The Rise of Technological Power in the South

The Rise of Technological Power in the South 10.1057/9780230276123 - The Rise of Technological Power in the South, Edi...

Author: Xiaolan Fu | Luc Soete | eds.

79 downloads 2095 Views 8MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

The Rise of Technological Power in the South

10.1057/9780230276123 - The Rise of Technological Power in the South, Edited by Xiaolan Fu and Luc Soete

Also by Xiaolan Fu EXPORTS, FOREIGN DIRECT INVESTMENT AND ECONOMIC DEVELOPMENT IN CHINA

Also by Luc Soete UNDERSTANDING THE DYNAMICS OF A KNOWLEDGE ECONOMY (co-editor) THE ECONOMICS OF THE DIGITAL SOCIETY (co-editor) TECHNOLOGY AND THE FUTURE OF EUROPEAN EMPLOYMENT (co-editor) THE ECONOMICS OF INDUSTRIAL INNOVATION (co-author) THE ECONOMICS OF GROWTH AND TECHNOLOGICAL CHANGE (co-author) WORK FOR ALL OR MASS UNEMPLOYMENT (co-author) THE ECONOMICS OF TECHNOLOGICAL CHANGE AND INTERNATIONAL TRADE (co-author) NEW EXPLORATIONS IN THE ECONOMICS OF TECHNOLOGICAL CHANGE (co-editor) TECHNICAL CHANGE AND ECONOMIC THEORY (co-author) TECHNICAL CHANGE AND FULL EMPLOYMENT (co-author) UNEMPLOYMENT AND TECHNICAL INNOVATION: A Study of Long Waves and Economic Development (co-editor)


The Rise of Technological Power in the South Edited By

Xiaolan Fu and

Luc Soete


Selection and editorial matter © Xiaolan Fu and Luc Soete 2010 Individual chapters © Contributors 2010 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6-10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2010 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN: 978–0–230–23840–4 hardback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. 10 9 8 7 6 5 4 3 2 1 19 18 17 16 15 14 13 12 11 10 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne


Contents List of Figures

viii

List of Tables

x

Notes on Contributors

xiv

Introduction Xiaolan Fu and Luc Soete

1

Part I Policy, Strategy and Catch-up: Cross-Country Analysis 1

2

3

Innovation Strategies in Brazil, China and India: From Imitation to Deepening Technological Capability in the South Carl Dahlman

15

Economic Growth and Technological Capabilities in BRICS: Implications for Latecomers to Industrialization Deepak Nayyar

49

The Changing Geography of Innovation Activities: What do Patents Indicators Imply? Xuan Li and Yogesh A. Pai

69

Part II Policy, Strategy and Catch-up: Country Case Studies 4

5

6

China’s Catch-up and Innovation Model: A Case of the IT industry Xielin Liu

89

Science and Technology and Economic Growth in South Africa: Performance and Prospects David Kaplan

107

Market-Oriented Reforms, Domestic Technological Capabilities and Economic Development in Latin America Jorge Katz

125

v


vi

Contents

Part III Innovation Systems and Technological Capabilities 7

8

9

The Finance of Innovative Investment in Emerging Economies Jörg Mayer

147

A Comprehensive Model of Technological Learning: Empirical Research on the Chinese Manufacturing Sector Jin Chen, Xiaoyu Pu and Haihua Shen

170

The Innovation of SMEs and Development of Industrial Clusters in China Jinmin Wang

186

Part IV

10

Foreign Direct Investment and Technology Transfer

FDI, R&D and Innovation Output in the Chinese Automobile Industry Chen Fang and Pierre Mohnen

11 The Role of FDI in the Development of Innovative Capacity: The Case of Russian Companies Juha Väätänen, Daria Podmetina and Marina Aleksandrova 12

Human Capital and Technological Spillovers from FDI in the Chinese Regions: A Threshold Approach Miao Fu and Tieli Li

13 Transnational Corporations from Emerging Economies and South-South FDI Torbjörn Fredriksson

203

221

238

258

Part V Technology and Sustainable Development 14

15

Technological Competences in Sustainability Technologies in the BRICS Countries Rainer Walz Coordination, Convergence or Contradiction: Information and Communication Technologies for Integration and Development in Southern Africa and the Southern Cone Patience I. Akpan-Obong and Mary Jane C. Parmentier

281

300


Contents vii

16

Sustainability of Technology-intensive Social Innovation in India: The Role of Absorptive Capacity and Complementary Assets Xiaolan Fu and Christine Polzin

320

Conclusions: Science, Technology and Development – Emerging Concepts and Visions Xiaolan Fu, Luc Soete and Lina Sönne

341

Index

355


Figures I.1 1.1 1.2 1.3

3.1 3.2 6.1 8.1 8.2 8.3 9.1 12.1 13.1

13.2

13.3 13.4

13.5 14.1 14.2

R&D expenditure in 2005 (circles reflect amount of spending in US$ billions purchasing power parity (PPP)) GNI/per capita growth (1980–2008), per capita income (PPP)and size of economies (in nominal US$, 2008) FDI Inflows as a percentage of GDP in Brazil, China and India 1980–2006 Relative R&D expenditure and ratio of scientists and engineers per 1000 (expenditure in 2006 in US$ billions PPP) Weekly USPTO utility patent grants, 2006–8 US patent grants by technology category, 1981–99 Domestic learning and the evolution of local technological capabilities The model: Technological learning’s influence on innovation performance Original model of the research The adjusted model of influence factors The specialized wholesale market and the growth of local industrial clusters Confidence interval construction in double threshold model: All regions Number of entries from developing and transition economies among the list of Fortune Global 500 companies, 1990, 2005–9 Number of developing and transition economies with a stock of outward FDI valued at more than US$5 billion, 1980, 1990, 2000, 2006 and 2008 Outward FDI from the BRICS, 1995–2008 (US$ billions) Cross-border M&A purchases by companies based in developing or transition economies, 2000–6 (US$ millions) Oil and gas production of selected TNCs outside their home country, 2005 The influence of selected sustainability technologies on environmental themes Scheme of a system of sustainability innovations

3 19 24

33 80 82 141 178 180 181 196 247

259

260 260

263 265 283 286

viii


Figures

Results according to survey data from the WEF (2006) on general innovation conditions in the BRICS countries 14.4 Share of BRICS countries in patents and world exports for sustainability-related technologies, 2000–4 14.5 Specialization pattern of BRICS countries for sustainability technologies 14.6 Specializations of the BRICS countries within the analysed sustainability fields 15.1 An integrated theoretical model representing the interrelationships between regional integration, ICTs and development and socio-economic development

ix

14.3

288 290 291 294

303


Tables I.1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10

2.1

2.2 2.3 3.1 3.3 3.2 3.4 3.5 4.1

Overview of the BRICS countries, 2007 Basic economic indicators in Brazil, China and India Main country characteristics Acquiring foreign knowledge Literacy and enrolment rates by level of education 1980–2007 University enrolments 2007 – 12 countries with most students Higher educational attainment in non-OECD countries, against OECD R&D inputs and outputs Diffusion of technology Gross fixed investment as a percentage of GDP 1980–2007 Percentage shares of size of economy and population and of various estimates of technological capability for Brazil, China and India in the world Brazil, India, China and South Africa in the world economy: Share in world population and world GDP, 1820–2001 GDP, population and GDP per capita: Brazil, India, China and South Africa, 2000 and 2005 Growth performance of Brazil, India, China and South Africa, 1951–80 and 1981–2005 Top five patented technologies (domestic applicants), 2005 Comparison between applications from Toyota and Tsinghua University, 2005 Top five patented technologies (foreign applicants), 2005 Utility models, plant variety and designs in the definition of patents A comparison of US, EU, Brazilian and Indian legal provisions concerning the grant of patents Expenditure of in-house R&D and technology importation and assimilation (in 100 million RMB)

2 18 21 22 28 29 30 32 38 41

42

50 52 53 73 74 74 76 78 92

x


Tables xi

4.2 Ratio of R&D/sales in large and medium-sized companies 4.3 Overview of the Chinese IT industry 4.4 R&D outsourcing for universities and R&D institutes from large and medium-sized industrial enterprises 4.5 Selected M&A deals by Chinese firms in the IT industry (2001–5) 4.6 National S&T programmes (in 0.1 billion RMB) 4.7 Outline of foreign-related companies in the IT industry in China 4.8 Some indicators of catching up in the telecommunications industry 4.9 Sales and R&D expenditure of Huawei and ZTE (2001–5) 5.1 Funders and performers of R&D, 2004/05 (R millions) 5.2 Patents of South African origin granted by the US Patent and Trademark Office (USPTO), 1994–2007 5.3 Patents of South African origin in the PCT 1994–2007 5.4 Patent applications at the Companies and Intellectual Property Office (CIPRO), South Africa, 1994–2005 5.5 Number and share of South African and foreign patent applications (filed and granted) at CIPRO, 2000–2 and 2004–6 5.6 Patents filed in South Africa, by country of origin, 2004–6 5.7 Share of global high-tech exports 1992–2005, and share of high-tech exports in national exports 1992–2002: South Africa, Brazil and Argentina 5.8 Knowledge Economy Index (KEI): South Africa in comparative perspective, 1995–Latest Year 5.9 FTE researchers by sector, 1992–2004 6.1 Structural changes in Latin America, 1970–2002 6.2 Soya-bean production in Argentina, 1973–4 and 1993–4 6.3 The evolution of salmon farming in Chile, 1960–2000 6.4 R&D expenditure as a percentage of GDP 7.1 Sources of innovative investment finance, selected countries, 1999–2006 7.2 Sources of innovative investment finance, 2002–2006 7.3 Financial determinants of innovative investment 7.4 Sources of innovative investment finance, Brazil, 2003 7.5 Sources of innovative investment finance, China, 2003 8.1 Reliability analysis of variables 8.2 Effect decomposition and key factor identification

93 95 98 99 100 101 102 103 108 111 111 111

112 112

113 113 115 131 134 137 140 152 158 160 162 164 179 182


xii Tables

9.1 10.1 10.2 10.3

10.4

10.5 10.6 10.7 10.8 10.9 11.1 11.2 11.3 11.4 11.6 11.5 11.7 11.8 11.9 11.10 11A.1 12.1 12.2 12.3 12.4 12.5

The turnover of Yiwu China Commodities City Group Co., Ltd (RMB100 million) Explanatory variables introduced in the two generalized tobit models Data cleaning: Chinese automobile industry, firm-level data, 2002–3, 2005–6 Innovation indicators by type of ownership, Chinese automobileindustry, firm-level data, 2002–3 and 2005–6 Innovation indicators by industry, Chinese automobile industry, firm-level data, 2002–3 and 2005–6 R&D propensity R&D intensity for firms with R&D Innovation propensity Innovation intensity for firms with new products Generalized tobit estimation of R&D efforts in the Chinese automobile industry Companies with foreign participation, share of GRP 2005 (%) FDI as a share of gross fixed capital formation (GFCF) and GRP in 2004 (%) Industry sectors, R&D expenditure and foreign ownership Financial indicators R&D operations Test results on labour productivity Test results on R&D expenditure/sales New product development (NPD) Test results on NPD Test results on patents/R&D personnel The number of companies with foreign participation, 2005 Summary statistics of variables Likelihood ratio test for threshold effects Threshold estimates and their 95 per cent confidence intervals Threshold regression results for knowledgeproduction model: All regions and Mid-west China Human capital in China: Average promotion rates and proportion of educated workforce

191 208 209

210

211 212 213 213 214 215 222 223 228 228 229 229 230 231 231 232 235 246 247 247 248 250


Tables xiii

12.6 Percentage of provinces in the three regimes separated by thresholds 13.1 Number of companies from developing and transition economies among the Fortune Global 500 companies, 2007–9 13.2 FDI from developing and transition economies in selected LDCs, various years (%) 16.1 Main differences between business and social innovations 16.2 Three cases of social innovation

252

261 269 323 330


Contributors Patience Akpan-Obong is Assistant Professor at the School of Applied Arts and Sciences of Arizona State University (ASU), Polytechnic Campus. She holds a Ph.D. in Political Science from the University of Alberta, Edmonton, Canada (2003) and a Master’s in Journalism from Carleton University, Ottawa, Canada (1996). She did her undergraduate work in Mass Communication at The Polytechnic, Calabar, in Nigeria, before going to the University of Toronto for the Gordon Fisher Fellowship in 1994. Dr Akpan-Obong worked as a journalist in various positions with newspapers and magazines for several years in Nigeria and Canada. Before becoming a tenure-track Assistant Professor in the autumn of 2006, she had taught at ASU since 2003 as Adjunct Professor at both the West and Polytechnic campuses. Prior to her time at ASU, she taught at the University of Alberta in Edmonton, Canada, as a Sessional Lecturer. Dr Akpan-Obong has taught courses in globalization, Third World politics, social change, international relations, world politics, democratisation and women in contemporary society. Her research interests focus on the role of information and communication technologies in socio-economic development, the intersection of globalization and information technology, the politics of underdevelopment and gender and information technology. Dr Akpan-Obong’s research in these areas has been published in peer-reviewed academic journals and in book chapters. She also has two unpublished works of fiction and a collection of poems. Marina Aleksandrova is a researcher at Lappeenranta University of Technology, Finland, and works for the project ‘Innovativeness of Russian high-tech industries’. Her research interests cover corporate governance, privatization and nationalization in Russia, competitiveness and innovativeness, the role of government in the economy in transition and the development of state-owned companies and their role in the Russian economy. Jin Chen is Professor of Management at the College of Administration Management, Zhejiang University. He is Dean of the Research Centre for Science, Technology & Education Policy (RCSTEP). He is also the Deputy Director of the National Institute for Innovation Management, Zhejiang University. He received his Ph.D. in 1994, and in 1998 he was a Visiting xiv


Contributors xv

Scholar at the Alfred Sloan School of Management, MIT. He also visited SPRU at the University of Sussex, among other institutions. Dr Chen is a committee member of Education of the Chinese Academy of Engineering, Research Association of Science and Technology Policy of China. He is also a member of the Institute of Electric and Electronic Engineers, Inc. (IEEE) and the American Society for Engineering Education (ASEE). His research areas include technology and innovation management, strategic management, human resource management, sustained development and Chinese S&T policy. He is a Principal Investigator of a variety of research projects in China as well as abroad. The projects he is working on include managing complex production systems, managing radical innovation in China, the internationalization of technological innovation for Chinese enterprises, human resources issues for technical innovation and system dynamics and technology management. He has published several books and more than 180 papers on management of technology and innovation. Journal his work has appeared in include IEEE Transactions on Engineering Management, Technovation and the International Journal of Innovation and Entrepreneurship. He was awarded the ‘Excellent Young Teacher Award’ and the ‘Huo Yingdong Prize by the Ministry of Education, of the People’s Republic of China, as well as the ‘Excellent Youth Fund’ by the National Science Foundation of China. Carl Dahlman is the Henry R. Luce Professor of International Relations and Information Technology at Georgetown University’s Edmund A. Walsh School of Foreign Service. He earned a BA magna cum laude in International Relations from Princeton University and a Ph.D. in Economics from Yale University, and has also taught courses at Columbia University’s School of International and Public Affairs. He worked for more than 25 years at the World Bank, where he was Senior Advisor to the World Bank Institute and managed the Knowledge for Development (K4D) Program. He had previously served as Staff Director of the 1998–1999 World Development Report, Knowledge for Development, and as Resident Representative and Financial Sector Leader in Mexico between 1994 and 1997 – during one of the biggest financial crises to hit the nation. Before his position in Mexico, Dahlman led divisions in the World Bank’s Private Sector Development, and Industry and Energy Departments. He conducted extensive analytical work in major developing countries, including Argentina, Brazil, Chile, Mexico, Russia, Turkey, India, Pakistan, China, Korea, Malaysia, Philippines, Thailand and Vietnam. His publications include China and the Knowledge Economy:


xvi Contributors

Seizing the 21st Century; Korea and the Knowledge-Based Economy: Making the Transition and India and the Knowledge Economy; Leveraging Strengths and Opportunities. He is currently finishing a knowledge economy study on Mexico, working on a book on the challenge of the knowledge economy for education and training in China and collaborating with research teams in Finland, Japan and Korea to produce books on each country’s development strategies. Chen Fang is Assistant Professor at the Institute of Policy and Management, Chinese Academy of Sciences (CAS), a researcher at the CAS Center for China’s Innovation and Development and the CAS Center for Research and Training of China’s Intellectual Property Rights. Her research areas are industrial innovation, S&T statistics and policy evaluation. She has been working at the Institute of Policy and Management since July, 2008. She has a Ph.D. and an MA in Economics from the Renmin University of China and a BA in economics from the Henan Institute of Finance and Economics. Torbjörn Fredriksson is Officer-in-Charge of the Policy Issues Section at UNCTAD, and one of the main authors of the annual World Investment Report. Before joining UNCTAD, Mr Fredriksson was Head of Research at the Invest in Sweden Agency from 1996 to 2000. As such, he was part of the management team and in charge of economic and policy analysis related to FDI and TNC activities. He has also been Head of Section at the Ministry of Industry and Commerce in Sweden and has done research at the Industrial Institute for Economic and Social Research in Stockholm (IUI). Mr Fredriksson has published articles on multinational enterprises in the Journal of International Business Studies, the International Journal of Industrial Organization and the Journal of World Investment, among others. Miao Fu completed his undergraduate studies at Peking University and attained a Doctorate in Quantitative Economics from Huazhong University of Science and Technology. His work has mainly been in the areas of spatial econometrics, GIS modelling, sustainable development and technology spillovers. He is currently an Associate Professor of Economics at Guangdong University of Foreign Studies. Dr Fu has published 11 journal articles found in the Chinese Social Science Citation Index, three of which are in journals issued by the Chinese Academy of Social Sciences or associations at the national level. He is currently the principal investigator of an HSSR Project in econometrics funded by the Ministry of Education of China.


Contributors xvii

Xiaolan Fu is Director of the Sanjaya Lall Programme for Technology and Management for Development, and Fellow of Green-Templeton College, at the University of Oxford. She is also project co-leader of several EPSRC- and ESRC-funded projects and a Senior Research Associate at the Judge Business School of the University of Cambridge. Dr Fu is the author of Exports, Foreign Direct Investment and Economic Development in China and serves on the Editorial Board of Oxford Development Studies and the International Journal of Technological Learning, Innovation and Development. Her papers appear widely in major international journals. She received the European Commission Gate2Growth Academic Network 2005 ‘European Best Paper’ Award, and has conducted consultancy research for ILO, UNCTAD, UNIDO, UKTI and the Chinese government. Dr Fu serves on the Advisory Expert Group for the OECD Global Investment Forum and the Board of Directors of the Chinese Economic Association (UK/Europe). David Kaplan is Professor of Business–Government Relations and Professor at the Department of Economics, University of Cape Town (UCT). He holds a D.Phil. from the University of Sussex, an MA from the University of Kent and a BAB in Communications from UCT. He has extensive experience in working with and in government, particularly the Departments of Science and Technology (DST) and of Trade and Industry (DTI). He was Chief Economist at the DTI (2000–3), and is currently Chief Economist (part-time) at the Department of Economic Development and Tourism in the Western Cape. He has served for four years as a member of the National Advisory Council on Innovation (NACI). With support from the IRD in France, he established the South African Network of Skills Abroad (SANSA), a network to engage South Africans located abroad in utilizing their skills to support development in South Africa. At UCT he founded the Science and Technology Research Centre (STPRC) in 1994 (with the support of the IDRC, Canada) and, together with two colleagues, he established the Development Policy Research Unit in 1995. He was Director at the STPRC and the DPRU for much of the 1995–2000 period, before joining the DTI. Dr Kaplan is concerned with policyoriented research, mainly in the fields of industrial and technology policy, and particularly with how governments and business can collectively develop policies and strategies that can best enhance development. A central thrust of his research and teaching focuses on innovation within South African companies, and how it is affected by the policy environment.


xviii

Contributors

Jorge Katz was born in Buenos Aires, Argentina, where he received his first degree in Economics from the University of Buenos Aires. He also holds a D.Phil. in Economics from Nuffield College, the University of Oxford. His doctoral dissertation ‘Production Functions, Foreign Investment and Growth’, was published by North Holland Publishing Company in 1969. After returning to Argentina, he was appointed Professor of Economics at the University of Buenos Aires, where he lectured on Industrial Organization and Economic Development in Argentina for nearly two decades. During that period, he conducted a long-term research project for ECLAC and the Inter-American Development Bank, on the Development of Domestic Technological Capabilities in Latin America. The results of this project have been published in numerous books and articles in Spanish and English. In 1994 he was appointed Director of the Division of Production, Productivity and Management at ECLAC, Santiago de Chile, a post he held until his retirement from the UN System in 2003. Since then, he has been based at the University of Chile as Professor of Economic Growth and Innovation. His publications include 18 books on Latin American industrial and technological issues and almost 80 papers in Spanish and English internationally reviewed journals. Tieli Li holds a Doctorate in Science from the School of City and Environmental Science of Northeast Normal University, China. He is currently the Dean of the School of International Economics and Trade, at the Guangdong University of Foreign Studies, China. He is also a Professor, and has taught courses for undergraduate students – including microeconomics, macroeconomics, international trade, industrial economics and regional economics – as well as for postgraduate students – such as studies of trade within regional groups, studies of Sino-European Economic and Trade Relations and others. His current research interests are related to regional economic development issues, the relations between regional innovation capacity and economic development, the appraisal of regional innovation capacity, impacts of direct foreign investments on local technology progress, the relations between regional set-up and cities, the issue of centre versus periphery, studies of regional economic integration, studies of the economy and trade in Hong Kong, Macao and Taiwan, among others. Xuan Li is Programme Coordinator of the Innovation, Access to Knowledge and Intellectual Property Programme (IAKP) at the South Centre, Geneva. She is Director of the South Centre Distance Learning Course on ‘Intellectual Property Policy and Development’,


Contributors xix

jointly implemented with the United Nations Institute for Training and Research (UNITAR). She is a member of the editorial committee of the South Bulletin: Reflections and Foresights. She has also been working as a Consultant for a number of national and intergovernmental organizations, including the World Health Organization, the Food and Agriculture Organization, the Chinese Ministry of Commerce, the Chinese State Intellectual Property Office, the State Administration of Traditional Chinese Medicine and the World Trade Institute, among others. She has been working on international trade, intellectual property and development issues for the Asian Development Bank, ICTSD, and was the Lead Economist of the South Centre. She has also been actively involved in the Chinese National Strategic Formulation on Intellectual Property Rights. Xielin Liu is Professor and Director of the Research Centre of Management of Information and Innovation, Graduate University of the Chinese Academy of Science, Beijing. He has been a visiting fellow at the Sloan School of Management (MIT), the Hitochibashi University of Japan and the Stockholm School of Economics. Dr Xielin holds a Ph.D. from Tsinghua University, an M.Sc. from the Chinese Academy of Science and a B.Sc. from Peking University. He was also a Professor at the National Research Center for Science & Technology for Development (Ministry of Science and Technology), and acts as Vice-President of the Chinese Association of Science and Science & Technology Policy. His research areas are mainly innovation policy, management of technology and innovation. He has published several papers in Research Policy, Technovation, Journal of Management Studies and International Journal of Technology Management, and has also written eight books in the last ten years. Jörg Mayer holds a degree in Economics from the University of Freiburg (Germany) and a Ph.D. in Economics from the European University Institute in Florence (Italy). He served as an economist for the German Central Bank in Frankfurt, and also worked for the International Monetary Fund in Washington, D.C. and the Centre for European Policy Studies in Brussels. In 1991, he joined UNCTAD, and since 1997 he has been part of the team that prepares the organization’s Trade and Development Report. His main work areas include development strategies, the pattern of international trade and the interrelationships between trade, technology transfer, investment and growth, on which he has published several articles in refereed journals and chapters in edited books.


xx

Contributors

Pierre Mohnen is Professor of Microeconometrics at the School of Business and Economics of Maastricht University and Professorial Fellow at UNU-MERIT (Maastricht Economic and social Research and training center on Innovation and Technology). He is also Associated Fellow at CIRANO, the Montreal-based Center for Interuniversity Research and the Analysis on Organizations. He has a BA and an MA in Economics from the Catholic University of Louvain and a Ph.D. in Economics from New York University. He was Professor of Economics at the University of Québec in Montréal (UQAM) from 1984 to 2001. His research areas are the economics of production, applied econometrics, productivity and innovation. Deepak Nayyar is Professor of Economics at Jawaharlal Nehru University, New Delhi, and has previously taught economics at the University of Oxford, the University of Sussex and the Indian Institute of Management, Calcutta. He was also Vice-Chancellor of the University of Delhi (2000–5). Dr Nayyar was educated at St Stephen’s College, University of Delhi, and as a Rhodes Scholar he studied at Balliol College, University of Oxford, where he obtained a B.Phil. and a D.Phil. in Economics. He has received the VKRV Rao award for his contribution to research in economics, and his distinguished academic career has been interspersed with short periods in government. He worked for the Indian Administrative Service, and later as an Economic Adviser in the Ministry of Commerce, Chief Economic Adviser to the Government of India and Secretary in the Ministry of Finance. Dr Nayyar is Chairman of the Board of Governors of the World Institute for Development Economics Research, UNU-WIDER (Helsinki). He served on the Board of Directors of the Social Science Research Council in the United States (2001–7) and as Chairman of the Advisory Council for the Department of International Development, Queen Elizabeth House, University of Oxford (2004–7). He has been President of the Indian Economic Association and is also on the editorial boards of several professional journals. Dr Nayyar is a member of the National Knowledge Commission in India, as well as of the World Commission on the Social Dimension of Globalization. His research interests are primarily in the areas of international economics, macroeconomics and development economics. He has published papers and books on a wide range of subjects, including trade policies, industrialization strategies, macroeconomic stabilization, structural adjustment, economic liberalization, trade theory, macro policies, international migration and the multilateral trading system. In addition, he has written extensively on economic development in India, and globalization and development is an area of focus in his present research.


Contributors xxi

Yogesh A. Pai is currently working as Associate Fellow at the Centre for Trade and Development based in New Delhi. He has a LLM in IP Laws from the Cochin University of Science and Technology, India. His previous work is in the area of software patents, software copyrights, patent in standards, TRIPS and other international agreements. He is credited with a number of publications, and the article included in this book was part of his research at the South Centre, Geneva. Mary Jane C. Parmentier is a Lecturer at the School of Applied Arts and Sciences of Arizona State University (ASU), Polytechnic Campus. She received her BA in Spanish from Southern Connecticut State University in New Haven, Connecticut. She also holds an MA in International Relations from San Francisco State University in San Francisco, California, and a Ph.D. in International Studies from the Graduate School of International Studies at the University of Denver, in Colorado. Dr Parmentier has been at ASU’s Polytechnic Campus since 1999. She was a founding faculty member and continues to co-direct and teach in the Global Technology and Development (GTD) programme (an MS in Technology degree in the College of Technology and Innovation). Dr Parmentier teaches political science and international politics classes in Social and Behavioral Sciences (SBS), and is part of the SBS faculty team that is designing new academic programmes in the social sciences for the Polytechnic Campus. Dr Parmentier’s past research focused on religion and political change in North Africa, and her current work is in the area of technology and socio-economic and political development, particularly in Latin America. She is proficient in Spanish, French and North African Arabic, and has lived in Spain and Morocco. Daria Podmetina works as project manager at Lappeenranta University of Technology, Finland. Her main research project deals with the innovativeness of Russian companies. Her research interests also include emerging economies, internationalization (inward and outward FDIs, mergers & acquisitions, marketing strategies, exporting and importing), R&D, innovations, technology transfer and competitiveness issues. Christine Polzin is a coordinator for the Emergia Institute’s research project on biofuels since 2008. Her interests and expertise lie in sustainability strategies, energy and climate change. She worked as a Research Officer at the University of Oxford on sustainable business models for technology-intensive social innovations in developing countries. Christine conducted field research in Africa and Asia and worked on assignments for various organizations including GTZ, KfW and the


xxii

Contributors

German Foreign Office. She holds a Master’s degree in International Economics and Business Studies from Reims Management School and the European University Viadrina, as well as an M.Phil. in Development Studies from the University of Oxford. Xiaoyu Pu is currently a Ph.D. candidate at the School of Management, Zhejiang University. Her research focus is technology innovation management. Haihua Shen is a business development representative at the AFS Department, Dupont China. He got his Master’s in Technology and Innovation Management from the School of Management at Zhejiang University in 2006. Luc Soete is Director of UNU-MERIT (the United Nations UniversityMaastricht Economic and social Research and training centre on Innovation and Technology), which emerged out of the integration of UNU-INTECH (the United Nations University Institute for New Technologies) and the University of Maastricht Research Institute MERIT. He is also Professor of International Economic Relations (on leave) at the Faculty of Economics and Business Administration, University of Maastricht. Professor Soete was the founding director of MERIT, which he set up in 1988, and he oversaw the integration in 2005 of MERIT into UNU-INTECH to form the new research and training centre, UNUMERIT. He is a member of the Dutch scientific advisory body Adviesraad voor Wetenschap en Technologie (AWT). Before moving to Maastricht in 1986, he worked at the Department of Economics of the University of Antwerp (previously known as UFSIA), the Institute of Development Studies and the Science Policy Research Unit at the University of Sussex and the Department of Economics at Stanford University. Professor Soete completed his first degrees in Economics and Development Economics at the University of Ghent, Belgium, before obtaining his D.Phil. in Economics at the University of Sussex. His research interests cover the broad range of theoretical and empirical studies on the impact of technological change – in particular new information and communication technologies – on employment, economic growth and international trade and investment, as well as the related policy and measurement issues. Lina Sönne is a Researcher at UNU-MERIT, Maastricht, the Netherlands, focusing on pro-poor/inclusive/base of the pyramid innovation and entrepreneurship. She is currently completing a Ph.D. thesis concerned with how to finance entrepreneur-based innovation in rural


Contributors xxiii

India. Previously, Lina worked in the investment team of the Invention & Innovations Programme at the National Endowment for Science, Technology & the Arts (NESTA), London, UK. She has an M.Sc. in Business Economics from City University, London, as well as an MA (Hons) in European Studies. Juha Väätänen works as a Professor of International Operations and Transitional Economies at Lappeenranta University of Technology, Finland. His fields of expertise are international business operations, business environment in transitional economies, foreign direct investment, country and enterprise competitiveness and innovative capacity. His current research focuses on the innovativeness of Russian high-tech industries. Rainer Walz is the Deputy Head of the Competence Centre Sustainability and Infrastructure Systems, as well as Project Coordinator, at the Fraunhofer Institute for Systems and Innovations Research (FhGISI). He studied Economics and Political Science at the University of Freiburg and at Brock University, Ontario, Canada. As Research Fellow at the University of Wisconsin and as member of staff of the Enquête Commission ‘Protecting the Earth’s Atmosphere’ of the German Parliament, Dr Walz gained specific experience in international comparisons and the formulation of environmental protection policies and politics. In 1991, Dr Walz joined the FhG-ISI as Senior Scientist. He has been Deputy Head of the Competence Centre since 1996. His research areas and fields of publication include, among others, the economic effects of technological change and environmental policies, policies for climate and water protection and innovation for sustainable development. He lectures in the field of environmental economics. Jinmin Wang is Senior Lecturer in Strategic Management and International Business, Nottingham Business School, Nottingham Trent University, the United Kingdom. He is also a Lecturer at the College of Public Administration, Zhejiang University, Hangzhou, China. He was a Research Fellow at the Asia Research Centre, London School of Economics and Political Science in 2003–4. His research interests include innovation and the development of industrial clusters, trade and investment in the Asia-Pacific region and economic reform in China.


This page intentionally left blank


Introduction Xiaolan Fu and Luc Soete

The rise of the emerging economies, especially the Golden BRICS (Brazil, Russia, India, China and South Africa), is changing the landscape of the world economy. These economies have experienced fast economic growth in the past two decades (at least), and are emerging as important economic forces in the global economy. For the past three decades, the average annual gross domestic product (GDP) growth rate of China, for instance, has been as high as 9.8 per cent. This is much higher than the average 3.0 per cent annual growth rate of the world economy. In 2007, the annual GDP growth rate percentage was 13.0 in China, 9.1 in India, 8.1 in Russia, 5.4 in Brazil and 5.1 in South Africa. Again, this is much higher than the world average growth rate of 3.8 per cent in the same year. By 2007, China was ranked amongst the four largest economies in the world in terms of total GDP. The significance of the rise of the BRICS lies not in the pace and duration of economic growth; both Korea and Japan enjoyed a similar experience during the three decades after 1960 (Kaplinsky, 2006). What is significant, however, is the combination of this fast growth with the large size of the BRICS economies. In 2007, the BRICS countries accounted for 43 per cent of world population, 30 per cent of the global surface area, 13 per cent of world GDP and 12 per cent of global net foreign direct investment (FDI) inflow (Table I.1). The impact of the rise of the BRICS is likely to be much greater than that of the rise of the Asian Tigers. As one of the main drivers of national competitiveness, technological capabilities in these emerging economies have also grown significantly. The BRICS countries are catching up with the industrialized countries, especially in certain industries. In 2005, the total R&D expenditure in China was about a third of that in the EU as a whole. However, in terms 1


Table I.1

Overview of the BRICS countries, 2007

Population, total (millions) Surface area (sq. km) (thousands) GDP (current US$) (billions) GDP growth (annual %) Exports of goods and services (% of GDP) Imports of goods and services (% of GDP) Mobile phone subscriptions (per 100 people) Internet users (per 100 people) High-technology exports (% of manufactured exports) FDI (Balance of payments, current US$) (millions)

Brazil

Russia

India

China

South Africa

World

192 8515 1313 5.4 14

142 17,098 1290 8.1 30

1125 3287 1177 9.1 21

1318 9598 3206 13.0 42

48 1219 283 5.1 32

6610 133,946 54,584 3.8 28*

12

22

24

32

35

29*

63

115

21

42

88

51

35.2 12

21.1 7

7.2 5

16.1 30

8.3 6

21.8 18

34,585

55,073

22,950

138,413

5746

2,139,338

Note: *Data for 2006. Source: World Bank (2007).


