.388 15231.97 36014.4 4800.9 1339.124 141524.8 1,277,692 6039.492 1,764,944 9982.838 97892.93 6819.965 3905.811 3637.289 37011.11 6521.547 316868.7 133857.6 96363.63 49356.43 1,177,387 3,139,974 4020.27 12664.49 279347.7

Proportion of merchandise export to world merchandise export 0.35 1.18 0.11 0.05 3.55 0.03 0.90 0.06 0.14 0.06 3.65 0.24 1.78 0.19 0.04 0.09 0.34 1.05 0.06 0.08 0.07 0.02 0.01 0.76 6.22 0.06 12.05 0.03 0.23 0.03 0.02 0.03 0.29 0.05 0.51 0.73 0.55 0.68 4.87 8.23 0.03 0.03 1.86 (continued)

Annexures

176

S. no. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87

Countries Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe

GDP (current US million dollars) 18470.82 44024.59 2634.682 2619.406 293358.4 1718.266 30319.7 314264.9 45440.19 68328.91 119790.9 51666.17 6076.503 4652.574 29118.87 49095.18 65977.75 78726.03 40591.45 117473.5 6204.748 38899.86 116699.4 535070.5 9389.551 258154.7 259016.2 11163.66 88299.44 5068 13519.51 202545.8 50701.37 1,093,214 6,010,634 9239.419 783306.9 47036.42 6471.745 92418.15 3794.81 11737.77

Source: Based on UNCTADstat database

Proportion of merchandise export to world merchandise export 0.20 0.84 0.03 0.03 1.16 0.02 0.12 3.77 0.27 0.39 0.97 0.16 0.01 0.03 0.09 0.23 0.41 0.47 0.14 1.27 0.02 1.51 0.67 1.59 0.05 1.65 1.82 0.12 0.66 0.06 0.10 0.37 0.67 5.30 11.26 0.05 2.97 0.50 0.07 0.41 0.04 0.05

Annexures

177

 nnexure 5.2 Relation Between GDP and Merchandise A Export (2015)

S. no 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia

GDP (current US million dollars) 164779.4 117955.2 632343.4 376967.4 31125.85 194,466 511908.9 32997.7 1,772,591 12929.52 48952.96 28415.97 1,552,808 240796.4 11,158,457 292080.2 52958.37 32075.94 272,424 301307.8 67103.27 100176.8 315917.1 25850.21 4391.065 231960.2 2,425,204 13734.56 3,363,600 37156.08 194860.2 63794.22 8874.851 20364.83 121715.2 16779.6 2,116,239 861,934

Proportion of merchandise export to world merchandise export 0.23 0.20 0.34 0.92 0.07 0.20 2.51 0.05 1.15 0.04 0.16 0.02 2.47 0.38 13.74 0.22 0.06 0.07 1.41 0.58 0.06 0.11 0.12 0.03 0.00 0.36 3.06 0.04 8.03 0.06 0.17 0.06 0.01 0.05 0.60 0.03 1.61 0.91 (continued)

Annexures

178

S. no 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81

Countries Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay

GDP (current US million dollars) 398562.6 283,716 1,821,580 4,383,076 37517.46 63398.52 1,377,873 114054.1 296284.5 9746.922 11510.95 1,140,724 11757.94 100359.1 750318.1 173416.6 494582.6 386578.4 266458.3 52132.29 27714.12 190427.9 292449.1 477066.5 199121.7 177,956 653219.3 13632.55 292733.9 314571.2 1,192,955 82316.16 495694.4 677151.1 28392.77 395,168 25927.22 41198.52 717887.9 370296.3 2,858,003 18,139,554 53442.38

Proportion of merchandise export to world merchandise export 0.38 0.73 2.77 3.78 0.05 0.04 3.18 0.33 1.21 0.02 0.01 2.30 0.03 0.13 3.43 0.21 0.31 0.64 0.13 0.10 0.05 0.21 0.35 1.20 0.33 0.37 1.22 0.02 2.12 0.49 1.70 0.06 0.85 1.75 0.01 1.30 0.07 0.09 0.87 1.88 2.78 9.09 0.05 (continued)

Annexures

S. no 82 83 84 85 86 87

179

Countries USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe

GDP (current US million dollars) 1,934,940 344331.4 193241.1 165376.7 21255.23 13893.03

Proportion of merchandise export to world merchandise export 3.25 0.22 0.98 0.41 0.04 0.02

Source: Based on UNCTADstat database

 nnexure 5.3 Residuals from Regression of Merchandise A Export on GDP (1990)

S. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France

Residuals −0.22 −0.36 −0.46 0.33 −0.35 −0.46 2.57 −0.43 −0.54 −0.39 −0.35 −0.41 1.76 −0.29 0.36 −0.39 −0.42 −0.38 −0.24 0.28 −0.40 −0.40 −0.45 −0.44 −0.43 −0.03 2.65

Standard residuals −0.20 −0.32 −0.41 0.29 −0.30 −0.41 2.26 −0.37 −0.47 −0.34 −0.31 −0.36 1.54 −0.25 0.32 −0.34 −0.37 −0.34 −0.21 0.24 −0.35 −0.35 −0.40 −0.38 −0.38 −0.02 2.33 (continued)

Annexures

180 S. no. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71

Countries Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden

Residuals −0.39 7.28 −0.44 −0.45 −0.42 −0.43 −0.42 −0.25 −0.41 −0.70 −0.03 −0.12 0.12 1.55 0.10 −0.42 −0.44 0.74 −0.28 0.29 −0.41 −0.41 0.01 −0.43 −0.39 2.56 −0.28 −0.22 0.24 −0.41 −0.45 −0.42 −0.42 −0.33 −0.19 −0.16 −0.40 0.54 −0.43 0.97 −0.05 −0.16 −0.41 0.57

Standard residuals −0.34 6.40 −0.39 −0.39 −0.37 −0.38 −0.37 −0.22 −0.36 −0.62 −0.03 −0.11 0.10 1.36 0.09 −0.37 −0.39 0.65 −0.25 0.26 −0.36 −0.36 0.01 −0.37 −0.35 2.25 −0.25 −0.19 0.21 −0.36 −0.39 −0.37 −0.37 −0.29 −0.17 −0.14 −0.35 0.48 −0.38 0.85 −0.05 −0.14 −0.36 0.50 (continued)

Annexures S. no. 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87

181 Countries Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe

Residuals 0.75 −0.35 0.00 −0.40 −0.37 −0.57 0.11 2.18 −3.90 −0.41 0.61 −0.06 −0.39 −0.26 −0.41 −0.42

Standard residuals 0.66 −0.31 0.00 −0.35 −0.33 −0.50 0.10 1.91 −3.43 −0.36 0.54 −0.05 −0.34 −0.23 −0.36 −0.37

 nnexure 5.4 Residuals from Regression of Merchandise A Export on GDP (2015)

S. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Côte d’Ivoire

Residuals −0.35 −0.34 −0.58 0.19 −0.41 −0.40 1.68 −0.43 −0.61 −0.43 −0.34 −0.45 0.86 −0.25 5.05 −0.46 −0.44 −0.41

Standard residuals −0.31 −0.31 −0.52 0.17 −0.37 −0.36 1.51 −0.39 −0.55 −0.38 −0.30 −0.41 0.78 −0.23 4.54 −0.41 −0.39 −0.37 (continued)

Annexures

182 S. no. 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61

Countries Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines

Residuals 0.75 −0.10 −0.45 −0.42 −0.57 −0.44 −0.45 −0.27 0.81 −0.42 5.09 −0.42 −0.43 −0.44 −0.45 −0.42 0.05 −0.44 −0.40 −0.18 −0.37 0.06 0.97 0.08 −0.44 −0.47 1.71 −0.21 0.53 −0.45 −0.45 1.00 −0.44 −0.40 2.42 −0.38 −0.51 −0.11 −0.52 −0.39 −0.43 −0.39 −0.32

Standard residuals 0.68 −0.09 −0.40 −0.38 −0.51 −0.40 −0.41 −0.24 0.73 −0.38 4.58 −0.38 −0.38 −0.39 −0.40 −0.38 0.04 −0.40 −0.36 −0.17 −0.33 0.06 0.87 0.07 −0.39 −0.42 1.54 −0.19 0.48 −0.40 −0.40 0.90 −0.39 −0.36 2.17 −0.34 −0.46 −0.09 −0.47 −0.35 −0.38 −0.35 −0.29 (continued)

Annexures S. no. 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87

183 Countries Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Viet Nam Yugoslavia, SFR (former) Zambia Zimbabwe

Residuals 0.39 −0.27 −0.22 0.28 −0.45 1.45 −0.20 0.37 −0.45 0.02 0.79 −0.47 0.55 −0.41 −0.40 −0.12 1.15 0.22 −4.76 −0.45 1.36 −0.49 0.38 −0.17 −0.43 −0.45

Standard residuals 0.35 −0.24 −0.20 0.25 −0.40 1.30 −0.18 0.33 −0.41 0.02 0.71 −0.42 0.49 −0.37 −0.36 −0.11 1.03 0.19 −4.28 −0.40 1.22 −0.44 0.34 −0.15 −0.39 −0.40

 nnexure 6.1 Argentina: Standardized Residuals A from Regression of GDP Ratio on Trade Ratio (1990 and 2015)