BRICS as % of world 43 30 13

12

Introduction 3

of gross expenditure on R&D, China is now moving close to the EU average. In terms of research per 1000 employees, the Russian Federation in fact ranks higher than the EU average (Figure I.1). The extent of these activities in relation to technological capabilities raises several important research and policy questions. What are the innovation and technology policies that the BRICS countries have adopted? What role have they played in the catch-up process? What are the determinants of technological learning and innovations in these countries at the firm level? What are the roles of small and medium enterprises (SMEs)? Should a latecomer economy rely on foreign technology transfer or indigenous innovation for technology upgrading? What is the role of technology in confronting the challenges of environmental change and social division for sustainable development? The experiences of the BRICS have important implications for the world and will provide valuable lessons to other developing countries with regard to industrial, technological and trade policies.

12

Japan 131

Researchers per 1000 employment

United States 10

324

8 Russian Federation 17 6

EU27 231

4

China South Africa 4 71 14 India 24 Brazil

2

0 0

1

R&D expenditures in billions of current PPP (1) 2

3

4

GERD as % of GDP Figure I.1 R&D expenditure in 2005 (circles reflect amount of spending in US$ billions purchasing power parity (PPP)) Source: Tojo (2008).


4

Xiaolan Fu and Luc Soete

The BRICS are a group of the world’s largest developing and transition economies. Despite sharing the common features of large size and developing country status, the countries are diverse in their factor endowment, economic structure and development history and strategy. Cross-country international comparison will no doubt provide interesting findings, but this method alone may not be sufficient for in-depth analysis. Therefore, by using both cross-country comparisons and in-depth case studies of emerging economies, this book aims to: 1) explore the drivers of technological upgrading and catch-up in the emerging economies, especially in respect to technology and innovation policies, national innovation systems, the role of FDI and SMEs; 2) compare the similarities and differences between the emerging countries and draw policy and practical implications for other developing countries and 3) discuss emerging concepts and visions for future research in this area. The analyses are conducted at multiple levels, ranging from firm to sector, region and country level, and use both quantitative and qualitative research methods. This volume includes a selection of papers presented at the first Sanjaya Lall Programme Annual Conference in 2008, held at Oxford University. We used a three-tier scrutiny process, involving: 1) an initial selection by conference organizers; 2) a second-round internal review and selection by the editors and 3) a double-blind refereeing procedure. Seventy papers were chosen from a total of 150 submissions in the initial selection process. Around 30 papers were subsequently chosen in the second-round editorial selection. After receiving the referees’ reports, the editors decided to include 16 papers in the final volume. The book is divided into five parts. Parts I and II evaluate the role of innovation policy and strategy in the technological catch-up process of the emerging economies through cross-country comparative studies and in-depth country case studies, respectively. Part III investigates the role of national, regional and sectoral innovation systems and drivers of domestic technological capabilities. Part IV explores the role of inward and outward FDI in technology transfer. Part V discusses the role of technology for sustainable development in the emerging economies.

I.1

Policy, strategy and catch-up: Cross-country analysis

This part consists of three papers exploring the role of innovation policy and strategy in the technological catch-up process of the emerging economies through cross-country comparative studies. In Chapter 1, Dahlman examines the innovation strategies of three of the BRICS


Introduction 5

(Brazil, India and China) and compares their economic performance over the last 25 years. The chapter then extracts some implications regarding the link between innovation strategy and economic performance. In light of a theoretical framework that innovation in developing countries should not be defined just in terms of a shifting global technology frontier but rather in terms of what is new inside a country (including any policies and mechanisms which draw on both global knowledge and the domestic R&D effort) the chapter finds that the innovation strategies of these three countries have been quite different. India has been the most autarkic until recently. China has drawn the most on global knowledge, although more recently it has invested massively in its own R&D. Brazil can be placed somewhere in between. It has been almost as closed as India to trade, but more open than India in terms of FDI. Moreover, Brazil is falling behind both other countries in its domestic R&D effort. The three countries have also had very different growth performances over the last 25 years. Findings from this comparative analysis indicate the importance of tapping into global knowledge and using it effectively, as well as the significance of education, reverse engineering, diasporas, competition, stable macro conditions and strong efforts for technology diffusion. In Chapter 2, Nayyar analyses the implications of the rise of technological capabilities in Brazil, India, China and South Africa for developing countries, with particular focus on latecomers to industrialization. Starting with a review of the development history of the BRICS economies and focusing on the analysis of the nature of technological development in these emerging economies (foundations, dilemmas and specificities), Nayyar argues that the emerging economies are characterized by specificities. In the national context, the size of the economy matters as it determines the number of scientists and engineers and the size of the domestic market. In the international context, which is shared by all countries, there are specificities that characterize the emerging economies in terms of their capacity to exploit available opportunities. Domestic firms in Brazil, India, China and South Africa have such capabilities, which domestic firms in other developing countries may not. Moreover, as the emerging economies are also late industrializers, it is possible that their technologies are more appropriate for countries in the developing world. Therefore, there is much to learn from the experience of technological development in the emerging economies, but such learning should seek to contextualize rather than replicate. Despite widely accepted evidence of the changing geography of innovation activities based on patent numbers, in Chapter 3 Li and Pai urge


6


caution in using the conclusion of the WIPO Patent Report that the sharp rise in the number of patents filed in North-east Asia (and particularly in China) indicates a changing geography of innovation. The authors point out that the drawback of such an international comparison not only relates to how to interpret properly the figures on patent filings and ‘resident patent filings’, but also to high heterogeneity in the value of patents. On the basis of detailed analyses in China and comparison of the legal framework in the EU, US, Brazil and India, this chapter shows that hasty generalization of the changing geography of innovation patterns should be avoided. A full assessment requires further econometric, classificatory and survey research, followed by interdisciplinary interpretation. The way forwards must therefore be to develop a proper set of indicators to monitor the changes in innovation capacities, especially those in the developing countries.

I.2 Policy, strategy and catch-up: Country case studies This part consists of 3 papers analysing the role of innovation policy and strategy in the technological catch-up process of the emerging economies using in-depth country case studies. Taking into account the need to recognize country-specific contexts in any analysis, Liu provides an in-depth case study of the catch-up strategy of China in Chapter 4. By comparing the history and contexts in which firms attempted to catch-up in the information technology (IT) industries in China and Japan and by examining the different models that they have adopted, this chapter suggests that the context in which Chinese firms are attempting to catch-up is fundamentally different from that facing earlier latecomers such as Japan. It argues that market size, market-oriented innovation, participation in a global alliance, open innovation, spillovers of FDI and the role of government are the main elements of Chinese catch-up strategy. In Chapter 5, Kaplan presents another interesting case study of the science and technology policy of South Africa. Through a critical review of two recent reports on science and technology development in South Africa, this chapter provides a high-level assessment of the innovation performance of the South African system, identifies the constraints on innovation performance, and examines the government’s strategic goals and aspirations for the South African science and technology (S&T) system as detailed in a ten-year plan (DST, 2007). The chapter concludes by highlighting the importance of human capital and argues that rather than using a policy of attempting to advance on all fronts,


Introduction 7

priority needs to be given to development of high-level skills. Instead of attempting to do everything at once, policy needs to be sequenced, with the first priority being expansion of high-level skills for innovation. In addition to the factors that identified by Liu and Kaplan in earlier chapters, structural change is another important driver of technological upgrading of countries. In Chapter 6, Katz examines the linkage between market-oriented reforms, domestic technological capabilities and economic development in Latin America. Market-oriented structural reforms in the past three decades and the globalization of economic activities have brought major changes to Latin America inducing the gradual phasing-out of many industries and institutions of the ‘inward-oriented’ period of industrialization. Their production structure now features many new sectors of economic activity closer to their natural comparative advantages. A modern sector of economic activity has emerged in natural resource-processing activities, as well as in ‘maquila’-type industries and in service sectors catering for local demands in areas such as telecoms, banking and financial services, water provision and sanitation services, and others. Yet such structural transformation has not been strong enough to incorporate the vast majority of the population. Thus, the modernization process has occurred handin-hand with a significant expansion in the gap between the rich and poor segments of society. This has clearly resulted in a dramatic sequel of frustration and despair that makes political governability an increasingly difficult issue. Katz argues that the lack of ‘initial entitlements’ resulting from low quality education, poor health services, insufficient provision of public goods which might ‘level the playing field’, and different forms of market failure have been instrumental in causing market-oriented structural reforms to fail to deliver a broader pattern of improvement in economic efficiency.

I.3

Innovation systems and technological capabilities

There are three chapters in this part. A sustained rise in innovation presupposes high rates of investment in intangible elements such as education and R&D. For private investment to take place, firms not only need an incentive in terms of expected future profits, but they also need to have access to reliable, adequate and cost-effective sources of finance to fund their investment. In Chapter 7, Mayer analyses the role of different sources of finance for innovative investment and looks at the experience of the BRICS in this regard. The analysis draws on enterprise data to provide statistical evidence on the role of different sources of


8


investment finance, and examines the role of different sources of investment finance for innovation. Research in this chapter also points to the different approaches employed in the BRICS countries with regard to the model and sources of finance for innovative investment. In addition to effective financial support for innovation activities, effective technological learning is another important determinant of the catch-up process, especially for firms in developing countries. Developing countries, for example, China, often find that the core technologies used in domestic firms are still in the hands of foreign companies. Technological learning is therefore essential for these firms in order to develop indigenous innovation capability in the context of upgrading the structure of manufacturing industries. In Chapter 8, Chen, Pu and Shen examine the relationship between technological learning, technology capability and innovation performance using data collected from a sample of 92 Chinese firms. Their results highlight the importance of technological learning sources, contents and levels such as internal communications; a focus on the absorption of tacit knowledge; the enforcement of organizational learning from both internal and external sources and the strengthening of the motivation mechanism for technological learning. In any innovation system, the SMEs play an important role as these firms constitute the most dynamic part of an economy. In Chapter 9, Wang explores the innovation mechanisms underlying the SME industrial clusters in China using a case study of the Yiwu socks cluster in Zhejiang Province. The research in this chapter suggests that the growth and development of the Yiwu socks cluster is a typical example of the evolution of autonomous organizations based on market expansion, technological innovation and regional economic development. The strategic linkage between horizontal co-operation and technological innovation through the local specialized wholesale market and informal social network has generated a special institutional arrangement that can overcome the inadequate innovation incentives for local SMEs and contributes to the upgrading of industrial clusters along the global value chain.

I.4

Foreign direct investment and technology transfer

Technology transfer through FDI has for a long time been regarded as an important channel for technology upgrading in latecomer economies. Governments of developing countries have made great efforts to attract FDI with the expectation that such investment will lead to the


Introduction 9

transfer of advanced technology to their local firms. However, evidence from existing empirical studies is mixed. Research in this area is therefore inconclusive. Lall and Urata (2003) argue that advocates of globalization seem to assume that the private interests of multinationals do not diverge from the social interests of the host countries. Despite the expectation that there will be a ‘trading market for technology’, researchers find that it is difficult to acquire state-of-the-art technology through inflows of FDI and imports, and that huge inflows of FDI may even weaken indigenous industrial and technological capabilities (Aitken and Harrison, 1999; Wang and Gao, 2005). Given the rise of the BRICS and their emphasis on a more open policy toward FDI, it is important to examine the role of this in technology transfer and technological upgrading in these countries, reflecting upon any implications for other latecomer economies. China has absorbed a huge inflow of FDI since the economic reforms of 1978 and the country in fact ranks as the largest recipient of FDI in the developing world. In Chapter 10, Chen and Mohnen analyse the determinants of, and the interrelationships between, innovation input and innovation output. In particular, the authors examine whether FDI had any influence on these two aspects of innovation using firm-level data relating to the Chinese automobile industry. They employ a generalized Tobit model to estimate both R&D and the share of innovative sales for 2002/2003 and 2005/2006. The findings show that firms with FDI are less R&D-intensive, but, when they do innovate in new products, they are more innovative in their products than domestically funded firms. In Chapter 11, Väätänen, Podmetina and Aleksandrova investigate the role of FDI in the development of innovative capacity in Russian companies. The study is based on the survey of 176 R&D-oriented Russian companies conducted in early 2008. The sample is composed of companies which are active in innovation or which represent an industry with high innovation intensity. The survey results show that the labour productivity of foreign-owned companies is 10 per cent higher than domestic companies. Surprisingly, foreign companies have lower R&D expenditure as a percentage of sales: 6.2 per cent against 6.5 for local companies. Contrary to expectations, there were no significant performance differences between foreign and domestic companies in terms of innovative capacity, as measured by new product development or patent activity. The authors argue that the potential effect of FD on the development of innovative capacity of Russian companies remains limited.


10 Xiaolan Fu and Luc Soete

For technology transfer though FDI to be successful, the absorptive capacity of local firms is crucial. A threshold level of human capital is often argued to be a necessary precondition if FDI is to promote technology upgrading and economic growth in developing countries (Balasubramanayam et al., 1996). In Chapter 12, Fu and Li examine the role of human capital as a determinant of FDI spillovers using a threshold approach. Based on the threshold regression and Chinese provincial panel data, they find double thresholds of 4.85 and 10.99 per cent for human capital, in terms of the percentage of workers with higher education: 4.85 per cent is the threshold that significantly mitigates the negative effects of FDI, while the most important threshold is 10.99 per cent, which changes the negative effects of FDI into positive spillover effects. This means the impact of FDI on local productivity growth depends on the absorptive capacity of human capital. In China, there are big discrepancies among the regions: some provinces do surpass the sign change threshold and so enjoy positive technology spillovers from FDI, but others do not. One of the most significant phenomena in the past century has been the rise of outward direct investment from the developing countries. The BRICS have been the leader of this recent surge. In Chapter 13, Torbjorn reviews some recent developments with respect to the growing importance of transnational corporations (TNCs) from the South and their overseas expansion. He argues that the current situation differs from that of the 1970s and 1980s. The scale of the phenomenon is much larger and both the geographical composition of flows and stocks as well as the drivers and determinants are different. He argues that the overseas expansion of latecomer TNCs opens up new possibilities for them to access knowledge and technology in foreign locations. It represents an important complement to other channels of technology transfer – such as licensing, imports and inward FDI. In combination with the observed trend towards more internationalization of R&D, this contributes to a strengthening of the interlinkages between national innovation systems.

I.5

Technology and sustainable development

This part includes four papers which address the role of technology in confronting the challenges of sustainable development, with an emphasis on climate change, social divisions and economic development in general. From a global perspective, the challenge posed by sustainable development is becoming increasingly urgent. In rapidly growing economies,


Introduction 11

knowledge transfer and technology co-operation are becoming important issues in their development process. Based on the heuristics of a system of sustainability innovation approach, Walz analyses, empirically, in Chapter 14 the importance of technological competences and absorptive capacity for sustainable technologies in the BRICS countries. The results show that sustainability-related research is mostly carried out within broader, more sector-oriented programs. With the exception of South Africa, this topic is still underemphasized in the BRICS countries. Developing technological competences in the relevant fields of sustainability is a key indicator of the absorptive capacity of sustainability technologies. International patents and successes in foreign trade indicate the extent to which a country is already able to ‘open up’ internationally. The resulting pattern shows the various strengths and weaknesses of the BRICS countries. The differences within the countries imply that the analysis must proceed on a technology-specific level. Furthermore, there is a strong need for strategic positioning of the countries and for coordination of the various policy fields involved. Technology upgrading for inclusive development is an important task for policy-makers and academic researchers. This in particular is a major challenge facing the emerging economies. In Chapter 15, AkpanObong and Parmentier present a framework for the study of information and communication technologies (ICTs) in which integration and development are considered to be interrelated processes. The authors explore possible convergence and coordination through a review and analysis of the policies regarding ICTs for development in Brazil and South Africa, comparing them with policies regarding ICTs for integration between their respective regions. The authors adopt an exploratory, qualitative approach and analyse ICT policies at both national and regional levels. Their research shows that while national and regional ICT policies share the goals of development, the emphasis on development policies varies considerably: even policies relating to similar issues are neither coordinated nor connected. The authors argue that research in this field needs to focus on processes and outcomes particularly with respect to compatibility, synergy and enhancements in human development as the ultimate goals of policy and its implementation. Given the important role that IT can play in the development process, sustainability of the ICT for development projects has been a major bottleneck. In Chapter 16 Fu and Polzin explore the determinants of the sustainability of technology-intensive social innovation with special emphasis on absorptive capacity and complementary assets. Through examination and comparison of a series of case studies from India


12


they find that absorptive capacity and complementary assets can be critical factors of success and sustainability for ICT-enabled development projects in developing countries. Moreover, customer freedom of choice enables social innovation to meet the needs of the grass roots and thereby enhances the vitality and sustainability of the technologyintensive social innovations.

References Aitken, B. and A. Harrison (1999), ‘Do domestic firms benefit from direct foreign investment? Evidence from Venezuela’, American Economic Review, 9 (3), 605–18. Balasubramanayam, V. N., M. Salisu, D. Sapsford (1996), ‘Foreign direct investment and economic growth in EP and IS countries’, Economic Journal, 106(434), 92–105. Department of Science and Technology (DST) (2007), ‘Innovation towards a knowledge-based economy. Ten-year plan for South Africa (2008–2018)’, Draft. Pretoria: DST. 10 July. Kaplinsky, R. ed. (2006), ‘Asian Drivers: Opportunities and Threats’, IDS Bulletin, 37 (1). Lall, S. and S. Urata (2003), ‘Introduction and overview’, in S. Lall and S. Urata (eds), Competitiveness, FDI and Technological Activity in East Asia, published in association with the World Bank. Cheltenham, UK: Edward Elgar. Tojo, Y. (2008), ‘New modes of innovation: Increasing role for emerging economies’, paper presented at the SLPTMD conference on ‘Confronting the challenge of technology for development’. Oxford: Oxford University. Wang, Q. and J. Gao (2006), ‘A retrospective analysis of the foreign capital attracting practice based on the “exchange-market-for-technology” strategy in china’, Journal of Hubei University (Philosophy and Social Science), 33(3), 261–4. World Bank (2007), World Development Indicators 2007. Washington, DC: World Bank.


Part I Policy, Strategy and Catch-up: Cross-Country Analysis




1 Innovation Strategies in Brazil, China and India: From Imitation to Deepening Technological Capability in the South Carl Dahlman

1.1

Introduction

There has been a strong interest on the link between invention, innovation and economic growth among both academics and policy-makers.1 Much of the earlier work focused on the link between R&D and economic growth. Over time, as there has been greater understanding of the complexity of the relationships, the focus has broadened to innovation and economic growth.2 However much of the analysis has focused on innovation as something that is new at the frontier of global technology. As a result, much of the effort has continued to be focused on the relationship between R&D, global innovation and economic growth. However innovation studies done at firm level in both European and developing countries have identified that the main sources of innovation, in the sense of technology that is new to the firm, is not its own R&D or even domestic R&D by universities or specialized institutes. The main sources are technology embodied in machinery and equipment, technical information and specialized inputs from suppliers of inputs and components, licensed technology, copying from other firms, hiring persons with specialized knowledge and only in some cases its own R&D or R&D services from others. The most important source varies by firm, sector and country, depending on the nature of the sector and the specific level of development of the firm and the country. Thus firms at the global technological frontier are more likely to report that their main source of innovation is their own R&D because they have already 15


16

Carl Dahlman

incorporated existing technology. Firms that are behind the frontier are not likely to have that much R&D capability and the most important source for them is existing technology obtained from others.3 This also holds for the sources of innovation for countries. Countries that are at the frontier need to invest more in R&D in order to increase their productivity. Countries that are far behind the frontier can innovate by acquiring knowledge that already exists. As they get closer to the frontier it makes sense for them to invest more in their own R&D.4 This implies that countries need to adjust their technological strategy to their level of technological development and that the strategies need to change over time as they get closer to the frontier. There has been an acceleration in the rate of creation and dissemination of knowledge.5 In addition, knowledge is increasingly becoming global. There has been a significant increase in the number of co-authored papers in many fields by persons from different countries and this internationalism also holds true for patents. More importantly, more than half of all global R&D is done by transnational companies. Spending by large multinationals is larger than the total R&D expenditure of all but the very largest developing countries who invest in it. For example, Toyota ($8.4 billion in 2007) and General Motors ($8.1) billion each spent more on R&D than India.6 In addition, a survey of the top 1000 companies doing R&D in the world shows they made on average 55 per cent of their innovation investment outside their home countries (Jaruzelski and Dehoff, 2008). Because the global stock of knowledge is increasing rapidly, innovation in the context of developing countries should be considered not just in terms of the creation of knowledge that is new to the world, but also in terms of products, services or forms of organization that are new to local practice, rather than just to global practice. Furthermore innovation can be new to the country, new to the sector or – at a micro level – new to the firm. Therefore it is useful to distinguish three sources of innovation. The first is acquiring technology that already exists abroad. The second is the domestic creation of relevant new knowledge. The third is the dissemination and effective use throughout the economy of this new knowledge, whether it has been created locally or imported from abroad. Doing any of these three requires certain levels of education. The type of education needed for adopting foreign technology and diffusing it is different to that needed for advancing the global frontier. For the first two, strong basic and secondary education systems and some higher education are necessary. To undertake R&D and really innovate


Innovation Strategies in Brazil, China and India

17

at the level of the world frontier, however, a strong tertiary education in science and engineering is necessary. From this broad perspective, the innovation strategies of Brazil, China and India – the three largest developing economies7 – have been quite different. The three countries have also had very different growth performance over the last 25 years. This paper examines their innovation strategies, compares their economic performance over the last 25 years and draws some implications about the link between innovation strategy and economic performance. The paper will also document the rising technological power of these countries in terms of R&D expenditure and outputs as well as in terms of the technology intensity of their trade. These three countries are also interesting in that two of them, China and India, have not been as negatively affected by the 2008/2009 financial and economic crises. While overall world growth slowed to 3.1 per cent in 2008 and the world economy is expected to shrink by 1.4 per cent in 2009, these two countries still had positive growth rates in 2008 and 2009 (see Table 1.1) and are expected to continue to grow faster than the rest of the world. This has important implications for other countries. The essence of the argument is that over the last 30 years China has grown faster than the other two countries studied here because it has been more aggressive at acquiring foreign technology, developing technological capability and diffusing technology. Brazil started significantly ahead of the other two in 1980 in terms of per capita income, adoption of foreign knowledge, education and R&D effort, but it has fallen behind in relative terms with respect to both other countries, and in some cases in absolute terms with respect to China (such as R&D and the diffusion of some technologies such as broadband). India’s performance has been in between the other two, but since 2003 its growth has accelerated to near-Chinese levels. Its better growth performance compared to Brazil between 1980 and 2000 has been largely the result of better macro stability and higher investment. However its acceleration in growth since 2003 has been caused by its more aggressive adoption of foreign technology and in particular its being able to leverage information technology services. Section 1.2 summarizes some of the key economic parameters of the three countries including size, economic indicators and various social and structural characteristics. Sector 1.3 analyses the extent to which they have acquired global knowledge through various modes. Section 1.4 reviews how they have built up education and R&D capabilities. Section 1.5 summarizes how they perform on the dissemination


18 Carl Dahlman Table 1.1

Basic economic indicators in Brazil, China and India

GNI (2007) GNI (2007 nominal US$ billions) GNI as share of global GNI (%) GNI/capita (nominal) GNI (billion PPP) GNI as share of global GNI (PPP) GNI per capita (2007 PPP) Growth of GDP (1980–2006) 1980–90 av. annual growth 1990–2000 av. annual growth 2000–7 av. annual growth 2008* 2009 estimate* 2010 estimate* People (2007) Population (millions) Population as share of global population Life expectancy at birth Poverty and inequality % below US$1.25/day poverty line (2005) % below US$2/day poverty line (2005) Gini coefficient (2004)

Brazil

China

India

1122 2.12 5860 1776 2.70 9270

3126 5.91 2370 7151 10.87 5420

1071 2.02 950 3083 4.69 2740

2.7 2.7 3.3 5.1 –1.3 2.5

10.2 10.6 10.3 9.0 7.5 8.5

5.8 5.9 7.8 7.3 5.4 6.5

192 2.90 72

1318 19.93 73

1125 17.02 65

7.8 18.3 55.0

15.9 36.3 41.5

41.6 75.6 36.8

Note: The GNI in PPP figures are based on new PPP series published in December 2007, which reduced the estimates for both China and India by 40 per cent. *IMF WEO (July 2009) Source: World Bank WDI 2009.

of technology. Finally, section 1.6 draws out some key conclusions and implications.

1.2 Characteristics of three countries China and India are the two most populous countries in the world, accounting for 20 and 17 per cent of the world’s population respectively (Table 1.1). Brazil is the fifth largest, accounting for almost 3 per cent of world population. China has grown faster for a longer period of time than either of the other two. As a result, although it was smaller than the other two in 1980, its total gross national income (GNI) is almost three times that of India or Brazil (see Figure 1.1). However, because of Brazil’s earlier start, that country’s GNI per capita is two and a half times that of China, which has a GNI per capita two and a half times


Average inflation adjusted GNI per capita growth rate from 1980–2008 (%)


19

14 United States Russian Federation China India Brazil Germany Japan

12 China, 3678 10 8

India, 1215

6

Japan, 4879 Brazil, 1411

4

United States, 14466

2

Russian Federation,1364

0 –10000

0

10000

20000

30000

40000

50000

60000

GNI/Capita in PPP in 2008 Figure 1.1 GNI/per capita growth (1980–2008), per capita income (PPP)and size of economies (in nominal US$, 2008)

that of India. In purchasing power parity (PPP) terms, China’s gross domestic product (GDP) is more than twice that of India, and India’s is almost 75 per cent greater than Brazil. Brazil actually had high rates of growth from the second half of the 1960s to until about 1981, when it was severely affected by the Latin American debt crisis and the second oil shock. It basically lost two decades of growth as a result of the major macro imbalances that ensued. Brazil only managed to stabilize its economy in the first decade of this century. However, even now, its growth rate significantly trails that of the other two countries and has been more negatively affected by the 2008/2009 global economic crisis (Table 1.1). India, on the other hand, had a more steady rate of growth of almost 6 per cent between 1980 and 2000 except for a severe financial crisis in 1991 which forced it to liberalize the economy and eventually put it on a higher growth path. In the last four years it has been growing at more than 8 per cent a year, nearly approaching Chinese rates of growth. All three are still developing countries, at three different levels of development. In terms of their GDP per capita Brazil is classified as a high-income developing economy, China as a lower-middle-income


20

Carl Dahlman

economy and India as a low-income economy.8 Moreover they still have significant pockets of poverty. In Brazil 18 per cent of the population still lives on less than US$2 a day,9 while that percentage jumps to 36 for China and 76 for India. However, while richer in per capita terms, Brazil has a much more unequal distribution of income as revealed by its higher Gini coefficient (bottom of Table 1.1). The three countries differ in their relative wealth, political system, economic structure, development strategy and degree of global integration, structure of trade and technological strategy. Table 1.2 presents a brief summary of the main characteristics of each country.

1.3 Acquisition of foreign knowledge The main means of acquiring foreign knowledge are trade, FDI, technology licensing, foreign education and training, use of the diaspora, copying and reverse engineering and accessing foreign technical information in print and, now, through the Internet. On all these counts, China has been more aggressive and systematic than Brazil or India. Trade Purchases of foreign products and services are a key way of gaining access to knowledge embodied in those goods and services. Exports of manufactured products also force countries to keep up with foreign technology trends because of the need to keep up with the competition. China began opening up to the world much earlier than India did and has become much more integrated into the global economy. Moreover, the speed with which China opened up is impressive. In 1980 imports and exports as a share of GDP were similar to Brazil’s and just slightly larger than in India. By 2007 trade was 76 per cent of GDP in China,10 compared with 46 in India and only 27 in Brazil (Table 1.3). By initially protecting its industries from imports, China developed basic technological capability. Then by opening up to FDI in special economic zones with near free trade status it was able to get access to world class technology and inputs. This worked very well and not only began to modernize China, but also provided much needed foreign exchange and employment. The number of these special economic zones was expanded from the initial four to 19 and then to many more.11 This programme was so successful in generating employment and foreign exchange that by the late 1990s China decided significantly to widen this free trade status by joining the World Trade Organization (WTO).



21

Table 1.2 Main country characteristics Characteristics Brazil

Upper-middle-income, democratic government Natural-resource rich on a per capita basis Large workforce with moderate growth Heavily industrialized but two-thirds of value-added in services Development strategy based largely on import substitution with early focus on technology development during the ‘miracle growth years’ Still relatively little integration through trade among large economies Country is primarily an exporter of primary commodities, although it has a well developed industrial sector, and had a more significant share of manufactures in its merchandise exports until the boom in commodity prices of 2001–8 led to more specialization in primary commodities Low investment to GDP ratio. Extensive use of FDI, but bulk of this is in the protected domestic markets and as part of mergers and acquisitions (M&A) Had lead in education, but gap has narrowed sharply Had lead in R&D but fell behind others on expenditure, publications and patenting, most R&D is still done by the government

China

Lower-middle-income, authoritarian government Natural-resource poor on a per capita basis Very large workforce but has made demographic transition and will have rapidly aging population in future Very heavily industrialized with greatest share of value-added in industry Development strategy originally based on autarkic development until gradual opening up into global economy starting in the late 1970s through an export led strategy. Most globally integrated through trade of large world economies Country is almost exclusively an exporter of manufactured products Started with labour-intensive manufactures, but is rapidly moving up the technology ladder Very high share of investment in GDP. Heavy use of FDI, originally 100 per cent oriented to export market, but gradually allowed into domestic market Was behind in tertiary education but has surpassed Brazil in general educational attainment and nearly caught up in tertiary enrolment rates Was behind in R&D but now leads by significant margin. Two-thirds of R&D already carried out by they productive sector

India

Low-income, joined lower-middle-income in 2007, democratic government Natural-resource poor Very large and rapidly growing population and workforce Agriculture still the largest employer, low share of manufacturing in valueadded, but largest share of value-added is already in services Development strategy based on autarkic development until gradual opening up and integration into global economy, starting in the early 1990s. Still remains one of most protected economies in terms of tariff and non-tariff barriers Country had been largely an exporter of labour-intensive manufactures. However since 2000 it has rapidly expanded it exports, particularly those facilitated by information technology Low investment rates until recently. Little use of FDI until recently Invested early in elite higher education, but neglected basic education and has fallen behind China at all levels of education Started with autarkic technology strategy and state-dominated R&D but has gradually opened to world and private sector has increased R&D spending


22

Carl Dahlman

Table 1.3 Acquiring foreign knowledge

Trade as a percentage of GDP 1980 1995 2007 Merchandise imports as a percentage of GDP 1980 1995 2007 Manufactured imports as a percentage of merchandise imports 1980 1995 2007 Average tariffs (%) 1990–2 Average simple tariff Average weighted tariff 1997 Average simple tariff Average weighted tariff 2006 Average simple tariffs Average weighted tariffs Average gross FDI/GDP 2000–5 Royalty and licence fee payments (US$ million) 1990 1997 As a percentage of GDP 2007 As a percentage of GDP Tertiary students studying abroad 2007* As percentage of students studying abroad As percentage of tertiary students in country

Brazil

China

India

22 16 27

21 44 76

15 23 46

9.8 7.0 9.6

– 18.1 29.8

7.5 9.7 18.4

41 71 64

– 79 68

39 53 46

25.1 26.7

42.9 40.6

81.8 83.0

11.9 14.8

17.8 20.9

30.0 27.7

12.3 6.8

8.9 5.1

17.0 13.8

3.4

3.2

0.9

54 529 0.064 2,259 1.72 21,556 0.77 0.4

0 543 0.060 8,192 2.56 421,128 15.03 1.9

72 150 0.039 949 0.80 153,312 5.47 1.1

Note: * The total number of tertiary students studying outside their home country was 2,800,470. Sources: World Bank, WDI various years, and UNESCO (2009) for students studying abroad.