S. no. 1 2 3 4 5 6

Countries Algeria Angola Austria Bahrain Bangladesh Belgium-Luxembourg

1990 0.015 −0.030 0.259 0.017 −0.062 −0.503

2015 −0.427 −0.079 0.030 −0.097 −0.074 −0.076 (continued)

Annexures

184 S. no. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Countries Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Cote d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius

1990 −0.793 −4.573 0.016 −0.073 0.031 0.908 −1.471 0.119 −0.177 −0.012 0.014 0.002 0.212 −0.013 −0.109 −0.261 0.008 0.011 0.240 1.341 0.011 0.169 0.019 0.156 −0.025 0.014 0.003 0.054 0.021 0.564 −0.003 −1.124 0.076 0.079 4.761 0.014 0.016 0.227 0.009 −0.269 0.003 0.003

2015 −0.825 −5.101 −0.107 −0.092 −0.107 0.326 −0.755 2.397 −0.104 −0.107 −0.105 −0.007 −0.082 −0.107 −0.168 −0.185 −0.105 −0.111 −0.021 0.754 −0.108 0.652 −0.098 −0.039 −0.093 −0.108 −0.108 −0.061 −0.105 0.472 0.009 −0.116 −0.035 0.365 1.918 −0.165 −0.080 0.213 −0.065 −0.284 −0.108 −0.117 (continued)

Annexures S. no. 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86

185 Countries Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1990 −0.605 0.013 0.018 −3.223 0.076 0.051 0.089 0.071 −0.022 −0.525 −0.514 0.026 −0.016 −0.128 −0.127 0.139 0.020 −0.125 0.018 −0.264 0.018 0.311 0.052 −0.045 −0.034 0.003 −0.042 −0.041 0.066 1.417 4.379 −0.898 0.131 −0.298 0.022 0.055 0.016 0.023

2015 −0.143 −0.107 −0.166 −0.271 −0.053 −0.012 0.035 −0.052 −0.576 −0.104 −0.263 −0.052 0.002 −0.041 −0.038 0.037 −0.118 −0.016 −0.126 −0.183 −0.070 0.080 −0.253 −0.149 −0.197 −0.583 −0.137 0.200 0.034 1.156 6.527 −0.499 0.675 −0.405 −0.449 −0.076 −0.102 −0.106

Annexures

186

 nnexure 6.2 Brazil: Standardized Residuals from Regression A of GDP Ratio on Trade Ratio (1990 and 2015)

S. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Cote d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy

1990 −0.0716 −0.0792 −3.0577 0.4932 −0.0145 0.0988 −1.1000 −0.3813 0.0751 0.0936 0.0610 0.6924 −0.9633 0.6590 −0.1227 −0.0059 0.0714 0.0595 0.3765 0.0155 −0.1332 −0.0788 0.0517 0.0657 0.3678 2.0389 0.0346 0.2437 0.0579 0.2740 0.0462 0.0502 0.0384 −0.1670 0.0797 0.9946 0.1740 −1.7502 0.1441 0.2336

2015 −0.335 0.023 −3.992 0.200 0.023 −0.007 −0.373 −0.713 0.083 0.089 0.083 0.523 −1.207 −3.328 −0.323 0.058 0.079 0.195 0.143 0.031 0.011 −0.100 0.073 0.075 0.115 0.910 0.078 −0.010 0.064 0.217 0.082 0.067 0.066 0.051 0.027 0.358 0.162 0.112 0.156 0.152 (continued)

Annexures S. no. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86

187 Countries Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay USSR (former) Venezuela, RB Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1990 5.5556 0.0410 0.0768 −0.0555 −0.1823 −0.2343 0.0568 0.0624 −0.1012 0.0697 −0.0455 −3.6949 0.1673 −0.1633 0.2437 0.1424 −0.0041 −1.1488 −0.3234 −0.0273 −0.1027 −0.1136 0.0846 −2.5947 0.0656 −0.3651 0.1947 0.5276 0.0885 0.2609 −0.1598 0.0835 −0.0611 0.0105 0.0585 0.3725 −0.0926 1.2258 −0.0704 −1.3252 2.3782 −0.8511 0.0816 0.0486 0.0734 0.0811

2015 2.098 0.057 0.116 −0.429 0.041 −0.309 0.073 0.076 −0.531 0.082 −0.105 −1.572 0.205 −0.584 0.112 0.252 −0.582 0.057 −0.356 0.151 0.292 −0.075 0.125 −0.249 0.063 −0.179 −0.032 −0.130 0.114 0.187 −0.028 0.084 −0.259 −0.171 0.044 0.342 −0.192 1.527 6.443 −0.671 0.645 −0.403 −0.502 0.089 0.086 0.077

Annexures

188

 nnexure 6.3 Paraguay: Standardized Residuals A from Regression of GDP Ratio on Trade Ratio (1990 and 2015)

S. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Cote d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India

1990 −0.4307 −0.1943 −1.2622 0.0314 −0.2036 −0.1642 −0.0380 −0.2357 −3.8353 −0.2047 −0.1780 −0.1929 0.6785 −0.4669 0.3124 −0.1567 −0.2006 −0.1935 −0.1390 −0.0136 −0.1992 −0.1961 −0.1494 −0.2035 −0.2081 0.0055 1.3641 −0.2009 1.7710 −0.2011 −0.0488 −0.2006 −0.2061 −0.2055 −0.1588 −0.2003 0.2726

2015 −0.198 −0.198 −1.494 −0.077 −0.230 −0.210 −0.036 −0.306 −2.624 −0.236 −0.221 −0.229 0.452 −0.744 3.492 −0.128 −0.220 −0.228 −0.124 −0.131 −0.215 −0.232 −0.145 −0.233 −0.240 −0.147 0.749 −0.242 1.014 −0.226 −0.193 −0.215 −0.238 −0.233 −0.190 −0.234 0.515 (continued)

Annexures S. no. 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

189 Countries Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States Uruguay

1990 −0.0430 −0.0299 −0.1394 1.1791 3.6723 −0.2039 −0.1968 −0.1593 −0.1816 −0.1573 −0.2067 −0.2065 −0.4211 −0.2069 −0.1742 −0.7191 −0.1422 −0.1660 −0.0285 −0.1482 −0.4238 −0.1975 −0.1452 −0.1214 −0.3040 −0.1513 −0.0299 −0.2013 −0.1800 −0.0433 0.4690 −0.1979 0.1474 −0.2184 −0.1911 −0.0853 −0.2031 −0.1912 0.0105 −0.1583 0.9335 6.5548 −0.3089

2015 0.132 −0.076 −0.120 0.327 1.593 −0.226 −0.213 0.194 −0.211 −0.137 −0.237 −0.241 0.128 −0.237 −0.214 −0.051 −0.163 −0.033 −0.065 −0.138 −0.263 −0.266 −0.110 −0.119 −0.211 −0.175 0.052 −0.243 −0.130 −0.123 0.130 −0.204 −0.034 0.029 −0.229 −0.108 −0.231 −0.250 −0.019 −0.093 0.990 7.400 −0.398 (continued)

Annexures

190 S. no. 81 82 83 84 85 86

Countries USSR (former) Venezuela, RB Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1990 0.8929 −0.1810 −0.2001 −0.1414 −0.2051 −0.1965

2015 −0.032 −0.109 −0.223 −0.177 −0.232 −0.239

 nnexure 6.4 Uruguay: Standardized Residuals A from Regression of GDP Ratio on Trade Ratio (1990 and 2015)

S. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Countries Algeria Angola Argentina Austria Bahrain Bangladesh Belgium-Luxembourg Bolivia Brazil Brunei Darussalam Bulgaria Cameroon Canada Chile China Colombia Costa Rica Cote d’Ivoire Czechoslovakia (former) Denmark Dominican Republic Ecuador Egypt El Salvador Fiji Finland

1990 −0.1658 −0.2059 −1.5541 −0.0177 −0.2051 −0.1706 −0.0243 −0.2262 −3.1321 −0.2054 −0.1819 −0.1956 0.4734 −0.3165 0.0267 −0.2808 −0.2035 −0.1952 −0.1972 −0.0450 −0.2041 −0.1999 −0.1828 −0.2038 −0.2086 −0.0134

2015 −0.183 −0.093 −2.476 0.013 −0.134 −0.075 0.097 −0.202 −3.672 −0.144 −0.138 −0.137 0.440 −0.443 1.170 −0.079 −0.149 −0.136 −0.028 −0.386 −0.125 −0.172 −0.091 −0.145 −0.148 −0.091 (continued)

Annexures S. no. 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69

191 Countries France Gabon Germany Ghana Greece Guatemala Guinea Honduras Hungary Iceland India Indonesia Iran Ireland Italy Japan Jordan Kenya South Korea Kuwait Malaysia Malta Mauritius Mexico Mongolia Morocco Netherlands New Zealand Nigeria Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Saudi Arabia Senegal Singapore South Africa Spain