This involved committing to a major programme of reducing tariff and non-tariff barriers and opening up to foreign investment not only in manufacturing, but also in financial and other service areas. Unlike China, which has substantially removed tariff and non-tariff barriers to trade as part of its joining the WTO, India is still one of the



23

most closed economies in the world (Table 1.3). In the 1950s India followed a very autarkic policy of self-reliance, depending initially mostly on massive capital goods imports from the Soviet Union, as did China at the time. However, unlike China, India maintained its strongly inward-oriented nationalist policy through the 1980s. It was only after the trade liberalization of the early 1990s that India began to open up more to foreign technology imports. There were also very strong restrictions on FDI and on the licensing of foreign technology. Brazil also had a relatively autarkic policy, but not as extreme as India’s. Brazil has a period of rapid economic growth during the ‘miracle years’ of 1965–80 when it continued a period of rapid import substituting industrialization. In the 1990s, even after India’s liberalization, Brazil still had lower tariff and non-tariff barriers than India or China (Table 1.3). However since joining the WTO, China has become more open. Furthermore, Brazil is still the least integrated into global trade of the three countries in terms of imports and exports as a share of GDP. This has meant that its industries have been, and continue to be, relatively protected from global competition. This has meant less pressure to innovate with new products and processes and lower costs. FDI The inflows of FDI into China have been several multiples of those into India and higher than those into Brazil (Table 1.3 and Figure 1.2). This is the result of several factors. First, China opened up its regulatory regime towards FDI more than ten years earlier and more broadly than India did. Second, China’s larger and richer market has been an important pull factor so it has surpassed even Brazil. Third, China has many cost advantages over India and Brazil, even though its labour costs are now higher than India’s. Transportation is more efficient, service infrastructure is more developed and the red tape for trade in physical products is less burdensome. As a result, China has been very attractive not just as a production platform for global markets, but also of production for the domestic market as this is the fastest growing in the world. This strong pull of producing in China has also permitted the government to encourage strong competition among foreign multinationals to bring their best technology when they locate in China, even though they are very aware of poor intellectual property rights (IPR) protection and the risk that their technology will be pirated. The most important contribution of FDI to China is not capital, since China has had a high savings and investment rate. Access to advanced technology and management through FDI has been much


24 Carl Dahlman

Percent of GDP

7 6 5

Brazil

4 3

China India

2 1 2006

2004

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

0

Year Figure 1.2 FDI inflows as a percentage of GDP in Brazil, China and India 1980–2006 Source: Computed from UNCTAD (2008).

more important. Equally crucial is entry into global markets as the foreign investors integrate their Chinese operations into their global supply chains (Gill and Kharas, 2007). Moreover the latter does not even require ownership of production plants in China, but just sourcing from China. An excellent example of this is Wal-Mart which sources more than $25 billion of goods from China directly into its retail stores without even using middlemen. India only began to open to FDI in the 1990s and then only slowly and selectively. As a result it got very small inflows. In the last five years India has liberalized FDI inflows and trade inflows but, as noted, both are still very small compared to those of China. Thus Indian industrial policy protected domestic industry for too long so it did not take advantage of the technology it could get from abroad, or the economies of scale and scope of pushing its firms to operate globally. Relatively little FDI has been attracted until recently because of high transaction costs and poor infrastructure. The exception has been in services related to software and information and communications technologies (ICT), which have not been constrained by any regulatory regime or lack of physical investment infrastructure. In Brazil, a large percentage of total FDI came in the period 1998– 2001 in order to exploit domestic natural resources or as part of the privatization of state-owned enterprises (SOEs).12 In addition, although the government put various export and local content requirements on FDI into manufacturing, this did not have as much leverage as China. The main reason is that it was not as attractive a location for exports or for the domestic market.



25

Technology licensing China has also been much more aggressive in licensing foreign knowledge through formal technology-licensing agreements than either Brazil or India. Although in 1990 India was paying more for technology licensing, by 1997 China’s payments were the largest. By 2007 Chinese royalty and fee payments were 8.6 times those of India and 3.6 times those of Brazil in absolute terms, and three and two times, respectively, those of India and Brazil as a share of GDP. Foreign education and training China has also been much more proactive in using foreign education as a means of getting access to foreign knowledge. In 2007 15 per cent of all tertiary students studying outside their home country were from China, versus 5.5 per cent from India and only 0.8 per cent from Brazil. Part of the difference is that China has a larger population. However even as a proportion of the tertiary students studying in each country, the ratio in China was almost twice that of India and five times that of Brazil (Table 1.3). Diasporas Both China and India have benefited enormously by drawing on their respective diasporas. Brazil has a very small diaspora so it has not had this as a major means of gaining access to global knowledge. China has done this more systematically and for longer than India. More than half of the FDI in China has come from Taiwan, Hong Kong and Singapore. These are economies which have had substantial experience of operating in the global market. They were already very plugged in to global supply chains. They initially moved their more labour-intensive operations into China. As China has moved up the technology ladder they have been moving more technology-intensive operations. This is particularly true for Taiwanese companies which are now putting some of their most advanced production facilities in to China. In addition, China has set up special high technology parks specifically targeted at attracting back experienced overseas Chinese to set up high-tech companies in China. Several of the more than 100 high-tech parks in China cater specifically to this diaspora. India has done this to a much smaller degree. Furthermore, China has also made a more sustained effort to attract back Chinese professors and former foreign students to staff the rapid ramping up of its tertiary education sector. India has had much less success in doing this because it is more constrained by regulations that do not allow its universities to pay professors competitive salaries.


26 Carl Dahlman

Copying and reverse engineering While there are no hard data on this, it is quite certain that China has been much better at doing this than India or Brazil. Greater access to foreign knowledge through all the formal channels listed above, higher levels of human and technological capital and a policy (now changing) of ignoring IPR laws have given China an advantage in copying and reverse engineering foreign technology. Several observations can be drawn from this comparison of the three countries. The first is the importance of acquiring global knowledge effectively. China has done this extremely well through all modes. India has not made much use of any of the channels except – more recently – trade, FDI and the diaspora connection. Brazil has made less use of trade and foreign education and the diaspora. Unlike India, though, it did attract high levels of FDI, but much of this was through acquisitions. Acquiring foreign knowledge through trade has perhaps been the most important aspect. China has imported lots of technology embodied in capital goods and components. In addition importing more foreign products also facilitates copying and reverse engineering. FDI is perhaps the second most important way of tapping knowledge. Foreign investment is a package that includes not just technology but also managerial and organizational knowledge. The second is that it is not just a matter of getting lots of FDI. It is important to use it effectively. China was able to do this because it was an attractive location for FDI. Initially the attractiveness was its low labour costs for export-oriented FDI. Subsequently it was the lure of the very large and rapidly growing domestic market. Both of these gave the government strong negotiating powers with respect to FDI. Initially the government forced companies to go into joint ventures with domestic firms and negotiated local content and training requirements.13 This greatly helped domestic firms to develop technological and management capability. For example, in the automotive industry the government managed to force both Honda and Toyota to do joint ventures with the same Chinese manufacturer. This allowed the Chinese company to use the best of both systems to develop its own brand and production. The government was able to negotiate this because of the attractiveness of the domestic Chinese market to the foreign manufacturers. Once the cost advantage of producing in China became apparent to both the government and multinational companies, the former relaxed the joint venture requirement in order to encourage foreign firms to bring in their best technology. The third is the importance of competition. One of the reasons Brazil did not get as much FDI as China was that most FDI to Brazil’s



27

manufacturing industries came to exploit the protected domestic market. In China it came initially to exploit a low-cost production base for export markets. Subsequently, when China allowed FDI for the domestic market there was a similar attractiveness in producing for the protected domestic market, although that quickly disappeared when China reduced tariff and non-tariff barriers as part of its commitments to joining the WTO. As a result, foreign companies then had to bring in their best technologies because they were competing with other foreign firms for both the export and domestic markets. A fourth is the importance of foreign education. During the period 1980–2000 China sent more than 300,000 students abroad. They acquired cutting-edge academic knowledge from many of the best universities in the world. Many stayed abroad and acquired important research and work experience. However many returned and brought back both academic knowledge and practical work experience. This has been important in building up Chinese academic institutions to train more generations of Chinese students as well as to run domestic enterprises. India and Brazil sent far fewer students abroad. The return flow of Indian students was smaller than that of the Chinese in absolute and in relative terms. In the last few years a much larger percentage of Indian students has been returning because of exciting job prospects with the rise of the high technology industry in India. The return flow of Brazilian students may be larger in relative terms, but is small in absolute numbers because so few students have gone to study abroad. A fifth is the importance of the diaspora beyond returning students. The diaspora has been critical for China’s very rapid development. The first export processing zones were set up close to Hong Kong and Taiwan as most of the FDI came from those two economies. Market knowledge and entrepreneurship from the overseas Chinese communities beyond Hong Kong and Taiwan has been very important. Many of the special high-tech industrial parks located in China were set up explicitly to attract back overseas Chinese. Indian success in the ICTenabled export services industry has also resulted to a large extent from the linkages to its diaspora in the high tech-area of the US and Europe. Initially India did not make much use of its diaspora. It was joked that non-resident Indians (NRIs) were not required Indians. However India eventually learned the importance of harnessing its diaspora and since then has make special efforts to attract its members back by giving special tax breaks and other fringe benefits. Brazil does not have much of a high-tech diaspora and has not made any significant efforts to attract its biaspora back.


28

Carl Dahlman

1.4 Development of domestic capabilities The development of domestic technologies has two principal components. The first is the basic education and specialized human capital required to undertake technological effort. The second is specific allocation of human capital and other resources to undertake R&D. 1.4.1 Education People need education and new skills to use new technologies. To produce new knowledge, they need a more specialized tertiary education. More educated people tend to adopt new technologies faster.14 This section will compare education across the three countries, first studying basic educational attainment. It will then cover secondary and higher education enrolment rates. Basic education As a very poor developing country even just 25 years ago, China had very low levels of education. However it has made massive investments in basic education and now has the highest literacy rate, as well as the highest average educational levels, of the three countries (Table 1.4). India still has very low literacy rates. Illiteracy is 46 per cent among women and 23 per cent among men. The country’s basic education system is still very poor with tens of millions of primary school-aged

Table 1.4

Literacy and enrolment rates by level of education 1980–2007

Literacy rate, population 15 and above (%) 1980 1995 2007 Average educational attainment of adult population (2000) Secondary education enrolment ratio (%) 1980 1995 2007 Tertiary education enrolment ratio (%) 1980 1995 2007

Brazil

China

India

74 83 91 4.88

66 81 93 6.35

41 53 66 5.06

34 45 105

46 67 76

30 49 55

11 11 54

2 5 23

5 6 12

Source: WDI various years.



29

children out of school. Brazil was ahead of both countries in all primary enrolment rates and literacy, but now it has a lower literacy rate and a lower overall educational attainment than China. Secondary education In 1980 China had the highest secondary enrolment rates of the three countries because of the importance the communist government placed on education as part of its industrialization strategy. Brazil had very low secondary enrolment rates for a country of its per capita income until the mid-1990s when the Cardoso government (1995–2003) focused on rapidly increasing secondary education.15 India has low secondary enrolment rates and has fallen significantly behind the other two countries. Low literacy and low secondary enrolment rates have been a brake on assimilating new technologies in India. Tertiary education China has undertaken a massive expansion of its tertiary education system since the late 1990s to make up for the havoc wreaked on the educational system after the Cultural Revolution (1965–75). By 2007 its enrolment rate reached 23 per cent and because of is large population it had more students at the tertiary level than the US: 25 million versus 17 million in the US and 13 million in India (Table 1.5).16 In addition 40 per cent were in engineering and sciences. India set up seven Indian Institutes of Technology starting in the 1950s and later several Indian

Table 1.5 University enrolments 2007 – 12 countries with most students Country

Number enrolled

China US India Russian Federation Brazil Japan Indonesia S. Korea Iran Ukraine Egypt Mexico

% of world enrolments*

2,534,600 1,775,900 1,285,300 937,000 527,300 403,300 375,500 320,900 282,900 281,900 259,400 252,900

16.8 11.8 8.5 6.2 3.5 2.7 2.5 2.1 1.9 1.9 1.7 1.7

Note: * Out of total global tertiary enrolments of 150,656,000. Source: UNESCO (2009), Table 8.


30 Carl Dahlman

Institutes of Management, which produced a critical mass of well educated English-speaking professionals who have been instrumental in India’s emergence in software and ICT-enabled services. Although India has expanded its tertiary enrolment rates, it has not done so as fast as China and is now considerably behind. Moreover, the quality of higher education is poor with the exception of those institutes mentioned above (which produce fewer than 7000 graduates a year), the Indian Institutes of Science and some of the regional engineering colleges. The low quality of tertiary education and the regulatory constraints on expanding high-quality institutions will be a major bottleneck to India’s continued rapid growth in knowledge-intensive services. Brazil was originally considerably ahead of both China and India in tertiary education and while its system has expanded the least, it is still slightly ahead of China in tertiary enrolment rates. China’s rapid ramping up of its tertiary education system means that by 2005 it was the country with the largest number of graduates from tertiary education: 70 million, compared to about 65 million for the US, 53 million for India and 42 million for Russia (Table 1.6). This large Table 1.6 OECD

Higher educational attainment in non-OECD countries, against Percentage of population aged 25–64 with tertiary degree

Indonesia (2003) Brazil China* Romania India* Thailand Malaysia (2003) Chile Argentina (2002) Peru (2002) OECD Philippines Israel Russian Federation (2003)

4.2 7.8 9.5 10.4 11.4 12.1 12.1 13.2 13.7 18.0 25.1 27.3 45.4 54.6

Of which: university degree 2.3 n/a 3.2 n/a n/a 9.1 n/a 10.3 9.1 8.9 18.9 14.0 29.4 20.8

Population aged 25–64 with tertiary degree (thousands) 4227 6655 70,336 1213 52,600 3997 1309 1055 2322 2014 171,553 8960 1376 42,238

Note: n/a = not available. * Overestimated, as all people with a tertiary degree are included. Source: OECD (2007), excluding countries with less than one million university graduates, available from Statlink at: http://dx.doi.org/10.1787/117384133584



31

critical mass of university-trained persons is a tremendous technological asset for the country. It is allowing China to acquire and rapidly diffuse foreign technology. The high percentage of graduates in science and engineering also gives it a tremendous advantage in undertaking R&D. 1.4.2 R&D A key element of domestic technological capability is R&D. Domestic R&D is necessary for acquiring global knowledge as well as for generating new knowledge. Although not all creation of knowledge is the result of formal R&D effort, R&D is usually part of the process. This section will compare the innovative efforts of the three countries in terms of two inputs (R&D and scientists and engineers) and two outputs (scientific and technical publications and patents) as well as some assessment of their commercialization of technology and protection of intellectual property rights. China has long had the largest number of scientists and engineers carrying out R&D among the three countries, and it has ramped up the number the most quickly. It has also increased expenditure on R&D as a percentage of GDP the most among the three countries (Table 1.7). Figure 1.3 puts the R&D expenditure and the relative intensity of scientists and engineers in R&D in perspective for the five main developed countries (the G5) and the BRIC, plus the Russian Federation and South Korea. The bubble graph displays two relative measures of R&D inputs: R&D expenditure as a share of GDP on the x axis and the number of scientists and engineers per million population on the y axis. Because knowledge, unlike other goods, is not consumed in its use, the absolute amount matters, so the point for each country is drawn proportionally to the absolute amount spent.17 In R&D spending China leads the countries under study. By the end of 2006, in PPP terms, China was the third largest spender on R&D in the world.18 This is the result of an explicit strategy by the Chinese government to go beyond acquiring global knowledge through copying, reverse engineering, FDI and technology licensing to actually investing in innovation on its own account. In 2005 the government announced a 15-year plan to increase expenditure on R&D to 2.0 per cent by 2010 and to 2.5 per cent (the average level of more advanced developed countries) by 2025. By 2006 it had already increased it to 1.42 per cent of GDP. In addition, as part of the global outsourcing trend, many multinational corporations (MNCs) are increasing their R&D work in developing countries, particularly China and India. By 2006 there were more than 750 MNC R&D laboratories in China and above 250 in India. It is estimated that the foreign share


32 Carl Dahlman Table 1.7 R&D inputs and outputs Indicator Researchers in R&D 1995 2006 R&D researchers per million population 1995 2006 Spending on R&D (US$ billions) US$ billion nominal 2007 US$ billion in PPP 2007 Spending on R&D (percentage of GDP) 1995 2006 Scientific and technical journal articles 1995 2005 Scientific and technical journal articles per million population 1995 2005 Patenting by residents in country 2005 2007 Patents granted by US PTO Average 1991–5 Average 2002–6 Patent applications granted by USPTO per million population 1991–5 average 2002–6 average Share in triadic* patent families 2005 Average annual rate of growth 1995–2005 Science parks and business incubators

Protection of IPR Survey 1 = least; 7 = most 2000 2009

Brazil

China

India

26,578 84,971

531,997 926,252

145,115 117,528

168 461

445 926

157 111

9.1 14.6

37.5 101.5

7.7 21.3

– 0.82

0.55 1.42

0.8 0.69

3471 9889

9261 41,596

9591 14,608

21.51 52.9

7.69 31.9

10.29 13.4

2757 3810

10,066 153.060

1545 4521

65 135

56 448

36 316

0.40 0.75 0.1 14.6 Few and few spin-offs

0.05 0.35 0.8 36.7 Many and many spin-offs

0.04 0.30 0.2 27.6 Few and few spin-offs

4.5 3.0

3.6 4.0

3.6 3.6

Note: * Triadic patents are those filed at the European Patent Office, the US Patent and Trademark Office (USPTO) or the Japan Patent Office, which all protect the same invention. Source: World Bank WDI for various years, except for triadic patent which is from OECD (2007), and protection of IPR which is from WEF (2009/2010) and WEF (2000).

of spending in business R&D in China is 25 per cent as a result of the rapid increase of R&D being done in China by foreign firms (OECD, 2008: 166). The survey of the largest private spenders on R&D cited earlier also shows that China and India are the first and second largest net


Scientists & engineers in R&D per million


33

14000 12000

Japan 142.2

10000

Usa 345.7

France 41.7

8000 Russia 19.6

6000

South Korea 35.6

Germany 68.7

Uk 36.3

4000 2000

China 87.5 23.2 india

16.8 brazil

0 −2000

Brazil

0

China

0.5

1

1.5

2

2.5

3

3.5

4

Spending on R&D as % of GDP India

Korea

Russia

Us

Japan

Germany

Uk

France

Figure 1.3 Relative R&D expenditure and ratio of scientists and engineers per million (expenditure in 2006 in US$ billions PPP) Source: Author’s calculation based on World Bank WDI 2008b, and Dutz (2007) for India.

importers of R&D by foreign companies in the sense that foreign MNCs do more in these countries, than companies from these countries do abroad (Jarulzelski and Dehoff, 2008). In India, the additional R&D investment by MNCs, as well as increased investment by the domestic private sector, has raised total R&D spending. This has concentrated in pharmaceuticals, ICT, electronics and automotive parts and has raised Indian R&D expenditure from a 20-year average of 0.88 per cent of GDP to 1.1 per cent in 2005.19 In terms of output of R&D, China and India had roughly the same number of scientific and technical publications in 1995 and roughly three times the number of Brazil. However, China has since more than quadrupled, and Brazil nearly tripled their numbers of scientific and technical publication, while the figure for India has increased only by about 50 per cent. In terms of domestic patenting efforts, in the two years studied China radically increased the number of patents given to domestic residents (Table 1.7) and became the country with the third highest number of domestic patents in the world, after Japan and the US (WDI, 2009, based on WIPO). Because national patent regimes differ significantly in what is considered patentable, it is useful to examine patenting activity in a major


34

Carl Dahlman

international market. In 1995 Brazil used to lead the three countries in terms of patents granted at the US Patent and Trademark Office (USPTO). However, China and India have dramatically increased their patenting since then and have averaged two to almost four times the number patented by Brazil over the period 2002–6. Taking triadic patents as a broader measure, in 2005 China led the other two countries. Among the 12 largest triadic patentees in the world in 2005, China had the highest average annual rate of growth (37 per cent), followed by India, which was second with 28 per cent (OECD, 2007). The creation of knowledge and patenting do not contribute to economic performance unless they are actually put into practice. Therefore knowledge has to be commercialized, particularly if it has been created in government research labs or universities as opposed to firms that can apply the technology directly. This requires an appropriate technological commercialization infrastructure. That includes: adequate IPR protection, technology-transfer offices at universities and research institutes, science/industrial parks, business incubators, early-stage technology finance and venture capital. China has developed the largest number of science parks and business incubators among the three countries. In addition it has the largest number of spin-off firms. In fact, it has got to a point where the government has had to tone down its strategy to promote so much spin-off activity in universities and get them to focus more on teaching and basic research.20 In terms of intellectual property protection, Brazil was ahead of China and India in the 1990s. China and India had to strengthen their regimes in order to fulfil their WTO requirements and now their regimes are considered even stronger than those of Brazil (Table 1.7). However, enforcement is weak and fines are low. This is considered a serious problem in China because of its greater capability in copying and replicating foreign technology and products. As a result of their growing R&D and technological capability all three countries have made major technological innovations. All three have had major technological achievements in agriculture, medicine and security (defence). They have also made many technological achievements in other areas. In Brazil some of the most notable advances include the development of aircraft technology by Embraer (the fourth largest aeroplane manufacturer in the world), development of the ethanol-based automotive fuel system and flexible fuel vehicles and deep oil exploration and drilling platform technology. In China they have included advanced rocket and satellite technology, nuclear weapons, advanced electric cars, solar and wind energy and many developments



35

in nano- and biotechnology and in the electronic hardware industry. In India they have included nuclear weapons and satellite technology, world class drug development by the strong pharmaceutical industry, software, electric cars and wind energy. While many of their innovations have to do with security and global competitiveness others have been developed to deal with specific local needs and conditions. These range from different medical needs and consumer preferences to services and products that are more appropriate for lower-income populations, particularly in India. A good example of the latter is the US$2500 Nano car for the masses. Several observations can be drawn from comparing the three countries. The first is the tremendous importance of education for acquiring and using knowledge and for developing new knowledge. High levels of literacy were part of the attractiveness of the low labour costs that attracted FDI into China. In terms of capabilities, people need to have not just basic literacy but numeracy and many job- and technologyspecific skills to be able to use the new technologies and production organizing techniques. Some of these need to be provided through an improved basic education curriculum. Others have to be provided through specialized training at vocational centres or even technology suppliers, or as part of firm-specific training. According to investment climate surveys undertaken at firm level by the World Bank, there is more firm-based training in China than in Brazil. and much more than in India.21 Again, this is an area with tremendous potential, but most developing countries have not developed the policies, institutions or capability to exploit it. Therefore it merits much more effort. Higher education in science and engineering is critical for developing new technology. The creation of a critical mass of highly educated engineers and Masters in Business Administration (MBAs) was critical for the successful development of India’s software- and ICT-enabled services. Investments in high-level human capital have been critical for Brazil’s islands of excellent in aeroplanes, deep oil exploration and agricultural research. The second observation is of the importance of copying and reverse engineering not just for acquiring foreign knowledge, but also for developing global frontier knowledge domestically. This has meant initially low enforcement of IPRs. As noted this has been a key element of China’s rapid catch-up. In India the explicit lack of protection of product patents was a key to the development of a strong indigenous pharmaceutical industry. Domestic firms were able copy foreign products and to produce them with slightly different processes. This had the


36 Carl Dahlman

double benefit of producing pharmaceuticals at low prices to improve the welfare of Indian citizens and developing a strong domestic pharmaceutical industry. By the time India extended process patents to pharmaceuticals in 2005, it has developed a strong domestic industry that could compete with foreign firms. Brazil has not made as much use of copying and reverse engineering. A third observation is that as countries or firms develop and get nearer the global frontier they need to put a lot of effort into developing new knowledge in order to remain competitive as other countries or companies catch up. Countries therefore have to place a premium on strengthening the whole institutional infrastructure in order to develop new knowledge. For these countries, this means placing emphasis on public and private research centres, university research and the training of scientists and engineers. For companies, it means scanning the world for relevant technical knowledge, developing strategic alliances with other firms who can contribute relevant knowledge, interacting closely with government and university laboratories to get access to relevant basic knowledge and making the effort to create new knowledge. Some of the critical policy issues here are on how the limited public resources are allocated and how effectively they are used. Unfortunately in most developing countries these very limited resources are not allocated or used very well. Therefore within this area a priority is to improve the allocation of public resources. This includes improving the definition of what areas the government should support as this is very critical when budgets are small. A second priority is controlling how effectively these resources are managed and what their contribution to the economy is. It is difficult to justify pure academic research in countries with pressing social and economic needs where more applied R&D can make an important contribution. Going beyond R&D done by the public sector, another key issue is how to get the private sector to do more R&D. The private sector needs to be encouraged to undertake more R&D, not only to be able to keep up to date with new developments and incorporate them, but to also carry out cutting-edge research in areas critical for their own competitiveness. Because of the problems of market failure in the appropriability of the returns to private R&D this calls for the creation of public support programmes such as matching grants and tax subsidies to stimulate R&D by the private sector. Furthermore, while in the first instance it makes sense to invest in the areas where developing countries already have a comparative advantage in order not just to maintain, but also to enhance, that advantage it is



37

also important for them to invest in new technological areas such as genetic engineering, biotechnology and nanotechnology. The public sector will have to play a greater role in carrying out this type of riskier and more uncertain research. It should be seen as part of an investment portfolio strategy of exploring new areas with potential high returns. Such investment is necessary in order to have the capability to move rapidly into into those areas that begin to show promising results. Obviously how much a country should invest in R&D and commercialization infrastructure will depend on its resources and its size. The richer and more developed it is in terms of institutions and human capital, the more it can do. However, even smaller poor countries have to have some capacity to create knowledge. At the very minimum they need some R&D capability to assess relevant global knowledge, to help negotiate and acquire it and to help adapt it to local conditions.

1.5

Dissemination of knowledge

The dissemination of knowledge takes place through the expansion of the enterprises that have developed it, their sale or transfer of that knowledge and various forms of imitation or replication by other firms or organizations. For the dissemination of knowledge, it is important to have appropriate mechanisms to educate potential users in the benefits of the technology. This often involves more than just providing technical information. In agriculture, for example, it involves showing the potential users the actual performance of the new technology in their domestic conditions. In manufacturing, much dissemination occurs not just by the expansion of the innovating firm but by the sale of the new machinery or other inputs that embody the new technology. Within any country there is a tremendous range of productivity in any sector. Therefore there are great returns to being able to raise average productivity to local best practice (or even better to global best practice by acquiring more knowledge from abroad). This requires stronger public policies focusing on dissemination and use, which includes standards and quality control legislation. It also requires an institutional infrastructure consisting of technical information services; extension services for agriculture, industry and services; productivity organizations; metrology standards; quality control institutions and industrial clusters. There are significant differences between the three countries in the speed of dissemination of knowledge. Table 1.8 presents data on two basic infrastructure technologies (electricity per capita and access to


38

Carl Dahlman

Table 1.8 Diffusion of technology

Electricity consumption (kWh/capita ) 1995 2006 Percentage of population with access to improved sanitation facilities 1990 2007 Life expectancy at birth 1980 2007 Under 5 mortality rate per 1000 1990 2007 Televisions per 1000 persons 1996 Percentage of households with television2006 Fixed and mobile phones per 1000 persons 1996 2007 Computers per 1000 persons 1996 2007 Internet users per 1000 persons 1997 2007 Fixed broadband Internet subscribers per 1000 persons 2004 2007 Fixed broadband Internet tariff US$/ month2008 ICT expenditure as a percentage of GDP 1999 2007 Cereal yields kg per hectare 1990–2 2005–7 High technology exports 1995 Value in billions % manufactured exports 2007 Value in billions % manufactured exports GDP per capita 1980 2007

Brazil

China

India

1610 2060

637 2041

339 503

71 77

48 65

14 28

63 72

67 73

54 65

58 22 289 91

45 22 252 89

117 72 64 53

112 840

51 700

15 250

18.4 161

3.0 57

1.5 33

0.42 35.2

0.02 16.1

0.01 7.2

12.4 35.4 47

16.5 50.4 19

0.6 2.8 6

5.8 5.8

4.9 7.9

3.5 5.6

1916 3206

4306 5322

1947 2464

4.5 18

26.9 21

2.4 10

9.3 12

337.0 30

4.9 5

1942 6840

204 2432

251 1046

Source: World Bank WDI various years.



39

sanitation), two broad indicators of health which represent the diffusion of better preventive medicine and basic health technologies (life expectancy and under five mortality rate), the penetration rates for four ICT-related technologies (televisions, fixed and mobile phones, computers and Internet access), indicators of the diffusion of technology in agriculture (cereal yields per hectare) and manufacturing (percentage of high-tech manufactured exports in total manufactured exports) and the broadest measure of diffusion of technology – per capita GDP. As the richer country, Brazil generally starts with the highest penetration rates and maintains them in most cases since the use of technology is correlated with income. This is the case with the two infrastructure technologies. However what is notable is the speed with which China nearly catches up in electricity use per capita and in access to sanitation even though its per capita income is still less than half that of Brazil. India’s progress is slower, but it is a poorer country. In the two technologies relating to health, China actually starts out with a better performance than Brazil, because the communist system put more effort into basic health than the mixed economies in the two other countries. Brazil basically catches up in both health areas, but although India does make some progress, it lags considerably behind China and Brazil. Brazil starts with higher penetration rates in all the ICTs except broadband because it is a more developed country. While India and China have not caught up, it is remarkable how rapidly they have narrowed the gap. With respect to broadband access, it is very impressive that in 2004 China already had higher broadband penetration rates than Brazil and that it increased the gap between then and 2007. The penetration rates of all the ICTs other than Internet access actually increased faster in India than in China. The rapid increase may be because prices are lower in India than in the other two countries. The very rapid rate of increase in China is also a result of the higher rate of investment in ICT there (Table 1.8). China and India start higher than Brazil in cereal yields per hectare. That is somewhat surprising, but may be related to higher labour input per hectare, as well as to much greater use of fertilizer. Average fertilizer use per hectare in China was 232 kg in 1990–2 and 315 kg in 2002–5 compared to 76 and 114 kg in India and 66 and 157 kg in Brazil over the same period (WDI, 2009). Higher fertilizer use can itself be taken as an indicator of faster diffusion of that agricultural technology. In terms of the manufacturing, the share of high-tech products, in 1995 was already higher in China than in Brazil, and twice as high as in India. Moreover, while it has fallen in both Brazil and India, it increased


40 Carl Dahlman

by half in China to 30 per cent of manufactured exports. Part of the reason for the much higher intensity of technology in China is that it imports high-tech components from Japan, Taiwan and Singapore and uses them in exports. In addition much of the high-tech exports are made by foreign firms. However it does illustrate how effectively China is using foreign technology. In terms of GDP per capita (in nominal US dollars) as the broadest measure of productivity, Brazil started with a level 9.5 times that of China and 7.7 times that of India. Over the last 27 years China’s per capita income has increased to 22 times its 1980 level, while India’s has increased to four times and Brazil’s to three times. Thus Brazil’s per capita income is now just 2.8 times that of China and the latter’s is now 2.3 times that of India. This is a result of China’s more systematic use of foreign technology as well as to its higher rate of investment. Various observations can be drawn from these country comparisons. The first is the importance of strong diffusion efforts. These have been quite systematic in China. They have included not only increased work in agriculture, but also special programmes such as Spark for rural innovations and Torch for high-tech innovation. In India and Brazil they have been mostly in agriculture the green revolution in India and the extensive research and dissemination efforts through Embrapa in Brazil. Although both Brazil and India have set up dissemination efforts in manufacturing and services, these have not been as systematic or successful as those of China. It is also likely that the more competitive environment in China has led to more rapid diffusion through imitation as well as rapid expansion by the more innovative firms. The second observation is the importance of high investment rates. Thirty years ago Brazil had a significantly higher per capita income and was ahead of China and India in virtually all areas except population. In less than three decades China and India have grown faster and have outdone it in economic size and total trade and have narrowed the gap in per capita income and most economic and social indicators. The better performance of China and India has resulted from more stable macro conditions22 and higher rates of investment. China has grown faster than the other two countries because its investment rate as a share of GDP has been from 40 to over 100 per cent higher than in the two other countries (Table 1.9). While it is true that a lot of this investment has been inefficient, it is also true that high rates of investment allow the embodiment of new technology. In the case of China 60–75 per cent of its imports (which have increased from 18 to 30 per cent of its GDP) are manufactured products, most of which are components



41

Table 1.9 Gross fixed investment as a percentage of GDP 1980–2007

1980 1990 2000 2007

Brazil

China

India

23 20 23 18

35 35 38 43

21 25 25 39

Source: World Bank WDI various years.

and capital goods. High investment rates are important not only in the sense of providing more capital per worker (capital deepening), but also in terms of permitting the use of capital that embodies new technologies, particularly if the country makes use of more efficient imported capital equipment, as has been the case in China. As noted, the rapid penetration rate of ICTs in China is also related to the higher rate of investment in ICT as a share of GDP than in the other two economies.