1990 1.1655 −0.2058 1.3674 −0.2019 −0.0648 −0.1993 −0.2067 −0.2062 −0.1898 −0.2011 0.2587 −0.0406 −0.1960 −0.1424 0.9537 4.0104 −0.2058 −0.1978 0.0815 −0.1857 −0.1644 −0.2113 −0.2071 −0.1292 −0.2075 −0.1747 −0.0750 −0.1481 −0.3266 −0.0845 −0.1512 −0.2045 −0.2841 −0.2239 −0.1493 −0.1708 −0.1268 −0.1623 −0.1342 −0.2022 −0.2045 −0.0609 0.3882

2015 0.868 −0.151 0.550 −0.131 −0.055 −0.126 −0.146 −0.143 −0.106 −0.142 0.566 0.237 0.053 −0.017 0.387 2.063 −0.133 −0.117 0.205 −0.104 −0.066 −0.146 −0.146 −0.238 −0.144 −0.122 −0.208 −0.086 −0.585 0.026 −0.022 −0.524 −0.136 −0.286 −0.038 0.060 −0.168 −0.071 0.153 −0.145 −0.034 −0.054 −0.142 (continued)

Annexures

192 S. no. 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86

Countries Sri Lanka Sweden Switzerland Syria Thailand Trinidad and Tobago Tunisia Turkey United Arab Emirates United Kingdom United States USSR (former) Venezuela, RB Vietnam Yugoslavia, SFR (former) Zambia Zimbabwe

1990 −0.2017 0.1021 0.0100 −0.1924 −0.0911 −0.2032 −0.2006 −0.0147 −0.2039 0.6993 7.0095 0.2739 −0.1694 −0.2012 −0.1555 −0.2057 −0.1987

2015 −0.111 0.028 −0.032 −0.137 −0.071 −0.177 −0.129 −0.060 0.014 0.974 7.366 0.587 −0.590 −0.217 −0.068 −0.140 −0.143

 nnexure 6.5 Top Ten Trading Partners in Food and Live A Animals of Argentina

Per cent to total world trade of food and live animals 1995 World (Trade value, $000) 7,207,433 Brazil 26.18 Netherlands 7.79 USA, P. Rico 7.00 Germany 5.33 Spain 5.12 Italy 5.04 Japan 3.80 Chile 3.56 United Kingdom 2.96 Paraguay 2.27 Others 30.95

Per cent to total world trade of food and live animals 2015 World(Trade value, $000) 23,579,681 Brazil 10.24 Viet Nam 7.01 Spain 4.73 Algeria 4.60 Indonesia 4.36 Chile 4.00 Netherlands 3.77 USA,PR,USVI 3.60 Italy 3.60 Malaysia 3.05 51.05

Source: Calculation based on World Bank database

Annexures

193

 nnexure 6.6 Top Ten Trading Partners in Food and Live A Animals of Brazil

Per cent to total world trade of food and live animals 1995 World(Trade value, $000) 9,976,137 Netherlands 16.07 USA, P. Rico 9.90 Japan 6.29 Russian Fed 5.16 Italy 5.13 Belgium-Luxembourg 4.76 Spain 3.54 Germany 3.45 Argentina 3.37 France+Monac 2.79 Others 39.55

Per cent to total world trade of food and live animals 2015 World(Trade value, $000) 45,956,368 Netherlands 6.86 USA,PR,USVI 6.24 Germany 4.49 China 4.37 Japan 4.33 Saudi Arabia 4.29 Russian Fed 4.01 Venezuela 3.92 Egypt 3.21 Hong Kong 3.17 55.10

Source: Calculation based on WITS database

 nnexure 6.7 Top Ten Trading Partners in Food and Live A Animals of Paraguay

Per cent to total world trade of food and live animals 1995 World (Trade value, $000) 152,170 Brazil 47.400 USA, P. Rico 8.390 Netherlands 7.725 Uruguay 6.902 Chile 6.191 Argentina 5.005 Spain 3.953 Italy 2.799 Venezuela 2.378 Germany 2.266 Others 6.991

Per cent to total world trade of food and live animals 2015 World (Trade value, $000) 2,972,274 Chile 19.07 Brazil 14.16 Russian Fed 12.31 Poland 4.49 Italy 4.26 Peru 3.86 Israel 3.37 Netherlands 3.35 United Kingdom 2.80 USA, PR, USVI 2.14 30.18

Source: Calculation based on World Bank database

Annexures

194

 nnexure 6.8 Top Ten Trading Partners in Food and Live A Animals of Uruguay

Per cent to total world trade of food and live animals 1995 World(Trade value, $000) 902,436 Brazil 44.62 United Kingdom 6.30 Israel 5.25 Netherlands 5.05 Germany 4.38 Argentina 3.66 USA, P. Rico 3.54 Spain 3.36 Peru 3.35 Italy 2.46 Others 18.02

Per cent to total world trade of food and live animals 2015 World (Trade value, $000) 3,453,428 China 17.02 Brazil 14.56 USA, PR, USVI 10.40 Venezuela 4.93 Netherlands 4.28 Free Zones 4.01 Israel 3.76 Turkey 3.00 Mexico 2.99 Peru 2.92 Others 32.13

Source: Calculation based on WITS database

 nnexure 6.9 Top Ten Trading Partners in Beverages A and Tobacco of Argentina (1995 and 2015)

Per cent to total world trade of beverages and tobacco 2015 1995 World (Trade value, World (Trade value, $000) 269,382 $000) Brazil 18.633 USA,PR,USVI Paraguay 15.680 United Kingdom Spain 10.549 Canada United Kingdom 8.393 China USA, P. Rico 8.374 Brazil Japan 5.238 Paraguay France+Monac 4.965 Uruguay Belgium-Luxembourg 4.596 Belgium Germany 4.433 Netherlands Uruguay 4.005 Chile Others 15.134 Others Source: Calculation based on WITS database

Per cent to total world trade of beverages and tobacco 1,136,945 28.13 7.45 6.88 6.79 5.04 4.38 3.75 3.46 3.29 2.53 28.29

Annexures

195

 nnexure 6.10 Top Ten Trading Partners in Beverages A and Tobacco of Brazil (1995 and 2015)

1995 World (Trade value, $000) Belgium-Luxembourg Paraguay USA, P. Rico Germany United Kingdom Japan Netherlands Italy Spain Egypt Others

Per cent to total world trade of beverages and tobacco 2015 World (Trade value, 1,272,508 $000) 17.462 Belgium 16.420 China 11.738 USA, PR, USVI 9.126 Paraguay 8.507 Russian Fed 4.380 Netherlands 3.259 Germany 2.373 Indonesia 1.933 Turkey 1.756 Viet Nam 23.046 Others

Per cent to total world trade of beverages and tobacco 2,328,589 17.05 11.36 10.02 5.93 5.80 5.06 4.60 3.68 3.23 2.04 31.24

Source: Calculation based on WITS database

 nnexure 6.11 Top Ten Trading Partners in Beverages A and Tobacco of Paraguay (1995 and 2015)

1995 World (Trade value, $000) Argentina France+Monac South Africa Cus.Un BelgiumLuxembourg Brazil USA, P. Rico Morocco Spain Germany Egypt Others

Per cent to total world trade of beverages and tobacco 2015 World (Trade 9350 value, $000) 23.963 Bolivia 17.799 Uruguay 12.662 Aruba

25.85 23.61 11.86

10.988

Panama

10.48

10.960 6.592 5.738 3.085 2.269 2.053 3.891

USA, PR, USVI Belize Spain Germany Belgium Trinidad Tobago Others

Source: Calculation based on WITS database

Per cent to total world trade of beverages and tobacco

5.23 4.10 3.66 2.07 1.77 1.62 9.75

Annexures

196

 nnexure 6.12 Top Ten Trading Partners in Beverages A and Tobacco of Uruguay (1995 and 2015)

1995 World (Trade value, $000) Brazil Paraguay Argentina Areas nes United Kingdom Sweden Japan Canada BelgiumLuxembourg France+Monac Others

Per cent to total world trade of beverages and tobacco 2015 World (Trade 9654 value, $000) 51.604 Paraguay 18.364 Brazil 16.640 Chile 9.675 Curacao 1.645 Panama 0.535 Free Zones 0.507 USA, PR, USVI 0.454 Aruba 0.353 Japan

63,895 52.60 18.62 5.69 5.60 2.98 2.88 2.63 2.15 0.86

0.083 0.139

0.76 5.23

Mexico

Per cent to total world trade of beverages and tobacco

Source: Calculation based on WITS database

 nnexure 6.13 Top Ten Trading Partners in Mineral Fuel A of Argentina (1995 and 2015)

1995 World (Trade value, $000) Brazil Chile USA, P. Rico Uruguay Paraguay Netherlands Colombia Bolivia Asia Othr.ns Spain Others

Per cent to total world trade of mineral fuel 2,167,499 34.013 27.816 14.926 5.201 4.370 3.628 2.841 2.268 1.856 0.695 2.387

Source: Calculation based on WITS database

2015 World (Trade value, $000) USA, PR, USVI Brazil China Paraguay Chile Curacao India Uruguay Bolivia Senegal Others

Per cent to total world trade of mineral fuel 1,443,685 26.18 19.31 12.15 11.32 9.52 7.21 5.01 3.40 1.28 0.89 3.72

Annexures

197

 nnexure 6.14 Top ten Trading Partners in Mineral Fuel A of Brazil (1995 and 2015)