1.6

Conclusion

The main conclusion that can be drawn from this comparison, besides the fact that a country’s technology strategy needs to change according to its stage of development, is that all three countries have developed considerable technological capability. Over the last two decades there has been a significant increase in the technological capability of many developing countries. The three countries covered in this chapter account for a large part of that increase in capabilities, although there are important differences among them. One way to get some sense of the relative strengths of the three countries is to compare various measures of technological capabilities to the size of their economies and population in the world (Table 1.10). Some of most notable points are the as follows. China has a larger share of merchandise exports in the world than its share of world GNI, although the figure is not as large as its share of world GNI in PPP terms. More significant, however, is that its share of world high-tech manufactured exports is more than twice its share of merchandise exports. This clearly indicates that it already has a comparative advantage in hightech exports. Although China imports many of the components for these high-tech exports from other Asian economies (such as Japan,


42

Carl Dahlman

Table 1.10 Percentage shares of size of economy and population and of various estimates of technological capability for Brazil, China and India in the world World GNI 2007 GNI in PPP 2007 Merchandise exports 2007 High technology manufactured exports 2007 Commercial service exports 2007 Computer information communications and other commercial services 2007 Population 2007 Tertiary students 2007 Researchers in R&D 2007* R&D expenditure in 2007** Scientific and technical journal articles 2005 Patent applications filed by residents of country 2007 Triadic patent families 2005

Brazil

China

India

2.1 2.7 1.2 0.5 0.7 0.8

5.9 10.9 8.7 20.3 3.6 3.7

2.0 4.7 1.0 0.3 2.7 6.4

2.9 3.5 0.4 1.2 1.4 0.4

19.9 16.8 19.8 8.6 5.9 15.1

17.0 8.5 2.0 1.8 2.1 0.4

0.1

0.8

0.2

Notes: * Total number of researchers in world estimated by author from information in WDI 2009, but figures given for India appear to be low in original source. **Estimates made by author using R&D spending as a percentage of GDP from WDI and 2007. GNI figures for various regions countries and world. The R&D as a percentage of GDP is clearly underestimated in WDI for Brazil and India so their shares in the world are underestimated.

Hong Kong, Singapore, Taiwan and others), there is trend for many of the producers of these components to set up plants in China because of the rising capabilities and economies of scale of producing in China (see Gill and Kharas, 2007). Brazil’s shares of world merchandise and high-tech exports are much lower than its share of world GNI in nominal or absolute terms. The same is true for India, whose share of high-tech exports is even smaller than Brazil’s. However, India has a share of service exports relatively higher than its share of global GNI. Moreover its share of computer information, communications and other commercial services (which accounts for 74 per cent of its total commercial service exports) is much greater than it share of global GNI (even in PPP terms). Thus India clearly has developed a comparative advantage in the export of ICT-enabled services. In a sense, China has become the manufacturing workshop for the world. On the other hand, India’s strength has been in ICT-enabled services. Therefore in terms of the two unbundlings referred to by Baldwin (2006), China has benefited the most from the unbundling of



43

production, while India has benefited the most from the unbundling of tasks, facilitated by the rapid advances in ICT. Brazil is very weak in these types of exports and in commercial service exports in general and it has not taken advantage of either of the two unbundlings. Brazil’s relative export strength is in resource-based merchandise exports because of its strong natural resource endowments, but it is still not a very exportoriented economy. However, it has made impressive improvements in agricultural productivity and has developed some islands of excellence, including the export of aeroplanes, the development of alternative fuels and deep sea oil exploration platforms among others. Tertiary students can be taken as a quick proxy for investments in high-level human capability. On this indicator, Brazil does relatively better than China or India because it has a greater figure for tertiary students than for population. However, what is impressive on this measure it that because of its high ramping up of tertiary education, for China this figure is already approaching that for its share of global population. This is a very important long-term asset for making effective use of knowledge as well as for creating new knowledge. In addition, as is true for other knowledge assets, absolute size matters since knowledge is not consumed in its use. Although the quality of tertiary education in China and India is not yet up to standards of the developed economies, both countries are working on improving the quality and relevance of tertiary education. Investment in education will greatly leverage the capabilities of their large workforces and give them strong competitive advantages. The data on R&D expenditure in the table underestimate those of India and to some extent Brazil as they are not up to date. However even if some adjustments are made to these numbers it is clear that China is ahead of Brazil and India in R&D expenditure. This can also be seen in the much larger number of researchers in R&D in China where its share of the world figure is already proportional to that for its very large population. China’s greater strength in R&D can be seen in the much larger share of scientific and journal articles where, again, absolute numbers matter. It can also be seen in the much larger share of patent applications filed by its residents. When patent regimes are standardized by taking triadic patents, China’s share falls very significantly. However, its share of triadic patents is still proportionally greater relative to its population than those of Brazil and India. Moreover, as pointed out before, the average annual rate of growth of its triadic patents is 37 per cent, which is the highest among the countries registering triadic patents. India is second with an average annual rate of growth of


44

Carl Dahlman

28 per cent. It is clear that both China and India are well on their way to becoming major technological powers. Moreover, since China and India have been expanding their participation in the world economy rapidly they are going to have strong positive and negative effects on the rest of the world. The positive effects are additional growth poles. This will be manifested especially through additional demands for raw materials and commodity inputs. Until the economic crisis, this was manifested through increased prices for commodities. This was particularly strong in terms of increased prices for energy. This of course is a windfall for commodity exporters, including rich countries like Australia and Canada, middle-income economies including Brazil and poor countries with good natural resources in Africa. The negative side however is increasing competition in exports of manufactures. This is affecting producers of labour-intensive exports – for example exporters of textile products from Africa and the ‘maquila’ industries from Mexico and the Caribbean. However, China and India are moving up the technology ladder quite rapidly. In addition, by investing more in education and R&D they will be leveraging their large labour forces and putting a lot of pressure on relative wages through international competition. Furthermore, their continued rapid growth will put additional pressure on limited resources and on the capacity of the environment to absorb greenhouse gases. These are major issues that will require global attention. Improved technology can be part of the answer. The increased technological capability of these two countries can play an important role in finding more sustainable development strategies. However technology will not be enough. Addressing these issues will require concerted action on the part of the world community as well as major technology, financial and regulatory agreements.

Notes This paper is dedicated to the memory of Sanjaya Lall, who wrote prolifically on innovation issues, including on these three countries, and on their implications for other countries – see for example Lall (2003, 2004) and Lall and Weiss (2005). For an early review see Pavitt and Walker (1976), for recent reviews see Fagerberg et al. (2006) and Auerswald and Branscomb (2008). 1. For an early review see Pavitt and Walker (1976), for recent reviews see Fagerberg et al. (2008) and Auerswald and Branscomb (2008). 2. The Organisation for Economic Co-operation and Development (OECD) has gone through three phases. In the 1970s and 1980s it focused on science policy reviews. In the 1990s it expanded its focus to innovation studies and developed the Oslo Manual for the collection of innovation information.



3. 4.

5.

6. 7. 8. 9.

10.

11.

12. 13.

14.

45

More recently it has expanded its focus to consider ‘broader framework conditions’ that affect innovation. The third revision of the Oslo Manual in 2005 added organizational innovation as well as an appendix based on the Bogota Manual developed by Latin American researchers in RICYT (Ibero-American Network for Research on Science and Technology) collecting innovation in developing countries. See OECD (2005b). For useful summary on the evolution of innovation studies see Soete (2008). See the analysis carried out based on the European Community Innovation Surveys available in European Commission (2008). Aghion and Howitt have developed a U-shaped model of innovation which argues that firms that are close to the frontier tend to undertake more R&D whereas firms that are far from the frontier do not. In another paper Aghion and Howitt (2006) they develop the idea that Europe was not able to reduce the productivity gap with the US because it was locked in for too long on imitating and not enough on pushing back the frontier. However, that does not explain why some firms and some countries do actually manage to catch up even when coming from far behind. This can be seen, for example, in the increase in patenting activity, the proliferation of new products and processes, the shortening of product life cycles and rapid advances in ICT. See Jaruzelski and Dehoff (2008) for a listing of the global top 20 firms and an analysis of the spending of the top 1000 firms. Largest in terms of GNI or GDP in 2007. In 2007 India was reclassified by the World Bank as a lower-middle-income country as it just barely made the cut-off of $935 per capita. This is because Brazil has one of the most unequal distributions of income in the world as revealed by a Gini coefficient of 0.57 compared to 0.47 for China and just 0.37 for India. China’s high degree of trade integration was second only to Germany’s among the world’s large economies in 2007, and it will surpass Germany soon For a good analysis of China’s progressive entry into the global system see Naughton (2007), chapter 16: ‘International Trade’, and chapter 17: ‘Foreign Investment’, 375–424. Between 1970 and 1980 while China and India were not receiving any FDI, inflows to Brazil averaged about 1 per cent of its GDP (UNCTAD, 2008). Motorola, for example, was forced to develop an extensive training programme for the management of the 1000 largest Chinese SOEs (Dahlman, Zeng and Wang, 2007). Studies tracing the relationship between the stock of education and the long-term level of GDP find that a one-year increase in average educational attainment raises the level of output per capita by between 3 and 6 percentage points. Studies examining the relationship between the stock of education and the rate of growth of GDP find that an increase of one year of education raises the growth rate of GDP by around 1 percentage point. The cumulative impact of a 1 percentage-point increase in the rate of growth soon exceeds the one-time increase in output. Rising labour productivity accounted for at least half of GDP per capita growth in OECD countries from 1990 to 2000 (OECD 2005a). However, research suggests that there are


46

15.

16.

17.

18.

19. 20. 21.

22.

Carl Dahlman diminishing effects of growth above an average of seven and a half years of education. See Krueger and Lindale (2001). See Rodriguez, Dahlman and Salmi (2008) for an analysis of how the poor quality of basic and secondary education has been a brake on the ability of the Brazilian economy to absorb existing knowledge. There are, however, major differences in quality of the students in the three countries. It is significantly lower in the two developing countries because of the use of outdated textbooks and poor quality assurance systems. In addition in China, the very rapid ramp up of the educational system (increasing new entrants by 50% starting in 1998) has lead to many quality problems from the sheer difficulty of expanding the number of qualified teachers to meet the rising enrollments. Once knowledge is produced, it is potentially available to all, provided that it is made public in say a scientific or technical journal, or in a patent. However, insufficient education to understand the article or IPR restrictions may limit its actual use. In the old PPP series, China’s R&D expenditure surpassed Japan’s by the end of 2006, as was in fact reported by in OECD (2007). However, the new PPP series released in December 2007 reduced the PPP estimates for GDP in China and India by 40 per cent each. For more detailed information on R&D spending and innovation in India see Dutz (2007). See Xue (2007) for more details on this. Except for in IT firms in India where at any given time up to one-third of all the workers are taking some sort of training to keep on top of the fastchanging technology. China had has relatively stable macro conditions, including low inflation, low interest rates and a stable exchange rate. This has made for a predictable business environment which both domestic and foreign investors like. India has also had relative stable macro conditions since 1980, except for the financial crisis at the end of the 1980s that culminated in the financial crisis of 1991. Brazil on the other hand had a very unstable macro situation between 1980 and the early 2000s when it finally managed to attain macro stability.

References Aghion, Philippe and Peter Howitt (2006), ‘Appropriate growth policy: A unifying framework’, Journal of the European Economic Association, 4 (2/3), 269–314. Auerswald, Philip and Lewis Branscomb (2008), ‘Research and innovation in a networked world’, Technology and Society, 30, 339–47. Baldwin, Richard (2006), ‘Globalization: The great unbundling(s)’, paper contributed to event on Globalization Challenges to Europe and Finland organized by the Secretariat of the Economic Council, Prime Minister’s Office (June). Dahlman, Carl, Douglas Zhihua Zeng and Shuilin Wang (2007), Enhancing China’s Competitiveness through Life Long Learning. Washington, DC: World Bank.



47

Dutz, Mark A. (ed.) (2007), Unleashing India’s Innovation: Towards Sustainable and Inclusive Growth. Washington, DC: World Bank. European Commission (2008), European Innovation Scoreboard 2007 – Interactive benchmarking. http://www.proinno-europe.eu/ Fagerberg, J.D, D. Mowery and R. Nelson, eds (2006), Handbook of Innovation. Oxford: Oxford University Press. Gill, Indermit and Homi Kharas (2007), An East Asian Renaissance: Ideas for Economic Growth. Washington, DC: World Bank. IMF (2009), ‘World Economic Outlook July Update’. Washington, DC: IMF. Krueger, A.B and M. Lindale (2001), ‘Education and growth: Why and for whom?’ Journal of Economic Literature, XXXIX. Jaruzelski, B. and K. Dehoff (2008), ‘Beyond borders: The global innovation 1000’, in Booz Allen Hamilton, Strategy and Business (53). Lall, Sanjaya (2003), ‘Reinventing industrial strategy: The role of government policy in building industrial competitiveness’, Working Paper No. 111 (October). Oxford: Queen Elizabeth House. Lall, Sanjaya (2004), ‘Industrial success and failure in a globalizing world’, International Journal of Technology Management and Sustainable Development, 3 (3), 189–213. Lall, Sanjaya and John Weiss (2005), ‘China’s competitive threat to Latin America: An analysis for 1990–2002’, Working Paper No. 120. Oxford: Queen Elizabeth House. Naughton, Barry (2007), The Chinese Economy: Transitions and Growth. Cambridge, MA: MIT University Press. Organisation for Economic Co-operation and Development (OECD) (2005a), Education at a Glance. Paris: OECD. Organisation for Economic Co-operation and Development (OECD) (2005b), Oslo Manual: Guidelines for Collecting and Interpreting Innovation Data. Paris: OECD. Organisation for Economic Co-operation and Development (OECD) (2007), Science Technology and Industry Scoreboard. Paris: OECD. Organisation for Economic Co-operation and Development (OECD) (2008), Science Technology and Industry Outlook. Paris: OECD. Pavitt, Keith and William Walker (1976), ‘Government policy towards industrial innovation: A review’, Research Policy, 5, 11–97. Rodriguez, Alberto, Carl J. Dahlman and Jamil Salmi (2008), Brazil: Knowledge and Innovation for Competitiveness. Washington, DC: World Bank. Soete, Luc (2008), ‘Science, technology and development. Emerging concepts and visions’, Maastrict University UNU-MERIT Working Paper Series, 2008–001. UNCTAD (2008), Handbook of Statistics. Geneva: UNCTAD. UNESCO (2009), Global Education Digest. Montreal: UNESCO Institute of Statistics. World Bank (2008a), ‘Knowledge assessment methodology 2008’, see www. worldbank.org/kam. World Bank (2008b), World Development Indicators 2008. Washington, DC: World Bank.


48

Carl Dahlman

World Bank (2009), World Development Indicators 2009. Washington, DC: World Bank. World Economic Forum (WEF) (2009/10 and 2000), Global Competitiveness Report. Geneva: WEF. Xue, Lan (2007), ‘The changing roles of universities in China’s national innovation system’. PowerPoint presentation at World Bank Conference in Prague.


2 Economic Growth and Technological Capabilities in BRICS: Implications for Latecomers to Industrialization Deepak Nayyar

2.1

Introduction

This paper attempts to analyse the implications of the rise of Brazil, India, China and South Africa for developing countries, situated in the wider context of the world economy, with some focus on latecomers to industrialization. Section 2.2 highlights the economic significance of Brazil, India, China and South Africa in the world economy at the beginning of the twenty-first century, in comparison with the past, and outlines their performance in terms of economic growth during the second half of the twentieth century. Section 2.3 asks whether these countries could be the new engines of growth for the world, beginning with history and statistics to touch upon the underlying economic causation. Section 2.4 examines the possible impact of rapid growth in the four economies on developing countries, which could be complementary or competitive and, on balance, positive or negative. Section 2.5 analyses the nature of technological development in these emerging economies – foundations, dilemmas and specificities – to explore the lessons for late latecomers to industrialization.

2.2 Brazil, India, China and South Africa in the world economy The Emerging significance of Brazil, India, China and South Africa in the world economy must be situated in a historical perspective. Table 2.1, 49


50 Deepak Nayyar Table 2.1 Brazil, India, China and South Africa in the world economy: Share in world population and world GDP, 1820–2001 Year

Brazil

India

Percentage share of world population 1820 0.4 20.1 1870 0.8 19.9 1913 1.3 17.0 1950 2.1 14.2 1973 2.6 14.8 2001 2.9 16.5 Percentage share of world income 1820 0.4 16.0 1870 0.6 12.1 1913 0.7 7.5 1950 1.7 4.2 1973 2.5 3.1 2001 2.7 5.4

China

South Africa

36.6 28.1 24.4 21.7 22.5 20.7

0.1 0.2 0.3 0.5 0.6 0.7

32.9 17.1 8.8 4.5 4.6 12.3

0.1 0.2 0.4 0.6 0.6 0.5

Note: The percentages in this table have been calculated from estimates of population and GDP in Maddison (2003). The data on GDP are in 1990 international Geary-Khamis dollars, which are PPPs used to evaluate output that are calculated based on a specific method devised to define international prices. This measure facilitates inter-country comparisons over time. Source: Maddison (2003).

which is based on estimates made by Maddison (2003), presents evidence on the shares of Brazil, India, China and South Africa in both world population and world income for selected years during the period from 1820 to 2001. The table shows that, in 1820, these four countries accounted for 57 per cent of world population and almost 50 per cent of world income. There was a dramatic change over the next 150 years. In 1973, the share of these four countries of the world population was significantly lower at about 40 per cent but their share in world income had collapsed to less than 11 per cent, which was a small fraction of what it was 150 years earlier. The next 30 years witnessed some recovery. While the share of the four countries in world population remained at about 40 per cent, their share in world income had risen to almost 21 per cent in 2001. These aggregates reveal the essential contours, but also conceal some aspects of the story. There are similarities between China and India, just as there are similarities between Brazil and South Africa. But there are also significant differences between the two sets of countries. For much of the time, India and China had dominant shares.1 From 1820 the share of India and China in world population declined


Implications for Latecomers to Industrialization

51

steadily until 1973 but, over the same period, the decline in their share of world income was much more pronounced. Consequently, during the period from 1820 to 1973, there was a sharp increase in the asymmetries, or disproportionalities, between the shares of India and China in world population and in world income. The partial recovery in their share of world income during the period from 1973 to 2001 has reduced the asymmetry but the disproportionality remains significant. For much of the time, the shares of Brazil and South Africa were far smaller. But there were also other important differences. For one, the shares of Brazil and South Africa in both world population and world income increased, even if slowly for some of the time, throughout this period. For another, the shares of Brazil and South Africa in world population and in world income were symmetrical and proportional throughout this period. It is possible to juxtapose this past with the present. Table 2.2 outlines a profile of Gross Domestic Product (GDP), population and GDP per capita in the BRICS as compared with developing countries, industrialized countries and the world, in 2000 and 2005. It shows that the population of the world is more than 6 billion, of which a little less than 1 billion is in the industrialized countries, somewhat more than 5 billion is in the developing countries and more than 2.5 billion is in the four countries under study. Thus, 40 per cent of the population of the world and 50 per cent of the population of developing countries lives in the BRICS. There are two sets of figures on GDP and GDP per capita: at constant prices with market exchange rates and in terms of purchasing power parities. Consider each in turn. At market exchange rates, between 2000 and 2005, the share of the four countries increased from 7 to 9 per cent of world GDP and from 39 to 43 per cent of GDP in developing countries. Over the same period, at market exchange rates, GDP per capita in Brazil and South Africa was more than double, GDP per capita in China was about the same, while GDP per capita for India was less than half of the average GDP per capita in developing countries. It is worth noting that the BRICS are far below the level of GDP per capita in the industrialized countries and significantly below GDP per capita in the world as a whole. The picture is somewhat different if the comparison is in terms of purchasing power parity (PPP). Between 2000 and 2005, the share of the four countries increased from 20 to 24 per cent of world PPP-GDP and from 48 to 52 per cent of the PPPGDP of developing countries. It would seem that for Brazil, India, China and South Africa taken together, these shares in world income are now much more symmetrical with their share in world population. Over the


Table 2.2 GDP, population and GDP per capita: Brazil, India, China and South Africa, 2000 and 2005 GDP (US$ billion)

Population

(US$ per capita)

(million)

PPP-GDP (US$ billion)

(US$ per capita)

Country

2000

2005

2000

2005

2000

2005

2000

2005

2000

2005

Brazil India China South Africa Total Above Developing countries (BRICS as a percentage of developing countries) Industrialized countries World (BRICS as a percentage of world)

602 460 1198 133 2393 6058 (39.5)

670 644 1890 160 3364 7813 (43.1)

3461 453 949 3020 – 1191 –

3597 588 1449 3406 – 1440 –

174 1016 1263 44 2496 5085 (49.1)

186 1095 1305 47 2632 5427 (48.5)

1251 2402 4973 386 9011 18818 (47.9)

1393 3362 7842 463 13061 25322 (51.6)

7193 2364 3939 8764 – 3701 –

7475 3072 6012 9884 – 4666

24,542

27,148

27,304

29,251

899

928

25,157

27,898

27,988

30,058

31,756 (7.5)

36,352 (9.3)

5241 –

5647 –

6060 (41.2)

6438 (40.9)

45144 (20.0)

54573 (23.9)

7450 –

8477 –

Notes: GDP and GDP per capita are measured in constant 2000 US$. PPP-GDP and PPP-GDP per capita are measured in constant 2000 international dollars. Source: World Bank (2007).


–


53

same period, in PPP terms, GDP per capita in Brazil and South Africa was more than double the GDP per capita of developing countries and close to the world average. In comparison, GDP per capita in China moved ahead of GDP per capita of developing countries, whereas GDP per capita in India was about two-thirds of GDP per capita of developing countries. This snapshot picture situates Brazil, India, China and South Africa in the world economy at the present. But the observed reality has been shaped by their economic performance in the past. Table 2.3 sets out rates of growth in GDP and GDP per capita, during the period 1951–80 and 1981–2005 for Brazil, India, China and South Africa, in comparison

Table 2.3 Growth performance of Brazil, India, China and South Africa, 1951–80 and 1981–2005 Comparison with Regions and Country-Groups (per cent per annum)

GDP Brazil India China South Africa Asia Latin America Africa Developing countries Industrialized countries World GDP per capita Brazil India China South Africa Asia Latin America Africa Developing countries Industrialized countries World

1951–80

1981–2005

6.78 3.57 5.03 4.48 6.28 4.69 4.12 4.84 4.40 4.77

2.04 5.79 9.73 2.01 4.06 2.26 2.97 3.04 2.50 2.95

3.85 1.40 3.01 1.85 2.90 2.11 1.66 2.19 3.50 2.40

0.33 3.83 8.51 0.04 1.63 0.44 0.39 0.80 1.96 0.99

Source: Nayyar (2008b).


54

Deepak Nayyar

with regions within the developing world, developing countries, industrialized countries and the world economy.2 A study of Table 2.3 clearly shows that growth in GDP and GDP per capita in the period 1981–2005 was much slower than it was during 1951–80. This was also true for the world economy, for industrialized countries and for developing countries. Growth in GDP was in the range 4–5 per cent per annum during 1951–80 and in the range of 2–3 per cent per annum during 1981–2005 almost everywhere, except Asia where it was 6 and 4 per cent per annum respectively. Growth in GDP per capita slowed down considerably even in the industrialized countries, from 3.5 to 2 per cent per annum, but the slowdown was more pronounced for developing countries, from 2.2 to 0.8 per cent per annum. In Latin America and Africa, during the 1981–2005 period, growth in GDP per capita was less than 0.5 per cent per annum, while Asia fared better, at more than 1.5 per cent per annum. The economic performance of Brazil, India, China and South Africa presents a mixed picture, which does not quite conform to the trends observed in the aggregate. The most striking contrast is that China and India were the clear exceptions to this worldwide slowdown in growth. In both countries, growth rates in the second period were much higher than the perfectly respectable growth rates of the first period. So much so that, between 1951–80 and 1981–2005, average annual growth in GDP per capita almost trebled in both China and India. This was attributable in part to higher GDP growth rates and in part to lower population growth rates. Unlike China and India, however, Brazil and South Africa were a part of the worldwide slowdown in growth. In both countries, growth rates in the second period were much lower than the impressive growth rates of the first period. So much so that, during 1981–2005, average annual growth in GDP per capita was almost negligible in both Brazil and South Africa. But it is also worth noting that, during the 1951–80 period, average annual growth in both GDP and GDP per capita in Brazil was significantly higher than that in China and India. The growth performance of South Africa during 1951–80 was also better than that of India although it did not quite match that of China.

2.3

Engines of economic growth in the world

Globalization is associated with increasing economic openness, growing economic interdependence and deepening economic integration in the world economy. In such a world, growth prospects would be



55

significantly influenced, if not shaped, by the growth performance of leading economies. The discussion that follows asks whether Brazil, India, China and South Africa could be new engines of growth and considers the underlying causation and mechanisms.3 2.3.1 Engines of growth History provides obvious examples. The UK in the nineteenth century and the US in the twentieth century were engines of growth for the world economy. Statistical analysis for the period since the early 1960s provides confirmation.4 It is widely accepted that GDP growth in the US leads GDP growth in the world. A statistical analysis of long-term trends in economic growth, for the period 1963–2001, with five-year moving averages for both sets of growth rates, yields a correlation coefficient of 0.82, while a simple lead-lag analysis shows that the US economy leads the world economy by one year. Available evidence also reveals that developing countries, excluding China, follow trends in world economic growth and, thus, trends in economic growth of the US. It is worth noting that economic growth in developing countries follows economic growth in the US with a lag but with more pronounced swings in cyclical ups and downs. In reflecting on the future, is it possible to think of Brazil, India, China and South Africa as engines of growth for the developing world, even if not for the world economy? The answer depends, in large part, on the size of the four economies and their rates of growth. There are some pointers in recent experiences. Statistical analysis shows that, since 1980, the Chinese economy has also led world GDP, with a lag of one or two years, although the correlation coefficient is much smaller than that for the US (see United Nations, 2006: 22–3).5 This is not surprising. First, in 2005, China accounted for 5 per cent of world GDP at market exchange rates and 14 per cent of world GDP in PPP terms. Second, GDP growth in China has been about 9 per cent per annum for 25 years. By the same criteria, India is not an engine of growth, at least not yet. This is also not surprising. Its economic size is smaller than that of China and its growth rate is not as high. First, in 2005, India accounted for only 2 per cent of world GDP at market exchange rates and 6 per cent of world GDP in PPP terms. Second, GDP growth in India has been about 6 per cent per annum for 25 years. Even so, India is a potential engine of growth in terms of both attributes. Brazil presents a mixed picture. Its economic size is significant. In 2005, Brazil accounted for 2.5 per cent of world GDP both at market exchange rates and in PPP terms. But GDP growth was just 2 per cent per annum during the 1981–2005 period.


56 Deepak Nayyar

Brazil has the economic size but not the growth rate to drive the world economy. South Africa provides a sharp contrast. In 2005, it accounted for only 0.6 per cent of world GDP at market exchange rates and 0.8 per cent of world GDP in PPP terms. And GDP growth was only 2 per cent per annum in 1981–2005. Clearly, South Africa meets neither criteria, whether size or growth. Of course, given their economic size in relation to most developing countries, the four countries together could be a possible engine of growth for the developing world, but that would depend on the degree and the nature of linkages. Rapid economic growth in leading economies drives economic growth elsewhere in the world by providing markets for exports, resources for investment, finances for development and technologies for productivity. The classic examples – the UK in the nineteenth century and the US in the twentieth century – provide confirmation of the suggested economic causation and the possible transmission mechanisms. Indeed, during their periods of dominance in the world, both the UK and the US were engines of growth, in so far as they provided the rest of the world not only with markets for exports and resources for investment, but also with finances for development and technologies for productivity. Moreover, despite its diminished dominance, the US economy continues to be an engine of growth for the world. At this juncture, China is not quite an engine of growth in every dimension. Economic growth in China provides a stimulus to economic growth elsewhere, in large part, as a market for exports. So far, India and Brazil cannot be characterized as engines of growth in any dimension, perhaps not even as markets for exports. But, along with China, India and Brazil have some future potential in terms of markets for exports, resources for investment and technologies for productivity. South Africa is the obvious outlier in this picture, although it could provide some impetus to the growth process in Africa if linkages develop appropriately. In this context, it is worth noting that the four economies could, in times to come, provide a significant impetus to the growth process in their respective regions. What is more, even if the BRICS cannot be a substitute for the US as an engine of growth for the world economy, they could be an important complement to the older engine in driving global growth. 2.3.2

Causation and mechanisms

The economic causation outlined above is necessary but not sufficient. The overall effects of economic growth in leading economies on economic growth elsewhere depend upon: 1) whether such growth is



57

complementary or competitive, 2) whether the direct effects are reinforced or counteracted by the indirect effects and 3) whether, on balance, the impact is positive or negative.6 In principle, economic growth in leading economies may be complementary or competitive to economic growth elsewhere. It may be complementary in so far as it increases the demand for exports but it may be competitive in so far as it develops alternative sources of supply. It may be complementary if it provides resources for investment or finances for development but it may be competitive if it pre-empts such resources for investment or finances for development. It may be complementary if it provides technologies to others but it may be competitive if it stifles the development of technologies elsewhere. This distinction between the complementary and the competitive aspects is widely recognized. However, the distinction between direct effects and indirect effects is less clear because the latter sometimes are difficult to discern let alone measure. In situations where direct effects are complementary, indirect effects could be reinforcing if complementary but counter-acting if competitive. Some examples might be illustrative. The direct effects may be complementary if the leading economies, say the BRICS, provide cheap wage goods to other developing countries, but the indirect effects may be competitive if competition from firms in leading economies squeezes out local firms in other developing countries. The direct effects may be complementary if firms from these leading economies invest in other developing countries, but the indirect effects may be competitive if firms from industrialized countries relocate production and invest in Brazil, India, China or South Africa rather than in other developing countries. The direct effects may be complementary if these leading economies provide cheaper inputs for manufactured exports from other developing countries but the indirect effects may be competitive if competition from the leading economies squeezes manufactured exports out from other developing countries in the markets of industrialized countries. In principle, then, the impact of economic growth in leading economies on economic growth elsewhere, in different spheres, could be positive, negative or some combination of both. Therefore, on balance, such impacts can be either positive or negative. The outcomes may differ across space and change over time so that generalizations are difficult.7 The main mechanisms of interaction, through which outcomes would be shaped, are international trade, investment and finance.8 In this context, however, it is worth noting that domestic developments within such large countries could also have international consequences.


58 Deepak Nayyar

For instance, macro-economic policies in China, India and Brazil, once they become leading economies, may exercise an important influence on economic growth elsewhere. If such policies are counter-cyclical, which has been the case for the US, they would be supportive of economic growth elsewhere. But if these policies are pro-cyclical, which is common in developing countries, they could be disruptive for economic growth elsewhere. Similarly, exchange rates and interest rates in leading economies could exercise a significant influence, either positive or negative, on economic growth elsewhere in the world. For example, the undervalued exchange rate in China, which has persisted for quite some time, constrains prospects for labour-intensive manufactured exports from other developing countries, thereby limiting the potential demand stimulus to economic growth that could be provided by exports. Similarly, the combination of a high interest rate and a strong exchange rate, which has been the case for some time in both India and Brazil albeit for different reasons, pre-empts possible foreign capital inflows thereby limiting the potential external finance necessary to support economic growth in other developing countries.

2.4 Economic growth in BRICS: Impact on developing countries During the first quarter of the twenty-first century, economic growth in Brazil, India, China and South Africa could exercise a significant influence on prospects for developing countries. This impact could be either positive or negative.9 The emergence of BRICS in the world economy could have a positive impact on developing countries if it improves terms of trade, provides appropriate technologies and creates new sources of finance for development, whether investment or aid. Consider each in turn. It is clear that, for some time to come, the positive impact on developing countries would be transmitted through an improvement in their terms of trade.10 Rapid economic growth in China and India is bound to boost the demand for primary commodities exported by developing countries. The reasons are simple enough. Both China and India have large populations. But that is not all. In both countries, levels of consumption per capita in most primary commodities are low, while income elasticities of demand for most primary commodities are high. This burgeoning demand will almost certainly raise the prices of primary commodities in world markets and thereby improve the terms of trade for developing countries. It would benefit Brazil and South Africa



59

as exporters of primary commodities, while the revival of growth in Brazil and South Africa, when it happens, would reinforce this process. What is more, China already is, while India and Brazil are likely to become, sources of manufactured goods in the world market. Such manufactures, particularly wage goods but also capital goods, from China, India and Brazil are likely to be cheaper than competing goods from industrialized countries. At the same time, Brazil and South Africa could provide cheaper natural-resource-based manufactures. This would also improve the terms of trade for developing countries. The positive impact of the BRICS on developing countries through the other potential channels of transmission is not as clear. We do not yet have either the evidence or the experience. In principle, it is possible that they would develop technologies that are more appropriate for the factor endowments and the economic needs of developing countries. But it is too early to come to a judgment on this matter. Similarly, Brazil, India, China and South Africa are potential sources of finance for development. Their foreign aid programmes, particularly in Africa, constitute a modest beginning.11 But their contribution in terms of foreign direct investment (FDI) is limited so far. The emergence of the four countries in the world economy could also have a negative impact on developing countries if these economies provide developing countries with competition in markets for exports or as destinations for investment. Consider each in turn. At this juncture, China is clearly the largest supplier of labour-intensive manufactured goods in the world market. Even if not as large as China, India is also a significant supplier of labour-intensive manufactured goods in the world market. Brazil and South Africa are important suppliers of natural-resource-base manufactures. Brazil, India, China and South Africa are emerging suppliers of capital goods. There can be little doubt that manufactured exports from them span almost the entire range of manufactured exports in which other developing countries could have a potential comparative advantage. Hence, it is plausible to argue, if impossible to prove, that on balance the four countries could possibly have a negative impact on manufactured exports from other developing countries which have to compete with them for export markets in industrialized countries.12 This can change if and when China and India leave their place in the international trade matrix, in much the same way as latecomers to industrialization in Asia such as Japan, Korea, Hong Kong, Taiwan and Singapore left their place in the market for simple labour-intensive manufactures to countries that followed in their footsteps. It is not likely, at least in the medium term, because


60

Deepak Nayyar

both China and India have large reservoirs of surplus labour at low wages not only in the rural hinterlands but also in the urban informal sectors. Brazil and South Africa may not have such large reservoirs of surplus labour, but given their abundance of primary commodities and natural resources, it is not likely that they would leave their place in the market for processed products or resource-based manufactures to other developing countries that have similar endowments from nature. Available evidence shows that Brazil, India, China and South Africa absorb a significant proportion of inward FDI in developing countries both in terms of stocks and flows.13 Given that China, India and Brazil are now among the most attractive destinations for transnational firms seeking to locate production in the developing world, it is once again plausible to suggest, if impossible to prove, that FDI in China, India and Brazil might be at the expense of developing countries. South Africa may not have the same attraction as a destination, but it might draw FDI that could have gone to developing countries in Africa. At the same time, the share of China, India, Brazil and South Africa in outward FDI in the world economy, as it is from developing countries, is modest in both stocks and flows, indicating that firms from these four countries do not compensate with FDI in other developing countries.14 The less discernible but more significant negative impacts of the four economies, particularly China, on developing countries are the implicit barriers to change in the traditional division of labour and specialization in production. First, China and India might pre-empt opportunities for other developing countries to industrialize through exports of labour-intensive manufactures, which is attributable to their surplus labour and low wages situations that might continue for some time to come. Second, Brazil and South Africa might pre-empt opportunities for other developing countries to industrialize through agro-based or resource-intensive manufactures, which is attributable to their abundance in primary commodities and natural resources. But this is no more than a plausible hypothesis about possible future developments and cannot be tested. The problem has, however, surfaced in one dimension. China’s present division of labour with the developing world, reflected in the composition of trade flows, is no different from the old North-South pattern of trade, in so far as Chinese imports from the developing world are largely primary commodities while its exports to the developing world are largely manufactured goods.15 China’s trade with countries in South-east Asia is the exception to this rule. But Chinese trade with, and investment in, Africa confirms even more closely to this



61

caricature neocolonial pattern.16 Such traditional patterns of trade, it should be recognized, can neither transform the structure of production in developing countries nor make for a new international division of labour. Indeed, such trade can only perpetuate the dependence of developing countries on exports of primary commodities without creating the possibility of increasing value-added before export or entering into manufacturing activities characterized by economies of scale. Such path-dependent specialization can only curb the possibilities of structural transformation in developing countries. Trade with China can sustain growth and support industrialization in developing countries only if there is a successful transition from a complementary to a competitive pattern of trade, so that inter-sectoral trade is gradually replaced by intra-sectoral or intra-industry trade and specialization.

2.5 Technological development in BRICS and lessons for latecomers There are three sources of growth in emerging economies: labour absorption, capital accumulation and productivity increase. Theory and experience suggest that the contribution and the relative importance of each of these sources changes at successive stages of development. In countries that are latecomers to industrialization, the story can be set out in terms of stylized facts. To begin with, there is absorption of surplus labour from the rural sector outside agriculture into manufacturing at existing levels of wages and productivity. The process can be described as the extensive margin of labour absorption. At a later stage, there is a transfer of labour from low productivity to high-productivity occupations in the manufacturing sector while, at the same time, there is an increase in the average productivity of labour in the agricultural sector. Real wages rise in both sectors. The process can be described as the intensive margin of labour use. Capital accumulation performs a supporting role in the first stage and a leading role in the second stage. In the later stages of industrialization, productivity increase is the main source of growth. Such productivity increase depends, in part, on capital accumulation, but is ultimately determined by innovation capabilities and technological development. In an analysis of technology for industrialization in developing countries, Sanjaya Lall made an important distinction between firm-level technological capabilities at a micro level and national technological capabilities at a macro level.17 Firms operate not on a production function but at a point so that their technical progress, building upon their


62 Deepak Nayyar

own efforts, experience and skills, is localized around that point (see Atkinson and Stiglitz, 1969).18 Thus, evolutionary theories, which stress investment capabilities, production capabilities and linkage capabilities, provide a far more plausible explanation of firm-level technological change, which is a continuous process of absorption, learning and innovation.19 However, technological development in firms at a micro level is also shaped by technological capabilities in the economy at a macro level. National technological capabilities are the outcome of a complex interaction between incentives, capabilities and institutions. Each may suffer market failure and so require corrective intervention. Such interventions, which need careful formulation and application, are necessary for industrial success.20 It is exceedingly difficult, if not impossible, to provide a systematic analysis of the nature of technological development in emerging economies. Even so, if the object is to examine the implications for other latecomers to industrialization, it is important to consider the foundations of technological capabilities, the common policy dilemmas which surface and the specificities that characterize the emerging economies. The level of technological development and the capacity for innovation can be, and often are, country-specific, sector-specific or context-specific. Yet, there are discernible similarities in the essential foundations of technological capabilities among emerging economies. First, each of the BRICS created the initial conditions with an emphasis on higher education and science research in the early stages of their industrialization. This development of human resources laid the foundations of capabilities in individuals at a micro level. Second, each of them recognized the importance of learning by doing through import substitution in manufactured goods with some special effort to establish a capital goods sector. Such learning was critical to their endeavours to industrialize, for it created technological capabilities in firms at a meso level. Third, each of the four countries introduced some institutional mechanisms for catch-up in industrialization. Such mechanisms were neither mutually exclusive nor exhaustive. What is more, the relative importance of such mechanisms differed across countries and changed over time. There was an attempt to foster imitation and to leapfrog on the part of domestic firms, sometimes with explicitly or implicitly lax systems for the protection of intellectual property rights (IPR). This was often juxtaposed with a proactive technology policy in the form of strategic interventions by the government. Economic policies in the sphere of international trade and international investment were used to promote the insertion of domestic firms into global-value chains. Once domestic firms became competitive in world markets,



63

policy regimes were modified so that the acquisition of foreign firms also became a means for the acquisition of foreign technology.21 Such catch-up mechanisms created national technological capabilities at a macro level. The industrialization experience of emerging economies suggests that each of them faced some common policy dilemmas. It was recognized that an industrializing economy must be able to make a transition from importation to diffusion and innovation, at least in some sectors, so that the acquisition of technology through imports is, after a time, followed by the development of domestic technological capabilities. This meant striking a balance between imports of technology and indigenous technological development. There were situations where technologies were imported for particular sectors at a point in time but the import of such technologies was followed by stagnation rather than adaptation, diffusion and innovation at home. At the same time, in many instances, the indigenous development of technology did not lead to widespread diffusion, let alone technological upgrading. Although the underlying reasons are complex, it is clear that in such situations market structures and government policies did not combine to provide an environment that would encourage the absorption of imported technology and speed up the development of indigenous technology, or to create a milieu that would be conducive to diffusion and innovation. Even so, an open regime for the import of technology was not an answer, for the discipline of the market cannot restrain the recurrence of such imports by domestic firms time after time. Such firms are much like the schoolchild who can find someone else to write the examinations for him or her year after year and thus never learns. Domestic technological capabilities may not emerge either because there is no incentive to learn (imports are possible) or because they are stifled (imports are better). The problem may be accentuated in sectors where technical progress is rapid and obsolescence is high. Each of the emerging economies resolved this dilemma, in different ways, by striking a balance with a guiding and supportive role for the state.22 This meant planning for selective acquisition of technology where it was to be imported, setting aside resources for technology where it was to be produced at home or even deciding to opt out of a technology where it was not needed. For this purpose, it was necessary to formulate a policy regime for the import of technology, allocate resources for R&D and evolve government procurement policies. In the absence of such strategic intervention by the government, it would not have been possible for the emerging economies to develop technological capabilities at the meso and macro levels.


64

Deepak Nayyar

Striking a balance between imports of technology and indigenous technological development is a dilemma that confronts most latecomers to industrialization. But that is not all. There are two other common dilemmas. First, it is difficult to foster a culture of R&D in domestic firms. This culture does not emerge on its own. It needs incentives and disincentives, embedded in industrial policy, for start-up and scale-up. Second, it is difficult to develop synergies between science and industry that transform scientific knowledge into marketable products. This, too, does not happen on its own, but needs institutional mechanisms to build bridges between the two worlds. The emerging economies have made a modest beginning in addressing these dilemmas but the process is far from complete. Generalizations across countries are obviously difficult. The emerging economies are characterized by specificities. Some are national. Some are international. In the national context, the size of the economy matters. First, it determines the number of scientists and engineers. Second, it determines the size of the domestic market. Market size facilitates the realization of economies of scale, thus bringing about a cost reduction, just as a reduction in costs, hence prices, induces a demand expansion. In an industrializing economy, a rapid increase in the share of manufacturing in total output may then be associated with a steady decline in average costs over time. The underlying factors would be dynamic economies of scale, which are a function of cumulative past output or cumulative production experience, and an increase in labour productivity. The cost reduction, passed on to consumers in the form of lower prices, would stimulate demand expansion in domestic, as well as foreign markets. This cumulative causation, which is complex but good, is often termed the ‘Kaldor-Verdoorn Law’.23 Large domestic markets in the emerging economies, even if income levels of a large proportion of their population are low, could be conducive to innovations in products that have a mass market among low-income consumers. In the international context, which is common to all countries, there are specificities that characterize the emerging economies in terms of their capacity to exploit available opportunities. Some countries, which have high skills and low wages, are attractive choices for relocation of R&D by transnational firms. India is the obvious example. Some countries are more attractive destinations for transnational firms that seek to relocate production because of high wages or stringent regulations in their home countries. In this respect, India and China are attractive choices for relocation. Some countries have easier access to global



65

finance for the international relocation of R&D or industrial production. The emerging economies, in particular Brazil, India and China, do have such access. Some countries have a much greater capacity for the acquisition of technology through the acquisition of firms in industrialized countries. Domestic firms in Brazil, India, China and South Africa have such capabilities, which domestic firms in other developing countries may not. The preceding discussion on technological development in emerging economies, which draws upon the experience of Brazil, India, China and South Africa, suggests some lessons for other developing countries that seek to follow in their footsteps. Moreover, in conclusion, it is worth highlighting some possibilities and implications for latecomers to industrialization. In so far as the emerging economies are also late industrializers, it is possible that their technologies are more appropriate for countries in the developing world. It is plausible to suggest, if impossible to prove, that this is the case in terms of factor endowments because, unlike the industrialized countries, the emerging economies are capital-scarce and labour-abundant. However, their technologies might not always be the most appropriate for smaller developing countries as their scale of production and size of market is much larger. What is more, their technologies may also be context-specific. Even so, the possibilities of learning from the experiences of the four countries are significant. The lessons are clear enough: 1) create the initial conditions in terms of education and research, 2) foster learning by doing in the manufacturing sector, 3) establish catch-up mechanisms through policies and institutions that are conducive to the development of technological capabilities and 4) think about technology policy from a long-term perspective. Taken together, these would provide the essential foundations for technological development. But there is more to learning from these countries. There are two dimensions that deserve emphasis: managing the common policy dilemmas and recognizing the importance of specificities. In sum, there is much to learn from the experience of technological development in the emerging economies, but such learning should seek to contextualize rather than replicate.

Notes 1. The dominance was even greater earlier. During the period from 1000 to 1700, China and India, taken together, accounted for 50 per cent of world population and 50 per cent of world income. And, 2000 years ago, in AD1, China and India accounted for almost 60 per cent of world population and world income. For a more detailed discussion, see Nayyar (2008).


66

Deepak Nayyar

2. The evidence in Table 2.3, and the discussion that follows, draws upon earlier work of the author. See Nayyar (2008a). 3. The following discussion draws upon Nayyar (2009). 4. For a more detailed discussion on the statistical evidence and analysis cited in this paragraph, see United Nations (2006). 5. See United Nations (2006), pp. 22–3. 6. For a detailed discussion, see Kaplinsky (2006). The literature on this subject is limited. But the implications and consequences of rapid growth in China and India for the developing world are analysed in Kaplinsky and Messner (2008). 7. There is a clear need for more systematic research on the subject, on which information and understanding are both limited. 8. It would mean too much of a digression to consider these mechanisms here. For an analysis, and a discussion of the issues, see Nayyar (2009). 9. The discussion that follows is based on recent work of the author published elsewhere. See Nayyar (2008; 2009). 10. This proposition is stressed by Kaplinsky (2006), Rowthorn (2006) and Singh (2007). 11. For a discussion on China’s trade with, and aid to, Africa see Toye (2008). 12. Kaplinsky and Morris (2008) show that China’s emergence as a large exporter of manufactured goods in the world economy poses severe problems for export-oriented growth in sub-Saharan Africa, particularly in textiles and clothing. 13. During the period 2000–5, BRICS accounted for 20–5 per cent of the inward stock, and for about 37 per cent of inward flows, of FDI in developing countries (Nayyar, 2009). 14. For a detailed discussion, as also statistical evidence, see Nayyar (2009). It is worth noting that, during the period from 2000 to 2005, BRICS accounted for about 13 per cent of the outward stock, and about 10 per cent of outward flows, of FDI from developing countries. 15. An unpublished study by Rhys Jenkins and Chris Edwards, cited in United Nations (2006: 22), on China’s trade with 18 developing countries (six in Asia, six in Africa and six in Latin America) shows that countries that had significant trade with China were exporting mostly, agricultural or extractive, primary commodities. A study on China’s economic interaction with Latin America and the Caribbean also confirms the traditional pattern of trade, importing mostly primary commodities and exporting mostly manufactured goods (Inter-American Development Bank, 2005). Another study on the impact of China’s trade with, and FDI in, Latin America and the Caribbean shows that there are winners and losers that can be identified: sectors and countries producing primary commodities are the winners, while those producing or exporting manufactured goods are the losers (Jenkins et al,. 2008). 16. Chinese imports from Africa are mostly primary commodities while an overwhelming proportion of Chinese exports to Africa are manufactured goods. This composition, particularly in imports, is far more skewed than for China’s trade with developing countries or the world economy as a whole. Similarly, the composition of Africa’s exports to, and imports from, China is far more skewed than Africa’s trade with developing countries or



17. 18. 19. 20. 21.

22.

23.

67

the world economy as a whole. For a detailed discussion on this proposition, with supporting evidence, see Nayyar (2009a). For a detailed discussion, which also provides a succinct survey of the literature, see Lall (1992). See Atkinson and Stiglitz (1969). Such thinking, in terms of evolutionary theories, was developed by Nelson and Winter (1982). This argument is developed, at some length, by Lall (1992). See also Lall (1990, 1991). This has been an important factor underlying foreign acquisitions by Indian firms during the 2000s. For an analysis, and supporting evidence, see Nayyar (2008b). Such an economic role for the state was critical for the development technological capabilities in South Korea. See, for example, Amsden (1989) and Chang (1996). For a detailed discussion, see Nayyar (1997).

References Amsden, A.H. (1989), Asia’s Next Giant: South Korea and Late Industrialization. New York: Oxford University Press. Atkinson, A.B. and J.E. Stiglitz (1969), ‘A new view of technological change’, Economic Journal, 79 (4), 573–8. Chang, H.J. (1996), The Political Economy of Industrial Policy. London: Macmillan. Inter-American Development Bank (2005), The Emergence of China: Opportunities and Challenges for Latin America and the Caribbean. Washington, DC: IADB. Jenkins, R., E.D. Peters and M.M. Moreira (2008), ‘The impact of China on Latin America and the Caribbean’, World Development, 36 (2), 235–53. Kaplinsky, R. ed. (2006) Asian Drivers: Opportunities and Threats. IDS Bulletin, 37 (1). Kaplinsky, R. and D. Messner (2008), ‘The impact of the Asian drivers on the developing world’, World Development, 36 (2), 197–209. Kaplinsky, R. and M. Morris (2008), ‘Do the Asian drivers undermine exportoriented industrialization in SSA?’ World Development, 36 (2), 254–73. Lall, S. (1990), Building Industrial Competitiveness in Developing Countries. Paris: OECD Development Centre. Lall, S. (1991), ‘Explaining industrial success in the developing world’, in V.N. Balasubramanyam and S. Lall (eds) Current Issues in Development Economics. London: Macmillan. Lall, S. (1992), ‘Technological capabilities and industrialization’, World Development, 20 (2), 165–86. Maddison, A. (2003), The World Economy: Historical Statistics. Paris: OECD Nayyar, D. (1997), ‘Themes in trade and industrialization’, in Deepak Nayyar (ed.) Trade and Industrialization. Delhi: Oxford University Press. Nayyar, D. (2008), ‘The rise of China and India: Implications for developing countries’, in P. Artesis and J. Eatwell (eds) Issues in Economic Development and Globalization. London: Palgrave.


68

Deepak Nayyar

Nayyar, D. (2008a), ‘Learning to unlearn from development’, Oxford Development Studies, 36 (3), 259–80. Nayyar, D. (2008b), ‘The internationalization of firms from India: Investment, mergers and acquisitions’, Oxford Development Studies, 36 (1), 111–31. Nayyar, D. (2009), ‘China, India, Brazil and South Africa in the world economy: Engines of growth?’ in Amelia Santos-Paulino and Guanghua Wan (eds) Southern Engines of Global Growth. Oxford: Oxford University Press. Nayyar, D. (2009a), ‘The emerging Asian giants and economic development in Africa’, IPD Task Force on Africa. Columbia University: New York. Nelson, R.R. and S.J. Winter (1982), An Evolutionary Theory of Economic Change. Cambridge, MA: Harvard University Press. Rowthorn, R. (2006), ‘The renaissance of China and India: Implications for the advanced economies’, UNCTAD Discussion Paper No. 182, October. Geneva: UNCTAD. Singh, A. (2007), ‘Globalization, industrial revolutions in India and China and labour markets in advanced countries: Implications for national and international economic policy’, Working Paper No. 81, Policy Integration Department, March. Geneva: ILO. Toye, J. (2008), ‘China’s impact on Sub-Saharan African development: Trade, aid and politics’, in P. Artesis and J. Eatwell (eds) Issues in Economic Development and Globalization. London: Palgrave. United Nations (2006), ‘Diverging growth and development’, World Economic and Social Survey. New York: United Nations. World Bank (2007), World Development Indicators 2007. Washington DC: World Bank.


3 The Changing Geography of Innovation Activities: What do Patents Indicators Imply? Xuan Li and Yogesh A. Pai

3.1

Introduction

Innovation in the global marketplace is at the core of the twenty-first century knowledge-based economy (Schumpeter, 1980: 66). Innovation is in itself a fuzzy concept and measuring it is more difficult (Godin, 2008). Innovation may encompass the invention of products and processes coupled with their commercial exploitation. Measuring innovation performance can be vital in arriving at a formal link between innovation performance and economic growth (Freeman and Soete, 2007). Thus multiple indicators, along with patent statistics, can be used in assessing innovation performance (Lanjouw and Schankerman, 1999). The use of patent statistics in connection with innovation has a tradition (OECD, 2005). A brief literature review suggests that at least six attributes of innovative activities could be evaluated through patent data (Oltra and Kemp, 2007). The first attribute is the level of innovative activity (Griliches, 1990; Popp, 2005). Second, the patent data can illustrate the types of innovation and technological competencies of organizations (Oltra and Kemp, 2007). Third, the patent data can be used to indicate the technological strengths of nations. Fourth, patent data can be used to measure technology diffusion, as patent data are available from many different countries to track patterns of diffusion. Fifth, patent data are a good source for studying innovation instigators and of networks of innovators. From the bibliographic data on a patent, researchers can gather the identity and home country of the inventor and of the assignee (or the applicant). Such information enables researchers to identify the sources of innovation in terms of patenting organizations. Sixth, patent data can indicate technological spillovers and knowledge relatedness. There have been various attempts 69


70 Xuan Li and Yogesh Pai

to conceptualize relatedness among technological fields and to find appropriate measures for knowledge spillovers. Various methodologies have been proposed on the basis of patent data. Presently, great weight has been placed on patent counts in cross-country comparisons and in assessing national innovation capacity. For example, the 2007 and 2008 World Intellectual Property Organization (WIPO) Patent Reports (WIPO 2007; 2008) highlight the changing geography of innovation with the highest patenting growth rates in North-east Asia and, particularly, the sharp rise in patent filings in China. Some of the conclusions in the report are based on the assumption that patent applications are a critical indicator of innovative activity and that resident patent filings are a reliable proxy measure of such underlying activity in a country. Current practices and interpretations of measurements of innovation suggest that there is a heavy emphasis on patents as indicators of innovation in both national and international settings (Li, 2008: 4–6). It is recognized that the WIPO patent reports are valuable in understanding the use of the patent system in both developed and developing countries, including its internationalization. However, interpretation of such statistics as indicative of innovation performance by relying on resident patent activities, and for assessing cross-country innovation performance, does not present much conceptual clarity, especially in the context of the rise in patent filings in developing countries. Hence, the focus of this chapter will be on the justifiability of using patent statistics for measuring innovation performance in light of WIPO World Patent Reports. In particular this chapter looks at why patent statistics are weak indicators of innovation output, especially in a developing-country context, emphasizing especially China and other emerging economies. This chapter is structured as follows. Following the brief introduction, Section 3.2 will analyse the problems of patent counts as a proxy for geography of innovation performance, with specific reference to patent filings in China from the perspectives of both patenting quantity and quality. Section 3.3 considers the problems concerning patent counts as a proxy for cross-country innovation performance and analyses the underlying sources that affect the proper interpretation of patent application from a legal perspective. The conclusions are in Section 3.4.

3.2 Problems of patent counts as a proxy for geography of innovation performance There are fundamental problems in using patent counts for national innovation performance. The great weight placed on patent counts for


The Changing Geography of Innovation Activities

71

the purposes of assessing national innovation capacity has inherent limitations. First, the patent process involves an applicant filing a patent document with the patent office of the country in which he is seeking protection for his invention. The patent document is a rich mine of information on the invention it covers – information which can be used directly in constructing statistical indicators. This information pertains to the technical features of an innovation, claims, drawings, abstract, specifications, class of invention, cited patents and other documents as prior art and so on. Apart from technical features, the patent document contains information regarding the inventors, assignee, applicants, history of the application, priority date, country filing date, date of publication, date of denial or withdrawal, date of grant, date of termination due to non-payment of renewal fees, etc. However, it contains no information regarding the exact location of the invention. The inventor’s/ applicant’s address is not helpful in understanding the place of innovative activity. Second, WIPO considers resident patent filings a reliable proxy measure of underlying innovative activity in a country. The WIPO Patent Report highlights the fact that in China patent filings by residents increased by more than eight times between 1995 and 2005, and concludes that China is one of the most innovative countries as it has the highest growth rate for resident patent filing (+42.1 per cent). According to the WIPO Patent Reports, resident patent applications mean those where the first-named applicant or assignee is a resident of the state or region concerned, while non-resident patent applications are filed by applicants outside the relevant state or region. However, a large number of patent applications filed by foreign companies’ branches in developing countries are counted as domestic patents. The mixture of foreign and local patent filings is misleading and thus affects the objective assessment of the real level of local innovation capacity in these countries (Li, 2008: 21).For instance, a resident patent in China is not defined purely as a patent filed by Chinese nationals, but includes patents filed by foreigners. This is because equity joint venture, contractual joint venture and wholly foreign-owned enterprises are considered as domestic enterprises according to Chinese law. Accordingly, the patents filed by these foreign-owned enterprises are counted as domestic patents (Li, 2008: 19–20) While inventive activities take place in a particular geographical location, the inventor may file the first application (resident patent) from any patent office across the world. This creates at least theoretical problems in terms of identifying innovation, and may at times lead to double counts. There is no practical methodology



fundamentally establishing that innovation accrued within a particular geographical location is always attached to the first filing of resident patents. This is an inherent limitation in attaching importance to patent counts as being reflective of the geography of innovation within a particular location. Third, a country’s innovation capacity may be overstated by patent filing without classification. In China, patents are classified into the following three categories: 1) utility models, 2) industrial design and 3) invention patents. According to the Chinese patent regime, no substantive patent examinations are required for the utility model and industrial design categories. This implies that reaching a technical threshold is not necessary for an applicant to be granted a ‘patent’. A closer look at the composition of patents in 2005, and increases in patent applications in China over the last five years, demonstrates that: 1) utility models and industrial designs, with a total of 302,937, have a majority share in the total number of patent applications of 476,264, that is, 64.1 per cent, and 2) all three categories of patents experienced fast growth, of which industrial designs (163,371) grew by 47.4 per cent compared to 2004, utility models (139,566) grew by 23.7 per cent and invention patents (173,327) grew by 32.2 per cent. Therefore, the total number of utility models and industrial designs outweighed the number of invention patents and the growth rate in industrial design is higher than that of invention patents, so the sharp rise in patent filings in industrial design and utility models does not represent a significant improvement in innovation capacity in China (Li, 2008: 20–2). A survey of the State Intellectual Property Office’s (SIPO) annual reports amply illustrates this fact. Further, there are issues concerning the qualitative value of patent counts and the difference this makes to understanding innovation capacities. It must be noted that a patent count is not same as the value it encompasses. Across the three categories, the values of patents differ substantially. Invention patents have the highest added value. Within the same category, depending on the subject matter of the invention, the scope of patent, a basic patent or surrounding patents, the importance of the particular patent in the value chain, etc. varies substantially. This has implications for the net total outcome of the value assigned to resident patent filings, which, even in terms of quantity, may not be of higher value. While it is understood that the mere counting of patents at any level of aggregation does not provide good value indicators, it is remarkable to note that the intrinsic value of innovations varies significantly, either within the same field of technology or among different fields, and thus that single patent counts, which weigh all patents equally, lead to



73

skewed analysis of the value attached to patents. Among the invention patent applications, resident ones are dominant in the following areas: traditional Chinese medicine with 98 per cent, soft drinks with 96 per cent, food products with 90 per cent and Chinese-language computer inputs with 79 per cent. Non-resident applications are dominant in the high-tech areas as follows: wireless transmission with 93 per cent, mobile communications with 91 per cent, television systems with 90 per cent, semiconductors with 85 per cent, Western medicine with 69 per cent and computer applications with 60 per cent. It is clear that patent applications by non-residents are mostly in high-end and high-value-added areas. In terms of Chinese invention patents, applications for 2005 from both domestic and foreign inventors reveals that the top five sectors of technical innovation were: 1) computers, 2) telephone and data transmission systems, 3) natural products, 4) fermentation technology and 5) computer peripherals (Li, 2008: 23). Comparison of domestic and foreign applicants in China shows substantial differences. In 2005, Chinese inventors filed most applications in natural products, closely followed by digital computers while foreign applicants focused on digital computers and telecoms patents. (See Tables 3.1 and 3.2) To demonstrate the basic patents and surrounding patents, Table 3.3 shows the number and content of the patent applications filed by Toyota, Japan, and Tsinghua University, China, in 2005. It can be seen from the table that the applications filed by Toyota were concentrated in the key areas of drive systems and hybrid car control systems, while Tsinghua University’s applications were widespread, covering several areas including core technologies of hybrid power systems and controllers, but also Table 3.1 Top five patented technologies (domestic applicants), 2005 Rank

Items

Term

Industry

1

8196

B04

2 3

6355 4718

T01 W01

4

3497

D16

5

2976

D13

Natural products and polymers Digital computers Telephone and data transmission systems Fermentation industry Other foods, food treatment including additives



Table 3.2 Top five patented technologies (foreign applicants), 2005 Rank

Items

Term

Industry

1 2

12,985 8655

T01 W01

3

7207

L03

4

6629

W02

5

6336

U11

Digital computers Telephone and data transmission systems Electroconductors, etc. Broadcast, radio and line transmission systems Semiconductors, materials and processes

Table 3.3 Comparison between applications from Toyota and Tsinghua University, 2005 Applicant Toyota

Amount 8

Title Drive unit for hybrid electric vehicle (HEV) Control device for HEV equipped with transmission Control system for HEV Drive system and method for parallel HEV Hybrid drive unit for vehicle

Tsinghua University

6

Standardization method for HEV controller Automatic gear system for parallel HEV Test and research system for fuel cell hybrid drive system Multi-energy power unit assembly controller for HEV Hybrid drive system for fuel cell vehicle

peripheral systems such as control calibration and testing platforms. In the area of hybrid cars, developed countries such as Japan and the US have taken a lead in research and have dominated the field in many of the core technologies, while China’s research work has only recently taken off and it has few comparative advantages in core technologies except in peripheral technologies.



75

There can be considerable differences attached to patent value based on the difference between product and process patents. As demonstrated earlier, the patent applications under A61K for medical, dental and toiletry purposes ranked number one in all applications, of which 80 per cent were domestic inventions. Traditional medicinal knowledge (TMK) patents constitute the major components under IPC-A61K. Given the inherent difficulties in determining the structure of TMK, only ‘product-by- process’ applies, which has limited market power (Li, 2007). However, within the WIPO scheme of interpretation of patent counts, there is currently no differentiation made concerning the value assigned to each patent. Thus there can be a disparity in the levels in patents based on certain parameters. Various methodologies can be used in attaching value to patents, prominent among them being the licensing of revenue flows.

3.3 Problems of patent counts as proxies for cross-country innovation performance Conceptually, the benefit of a cross-country comparison approach is that it allows assessment of the relative share of various countries in innovation in a given national technology market. However, differences in the design and application of patent systems, which are left to individual countries, result in under- or overestimation of the technical innovative capacity and confuse the cross-country geography of innovation substantially. First, there is considerable divergence in invention patents and utility models across the world. These differences in the design of patent systems are primarily caused by gaps in the current international legal framework. The problem that eventually occurs in the case of patent statistics is that sometimes utility patents may be included within the general patent system. While the WIPO Patent Report validly makes this point, it goes no further in excluding them from its analysis. The methodology used in the report is not suggestive of the interpretation that utility patents have been excluded for all purposes while collecting statistics on patents. In some cases incremental inventions, which should form part of the utility models, may have been included in standard patents for invention statistics. The problems in understanding patent statistics can be illustrated through a Chinese example. During 2007, SIPO received 694,153 patent applications covering three types of patent: invention, utility and design. While this number may at first sight seem huge, the overwhelming



number of applications, pegged at almost 450,000, was for utility and design patents, which are not subject to substantive examination. Only standard invention patents receive full scrutiny involving patentability thresholds; 245,161 of these were submitted during 2007. Thus patent statistics from one country can be skewed as a result of differences in the design of the patent system, and primarily the question arises as to which patent statistics will be considered for making cross-country comparisons. Second, countries are free to enact specific patent-like legislation, or separate industrial design legislation. In the case of patent-like legislation, there is always the possibility of including these with standard patents for inventions statistics and thus giving a non-comparable picture of innovation geography. In many countries, designs are also protected through copyright. To illustrate the difference in the design of incentive mechanisms, which has immense implications for counting patent statistics among one of the categories in cross-country comparisons, a brief summary is provided in Table 3.4. Third, international comparisons based on patent applications are questionable because of the different propensity to patent and the Table 3.4

Utility models

Plant Patent Act

Design patents

Utility models, plant variety and designs in the definition of patents US

EU

Brazil

India

China

No (not to be confused with ‘utility patents’ which are synonymous with patents for inventions) Yes (both utility patents and an exclusive Plant Patent Act are in place provided they satisfy the different criteria required under different laws) Yes

No (allowed in a few countries based on national legislation, but not harmonized by EU law)

Yes

No

Yes

No

No

No

No

No

No

No

Yes



77

different value of patents across countries. The characteristics of national patent systems in terms of protectable subject matter, examination standards, scope of right and the first-to-file versus the first-to-invent system substantially influence patent propensity. These differences in patent regulations make it difficult to ascertain local innovation capacities and compare patent application counts across countries. Table 3.5 demonstrates the presence or absence of legal provisions, which forms the genesis of divergence in patent law. As seen above, four patent regimes are analysed to illustrate how different interpretations are possible. This leads to a divergence in the granting of patents, as seen in the comparison of two developedcountry patent regimes and two developing-country patent regimes in the four vital areas of: definition of invention, definition of field of technology, definition of patent law thresholds and disclosure requirements. Fourth, there is a further problem with international comparisons based on national patent counts which results from the high heterogeneity in the value of patents resulting from different standards of patent examination. As the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) leaves considerable flexibility for national patent regimes to adopt a patent law threshold, different patterns of examination can be adopted by different patent offices. There is a general trend in granting patents of lowering the patent threshold requirements of utility and inventive step (Federal Trade Commission, 2003). An associated outcome is questionable patents. For instance, the surge in patenting as a result of decisions of the US Federal Circuit Courts which opened up the way for a more liberal patenting framework could explain the rise in the grant of ‘questionable patents’. As the significance of patent counts is limited, placing patents with skewed values on an equal footing is problematic. The Manual of Patent Practice and Procedures reveals how different standards of patent examination and court interpretations can lead to varying patent granting levels in developed and developing countries. To illustrate, ‘prior use’, which can lead to lack of novelty is unique to the US law and restricted to use within the country.1 As TRIPS has not defined novelty, it leaves enough leeway for a favourable standard to be set. It leads to a higher degree of patent filing and granting, for instance in the area of traditional knowledge and use of genetic resources, since many applications of such knowledge (which have not been made known through publication) have been made outside the US. The inventive step, popularly the synonym for ‘non-obviousness’


78

Xuan Li and Yogesh Pai

Table 3.5 A comparison of US, EU, Brazilian and Indian legal provisions concerning the grant of patents

Definition of invention

Definition of patent law thresholds (i.e. novelty, utility and inventive step)

Other exclusions from patent eligibility

United States (national)

EU (regional)

Brazil (national)

India (national)

35 US Code 100(a) defines the term ‘invention’. 35 US Code 101 defines patentable inventions 35 US Code 102 defines conditions for patentability, novelty and loss of right to patent. 35 US Code 103 defines conditions for patentability, non-obvious subject matter.