1995 World (Trade value, $000) USA, P. Rico Netherlands Antilles Argentina Paraguay Uruguay United Kingdom Netherlands Nigeria Italy Colombia Others

Per cent to total world trade of mineral fuel 410,643 24.616 21.825 9.498 9.352 7.482 4.648 3.062 2.590 2.355 2.286 12.287

2015 World (Trade value, $000) China USA, PR, USVI Uruguay Chile India St. Lucia Bahamas Netherlands Curaçao Singapore Others

Per cent to total world trade of mineral fuel 13,747,977 30.13 15.97 9.56 8.19 8.06 4.88 4.10 3.71 3.62 3.03 8.75

Source: Calculation based on WITS database

 nnexure 6.15 Top Ten Trading Partners in Mineral Fuel A of Paraguay (1995 and 2015)

1995 World (Trade value, $000) Argentina

Per cent to total world trade of mineral fuel 2015 1859 World (Trade value, $000) 100 Brazil Argentina Bolivia Uruguay USA, PR, USVI Qatar Russian Fed Mexico Montserrat

Source: Calculation based on WITS database

Per cent to total world trade of mineral fuel 79.38 19.79 0.43 0.33 0.04 0.01 0.00 0.00 0.00

Annexures

198

 nnexure 6.16 Top Ten Trading Partners in Mineral Fuel A of Uruguay (1995 and 2015)

1995 World (Trade value, $000) Argentina Brazil Sweden Areas nes

Per cent to total world trade of mineral fuel 22,031 84.080 12.344 3.522 0.055

2015 World (Trade value, $000) Argentina Paraguay USA, PR, USVI Malaysia Spain

Per cent to total world trade of mineral fuel 21,911 0.58 0.39 0.02 0.00 0.00

Source: Calculation based on WITS database

 nnexure 6.17 Top Ten Trading Partners in Manufactured A Goods of Argentina (1995 and 2015)

1995 World (Trade value, $000) Brazil USA, P. Rico Chile Italy Hong Kong Uruguay Japan Netherlands Paraguay Spain Others

Per cent to total world trade of manufactured goods 2015 World (Trade 2,870,351 value, $000) 18.522 Brazil 13.836 USA, PR, USVI 6.571 Chile 6.057 Uruguay 5.234 Mexico 4.440 Paraguay 4.016 Croatia 3.801 China 3.212 Japan 2.289 Canada 32.020 Others

Source: Calculation based on WITS database

Per cent to total world trade of manufactured goods 3,287,703 17.31 17.20 7.95 5.58 5.03 4.96 3.50 3.48 2.99 2.98 29.03

Annexures

199

 nnexure 6.18 Top Ten Trading Partners in Manufactured A Goods of Brazil (1995 and 2015)

1995 World (Trade value, $000) USA, P. Rico Japan Argentina Italy Netherlands Chile Germany Korea Rep. Paraguay United Kingdom Others

Per cent to total world trade of manufactured goods 2015 World (Trade 11,612,603 value, $000) 20.532 USA, PR, USVI 10.522 Argentina 8.421 Netherlands 3.845 China 3.449 Mexico 3.371 Japan 3.295 Italy 3.184 Germany 2.799 Colombia 2.647 Chile 37.935 Others

Per cent to total world trade of manufactured goods 24,877,081 24.48 9.57 9.56 8.40 3.18 2.90 2.86 2.48 2.26 2.09 32.23

Source: Calculation based on WITS database

 nnexure 6.19 Top Ten Trading Partners in Manufactured A Goods of Paraguay (1995 and 2015)

1995 World (Trade value, $000) Brazil Argentina USA, P. Rico France+Monac Italy Uruguay Chile Hong Kong Asia Othr.ns United Kingdom Others

Per cent to total world trade of manufactured goods 2015 World (Trade 134550.8 value, $000) 22.813 Brazil 18.752 Italy 13.497 Argentina 11.894 Thailand 11.101 USA, PR, USVI 4.762 China 4.066 Uruguay 2.741 Bolivia 2.317 Mexico 2.056 Germany 6.002 Others

Source: Calculation based on WITS database

Per cent to total world trade of manufactured goods 323,872 29.10 26.08 13.81 5.39 5.18 3.69 3.64 2.15 2.15 1.79 7.02

Annexures

200

 nnexure 6.20 Top Ten Trading Partners in Manufactured A Goods of Uruguay (1995 and 2015)

1995 World Brazil Argentina USA, P. Rico Germany Hong Kong Malaysia Italy Canada Asia Othr.ns United Kingdom Others

Per cent to total world trade of manufactured goods 388,209 24.982 14.537 11.531 6.498 5.286 4.442 3.436 2.992 2.513 2.455 21.328