Not defined by the statute.

Not defined by the statute.

Section 2(j) defines ‘invention’.

Article 52(1) prescribes the legal threshold. Articles 54, 56 and 57 define novelty, inventive step and industrial application in relation to an invention.

Section 2(1) (ja) defines ‘inventive step’. Section 2(1) (ac) defines the term ‘capable of industrial application’. Section 2(l) defines the term ‘new invention’.

No specific statutory exclusions

Article 52(2), (3) and (4) lists the exclusions from patent eligibility.

Article 8 prescribes the legal threshold. Article 11 defines novelty in relation to inventions. Article 13 defines when an invention shall be deemed to involve inventive activity. Article 15 defines industrial applications in relation to inventions and utility models. Article 10 provides a list of items that are not considered as inventions. Article 18 provides for non-patentable inventions and utility models.

Section 3 defines what are not inventions within the meaning of the Act.

Continued


The Changing Geography of Innovation Activities Table 3.5

79

Continued

Disclosure of invention

Miscellaneous

United States (national)

EU (regional)

Brazil (national)

India (national)

35 US Code 112 defines the content of the specification which should accompany a patent application. 35 US Code 104 deals with foreign inventions. 35 US Code 105 deals with inventions in outer space.

Article 83 defines the standard of disclosure needed for the grant of European patents. –

Articles 24 and 25 define the standard of disclosure in relation to an invention.

Section 10(4) lays down the standard of disclosure in relation to a patentable invention. Section 4 provides that inventions relating to atomic energy are not patentable.

–

in the US, lays down a standard (PHOSITA) where the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which the said subject matter pertained. This definition has been open to several interpretations by the courts in the US and has been tightened or loosened at different points in time. The PHOSITA test may depend on how one defines the art (that is, the area of technology) and the level of skill of such a person. In KSR International Co. v. Teleflex Inc. et al.,2 Justice Kennedy remarked, ‘A person of ordinary skill is also a person of ordinary creativity, not an automaton.’ This slightly unsettles the US Patent and Trademark Office’s (USPTO) position, which grants patents to inventions based on the low-level application of the PHOSITA test. In fact, USPTO came up with new guidelines in October 2007, based on the Supreme Court’s decision in KSR (USPTO, 2007b). The guidelines offer a more restrictive approach and have slightly tightened the non-obviousness standards for the examination of patents. There is evidence that that fewer patents have been issued since the application of the KSR test by USPTO (see Figure 3.1). This


80

Xuan Li and Yogesh Pai 4000

3000

2000

1000

0 0

4

8

12

16

20 2006

24

28 2007

32

36

40

44

48

52

2008

Figure 3.1 Weekly USPTO utility patent grants, 2006–8 Source: Promote the Progress: The Patent Education Portal (2009).

clearly shows that on average the USPTO granted fewer patents in 2008 than it did in 2006. The USPTO accepts that there has been a dramatic decrease in the allowance rate (the percentage of patent applications) that is ultimately approved. The USPTO allowance rate dropped from 72 per cent in 2000, to 44 per cent in the first quarter of 2008. The office attributes this partly to the USPTO’s quality initiatives, but much is also attributed to the lack of quality in many applications that the USPTO receives.3 The EU tends to lower the patentability thresholds in certain sectors to facilitate their development. Through the European Patent Convention (EPC), the EU states with regard to novelty that an invention shall be considered to be new if it does not form part of the state of the art. The state of art comprises everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the European patent application. While prior use and prior publication are sufficient to kill novelty, the inventive step requires that the invention should be non-obvious to



81

a person skilled in the art in comparison to the prevailing state of the art. These magic words have given sufficient scope for wider interpretation, which has subsequently allowed the European Patent Office (EPO) to grant patents by lowering the inventive step criterion. An invention is considered to have industrial application if it can be of use in any kind of industry, including agriculture. In the case of software, minor software innovation became patentable by fixing it to a tangible medium with this loose interpretation. Recently, the EPO Technical Board of Appeals has invited open comments from various stakeholders concerning the merit of lowering the patentability thresholds for software-related inventions. This will certainly have consequences for how patents in the software area will be issued, in turn raising issues as regards counting. Fifth, changes in the law can substantially alter the statistical outcomes in any particular country. Thus cross-country comparison without recourse to the specificities of any changes can overly exaggerate patent statistics. For example, during the 1980s the US courts opened the gates to patentability of software and biotechnology inventions. This has consequently had implications for the rates of patent being filed in these areas. Further, some judicial developments in the second half of the 1990s led to the inclusion of patents for business methods. This had implications for the number of patents being granted. Furthermore, the Bayh-Dole Act of 1980 changed the academic patenting landscape so that federally funded institutes relied more on the patent system (Scherer, 2007). For example, the new rules allowed scientists, institutions and universities to own any patents resulting from their publicly funded research, making China the latest of many countries to introduce a Bayh-Dole-style intellectual property (IP) regime. Figure 3.2 illustrates how sudden changes in the law can have upsetting effects on the statistical outcomes of patents. It shows a sudden increase of US patent grants in the area of computers and communications in 1998. This was a result of the decisions of the US federal courts to reduce the patentability criteria for software-related inventions through the decision in Re Alappat in 1994.4 Sixth, there are challenges posed by differences in the standards of examination adopted by patent examiners. Are all patent examiners equal? This is a practical question which has come under greater evidence-based scrutiny in the recent past (Cockburn et al., 2002). Evidence was compiled from data gained through interviews with administrators and patent examiners at the USPTO, and a dataset of


82 Xuan Li and Yogesh Pai 35000 Computers and communications

30000 25000

Mechanical Other

20000 Chemicals

15000 10000

Drugs and Medicine

Electronics

5000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

0

Figure 3.2 US patent grants by technology category, 1981–99 Source: NBER Patent Citation Database, author’s calculations

patent examiners was analysed with reference to patent outcomes. The main finding was that patent examiners and the patent examination process are not homogeneous; examiners whose patents tend to be more frequently cited tend to have a higher probability of receiving a court’s invalidity ruling. Thus these conclusions have some relevance for the statistical outcomes. Therefore, conceptually it needs to be re-emphasized that those inventions which possess a very high degree of patent law threshold may or may not form part of the patent scheme in different countries. In this regard, the guidelines issued by patent offices can be important. Logistically, the guidelines are important in the sense that an Act cannot be expected to explain in greater detail the nuances to be followed, especially with regard to defining the type of claim format which could be allowed for each type of patenting activity. Thus the guidelines have the potential to blur definitional requirements and standards prescribed under the Act and thus they may be important in understanding what type of grants are allowed and disallowed. This is sufficient to suggest that different countries may follow different procedures in their patent offices, which can lead to denial or grant of a patent right. This has implications for understanding cross-country comparisons of the geography of innovation if successful patent count does indeed



83

point to it. Hence, patent counts must take the above nuances into consideration when ascertaining the geography of innovation, especially in developing countries.

3.4 Conclusions This article argues that the use of patent statistics as indicators for measuring innovation in developing countries and for the purposes of cross-country comparison of innovation activities should be made with caution. The correlation between the sharp rise in patent filings and national innovation capacity is not straightforward. Confusion resulting from domestic and non-resident patent filing and the mixture of counts between invention patents, industrial designs and utility patents, contribute to the over- or understatement of national innovation capacity. Moreover, there are emerging issues of patent quality– patent fees vis-à-vis firms’ strategic behaviour in patent acquisition and patent filing strategies in certain sectors – that make such a comparison weak. In the case of cross-country comparisons, patent registration containing no information indicating the place of innovative activity is fundamental. Differences in the design of patent systems have immense implications for cross-country comparison. The adoption of multiple indicators in measuring innovation performance in nonOECD (Organisation for Economic Co-operation and Development) countries is preferable. To facilitate appropriate policy responses, developing countries should be cautious in reading the WIPO interpretation of the counts on patent filings. Furthermore, it is necessary to make efforts to develop a proper set of indicators to monitor the changes in innovation capacities in each and every place, especially in developing countries.

Notes Dr. Xuan Li, Coordinator, Innovation and Access to Knowledge Programme (IAKP), South Centre. Mr. Yogesh Pai, Associate Fellow, Centre for Trade and Development, New Delhi, India. 1. USC 102 stipulates conditions for patentability; novelty and loss of right to patent. It states: ‘A person shall be entitled to a patent unless – (a) the invention was known or used by others in this country, or patented or described in a printed publication in this or a foreign country, before the invention thereof by the applicant for patent, or ... .’ (Emphasis not in original).


84 Xuan Li and Yogesh Pai 2. 550 U.S. 398 (2007). 3. See http://www.uspto.gov. 4. 33 F.3d 1526, 31 USPQ2d 1545.

References Cockburn, Iain M., Samuel S. Kortum and Scott Stern (2002), ‘Are all patent examiners equal? The impact of examiner characteristics on patent statistics and litigation outcomes’, NBER Working Paper No. 8980. Federal Trade Commission (2003), ‘To promote innovation: The proper balance of competition and patent law and policy’, Report by the Federal Trade Commission, October. Freeman, Christopher and Luc Soete (2007), ‘Developing science, technology and innovation indicators: What we can learn from the past’, UNU MERIT Working Paper Series, # 2007–001. Godin, Benoît (2008), ‘The rise of innovation surveys: Measuring the fuzzy concept’, Working Paper 16, Project on History and Sociology of STI Statistics, Montreal, Canada. Griliches, Z. (1990), ‘Patent statistics as economic indicators: A survey’, Journal of Economic Literature, 28, 1661–707. Lanjouw, Jean O. and Mark Schankerman (1999), ‘The quality of ideas: Measuring innovation with multiple indicators’, NBER Working Paper 7345. Li, Xuan (2007), ‘Inadequacy of patent regime on traditional medicinal knowledge – A diagnosis of 13-year traditional medicinal knowledge patent experience in China’, Journal of World Intellectual Property, 10 (2), 125–48. Li, Xuan (2008), ‘Patent counts as indicators of the geography of innovation activities: Problems and perspectives’. Geneva: South Centre. OECD (2005), Oslo Manual: Guidelines for Collecting and Interpreting Innovation Data. Paris: OECD. Oltra, Vanessa and René Kemp (2007), ‘Patents as a measure for eco-innovation, University of Bordeaux and UNU-MERIT’, 16 June. Popp, D. (2005), ‘Lessons from patents: Using patents to measure technological change in environmental models’, Ecological Economics, 54 (2–3), 209–26. Promote the Progress (2009), ‘New on metrics – utility patent grants down 13% as compared to 2006’, 17 November. Available at http:// promotetheprogress.com/new-on-metrics-utility-patent-grants-down-13as-compared-to-2006/757/ Scherer, F.M. (2007), ‘The political economy of patent policy reforms in the United States’, RWP07–042 KSG Faculty Research Working Paper Series. Schumpeter, J.A. (1980), The Theory of Economic Development. London: Oxford University Press. USPTO (2007a), ‘Changes to practice for continued examination filings, patent applications containing patentably indistinct claims, and examination of claims in patent applications’. Available at: www.uspto.gov/web/offices/pac/ dapp/opla/.../ccfrchanges.pdf. USPTO (2007b), ‘USPTO publishes examination guidelines for determining obviousness in light of the Supreme Court? KSRv Teleflex decision, Press



85

release 07-43, 10 October. Available at: http://www.uspto.gov/news/ pr/2007/07-43/jsp WIPO (2007), ‘WIPO patent report: A statistical review’. WIPO (2008), ‘WIPO patent report 2008’




Part II Policy, Strategy and Catch-up: Country Case Studies




4 China’s Catch-up and Innovation Model: A Case of the IT industry Xielin Liu

4.1

Introduction

In the decades during which the topic of catching up has received attention, two approaches and explanations have emerged that explain the catch-up process in the world. One approach is based on growth accounting. Using patents as an index for innovation capability, it has been used to find the key factors that determine catching-up processes in developing or newly developed countries. Furman and Hayes (2004) try to analyse this question in a qualitative way. Based on the growth model of Romer (1990), the theory of national competitive advantage (Porter, 1990) and national innovation systems (Nelson, 1993), they build a new framework and find that the factors behind catching up or standing still are the development of innovation-enhancing policies and infrastructure and the ever-increasing financial and human capital investment in innovation. However, their approach cannot explain country-specific factors in catching up. How can those factors work together to achieve the fast catch-up which has been achieved by some countries such as China? The second approach is more historical and dynamic. Freeman (1987) identifies government regulation, shop-floor innovation and social institutions such as life employment and Keiretsu (a group of firms with interlocking business relationships and shareholdings) as the key factors that account for Japan’s catch-up. For Kim (1997), learning is the key activity for Korean companies in order to master advanced technology. Lee and Lim (2001) try to explain Korea’s catch-up process in terms of technological regimes. Hobday (1995) used the ladder of catching up: original equipment manufacturer – original brand manufacturer – original design manufacturer (OEM – OBM – ODM) 89


90 Xielin Liu

to explain the emergence of Taiwan and other economies. But some researchers do not agree with this. For example, in World Bank reports, the market is more important than the government in explaining the Asian catch-up (World Bank, 1993). Not all countries have the opportunity or ability to capitalize on the chance to catch up (Fagerberg, 1988). For a developing country, it is not easy to proceed from the stage of imitation to the stage of innovation. As Bell and Pavitt (1993) point out, just installing large plants with foreign technology and foreign assistance will not help in the building of technological capability. Under its planned economy system, China failed to catch up. The country imported technology from the West heavily, but did not realize the transition from imitation to innovation until the 1950s–70s. China and the former Soviet Union used reverse engineering like Japan, but in China and the Soviet Union, much of the responsibility for diffusion and development rested in central research institutes rather than in large industrial firms, as in the case of Japan (Freeman, 1988: 336–7). Since China’s opening up and adoption of the market economy system, the economy has experienced a period of rapid catching up. Its emergence on the world stage has now become a very controversial case for economists to understand. The purpose of this chapter is to identify the key features of the Chinese catch-up process and to propose an alternative model of catch-up that takes into account key features of the current environment at the industry level. The remainder of the chapter is organized as follows. In Section 4.2 we briefly review the catch-up processes of Japan during the 1960s–80s to highlight key features of its model. In Section 4.3 we introduce our framework that is grounded in China’s experience, and use IT to describe China’s on-going catch-up process. In Section 4.4, based on empirical cases, we illustrate the catch-up process in China’s IT industry and in Section 4.5 we discuss the general implications of our framework for both research and policy-making.

4.2

The new environment for catch-up

4.2.1 Japan’s experience During the era in which Japan caught up, manufacturing technology was considered to be the core of an industrial society. The Japanese business management system, including features such as life-time employment, the seniority system and efficient production; the main bank system and Keiretsu not only emerged from and matched the country’s


China’s Catch-up and Innovation Model 91

institutions, but also created highly efficient closed networks among related firms (Kondo and Watanabe, 2003: 327). This system works especially well in complex manufactures with high interdependencies, such as the automobile and machine tool industries. Under this system, Japanese manufactures develop new products based on their own inhouse technology and in-house procurement of manufacturing parts. Engineering skill is accumulated by rotation and life-time employment, and Keiretsu linkages increase stability. 4.2.2

China’s experience

Starting with a low level of technological capability, Chinese firms (led by the government) placed heavy emphasis on reverse engineering and technology imports in their strategic development. The source of that technology was the Soviet Union in the 1950s, shifting to Japan, the US and Europe from the 1970s. For example, in the 1980s, there were more than 100 imported colour television production lines in China, most of them from Japan. It was the same story in chemicals, steel and many other heavy industries. Such imports also had unexpected consequences. In the switch industry (communications equipment for landline telephone systems), for example, there were eight different equipment standards from seven countries in use in China in the 1980s because of the diverse sources of technology. In each of the industries where China was using technology imports to develop technologies, the government’s efforts were frustrated by a recurring pattern of ‘lag, import, lag again, import again’. Three main factors led to this outcome. First, there was a gap between technology users and technology creators. Up to the 1980s, the large state-owned enterprises (SOEs) were the main technology users, but they had no incentive to master manufacturing technologies in order to innovate. The research institutes were supposed to be the technology creators, but they were too far away from production sites and were administered by a different part of the government than the SOEs. Furthermore, as economic and enterprise reforms progressed from the 1980s, there was even less coordination among them, leading to the disintegrated national innovation system described by Liu and White (2001). Second, Chinese enterprises spent little money on assimilating the imported technology (Table 4.1). They did not have a system similar to the ‘shop floor as laboratory’ system of Japanese firms. Corporate R&D labs were undertaking primarily maintenance work and perhaps quality control activities instead of activities that would improve or create new processes. Furthermore, general managers of these SOEs and government


92

Xielin Liu

Table 4.1 Expenditure of in-house R&D and technology importation and assimilation (in 100 million RMB)

Expenditure Expenditure on on R&D technology imports 1991 1993 1995 1998 1999 2000 2001 2002 2003 2004 2005 2006

58.6 95.2 141.7 197.1 249.9 353.6 442.3 560.2 720.8 954.4 1250.3 1630.2

90.2 159.2 360.9 214.8 207.5 245.4 285.9 372.5 405.4 367.9 296.8 320.4

Expenditure on technology assimilation 4.1 6.2 13.1 14.6 18.1 18.2 19.6 25.7 27.1 54.0 69.4 81.9

Ratio 1:1.54:0.07 1:1.67:0.07 1:2.55:0.09 1:1.09:0.07 1:0.83:0.07 1:0.69:0.05 1:0.65:0.04 1:0.66:0.05 1:0.56:0.04 1:0.39:0.06 1:0.24:0.06 1:0.20:0.05

Source: MOST. China Science and Technology Statistics Yearbook, 1991–2007.

officials cared more about the ‘hardware’ element of technology, such as equipment, rather than ‘soft parts, such as software, processes or people. Although they had the world’s best technology (i.e. hardware, plant and other equipment), they did not truly master information technology (IT) in the sense of being able modify and improve it. Their allocation of resources spent on R&D, technology imports and assimilation (Table 4.1) indicates their emphasis; enterprises typically spent US$5 for assimilation for every US$100 expended on equipment purchases. Third, although some Chinese enterprises did try to follow up technology imports with investment in internal R&D activities in order to learn and master the imported technology, their efforts were too small, especially compared to Japan. Even in 2005, the ratio of R&D to sales in large and medium enterprises remained below 1 per cent, far below that of developed countries, even though IT has been increasing steadily since 1994, but declining slightly since 2002 (Table 4.2). 4.2.3 Factors characterizing China’s development Although China has failed to make the transition from imitation to innovation in the way that Japan, South Korea and a few other countries and their firms have, it has been catching up economically since the 1980s. Gross domestic product (GDP) has grown at more than 9 per cent for more than 20 years, and is now second (in purchasing power parity (PPP) terms) only to the US (IMD, 2004). To explain this, we


China’s Catch-up and Innovation Model 93 Table 4.2 Ratio of R&D/sales in large and medium-sized companies Year

1991

1995

2000

2001

2002

2003

2004

2005

2006

R&D/sales

0.49

0.46

0.71

0.76

0.83

0.75

0.71

0.76

0.77

Source: MOST. China Science and Technology Statistics Yearbook, 1991–2007.

propose that China’s catch-up has followed a different model to that of Japan and Korea. First, globalization makes China one of the most important parts of the world and gives Chinese companies vast opportunities to access a lot of technology for local innovation. Owing to China’s open policy (when compared those of Japan and South Korea at a similar stage), companies in many industries can obtain technology from multiple sources and China’s large market has itself led to its becoming a market for international technology sourcing. The source of technology may be previously integrated manufacturers; for example, some watch manufacturers in Japan now sell t not only he final products, but also the core components (movements) to Chinese assemblers. Other sources of technologies are intermediate specialized technology providers. In the mobile phone handsets industry (discussed further below), there are many intermediate technology providers from Europe, Japan and South Korea. Many universities and high-tech small and medium enterprises (SMEs) have also become important technology providers. Finally, in industries where tacit knowledge is critical, the mobility of engineers is important. Second, the dynamics of technology play an important part in Chinese catch-up. Gerschenkron (1962) argues that targeting rapidly growing and advanced technologies is an advantage for catch-up countries. We argue that information technology is this kind of technology. IT has made the world smaller and changed the rules of the game for catch-up. When analysing windows of opportunity for developing country firms to catch up, Perez and Soete (1988: 475) argue that the life cycle of a technology system is more relevant than single product cycles because the knowledge, skills, experience and externalities of the various products within a system are interrelated and support each other. A third characteristic of the Chinese catch-up model is the extensive size of the local market and tough competition which have provided incentives for Chinese companies to develop product innovations as quick as possible. The Chinese economy is among the world’s fastest growing with an average annual growth rate of 9 per cent over the past


94 Xielin Liu

20 years. China has a relatively big market size. According to Chinese statistics, GDP per capita in 2005 was about US$1703 and in 2006 was US$2042 – higher than India’s (US$723), but lower than Russia’s (US$5129) and Brazil’s (US$3300) (NBS, 2006). However, China’s large population makes it one of the biggest markets in the world. In terms of total GDP, China is about seventh in the world, just behind Germany. Whether firms in developing countries can take advantage of international technology outsourcing or not is a problem mentioned in the literature, for example with regard to the ‘make-or-buy dilemma’ at the enterprise level (Pisano, 1990; Veugelers and Cassiman, 1999). Here, transaction costs are the key issue. When an international technology market emerges, transaction costs will gradually decline and firms in China would like to use strategies of technology outsourcing from the other countries to speed up their innovation (Cesaroni, 2004). Fourth, the learning capability of the Chinese in an open economy is an important factor for Chinese companies as regards catch-up. Openness allows Chinese firms to learn from importation, technology alliances and multinationals in China. Openness also gives Chinese companies the opportunity to learn from local universities and research institutes. Chinese companies spend about 26 per cent of their R&D expenditure on local universities and research institutes to do contract research for them (Liu, 2006). But the most important capacity is that Chinese companies can integrate market knowledge, take technology opportunities and form alliance capability rapidly. So, knowledge integration is the core capability of Chinese companies. The fifth characteristic of the Chinese catch-up process is the specific strategy of large companies. As Chinese companies often lack the necessary key technologies for innovation, cost advantage is a very common strategy in China. Some Chinese companies adapt a market-oriented innovation strategy to build up their capability (Liu, 2008). A few of them, mostly those supported by the government, take a technologyoriented innovation strategy. Last, the role of government is critical for some industries. Strong government support cannot make an industry or companies innovative. Large and direct subsidies for firms at product level have become impossible following China’s entry into the WTO. So there is a question of how intelligently the Chinese government can help the industry or companies to be innovative. Some Chinese researchers argue that a policy which is good for developed countries may not necessarily be good for developing countries. They strongly suggest the government should select strategic industries as a way to leverage innovation (Lu and Feng, 2004).



4.3

The catching-up experience in IT industries

In order to analyse the Chinese approach to catching up, we have selected the IT industry to see how the key factors shape the catch-up process in China. IT, led by the companies Huawei, Lenovo and ZTE, is regarded as the most innovative industry in China. 4.3.1 The role of market size and market-oriented product innovation We define innovation as a new combination of technology and the market in which both elements play an equally important role. The importance of the market, however, relates to its size, complexity and dynamics. In China, market-related knowledge is one of the key comparative advantages of multinationals. Market-oriented innovation is one of the key features of Chinese industrial catch-up (Liu, 2008), which stands in contrast to Japan and South Korea which experienced an incremental process where innovation was based on American companies’ radical product innovation. This market-oriented strategy is composed of three related elements: innovation for the low-end market, innovation for market niche and innovation in a fast-changing market or industry. The market size of the telecommunications industry can be seen from Table 4.3. It shows that in many sectors of the IT market, China has the fastest rate of growth and share of the world market, the size of which gives Chinese companies the chance to compete with multinationals in an open market. Innovation for low-end and local markets The low-end market is the biggest market in absolute size in China. However, It requires low-price technology. This was the case for Table 4.3 Overview of the Chinese IT industry

Fixed phone users Internet users Wireless phone users Sales of IT industry

2000

2006

Average annual growth (%)

145 million

371 million

21

33.70 million

131 million

31

85.0 million

449 million

40

607 billion

3800 billion (2005)

31

Share of the world 1/4 of the world 1/10 of the world

Number 3 in the world

Source: Calculated from Chinese Yearbook of Statistics of Information Industry.


96

Xielin Liu

Huawei, a leading telecommunications company in China that is challenging Ericsson and Cisco, based on a market-oriented innovation strategy. Huawei’s initial approach was to capture rural markets first and urban markets later so that it could avoid tough competition from multinationals in the cities. Huawei, established in 1987, first started as a distributor of the HAX switch produced by a Hong Kong company. Its first product was the C&C08 switcher with 2000 lines, for a small city in Zhejiang, a market neglected by the multinationals. According to Ren Zhengfei, CEO of Huawei, users and customers are the source of innovation for the company (Chen and Liu, 2003: 59). While multinationals may have technology advantages, local companies in developing countries have the advantage of local market knowledge. For example, multinationals did not offer the clamshell design of mobile phones which Asian customers prefer at the right time, and Chinese firms quickly captured that market; now, this design accounts for 80 per cent of the Chinese domestic market. 4.3.1.2 Innovation for market niche In order to make remote areas accessible for wireless communication, ZTE developed a special code division multiple access (CDMA) net called C++net, specifically for such areas. Lenovo, the leader in the computer industry, adapted a unique strategy: trade first, manufacture second and technology last. Lenovo’s key capability is its deep understanding of Chinese market needs, which allows it to design its PCs to appeal specifically to Chinese customers (Gold, Leibowitz and Perkins, 2001). As Xie and White (2005) observed: ...the PCs being sold by multinationals were not differentiated to match local customers in markets such as China, which were considered relatively minor in 1980s. Lenovo, in contrast, was designing products for different market segments, from banks and other large organizations to SMEs in the corporate market, and similarly diverse individual customer groups. Lenovo incorporated feedback and experience in user needs from its distribution channels and marketing department into design and innovation efforts in its businesslevel R&D centers. Thanks to this incremental and market-oriented product innovation Lenovo managed to replace IBM as the top computer in the Asian market.



4.3.1.3 Innovation in a fasting growing market and industry The IT industry is one of the typical fast-growing and changing industries in China and the world. The penetration rate of mobile phones in China is astonishing (Table 4.2). To survive in the strongly competitive Chinese market requires companies to be highly flexible. In China, a company needs to develop 40–60 new mobile handsets in one year (Liu, 2008). This never happens in other countries and forced many foreign mobile handset manufacturers, such as NEC, Siemens, Philips and others, to quit the Chinese market. Thus, compared with the Japanese model, which focused on shop-floor innovation, the Chinese model encouraged companies to become better at market-oriented innovation (Liu, 2008). 4.3.2 Open innovation and global alliance It is not easy for latecomer companies to spend huge amounts of money on R&D as Microsoft and IBM do. So, many companies in developing countries use an open innovation strategy (Chesbrough, 2003). Globalization of technology can be either a window of opportunity or a further burden, depending on whether the firm playing catch-up has made the technological effort to support the absorption, adaptation, mastery and improvement of technology or not (Archibuchi, 2003: 864). IT is one of fastest growing industries in China, which has contributed to narrowing the gap between China and US in the last two decades. Many companies have become leaders in the world such Lenovo, Huawei and ZTE, which is why it is a good industry case to test our propositions. 4.3.2.1 Global technology outsourcing and alliance The global technology outsourcing and alliance strategy can be discerned from the structure of expenditure across the many channels that companies use to source technology. Technology imports are the traditional source for Chinese firms, and will continue to play an important role. For a long time before 1999, industrial enterprises spent more money on technology importation than on their own R&D. Although R&D activities have received more attention since 1999, technology imports still matter very much for production. At the same time, Chinese companies use both domestic and global alliance strategies to develop technologies and products. For example, in their GoTa (global open trunking architecture) development, ZTE1


98 Xielin Liu Table 4.4 R&D outsourcing for universities and R&D institutes from large and medium-sized industrial enterprises

Total R&D expenditure (RMB billions) Funds for university (RMB billions) Share of total business R&D (%) Funds for R&D institutes (RMB billions) Share of total business R&D (%) Total outsourcing for domestic universities and R&D inst. (%)

2000

2001

2002

2003

2004

2005

2006

35.4

44.2

56.0

72.1

95.4

125.0

163.0

5.5

7.2

9.0

11.2

14.9

17.3

19.7

15.5

16.2

16.1

15.5

26.1

13.8

12.1

3.8

2.5

3.6

4.7

5.0

5.6

5.3

10.7

5.6

6.4

6.5

5.2

4.5

3.3

26.2

21.8

22.5

22.0

20.8

18.3

15.3

Source: MOST, China Science and Technology Statistical Yearbook, 2001–7. Beijing: Chinese Press of Statistics.

joined different alliances. In equipment development, they joined with North Telecom, China Capital Telecommunication and others. For end products, they formed alliances with South High-tech and other companies. For value-added services, they allied with more than 100 service provider companies. Table 4.4 shows that Chinese companies are spending more and more money on collaboration with universities and government research institutes. Huawei has formed joint laboratories with TI, Motorola, Intel, AGERE, ALTERA, SUN, Microsoft and NEC, as well as a joint venture with 3COM. 4.3.2.2 Global R&D centres Several Chinese firms have entered the global stage and embarked on strategies to address their weaknesses vis-à-vis world-wide competition. Huawei has set up five research institutes abroad, in Silicon Valley and Dallas in the US, Bangalore in India, and Russia. In Bangalore they have 800 software engineers, most of whom are locals. Huawei also has formed joint laboratories with TI, Motorola, Intel, AGERE, ALTERA, Sun, Microsoft and NEC, as well as a joint venture with 3COM. With this global distribution of R&D, Huawei has been able to develop lots of technologies related to WCDMA and software Versatile Routing


China’s Catch-up and Innovation Model 99 Table 4.5 Selected M&A deals by Chinese firms in the IT industry (2001–5)

Chinese bidder Holly group

TCL International

TCL International

BOE Technology Group

Lenovo group

Target foreign firm/unit

Industry

Philips Semiconductors, CDM hand-set reference design (US), 2001 Schneider Electronics AG (Germany), 2002

Telecommunications

Thomson SA, Television manufacturing unit (France), 2003 Hyundai display technology (South Korea), 2003 IBM PC Division (US), 2004

Electronics

Electronics

Electronics

IT

Source: From different news sources.

Platform (VRP) for the Internet Protocol Version 6 (IPV6) router. Lenovo have, after the acquisition of IBM’s PC division, three R&D centres in the world, one in USA, one in Japan and one in Beijing. 4.3.2.3 International acquisition International acquisition is another way of opening innovation in China (Table 4.5). Although this kind of strategy is very risky for Chinese companies, we have seen lots of such cases, such as Lenovo’s acquisition of IBM’s PC division or TCL’s case of acquisition of Thomsons’ television business. 4.3.4 The role of government In explaining the Japanese success, the former Ministry of International Trade and Industry (MITI) used to be considered important for Japanese industrial catch-up. In Korea, Lin and others also regard the role of government as being very important, for example through its provision of public knowledge and picking of successful strategies. China is a transition economy. Therefore, at the current stage, the government will play an important role in industrial catch-up. Targeting progressive industry is a government strategy for catch-up. Most developing countries will provide more resources for new and high technologies than for traditional industries. IT is the industry that all Asian countries have tried to develop fast. History shows that this kind of targeting of a dynamic industry can provide good returns.


100 Xielin Liu Table 4.6 National S&T programmes (in 0.1 billion RMB) 1996

2000

2001

2002

2003

–

5

6

7

8

9

25

35

45

Basic research (started in 1997, called ‘973’) High-tech R&D programme (started in 1986, called ‘863’) Key technologies R&D programme Torch programme (for companies in high-tech zones) Spark programme (for rural technologies based on SMEs) Key S&T diffusion programme

4.5

–

2004

2005

2006

10

13.5

35

50.1

24.7

5.2

10.3

10.6

10.6

12.5

14.6

14.7

30.0

0.51

0.5

0.5

0.5

0.5

0.5

0.5

1.1

0.39

0.4

1

1

1

1.05

1.17

4.02

0.19

0.2

0.2

0.2

0.2

0.23

0.23

–

Source: MOST (2007a).

Government science and technology programmes and five-year or long-term plans are the key policy tools in China. Table 4.6 provides a brief overview of the main programmes controlled by the Ministry of Science and Technology (MOST). A key issue is to what extent the government should interfere in the catch-up process, especially as market institutions become more and more important in resource allocation and since China has become a member of the World Trade Organization (WTO). For China’s IT industry, the government first used the strategy of market for technology (Lu and Feng, 2004). In the first ‘Law of Sino-Foreign Equity Joint Ventures’ adopted in 1979, Article 5 states that foreign companies should use advanced technology and equipment in their joint ventures with Chinese companies. For example, the government



required the joint venture of Shanghai Bell to manufacture large-scale integrated chips in China. At the same time, as most of the Chinese counterparts in the joint venture were state-owned at the time, mostly regional telecommunications equipment companies of the Ministry of Post and Telecommunication (MPT). So, MPT would sometimes use this advantage to ask Shanghai Bell to form R&D consortia with domestic companies. However, the direct result of this strategy for knowledge transfer is debatable. Nevertheless this strategy provided Huawei and ZTE with some market space that enabled them to survive in the early days. Later on, after China became a member of the WTO, this strategy came to end. The government also directly supports a technology called TD-SCDMA, developed by a former government research lab called Datang Telecommunication Research Academy. The support includes frequency allocation, alliance setting, testing and first launching. All these measures sent a strong signal that TD-SCDMA technology is now an authorized technology for future 3G markets. 4.3.5 Spillover of FDI and learning IT is the most FDI-intensive industry in China. In 2006, foreign-related companies contributed 76.98 per cent of total value added in the IT industry in China (Table 4.7). FDI has offered Chinese companies good opportunities to learn the latest technology on a global level. By way of joint venture agreements, many Chinese counterparts acquired production knowledge, from knowledge of assembling and Table 4.7 Outline of foreign-related companies in the IT industry in China

Number of firms Share in the IT industry (%) Value added (0.1 billion RMB) Share in the IT industry (%) Total employees (1000s) Share in the IT industry (%) Total profits Share in the IT industry (%)

2004

2005

2006

4780 38.51 3874.33

6480 40.48 5025.87

6922 40.82 6277.91

78.85

75.01

76.98

2729

3730.4

4328.3

62.03

67.76

69.10

782.06

822.02

1049.29

77.88

76.53

75.81

Source: Ministry of Information Industry. Yearbook of statistics of China’s Information Industry (electronic version).


102

Xielin Liu

testing to knowledge on manufacturing of circuit boards, quality control and manufacturing information systems. They also gained knowledge on maintaining, servicing and training from their parent companies. Mu and Lee (2005: 15) state that Shanghai Bell has established many maintenance centres, widely circulated information about their System-12 and trained many qualified engineers since its operation in China started. However, multinationals usually do not transfer their key technologies. Even though they set up their R&D centres in China, these will not have much direct contact with the company in China. An important type of knowledge spillover concerns managerial knowledge, including marketing and human resources. In the earlier days, domestic companies had limited knowledge beyond some on production. The entry of multinationals gave Chinese companies lots of good opportunities to learn. A survey on the source of knowledge from domestic companies in the end of 1990s showed that there was a strongly demonstable effect of FDI. The Chinese regard information on multinational products, product exhibitions and specialized journals as their most important sources of knowledge on technology and innovation. They will also obtain knowledge of which products are profitable, so that Chinese companies can import first and practise reverse engineering later on (Gao, 2004). 4.3.5.1 Performance of catch-up IT is the largest industry in China now and it is having a widespread impact on Chinese economy and society. IT gave birth to new and now giant companies such as Huawei, Lenovo and ZTE. Huawei generates 70 per cent of its sales on the international market and has become the third largest hardware company on the European market in 2007. Table 4.8 Some indicators of catching up in the telecommunications industry

Key products

International commercial time

Analogue programme 1965, Bell switcher Digital switcher 1970, France GSM base station 1991, Ericsson CDMA base station 1995, Qualcomm WCDMA base station 2001, Ericsson

Commercial time in China Time gap 1986, Great Dragon

21 years

1989, ZTE Huawei, ZTE, Datang, 1997–9 2001, ZTE 2003, Huawei

19 years 7 years 6 years 2 years

Source: Zhang and Liu (2008), 353.


China’s Catch-up and Innovation Model 103 Table 4.9 Sales and R&D expenditure of Huawei and ZTE (2001–5)

ZTE

Sale R&D R&D/sale (%)

2001

2002

2003

2004

2005

2000–6

10.92 0.79 7.2

– – –

17.45 1.33 7.6

22.70 2.25 9.9

21.58 1.96 9.1

– – –

Total invention patents applications Huawei

Sale R&D R&D/sale (%)

16.23 3.05 18.8

17.21 3.06 17.8

21.67 3.18 14.7

Total invention patents applications

3400 31.52 3.97 12.6

46.97 4.75 9.6

– – – 11,894

Source: ‘China top 100 IT companies’. www.ittop100.gov.cn.

Chinese companies quickly narrowed the technology gap with that of companies in developed countries (Table 4.8). In a sense, the industry provides a good example for catching up and sheds light on how to catch up in an open, dynamic and global world. At the same time, both Huawei and ZTE have spent a large proportion of their funding on R&D and have become the leaders in invention patents in the Chinese IT industry (Table 4.9).

4.4

Conclusion and discussion

China has narrowed the technology gap with developed countries since its opened up and adapted to the market economy system. Not all countries manage to achieve the same speed in catching up, which in any case is not easy for such a large country. So the Chinese experience can provide other developing countries, especially other BRICS countries, with some lessons to learn and share. First, a large home market can be a powerful leverage for companies in an open economy to compete with multinationals from around the world. A market-oriented innovation strategy is the most common and effective strategy in winning the catch-up war. This is what Japanese companies’ shop-floor innovation did 30 years ago in order to catch up with the US. Multinationals tend to care about the high-end, highquality market. This gives Chinese companies windows of opportunity to catch up. So, innovation aiming at the rural, low-end market is the passport for Chinese companies to win in the world market. Huawei, for example, is now the third biggest hardware provider in EU market and it can provide 3.5G technology. Some 70 per cent of its revenue in


104

Xielin Liu

2007 came from the international market. ZTE has a relatively similar market performance. Second, opening up and globalization can give companies in developing countries the opportunity to use global knowledge and alliances. But learning capabilities and appropriate innovation strategy are the keys to the goal of catching up. Traditional infant industry protection theory will not work in an open economy. We found that government market protection cannot promote innovation capabilities. Lenovo, Huawei and ZTE are not owned by the state and did not benefit from government protection. Yet, some national technology programmes, the promotion of linkages between industry, universities and government research institutes (GRI) can greatly support Chinese companies in entering and succeeding in the dynamic IT industry. In the case of the TD-SCDMA technology for 3G markets, government help for the industry came in the form of R&D, bank loans, market access, standard setting and purchase. However, this kind of protection only made Datang rely on the government for further support with little progress being achieved in terms of innovation and service. We also found that FDI can be a positive factor for catch-up in developing countries by providing frontier technology and the diffusion of knowledge. We observed positive spillovers from FDI in three cases: Shanghai Bell and others in fixed phone switches; Nokia, Ericsson, Motorola and Qualcomm in GSM and CDMA and Siemens in TD-SCDMA. For example, many key engineers in domestic companies were first trained in joint ventures, In a dynamic and advanced industry like IT, FDI can be a very important factor for technology transfer and catch-up. The final important lesson is that companies have to learn and spend on R&D themselves in this catching-up process. Huawei and ZTE have never regarded themselves just as makers of low-priced products. They continue spending on R&D and even go overseas to set up R&D centres. They also use alliances to speed up their innovation processes. Now, the innovation model is about to change. In order to have high economic growth, regional governments that control the prices of land, resources, labour and pollution have kept them at a low level. Chinese companies take advantage of this policy but spend less money on R&D and innovation. As the costs of labour, land, resources and pollution will increase in the future, this will bring China to a turning point towards a new path of development where some companies will die out because of rising costs, and others will survive with more innovative capability.



Note 1. ZTE is the second largest telecommunications equipment manufacturer in China. It was established in 1985, with sales of RMB34 billion in 2007.

References Archibugi, D. and C. Pietrobelli (2003), ‘The globalization of technology and its implication for developing countries: Windows of oppurtunity or further burden’, Technological Forecasting and Social Change, 70, 861–83. Bell, M. and K.L.R. Pavitt (1993), ‘Technological accumulation and industrial growth: Contrasts between developed and developing countries’, Industrial and Corporate Change, 2, 157–210. Cesaroni, F. (2004). ‘Technological outsourcing and product diversification: Do markets for technology affect firm’s strategies?’ Research Policy, 33, 1547–64. Chen, D. and L. Liu (2003), The Truth of Huawei (in Chinese). Beijing: Modern China Press. Chesbrough, H.W. (2003), Open Innovation, the New Imperative for Creating and Profiting from Technology. Cambridge, MA: Harvard Business School Press. Fagerberg, J. (1988), ‘Why growth rates differ’, in G. Dosi, C. Freeman, R. Nelson et al., Technical Change and Economic Theory. London: Pinter, 432–57. Freeman, C. (1987), Technology Policy and Economic Performance. London: Pinter. Freeman, C. (1988), ‘Japan: A new national system of innovation?’ in G. Dosi, C. Freeman, R. Nelson et al. (eds) Technical Change and Economic Theory. London: Pinter. Furman, J.L. and R. Hayes (2004), ‘Catching up or standing still? National innovative productivity among “follower countries”, 1978–1999’, Research Policy, 33, 1329–54. Gao, S. (2004), ‘FDI and the technological progress in Chinese telecommunication equipment industry’, in Xaiozhuan Jiang (ed.) The Restructuring of S&T Resources and Chinese Industrial Technological Competence in the Process of Globalization. Beijing: Chinese Press of Social Science. Gerschenkron, A. (1962), Economic Backwardness in Historical Perspective. Cambridge, MA: Harvard University Press. Gold, A. R., G. Leibowitz and A. Perkins (2001), ‘A computer Lenovo in the making’, The McKinsey Quarterly, 3, 73–83. Hobday, M. (1995), Innovation in East Asia: The Challenge to Japan. Brookfield, VT: Edward Elgar. IMD (2004), World Competitiveness Yearbook. Lausanne: IMD. Kim, L. (1997), Imitation to Innovation. Boston, MA: Harvard Business School Press. Kondo, R. and C. Watanabe (2003), ‘The virtuous cycle between institutional elasticity, IT advancement and sustainable growth: Can Japan survive in an information society?’ Technology in Society, 25, 319–35. Lee, Keun and C. Lim (2001), ‘Technological regimes, catching up and leapfrogging: Findings from the Korean industries’, Research Policy, 30, 459–83. Liu, Xielin (2006), ‘Dynamic innovation system of China’, Background paper for OECD.


106

Xielin Liu

Liu, Xeilin (2008), ‘Globalization, catch-up and innovation (in Chinese). Beijing: Science Press. Liu, Xielin and S. White (2001), ‘Comparing innovation systems: A framework and application to China’s transition context, Research Policy, 30, 1091–114. Lu, Feng and Feng Kaidong (2004), ‘Policy for developing domestic autonomous proprietary automobile industry’, Working Report of Ministry of Science and Technology (In Chinese). February. Mu, Q. and K. Lee, (2005), ‘Knowledge diffusion, market segmentation and technological catching-up: The case of the telecommunication industry in China’, Research Policy, 34, 759–83. National Bureau of Statistics (2006), China Yearbook of Statistics. Beijing: China Press of Statistics. Nelson, R.R., ed. (1993), National Innovation Systems: A Comparative Analysis. Oxford: Oxford University Press. Perez, C. and L. Soete (1988), ‘Catching up in technology: Entry barriers and windows of opportunity’, in G. Dosi, C. Freeman, R. Nelson et al. (ed.) Technical Change and Economic Theory. London: Pinter, 458–79. Pisano, G. (1990), ‘The R&D boundaries of the firm: An empirical analysis’, Administrative Science Quarterly, 35, 153–76. Porter, M.E. (1990), The Competitive Advantage of Nation. New York: Free Press. Romer, P. (1990), ‘Endogenous technological change’, Journal of Political Economy, 98, 71–102. Veugelers, R. and B. Cassiman (1999), ‘Make and buy in innovation strategies: Evidence from Belgian manufacturing firms’, Research Policy, 28, 63–80. World Bank (1993), The East Asian Miracle: Economic Growth and Public Policy. New York: Oxford University Press. Xie, W. and S. White (2005), ‘Capability building in Lenovo: Strategic choice and path dependency’, working paper.


5 Science and Technology and Economic Growth in South Africa: Performance and Prospects David Kaplan

5.1

Introduction

The main contribution to economic growth in South Africa prior to the early 1990s resulted from factor accumulation – principally capital, but also labour. Technological progress, as measured by total factor productivity (TFP) growth effectively made no contribution. From the 1990s, however, the economy shed labour such that labour made a negative contribution to growth. With low levels of investment, capital made a much smaller, albeit positive, contribution to growth. TFP growth, by contrast, became the major source of growth (Fedderke, 2005: 10). The contribution of TFP to economic growth in South Africa receives little recognition. One significant illustration of this is the recent international panel that examined South Africa’s policies for growth and, specifically, its aim to achieve both a more ambitious growth target and more inclusive employment generating growth. A number of proposals were made by the panel (for a summary see Hausmann, 2008). However, the panel made no study of, or recommendations in respect of, innovation or technological change more broadly – despite this being the principal contributor to growth. This chapter examines South Africa’s technological performance. Section 5.2 provides a high-level assessment of the innovation performance of the South African system. Section 5.3 identifies the binding constraint on innovation performance. Section 5.4 examines the government’s strategic goals and aspirations for the South African science and technology (S&T) system as detailed in a ten-year plan (DST: 2007). Section 5.5 concludes with some broad proposals for future policy directions. 107


108 David Kaplan

5.2

South Africa’s innovation performance

The S&T system faced a variety of threats and constraints in the early 1990s. But there was no collapse. The system was stabilized and a substantive re-reorientation of the system towards the new goals of the first democratic government was successfully achieved. As the Organisation for Economic Co-operation and Development’s recent review of the S&T system expressed it, ‘... the most striking achievement of South Africa has simply been to defy the extremely poor framework conditions facing the innovation system in the early 1990s’ (OECD, 2007: 4). 5.2.1 Inputs Indeed, the system has grown steadily. Between 1993 and 2004, gross expenditure on R&D (GERD) almost doubled as GERD increased from 0.60 to 0.87 per cent. Since 2001, GERD as a share of GDP has increased by 5 per cent per annum. All performers of R&D have seen steady increases in spending. Of particular importance is that the share of business enterprises in R&D is high and this share has been increasing. Business now accounts for 56.3 per cent of R&D. This is a larger share of R&D than in most comparator countries (OECD, 2007: 5). Higher education (21.1 per cent) and the government (20.1 per cent) are each responsible for approximately one-fifth of R&D performed, with the not-for-profit sector making up the remainder. The business sector provides 45 per cent of total R&D funding. A total of 68 per cent of R&D in the business sector is carried out by local business and a further 18.4 per cent by foreign businesses. The government provides 32 per cent and foreign sources provide 15 per cent of total R&D funding. Both the government and the higher education sector are recipients of significant R&D funding from business and the foreign sector. The funders and performers of R&D are summarized in Table 5.1.

Table 5.1

Funders and performers of R&D, 2004/05 (R millions)

Funder Performer Business Government Higher Education

Business enterprise

Government

Foreign

Other SA

Total

– 4735 296 426

– 519 1726 1610

– 1280 312 241

– 430 178 257

– 6964 2512 2534

Source: DST, 2007a: 23.


Performance and Prospects

109

In 2004/5, there were a little fewer than 30,000 full-time equivalent (FTE) personnel engaged in R&D. At 2.7 researchers per 1000 employees, this is low relative to other developed countries but also relative to a number of countries with a comparable gross domestic product (GDP) per capita. Although South Africa devotes comparatively large resources to R&D, this is not reflected in a greater number of researchers employed. The reason is that South African research workers command significantly higher salaries than in comparable countries. Moreover, while the number of FTE researchers has increased, this rate of growth has been slow – only 7 per cent in the period 1992–2004. This growing commitment of resources has been accompanied by extensive policy experimentation. Grounded in the overall concept of the national system of innovation (NSI), policy design has drawn extensively on international experience and thinking. It has included policies to improve governance of the innovation system; the more effective functioning of key performers of S&T, especially the science councils; new mechanisms for funding R&D and innovation and development of new organizational arrangements and programmes to support R&D and to undertake R&D directly. While there is clearly considerable room for improvement, extensive institutional reform and ongoing evaluation have enhanced efficiency of organizations and increased inter-organizational cooperation. The institutions undertaking, financing and supporting innovation, in the main, function effectively. More resources combined with an effective, improving, institutional structure and innovative and directed policy changes might have been expected to yield significant results. However, the output indicators for the S&T system are disappointing. 5.2.2 Outputs 5.2.2.1 Publications In terms of publications, there has been only a slight increase in South Africa’s scientific publications, as listed by the Institute for Scientific Information, since 1994. In relative terms, South Africa’s global share has declined significantly from a peak of 0.7 per cent in 1987 to 0.48 per cent in 2003. In contrast, other comparator countries, such as Brazil, Taiwan, South Korea and India, starting from a lower base, have overtaken South Africa as their share of world publications has climbed steadily (Pouris, 2003: 425–56).


110

David Kaplan

5.2.2.2 Patents The situation with regard to South African patents registered abroad is broadly similar to that for publications. There has been a slight increase in patents since 1994, but no clear trend is evident. Indeed, the number of South African patents registered in the US declined significantly in the period 1998–2005. South Africa’s relative position in terms of patents has declined. The share of all foreign patents registered with the USPTO has consistently declined from 0.28 per cent in 1992 to 0.13 per cent in 2007. For utility patents, South Africa’s share declined from 0.3 per cent in 1992 to 0.11 per cent in 2007. A recent study of the patenting activity at the USPTO by the five most innovative South African universities concludes that their performance is well below that of other countries (Lubango and Pouris, 2007: 7). In terms of South African patents at the Patent Cooperation Treaty (PCT), the trend is less defined. The PCT became operative in South Africa only in March 1999. Since that date, there has been no clear trend and considerable fluctuation. However, as with South African patents at USPTO, the number was lower in 2007 than in 2001. Moreover, as with USPTO, South Africa’s share of all foreign PCT patents declined consistently – from 0.42 per cent in 1999 and 2000 to 0.26 per cent in 2007. Table 5.4 shows the number of patent applications at the local Companies and Intellectual Property Office (CIPRO). The number of patents remained constant in the period 1994–8 and, then, declined by a third. Numbers picked up again in 2001 but have since been slowly declining. The drop in 1999 can be attributed to South Africa becoming a part of the PCT system in March 1999. Overall, local patenting at about 270 patents per million population is low and, at least since 2001, shows a slowly declining trend. A recent study concluded, ‘... at least 50 per cent of the patent applications filed at CIPRO in this period (2000–2) were filed by foreign nationals, with the biggest contributors being USA and German nationals’ (Innovation Fund, 2007: 28). A clear and consistent trend is evident for the 2001–6 period: 1) the overall number of patents has been stagnant, 2) the number and the share of South African patents has been declining and 3) pari passu the number and, particularly, the share of foreign patents has risen significantly. A detailed breakdown of patents filed in South Africa by country reveals that every country increased the number of its patents between 2004 and 2006. However, patenting by South Africans declined in the same period, by nearly 12 per cent.


Table 5.2

Patents of South African origin granted by the US Patent and Trademark Office (USPTO), 1994–2007 1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

Utility patents

101

123

111

101

115

110

111

120

113

112

100

87

109

82

All patents

109

127

116

114

132

127

125

137

123

131

115

108

127

116

Source: USPTO, 2006; USPTO, 2006a. Data for 2007 supplied by Paul Harrison of the USPTO, 18 January 2008.

Table 5.3

Patents of South African origin in the PCT 1994–2007 1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

30

42

72

84

114

317

387

419

384

357

411

358

422

360

SA patents

Source: WIPO Statistics Database.

Table 5.4

Patent applications at the Companies and Intellectual Property Office (CIPRO), South Africa, 1994–2005

Total Patents

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

10,414

11,050

10.956

11,734

11,953

7838

7793

10,553

10,048

9955

10,396

10,044

Source: Data are for provisional and complete patent applications. 1994–7 from Teljeur (2003: 55); 1998–2005 from Innovation Fund (2007: 36).


112

David Kaplan

Table 5.5 Number and share of South African and foreign patent applications (filed and granted) at CIPRO, 2000–2 and 2004–6

Total South African Foreign Foreign percentage of total

2000

2001

2002

2004

2005

2006

7793 5204 2589 33

10,553 4985 5568 53

10,408 4721 5687 55

10,420 4587 5843 56

10,456 4328 6128 59

10,787 4058 6729 62

Source: Data for 2000–2 from Innovation Fund (2007): 28. Data for 2004–6 supplied by CIPRO.

Table 5.6

Patents filed in South Africa, by country of origin, 2004–6

Year

SA

US

UK

FR

DE

NL

CH

JP

Others

2004 2005 2006

4587 4328 4048

1953 2112 2269

506 529 555

275 284 429

716 782 858

223 193 270

387 373 434

268 313 277

1517 1443 1666

Source: Data supplied by CIPRO.

5.2.2.3 Royalty receipts and payments1 The technology balance of payments (TBP) reflects a country’s commercial transactions related to international technology and know-how transfers. It consists of payments made or received for the use of patents, licences, know-how, trademarks, designs and technical services. These receipts and payment are, generally, registered as royalties paid abroad and royalties received from abroad. In South Africa, consistent data for royalties received is available only from 2000. Analysis of the data regarding royalties from 2000 shows that in the 2000–7 period, royalties received from abroad increased by 58 per cent, amounting to a compound annual rate of 6.8 per cent. In the same period, royalties paid to abroad increased by 360 per cent, amounting to a compound annual growth rate of 20.1 per cent. Moreover, royalty payments greatly exceeded royalty income. In 2000, royalty payments were some ten times royalty income. In 2007, royalty payments were 30 times royalty income. The widening adverse technological balance of payments and, more significantly, the relatively slow growth of royalty receipts from a very low base are a further indication of South Africa’s weak overall performance in innovation.



113

5.2.2.4 Shares of global trade A country’s performance in global trade in industries where innovation is central to economic success is an important output indicator, particularly in respect of business sector R&D. In the period 1992–2005, South Africa’s exports of high-tech products grew by 9.5 per cent annually. While, at first sight, this appears impressive, it is lower than the global average (11 per cent), and well below that for developing countries (21 per cent). As a result, as with scientific publications and patents, South Africa’s share of global high-tech exports has declined. This cannot be attributed to South Africa’s dependence on commodity exports. Brazil’s share of global high-tech exports has increased, and the share of hightech exports in total national exports has risen significantly. South

Table 5.7 Share of global high-tech exports 1992–2005, and share of high-tech exports in national exports 1992–2002: South Africa, Brazil and Argentina

Share of global high-tech exports South Africa Brazil Argentina Share of high-tech exports in national exports South Africa Brazil Argentina

1992

2005

0.07 0.29 0.05

0.07 0.49 0.03

1992

2002

1.63 3.9 2.1

2.51 10.5 2.6

Source: COMTRADE, TIPS and own calculations.

Table 5.8 Knowledge Economy Index (KEI): South Africa in comparative perspective, 1995–Latest Year

South Africa Brazil Argentina Upper middle income World

1995

Latest year

6.08 5.14 6.41 6.38 6.41

5.64 5.50 5.49 6.50 5.93

Source: World Bank (2007: 7). Accessed at http://info.wordbank.org/etools/kam2/KAM_page7. asp 7 September 2007.


114

David Kaplan

Africa’s export performance was similar to that of Argentina. The share of high-tech products in South Africa’s total national exports is much lower than that of Brazil and marginally lower than Argentina’s. 5.2.2.5 Composite indicators There are a number of composite indicators that measure the ability to generate, adopt and utilize new knowledge. The Knowledge Economy Index (KEI) compiled by the World Bank, probably the most widely used, is based on the average of the normalized performance of a country or region in four areas – economic incentives, institutional regime, education and human resources and the innovation system and information and communications technologies (ICT). In terms of KEI, South Africa has declined since 1995. By contrast, other commodity-based exporters, such as Brazil, have seen a rise in the index, as has the upper middleincome countries group. Overall, South Africa is currently ranked 50 out of 140 countries, a decline of nine places since 1995. South Africa’s performance on KEI is, again, similar to that of Argentina. 5.2.2.6 Constraints on performance In summary, the resources committed to R&D in South Africa are commensurate with other countries at similar stages of development and have been increasingly significant. Moreover, business accounts for a very significant and rising share of expenditure on R&D. However, the number of personnel engaged in research is lower than for many comparable countries and has risen only slowly. This reflects the high price of skills engaged in research, which, in turn, is a consequence of the limited supply of the needed skills. The high level output indicators – publication counts, patents (local, US and PCT), royalty receipts and payments, shares of global trade and composite KEIs – all tell essentially the same story. Despite the injection of more resources and the introduction of a raft of new policies derived from international experience that have significantly improved the policy environment, at aggregate level, South Africa’s innovation performance is largely stagnant if not declining slightly.

5.3

Accounting for poor performance

The poor performance of the system, as evidenced consistently by all of the output indicators, strongly suggests the existence of system-wide constraints. There is an emerging consensus that the key system-wide constraint is a severe shortage of skills (Blankley and Kahn, 2005; Kahn,



115

2006; Kaplan, 2007; NACI, 2006). The OECD review of the South African S&T system provides further confirmation that the skill shortage is the binding constraint on further development. The OECD identifies a looming crisis in human resource development in two areas. The first is a large and growing engineering gap. ‘A very large gap appears to be opening up between the supply of design, engineering and related managerial and technical capabilities and the demand for such resources being generated by the increased rate of investment across the economy’ (OECD, 2007: 7). The second is the very limited supply of university graduates capable of undertaking research. Unless this is addressed, the entire innovation system will be constrained (ibid.). The gross enrolment ratio (GER) for secondary schools has increased only moderately, from 87 per cent in 2000 to 89 per cent in 2005 (Department of Education, 2002: 6; 2006: 7). The number of school leavers passing mathematics at higher grade has remained stable at some 20,000 annually since 1997, which is a pitifully low number. Overall, the national university participation rate declined, from 17 per cent in 1993 to 15 per cent in 2001. The government is aiming for an overall participation rate for higher education of 20 per cent by 2012, which would require the higher education system to increase in size by a third. University undergraduate enrolments and qualifications have expanded by some 3.6 per cent annually, but the growth for engineering (1.4 per cent) and natural sciences (2.4 per cent) has been much lower. Currently, only 20 per cent of all graduates are in science and engineering. While doctoral enrolments have increased significantly, overall, university doctoral graduations have grown by only 3 per cent annually (NACI, 2006: 64–5). A consequence of the slow growth in the supply of human capital combined with an economy that is growing more rapidly in a way that is becoming ever more demanding for skills is that the numbers undertaking research have virtually stagnated. FTE researcher numbers grew by only 7 per cent in the period 1992–2004, see Table 5.9.2 Table 5.9

FTE researchers by sector, 1992–2004

Sector

1992

2004

Business Government Higher Education

3395 2428 3631

4411 2342 3374

Total

9454

10,127

Source: NACI (2006: 57); OECD (2007: 51).


116

David Kaplan

The centrality of the constraint on the S&T system arising from the shortage of high-level skills evidenced by the macro numbers is confirmed at the micro firm level. A recent study of 20 successful high-tech firms found that the inability to acquire more high-level skills was the major factor inhibiting these firms from expanding (Kaplan, 2007).3 The evidence is clear. Any further expansion of the S&T system will be predicated on a significant expansion in the supply of skills. The OECD’s assessment is that, overall, to facilitate expansion of the system, the university system needs to expand by a third (OECD, 2007: 89). Since university education is the end of a long pipeline and the number of matriculants leaving school, especially in mathematics and sciences, is static, such an expansion will necessarily, under any circumstances, take a considerable time. In the absence of an expansion in the supply of skills, any additions to the number and breadth of the areas of scientific and technological development to be supported can succeed only by if scarce skilled resources to the new additions are attracted from somewhere else in the system. The case of the Pebble Bed Modular Reactor (PBMR) is indicative. The recent expansion of the PBMR development programme was only achieved through the PBMR obtaining many of the additional skills it required at the expense of innovation activities and organizations elsewhere (ibid.: 136). The OECD report drew attention to the existence of: ‘... multiple projects chasing thinly spread resources’ (ibid.: 135). In 2002, in the National Research and Development Strategy (NRDS), four new missions for S&T were proposed (DST, 2002). The OECD findings were that two of these new missions had not been implemented, namely technology related to resource-based industries and technology and innovation for poverty reduction. The lack of implementation in respect to technology and innovation for poverty reduction is particularly problematic, given the high levels and persistence of poverty in South Africa despite moderate economic growth. The NRDS regarded this mission as particularly critical in the areas of health and education (Kaplan, 2004: 281). The OECD concluded: ‘... there appears to be no overall framework which coordinates and sets priorities about what the innovation system is doing with respect to poverty reduction or where the main problems and limitations are located’ (OECD, 2007: 138). The failure of implementation is likely also to have reflected stretched organizational and human resource capacities. The constraint imposed on the system by the shortage of skills has already curtailed the output and productivity of the system as a whole



117

and limited the new technology missions introduced under the NRDS in 2002. Recognition of the current effect of the skills constraint is particularly germane in the context of ambitious proposals for the S&T system as outlined in the Department of Science and Technology’s (DST) ten-year plan.

5.4 The DST ten-year plan The DST has put forward a ten-year plan, 2008–14, for South Africa’s innovation system (DST, 2007). At the time of writing, the plan is in the form of a draft document for discussion. The plan sets out a vision of where South Africa needs to be in 2018 (ibid.: 5) and outlines a series of concrete goals to be achieved. The plan is, however, more than simply a bold vision. It calls for targeted interventions in pursuit of this bold vision. ‘Nations that have achieved accelerated growth in outputs and capabilities have acted decisively, targeting investments in areas of strategic opportunity.’ The plan outlines a bold stance, targeting specific outcomes for 2018 (DST, 2007a: 8). In similar vein, the plan’s report declares: ‘... an opportunity exists for bold interventions that will secure a greater share of global markets’ (DST, 2007: 5). 5.4.1 System expansion and goals The plan envisages that South Africa will undergo a major transformation towards a knowledge-based economy. This transformation will be evidenced through progress in a number of indicators. These indicators are 1) economic growth attributable to technical progress, at 30 per cent (against 10 per cent in 2002); 2) national income derived from knowledge-based industries; workforce employed in knowledge-based jobs; 3) proportion of firms using knowledge to innovate, at more than 50 per cent; 4) GERD/GDP, at 2 per cent (0.87 per cent in 2004); 5) a global share of research outputs, at 1 per cent (0.5 per cent in 2002); 6) high-medium technology exports as a percentage of all exports, at 66 per cent (30 per cent in 2002) and 7) number of South Africa-originated US patents, at 250 (100 in 2002). Achieving these goals would represent a very significant expansion of the system and a sharp reversal of the trend of South Africa declining in comparative global performance. To achieve this turnaround would require improvement in a series of enablers. These are 1) matriculants with sufficient expertise in mathematics and science to enter university, at nine per cent (3.4 per cent in 2002); 2) science, engineering and


118

David Kaplan

technology (SET) tertiary students as percentage of all tertiary students, at 30 per cent; 3) annual Ph.D. graduates, at 2200 (963 in 2002); 4) gross availability of SET graduates to economy, at 450,000 (235,438 in 2002); 5) number of FTE researchers at 20,000 (8,708 in 2002) and 6) total researchers per 1000 population (DST, 2007: 9). 5.4.2 Grand challenges Much of the plan is concerned with outlining five grand challenges. These are seen as a way of expanding the research agenda, in general, as well as achieving specific outcomes (ibid.: 34): Biotechnology Space science and technology Energy Climate change Human and social dynamics It is envisaged that there will be significant innovation over a wide area in respect of each of these challenges. The objectives for each are specified as a series of envisaged outcomes. Brief outlines of the range of activity in regard to each challenge and some of the key envisaged outcomes are given below. Biotechnology The government has spent more than R450 million on biotechnology over the last three years. Given a range of natural advantages, a supportive policy environment could allow South Africa to be positioned as ‘... a major producer in the pharmaceutical, nutraceutical, flavour, fragrance and bio-pesticide industries’ (ibid.: 13). A number of results are envisaged. Foremost, South Africa would ‘be one of the top three emerging economies in the global pharmaceutical industry, based on an expansive innovation system using the nation’s indigenous knowledge and rich biodiversity’ (ibid.:15). Space science and technology In October 2005, the Minister of Science and Technology announced a three-year satellite development project. From 2005 through 2008, the government will invest R26 million to secure a mission-ready small satellite and increase technological capacity. The programme is centred at the University of Stellenbosch. The DST ten-year plan proposes a national space agency. This has come to fruition with the announcement by



119

the Deputy Minister of Trade and Industry, of the South African Space Agency. The agency will coordinate activities in space sciences, earth observation, communication, navigation and engineering services. A core objective in this area is, ‘... to win a growing slice of the global satellite industry’ (ibid.: 16). Among other results, it is envisaged that, by 2018, South Africa will have ‘independent earth observation highresolution satellite data available for all of Africa from a constellation of satellites designed and manufactured in Africa, undertaken at least one launch from South African territory in partnership with another space nation, and have in place a 20-year launch capability plan’ (ibid.: 18). Energy South Africa currently undertakes significant research in energy. There are two main foci. The first is conversion of coal and liquid natural gas to oil. This research is undertaken by South African Synthetic Oil Ltd (SASOL), the former state-owned company. SASOL is the largest R&D performer in the business sector, with an annual research budget of some R500 million. The second focus is nuclear energy. Since 1993, Eskom, the state-owned electricity generator and distributor, has been developing the PBMR. This draws on an original German technology for nuclear power generation. It is being significantly improved, adapted and incorporated into a technologically new, Generation 4, pilot plant that is due to be completed by 2012. Government funding for PBMR is currently running at more than R1 billion annually. It is the largest research project and employer of skilled engineers nationally and contributes materially to the growing shortage of design and engineering skills (OECD, 2007: 135). DST envisages a series of major R&D initiatives in energy: clean coal technologies, nuclear, renewable energy and hydrogen and fuel cells. Among the aims envisaged are: [to] expand the knowledge base for building nuclear reactors and coal plants parts; source more than fifty per cent of all new capacity locally; a twenty-five per cent share of the global hydrogen infrastructure and fuel cell market with novel PMG catalysts and have demonstrated, at a pilot-scale, the production of hydrogen by water splitting, using either nuclear or solar power as the primary heat source. (DST, 2007: 21) Climate change South Africa will lead research in Africa on understanding climate change, the impact of these changes and how to mitigate adverse effects.