2015 World Brazil Argentina Mexico USA, PR, USVI Germany Thailand China Paraguay Viet Nam Chile Others

Per cent to total world trade of manufactured goods 686377.8 20.61 16.57 12.96 11.59 9.63 5.26 3.88 3.35 1.93 1.54 12.69

Source: Calculation based on WITS database

 nnexure 6.21 An analysis of the Relation Between the Actual A and Expected Trade (Trade Ratio and Ratio of the Gravity Model Estimates) Figures 1, 2, 3, 4, 5, 6, 7 and 8 show relation between trade ratio and ratio of the gravity model estimates for Argentina, Brazil, Paraguay and Uruguay for 1990 and 2005,1 respectively. These figures illustrate the relation between actual and expected flow of trade between individual MERCOSUR members with each of their 84 trading partners. In 1990 and 2005, as seen in Figs. 1 and 2, Argentina is trading more with the geographically proximate countries, such as Canada, the United States, Brazil, Chile and Uruguay. However, in 2005, Brazil recorded lower actual trade than the estimate with Argentina. Contrary to this, over the years a large group of economies with varying nature registered lower actual trade than the estimated flow with Argentina. It could be broadly inferred from the pattern of spatial distribution of residuals that countries proximate to Argentina are trading more as compared to far off economies. Figure 3 shows in 1990, unlike Argentina, a diverse set of economies registered higher trade than the estimation with Brazil. Countries constituting this group are 1  It is to be noted in 1990 and 2005, correlation between trade ratio and ratio of the gravity model estimates has come out to be significant for all MERCOSUR members. For further analysis, regression is worked out considering ratio of the Gravity model estimates as dependent variable and trade ratio as independent variable.

Annexures

201

Fig. 1  Residual Mapping: Argentina, 1990

Fig. 2  Residual Mapping: Argentina, 2015

the United States, Canada, Argentina, Peru, Colombia, Denmark, France, Ireland, Spain, Cote D’Ivoire, Ghana, Senegal, Zambia, Zimbabwe, USSR (former), India, Sri Lanka and Vietnam. A group of few countries, for example, Chile, Paraguay, Uruguay, Venezuela, Germany, Hungary, Italy, Poland, Japan, Iran, Saudi Arabia, South Korea and Thailand, traded less than the expected trade with Brazil. Figure 4

202

Fig. 3  Residual Mapping: Brazil, 1990

Fig. 4  Residual Mapping: Brazil, 2015

Annexures

Annexures

Fig. 5  Residual Mapping: Paraguay, 1990

Fig. 6  Residual Mapping: Paraguay, 2015

203

204

Fig. 7  Residual Mapping: Uruguay, 1990

Fig. 8  Residual Mapping: Uruguay, 2015

Annexures

Annexures

205

sheds light on the changing spatial pattern of relation between actual trade and the estimated flow of trade between Brazil and its trading partners. Economies with actual trade more than the expected trade with Brazil are the United States, Canada, Colombia, Peru and Uruguay, eastern European economies, Cote d’Ivoire, Ghana, Senegal, Zambia, Zimbabwe and New Zealand. Furthermore, economies such as the USSR (former), India, Argentina, Bolivia, Mexico and Paraguay recorded less actual trade than expected with Brazil in 2005. Figure 5 illustrates in 1990, large number of economies spread across America, eastern Europe, South and Southwestern Asia, Oceania, and also few of the African economies (Angola, Cameroon, Ghana, Nigeria and Senegal) registered more actual trade than expected with Paraguay. On the other hand, a handful of economies, Belgium-Luxembourg, Brazil, Chile, France, Germany, Great Britain, Italy, Portugal, Japan and South Korea, recorded less actual trade with the Paraguayan economy than the expected trade flow. In 2005, as shown in Fig. 6, Paraguay has increasingly recorded more trade than the expected with economies that are geographically or geopolitically close. In this context, it could also be noted here that distance still matters and concepts like complementarity, transferability and intervening opportunities still hold significance, however, depending upon the nature of economies trading. Figure 7 shows in 1990 a few developing countries, such as Argentina, Honduras, Panama, Cameroon, Cote d’Ivoire, Gabon, Ghana, Senegal, India, Pakistan, Mongolia, Indonesia and Vietnam, recorded more trade than the expected level with Uruguay. It is worth mentioning that the export basket of these economies is tilted towards primary commodities. On the other hand, the majority of countries spread across globe recorded lesser trade than the estimated trade flow. Argentina is the only MERCOSUR member to trade more than the estimate with Uruguay. This might mainly be on account of the fact that these two economies share very close economic and political ties that gave impetus to trade between them. In 2005, a more or less similar pattern of relation between Uruguay and its trading partners can be noted (Fig. 8).