120 David Kaplan

It is envisaged that South Africa will be ‘an internationally recognized science centre of excellence with climate change research and modeling capability, benefiting the entire continent and an internationally recognized centre of excellence focused on the Southern Ocean and its contribution to global change processes’ (ibid.: 22). Human and social dynamics DST will develop a long-term programme to increase basic understanding of human behaviour. This will entail increasing use of computer modeling and have cognitive, behavioural and social applications. There will also be research in paleoanthropology, archaeology and evolutionary genetics. ‘This research will provide evidence-based support for interventions in learning processes and education, indigenous knowledge systems and heritage legacy’ (ibid.: 23). No anticipated results are stipulated in respect of this challenge. 5.4.3 Enhancing innovation and human resources Finally, the plan addresses the issue of human resources. Achieving the targets identified for the system and for the grand challenges will require a very significant expansion in the provision of high-level skills. The Ph.D. rate will have to increase by a factor of five over the next 10–20 years. Other targets include 1) 210 research chairs at universities and research institutions by 2010 and 500 by 2018 (58 were in place in 2006), 2) a 2.5 per cent of global share of research publications (2006: 0.5 per cent) and 3) 2100 Patent Co-operation Treaty international applications originating in South Africa (2004: 418) (ibid.: 31). In addition to the five grand challenges, the plan provides for enhancement of innovation through creation of a technology innovation agency (TIA). An innovation chasm, namely a gap between the research that is produced locally and what the market demands was first identified in NRDS in 2002. The role of the TIA is to bridge that gap through funding, investment and venture capital support, brokering and complimentary services. 5.4.4 Concluding remarks The DST ten-year plan clearly specifies a very ambitious agenda for S&T over the next decade. It would add substantially to the areas of research activity that would receive public support and funding while maintaining all of the areas proposed in the NRDS in 2002. This would clearly compound the system problem of too few resources being too widely spread. The problem is exacerbated in that many of the grand



121

challenges are very capital intensive and require major injections of funding in order to reach critical mass.

5.5 Conclusion: Policy sequencing The OECD was concerned that too much was already being attempted, with the result that resources were being spread too thinly. As a consequence, it concluded that scales of activity were too limited to secure significant results. Noting that two of the technology missions established in 2002 by NRDS had not been implemented, the OECD recommended a complete re-examination of the priorities and consequent missions specified by NRDS. The import of the proposed re-examination is clear, namely to reduce the number of technology missions. By contrast, the DST plan emphasizes the importance of continuing to support all of the technology missions established by the NRDS. Similarly, the science missions (astronomy, palaeontology, antarctic and marine sciences, biosciences, social sciences, earth systems and environmental systems), early-stage research projects such as nanotechnology and increased R&D for conventional sectors that boost key national priority sectors such as agriculture and health, all of which were identified by the NRDS, should continue to receive support (ibid.: 30). In addition, the DST plan specifies a wide raft of new areas for support – the grand challenges. The OECD mission was also concerned that the research categories that qualified for state support were far too widely specified. The mission argued strongly that much more focus was needed to ensure identification of meaningful priorities. Areas of innovation qualifying for support should be more limited and areas that qualify should be more clearly and tightly specified. In regard to a number of the grand challenges proposed by DST, such as biotechnology and space, a large number of fields are identified while the specification of some of the grand challenges is vague and unfocused, in particular, human and social dynamics. The contrast is clear. The OECD has called for the S&T system to curtail the spread of its innovation activities as well as to be more focused and clearer in its specification of that focus. The DST plan envisages an S&T system that will significantly increase the spread of its innovation to a series of new activities, many of which are broadly specified. Moreover, it will do so without curtailing any of its existing activities. What of the human resource dimension, namely the critical shortage of skills for innovation and the implications for growth? There


122 David Kaplan

is evidence that R&D expenditure in South Africa has had a positive impact on growth in total factor productivity, which has, in turn, been the major source of economic growth, at least until recently. Human capital formation, specifically the output of school leavers and university graduates has been the second factor underpinning total factor productivity growth (Fedderke, 2005: 34). As illustrated above, all the output indicators show that increasing expenditure allocated to R&D is now having diminishing returns and that this results from critical shortages in the supply of skilled labour. Thus, the limited supply of skilled labour independently retards growth, but also does so through the impact it exerts on the output and productivity resulting from R&D expenditure. Any further expansion and improvement in performance of the NSI is predicated on significantly increasing the supply of skills. However, high-level skills are the end of a long educational and training pipeline. As a consequence, expansion in supply is almost always slow and incremental. Moreover, in the South African context, there is no recent historical experience or current indication to suggest that any significant expansion of that supply is likely in the short term. The number of Ph.D.s graduating, for example, has been expanding at only 3 per cent annually. To meet the targets specified by the DST would require an annual increase of nearly 30 per cent, some ten times the current rate of growth. Moreover, there are no indications that there is currently either the requisite number of students qualifying for entry into tertiary education or the capacity to massively expand tertiary education. Rather than a policy of attempting to advance on all fronts, priority needs to be given to the development of high-level skills. Rather than attempting to do everything at once, policy needs to be sequenced, the first priority being the expansion of high-level skills for innovation. Addressing this issue will be central to enhancing the productivity of the NSI and, by extension, to underpin higher rates of economic growth.

Notes 1. This section is based on data supplied by the South African Reserve Bank. Royalty payments and income are not currently published by the Bank. It, therefore, gave permission to the author to outline the data trends but not to publish the actual data. 2. The number of researchers in government and higher education has clearly fallen. There has been a moderate increase in business of 30 per cent over 12 years, but some of that probably reflects better survey coverage (Kahn, 2007: 10).



123

3. The OECD report similarly quoted senior managers of a mining engineering company, ‘We have a huge skill shortage. This company could do twice the export business if we had the people’ (OECD, 2007: 103).

References Blankley, William and Michael Kahn (2005), ‘The history of research and experimental development measurement in South Africa and some current perspectives’, South African Journal of Science, 101 (March/April), 151–6. Department of Education (2002), Education Statistics at a Glance in 2000. Pretoria: Department of Education (February). Department of Education (2006), Education Statistics at a Glance in 2005. Pretoria: Department of Education (November). Department of Science and Technology (DST) (2002), ‘South Africa’s National Research and Development Strategy’, Pretoria: DST, available at http://www. dst.gov.za/legislation_policies/strategic_reps/sa_nat_rd_strat.htm Department of Science and Technology (DST) (2007), ‘Innovation towards a knowledge-based economy. Ten-year plan for South Africa (2008–2018)’, Draft. Pretoria: DST (July 10). Department of Science and Technology (DST) (2007a), National Survey of Research and Experimental Development (R&D) (2004/05 Fiscal Year). Pretoria: DST. Fedderke, Johannes (2005), ‘South Africa: Sources and constraints of long-term growth, 1970–2000. The World Bank Informal Discussion Paper Series on South Africa. Paper 2005/01. Hausmann, Ricardo (2008), ‘Final recommendations of the international panel on ASGISA’ CID Working Paper No. 161 (May). Cambridge, MA: Center for International Development, Harvard University. Innovation Fund (2007) The State of Patenting in South Africa. Special Report. Analysis of the South African patent landscape (Mclean Sibanda). Pretoria: Innovation Fund Intellectual Property Management Office. Kahn, Michael (2006), ‘After Apartheid. The South African national system of innovation: From constructed crisis to constructed advantage?’ Science and Public Policy, 33 (2), March, 125–36. Kahn, Michael (2007), ‘Internationalisation of R&D: Where does South Africa stand?’ South African Journal of Science, 103 (January/February), 7–12. Kaplan, David (2004) ‘South Africa’s national research and development strategy: A Review’, Science, Technology and Society, 9 (2), 273–94. Kaplan, David (2007), ‘The constraints and institutional challenges facing industrial policy in South Africa: A way forward’, Transformation, 64, 91–111. Lubango, L.M. and A. Pouris (2007), ‘Industry work experience and innovation capacity of South African academic researchers’, Technovation, 27 (12), December, 788–96. National Advisory Council on Innovation (NACI) (2006), ‘The South African national system of innovation: Structures, policies and performance. Background report to the OECD country review of South Africa’s national system of innovation’, Pretoria: NACI (21 July).


124 David Kaplan Organisation for Economic Co-operation and Development (OECD) (2007), Review of South Africa’s Innovation Policy. Paris: Directorate for Scientific and Technology Policy OECD, 26–7 March. Pouris, Anastasios (2003), ‘South Africa’s research publication record: The last ten years’, South African Journal of Science, 99 (September/October), 425–8. Teljeur, E. (2003), Intellectual Property Rights in South Africa: An Economic Review of Policy and Impact. Johannesburg: The Edge Institute. United States Patent and Trademark Office (USPTO) (2006), Patents By Country, State and Year Utility Patents (December 2006). A Patent Technology Monitoring Team Report. Alexandria, VA: US Patent and Trademark Office, Electronic Information Products Division, Patent Technology Monitoring Team. http:// www.uspto.gov/web/offices/ac/ido/oeip/taf/cst_utl.htm United States Patent and Trademark Office (USPTO) (2006a) Patents By Country, State and Year – All Patent Types (December 2006). A Patent Technology Monitoring Team Report. Alexandria, VA: US Patent and Trademark Office, Electronic Information Products Division, Patent Technology Monitoring Team. http://www.uspto.gov/web/offices/ac/ido/oeip/taf/cst_utl.htm World Bank (2007) ‘Knowledge for Development (K4D)’. Accessed at http://info. worldbank.org/etools/kam2/KAM_page7.asp


6 Market-Oriented Reforms, Domestic Technological Capabilities and Economic Development in Latin America Jorge Katz

6.1

Introduction

Modern Growth Theory is not adequately equipped to illuminate the process of economic and institutional transformation that Latin American countries went through as they proceeded from ‘inward- oriented’, stateled macroeconomic policy regimes to ‘outward-oriented’, market-led models of macroeconomic management. Modern Growth Theory is specified in terms of an equilibrium algorithm, in which changes in institutions and in the structure of the economy, as well as the creation and destruction of production and technological capabilities that obtain in the transition from one macroeconomic policy regime to the other, are not part of the topics the model is set out to examine. New sectors of economic activity were built up in Latin America during the 1990s, while many ‘old’ industries were gradually phased out. Different forms of capital intensive, computer-based production organization technologies were brought on board by the larger firms in the economy displacing ‘old’ technologies and production organization routines. ‘Large’ firms reduced their degree of vertical integration, moving closer to the ‘in-bond’ assembly extreme of the production organization spectrum and relying more heavily on the external outsourcing of intermediate parts and components. On the other hand, most small and medium enterprises (SMEs) were not able to adapt well to the new rules of the game and lost their share of gross domestic product (GDP). Thousands of them closed down during the process – estimates being that around 8000 SMEs closed down in Chile, and more than 12,000 did so in Argentina, during the 1980s. 125


126

Jorge Katz

In spite of the major institutional and economic transformations each country in the region has gone through during the past three decades, we should notice that average labour productivity continues to lag significantly behind international standards. Estimates indicate average labour productivity in manufacturing tends to be around 40 per cent of that attained by US manufacturing industries. This is the situation in the richest countries in the region – Argentina and Chile – but the productivity lag is much larger in cases such as Ecuador, Paraguay and Bolivia. Such an assertion, however, is somewhat misleading. Average labour productivity pools together estimates for the (small) modern sector of the economy, whose productivity indicators are closer to those of more developed industrial nations, and the much lower productivity figures of the far larger section of the production structure which still operates well behind international standards. The modern sector in the economy comprises new production activities which were not present in the economy a few years ago (or were present employing much less sophisticated production technologies), including: 1) natural-resource-processing industries which now operate with highly automated, state-of-the-art production facilities; 2) high-productivity service industries including banks, telecoms, energy and tourism and 3) a few technology-intensive industries, such as aircraft design and construction in Brazil (Goldstein, 2005). The fraction of the workforce that belongs in the modern section of the economy is paid much higher real wages than the rest of the working population and has gradually developed consumption habits which are quite similar to those exhibited by the large majority of the citizens in more developed industrial nations. For the individuals in this section of society the question as to whether or not ‘convergence’ with more developed countries will ever take place constitutes a rhetorical question, in so far as their lifestyle and shopping basket are indeed quite similar (and in some cases better) to those attained by the average citizen of, for example, Madrid or Rome. The real problem, however, has to do with the other two-thirds of the economy. In this rather large segment of the workforce people still work in highly outmoded production facilities attaining only much lower levels of labour productivity than in the former group. It is this wide segmentation of the production structure that appears as a matter of concern at present, as it has given cause to new and more intractable forms of tension and exclusion in society, making economic and political governance an increasingly difficult proposition. In other words, market-oriented reforms and the process of globalization of the world economy have resulted in structurally more heterogeneous


Reforms and Domestic Technological Capabilities

127

economies in which a (small) fraction of the population enjoys much greater access to the artefacts of modernity while an extended majority of the people suffers from exclusion in crucial areas such as high-quality education or health services, let alone clean water and sanitation services, social security protection, adequate transport services, etc. In this chapter I intend to examine the relationship between growth and the transformation of the institutional and production structure of the economy. This is a difficult topic to examine on the basis of formal equilibrium models, as they simply lack the specification capable of throwing light upon the issues we want to illuminate here. Rather, I shall proceed on the basis of individual industry case studies which clearly indicate the extent to which the inception of new activities in the economy involves the co-evolution of institutional, economic and technological forces which retrofit with each other during the process of growth. It is such a process of co-evolution which we find difficult to grasp on the basis of received equilibrium growth models.

6.2 Structural change and economic development: Conceptual issues In the classical tradition, the process of economic development is associated with changes in the structure of the economy. Adam Smith described the process as resulting from more ‘roundaboutness’ in production organization and from increasing returns to scale deriving from production specialization. A growing economy is one that becomes more complex through time, with the creation of new sectors of economic activity and the entry of new, more knowledge-intensive, firms in the economy. Pari passu with the above, new institutions, skills and learning processes develop in the social setting. It is such processes that led Moses Abramovitz to speak about ‘immediate’ and ‘ultimate’ sources of economic growth. An expanding capital/labour ratio constitutes in his parlance an ‘immediate’ source of growth, while learning, the accumulation of domestic technological capabilities, institutional changes and the development of a denser ‘cluster’ of interacting agents are considered by him as the ‘ultimate’ expression of what economic development is all about. Given the above, it becomes apparent that the long-term performance of the economy should not be described exclusively in terms of an aggregate algorithm but should also pay attention to changes in the structure of the economy and in the institutional set-up in which the economy operates. Economic growth involves in a very profound way


128 Jorge Katz

new institutions, learning processes and the creation of social capabilities. Macroeconomic circumstances are of course important, but it is overly simplistic to believe that they are a sufficient condition for a successful process of social change to obtain. Many of the changes in the ‘ultimate’ sources of growth we are referring to come together with the inception of new production activities in the economy. As new activities are started up and production capacity expands, learning processes take place and new institutions and forms of social interaction among agents in the economy develop. Furthermore, the inception of new activities is associated with numerous market and non-market forms of interaction both among firms and between them and other organizations such as universities, engineering associations, government regulatory bodies, municipal authorities, etc., many of which do not even operate on the basis of conventional market rules. The process is surrounded by externalities and synergies which conventional market analysis is simply unable to perceive. The functioning of any given production structure involves much more than market-driven exchanges among firms, individuals and public sector agencies. Contrary to modern growth theory which takes the production structure as given and describes growth as if it was ‘an expanding balloon’ – using Harberger´s inspired metaphor (Harberger, 1998) in which the relative size of each part of the structure does not change as the size of the balloon increases – we notice that changes in the structure of the economy constitute a major part of the process we want to examine. It is such change that allows for more diversity in the production structure and for productivity growth, as well as for the gradual expansion of more knowledge-intensive production activities, including the production of capital goods and the provision of engineering services. The inception of a new activity in the economy normally constitutes the response to a quasi-rent, that is, an above-‘normal’ rate of return on capital. In other words, innovation is not the response to an equilibrium scenario and to factor payments being equal to their marginal contribution to GDP. Quite on the contrary, for a new production activity to emerge, monopolistic profit opportunities are required. After the new product or production process is brought to the market and monopoly rents are captured by the innovator, imitators follow through, induced by the perception of above-normal profits being present in the industry. A conventional ‘industry life cycle’ process then follows. New firms enter the market, innovative rents are gradually eroded and eventually



129

the industry converges to a ‘mature’ stage in which a long-term equilibrium point has been achieved. This dynamic process does not follow a single and universal pattern. The co-evolution of economic, institutional and technological forces provides space for alternative scenarios to develop. In some cases the industry life cycle dynamic is set in motion by a multinational corporation that brings in new technology, opening up new markets, training domestic labour and subcontractors and upgrading domestic engineering practices. In other cases the cycle is put in motion by a family-owned SME or a large domestic conglomerate. The industrial organization dynamic developing in each case – the nature of the barriers to new entry, the learning curve firms follow after entering the market, etc. – are bound to be different. The state, far from being a neutral agent, also plays an important role in the configuration of the sector, acting through its regulatory agencies, financial institutions, universities and municipal authorities, many of which provide public goods, finance and coordination assistance which allow market functioning. The recent inception of new industries in Latin America – such as salmon farming in Chile or genetically-modified (GM) soya beans and vegetable oil in Argentina – can be seen as interesting examples in which the above-mentioned co-evolution of economic, institutional and technological forces has played a major role in industry organization. Whereas the case of salmon farming in Chile basically tells the story of an industry originally populated by SMEs entering the market successfully coached by public sector agencies, the expansion of soya beans and vegetable oil production in Argentina appears, primarily, to be the outcome of a large multinational corporation – Monsanto – bringing a new and much more productive technological package – seeds and agrochemicals – to the economy, setting in motion a new and more profitable way of producing an ‘old’ product such as soya beans and vegetable oil out of it. As a result, the institutional setting in which each of these new activities developed varied a great deal as we shall show in the section 16.3. In the initial years of industry inception, growth was determined by the rate at which new production capacity could be built up. New firms entered the industry, constructing new plants, hiring labour, developing subcontractors and so forth. They made decisions on the basis of expected profits, which basically depended on a large potential market being there to be catered for, and on the sector-specific institutions which succeeded in capturing the benefits of innovation. Natural


130 Jorge Katz

comparative advantages and a new set of institutions securing appropriability of innovation benefits triggered the process of industry expansion. The expectation of above-average returns on investment induced the building up of new production capacity, and the rate at which such capacity could come on stream was mostly conditioned by the availability of financial resources, production know-how, trained labour and country- and-sector-specific institutional and regulatory conditions. Although an adequate macroeconomic background was an important component of the global picture, a much broader set of sector and microeconomic circumstances also played an important role as well conditioning the development path of the new industry. The entry of new firms gradually made these industries more competitive. The available information indicates that from inception to maturity the life cycle of the above-mentioned sectors was close to two decades. They are now mature oligopolies strongly inserted in global food chains that are largely controlled by large multinational firms. As far as the linkage is concerned between the inception of new industries in the economy and the rate of expansion of GDP we should notice that the aggregate rate of growth of the economy is strongly conditioned by the rate at which new activities enter the production structure. Each individual new activity could be imagined as following an industry life cycle of the sort described in previous paragraphs, eventually reaching a ‘plateau’ when the dynamic forces inducing further growth wear out and the industry exhibits signs of saturation. New sectors have then to replace its role as an engine of growth in the economy. When the rate of inception of new production activities is high we could also expect the dynamics of the aggregate growth process to be high. On the contrary, if the process of structural change and diversification slows down, we could expect the economy to become less dynamic. As Saviotti and Pyka (2004: 4) say: ‘A faster rate of growth of variety would lead to faster economic development.’ Each individual new activity could be expected gradually to become less dynamic as the industry life cycle wears out, but the growth of GDP will strongly depend upon the rate at which new production activities are started in the economy.

6.3 Economic, institutional and technological forces underlying the process of new industry inception We have argued that a major process of structural change occurred in Latin America after the various economies in the region adopted trade liberalization and market deregulation policies from the mid-1970s. Table 6.1


Table 6.1 Structural changes in Latin America, 1970–2002 Argentina

I II III+IV V Total ICE*

Brazil

Chile

Colombia

Mexico

1970 1996 2000 2002 1970 1996

2000 2002 1970 1996 2000 2002

1970 1996 2000 2002 1970

1996 2000 2002

13.2 10.9 47.8 28.1 100

26.0 8.3 41.6 24.0 100 32.3

12.3 3.0 46.2 38.5 100

14.4 14.6 43.4 27.6 100 17.3

9.9 7.2 62.1 20.7 100 14.3

8.6 7.4 65.3 18.7 100 18.0

6.7 6.1 71.7 15.6 100 25.3

16.2 6.8 37.8 39.2 100

25.6 7.3 43.4 23.7 100 18.9

26.5 8.9 41.5 23.1 100 27.6

11.4 5.5 58.3 24.9 100

10.4 1.9 59.7 28.0 100 40.1

10.5 2.3 60.7 26.5 100 27.3

10.0 1.9 61.9 26.2 100 33.5

10.1 6.5 55.4 28.1 100 19.4

8.7 4.9 57.0 29.4 100 29.9

9.0 6.5 57.1 27.3 100 30.9

12.0 8.4 43.2 36.4 100

Notes: * Index of structural change with reference to 1970. I Engineering-intensive industries. CIIU 381,382,383,385; II Vehicles (CIIU 384) III+IVNatural-resource-intensive industries. (CIIU 311,313,314, 341, 351, 354, 355, 356, 371, 372) V. Labour-intensive industries. (CIIU 321, 322, 323, 324,331, 332, 342, 352, 361, 362, 369, 390) Source: PADI, ECLAC (UN).


16.4 18.8 39.1 25.8 100 22.1

15.6 18.6 40.8 25.0 100 22.5

132 Jorge Katz

provides an indication of the extent to which such structural transformation took place. The numbers refer to the relative share of various industry aggregates in manufacturing production in 1970, 1996, 2000 and 2002. In the cases of Chile and Argentina the restructuring of the economy involved a significant contraction of metalworking industries producing machinery, equipment and capital goods, and the concomitant expansion of natural-resource-processing activities, including agricultural commodities, minerals, fruit and wine, aquaculture and forestry products. In other words, both countries restructured in the direction of their (static) natural comparative advantages, phasing out engineeringintensive industries which developed under tariff protection during the ‘inward-oriented’ incentive regime of the immediate post-war years. Unlike the above, Mexico and the Caribbean countries proceeded in a quite different direction. Their economies restructured in the direction of ‘in-bond’ assembly industries – ‘maquiladoras’ – catering for US markets in areas such as vehicles, clothing and electronic products, including computers, colour televisions and video recorders. Such industries expanded on the basis of externally designed state-of-the-art manufacturing facilities producing the above-mentioned products, for which 95 per cent of the intermediate parts and components were imported. In order to proceed further into the exploration of the implications of the structural transformation exhibited by Argentina and Chile mentioned above, and of the extent to which such transformation involved the co-evolution of economic, technological and institutional forces retrofitting on each during the process of growth, we now turn to two individual case studies illuminating these aspects. 6.3.1 GM soya beans and vegetable oil production in Argentina The production of GM soya beans and vegetable oil is emblematic of the process of structural transformation the Argentine economy went through over the past two decades. The production of GM primary products began in the world in the mid-1990s. By 2002 there were nearly 60 million acres of genetically modified primary products under cultivation, with soya beans and maize being by far the two more important. Argentina has now well above 15 million acres of GM soya beans under cultivation. The transition from ‘conventional’ to GM soya beans involved major changes in production organization, technology and institutions in both the agricultural and the manufacturing spheres of the activity. It also



133

triggered new forms of interaction between firms, universities and public sector laboratories, as we shall briefly discuss in the next few pages. Consider first the impact of the introduction of GM soya beans on the agricultural sector. ‘Cero tillage’ (siembra directa), and ‘contract agriculture’ (agricultura de contratos) now dominate the scene. These two major technological changes induced equally important changes in the organization of production at the farm level as well as in sector-wide institutions. The ‘traditional’ farmer is no longer the central agent responsible for production decisions. His role has been transferred to large subcontractors who take responsibility for planning and organizing production, bringing with them seeds, agrochemicals and pesticides as well as funding for each agricultural season. Production now takes the form of ‘risk-contracts’, with financial intermediaries playing a central role in the organization of production. Monsanto and a few other large multinational corporations (MNCs), along with a small number of local firms, control the technological package upon which the industry operates. This is clearly different from the institutions that prevailed in the sector during the so called ‘green revolution’ of the 1960s. In those years agricultural technology was basically a public good’ created and disseminated by The National Institute for Agricultural Technology (INTA). Technological change was more related to mechanical improvements in machinery and equipment as well as to fertilizers. In the transition to GM soya beans technological changes have become more associated with genetic manipulation and biotechnology, both areas largely controlled by large MNCs and subject to patent protection. Many new institutions have developed in Argentina in association with the diffusion of GM soya beans. It is believed, for example, that as much as one-third of the seeds used in any given agricultural campaign are seeds coming from the previous agricultural season, which are used with complete disregard for intellectual property rights. Monsanto did not adequately protect its GM technology for soya-bean production in Argentina – in what should be considered as a major mistake in its business strategy – and is presently trying to regain market control by threatening legal action against Argentine exports of soya beans to Europe. The company has not yet succeeded in having its case accepted by European legal courts. Equally interesting from the institutional point of view is the fact that the Brazilian government tried to delay the diffusion of GM soya-bean production for a number of years, but finally failed in its attempt and as a result GM soya bean seeds are being


134 Jorge Katz

smuggled in large quantities from Argentina into Brazil, inducing a de facto expansion of the area under cultivation. Furthermore, as the expansion of GM soya-bean production involved rapid diffusion of the so called ‘cero tillage’ model of organizing production, a major change in land utilization has occurred pari passu the use of the new technology. An additional crop can now be obtained from the same piece of land in one calendar year, and this is having a significant impact on agricultural rents. On the other hand, high rents in GM soya-bean production have induced landowners to substitute for other production activities, such as dairy production and forestry products, with such negative consequences as desertification and reduced biodiversity. We hereby notice that the expansion of soya-bean production has had a number of ramifications, many of which have not as yet been carefully examined. Turning now to the manufacturing side of the case we notice that significant economic and institutional changes occurred in Argentina during the 1990s as a result of new highly capital-intensive catalytic production facilities replacing ‘old’ vegetable oil production facilities of the 1970s. Vegetable oil production is now a highly concentrated industrial activity dominated by a small number of large local companies and foreign conglomerates. As Table 6.2 indicates, the transition to state-ofthe-art technologies has involved a ten-fold expansion in labour productivity in the local vegetable-oil industry. It is also clear that the new technology made significant savings in terms of labour, compared to the old technology. Recent studies indicate that the expansion of the industry has also exerted a significant impact upon domestic R&D efforts in areas related to seed production, agrochemicals and pesticides. Looking at the expansion of the local biotechnology industry Bisang et al. have recently reported that some 80 firms have entered the field, exploring a wide range of new areas related to genomics and modern biotechnological

Table 6.2

Soya-bean production in Argentina, 1973–4 and 1993–4

Number of plants 1973–4 1993–4

67 59

Employees 6895 4934

Production volume (millions of tons) 1740 12,196

Tons per plant

Tons per employee

26 207

252 2472

Note: Less firms, more capital intensive, more efficient.



135

processes. According to them these firms have an annual turnover of around U$S 350 million, employ nearly 5000 people, spend close to 5 per cent of their sales on R&D activities – nearly U$S 18 million – and employ around 600 people in technology-generating activities. Approximately 80 per cent of these firms are small family enterprises, maintaining an active interaction with public sector R&D agencies 6.3.2

Salmon farming in Chile1

The process through which Chilean salmon farming firms attained international competitiveness took the best part of two decades, a period in which many new companies – both domestic and foreign – entered the industry, sector-specific institutions and skills developed and professional management took over an originally quasi-artisanal industry thus significantly altering the organization of production and international marketing practices. As a result of the cumulative impact of these changes Chile gradually acquired ‘world-class’ status as one of the three major salmon-farming countries in the world, alongside Norway and Scotland. One-third of the world demand for fresh salmon is now met by Chilean-based companies. Salmon farming in Chile can be described as having evolved through three quite different ‘stages’. In each of these stages the actors and the problems that had to be solved changed quite significantly. The first was an inception stage in which salmon farming was successfully introduced and adapted to the Chilean environment, almost entirely starting from imported genetic material. In this stage, research, trial and error and learning appear as major factors explaining both individual firm behaviour and the starting up of a new production activity. Teething problems were proverbial during that period, at both individual firm and industry levels. The Chilean government played an important role during these years, inducing the building up of new production capacity and the transfer of the required technology through Corfo and Fundacion Chile. In the second stage, the industry rapidly increased in size and complexity with the entry to the market of many intermediate input suppliers and service firms and the building up of a strong, sector-specific industrial cluster. The role of the public sector changed significantly during this period, taking a step back from its role as a proactive dynamic agent and instead taking an active role on the regulatory front. In the third stage, a major transformation in industrial structure resulted from through a massive arrival of foreign direct investment (FDI), Mergers and Acquisitions (M&A), changes in plant ownership and a rapid process of internationalization.


136 Jorge Katz

In a short period of less than 20 years Chilean salmon exports increased from less than US$50 million in 1989 to around US$2.4 billion in 2006. Salmon exports now account for close to 5 per cent of Chile’s total exports. From an almost negligible participation in world markets – 2 per cent in 1987 – Chile now caters for nearly onethird of world demand. As Table 6.3 indicates, the process involved the co- evolution of a large number of economic, technological and institutional forces that retrofitted onto each other during the course of time. Public organizations, foreign companies and a large number of SMEs participated in the early years of the industry’s inception. Although the public sector played a major role in the local inception of this new production activity it is clear that a new generation of Chilean entrepreneurs emerged in association with salmon farming and became the driving force motorizing the process of expansion. Regulatory and sanitary activities – such as fishing and cultivation permits, environmental-impact monitoring, control of salmon-eggs imports, etc. – were adequately performed by government agencies such as Sernapesca, Conama and others. The required regulatory infrastructure supporting the above activities was set in place in the late 1970s and 1980s (Aquanoticias, November 1997). Production practices in the early years of the industry were quasi-artisanal. Salmon food, for example, was prepared daily by each company from of fresh raw materials. The conversion rate from salmon food to the finished product was in the region of 3:1, that is, 3 kg of fresh food per 1 kg of salmon. This is three times larger than the input/output coefficient the industry exhibits today, suggesting that major productivity improvements have been attained during the process of expansion and that major learning has been attained at the individual firm level. (Aquanoticias, July 1997: 24). Many examples of this sort can be cited in relation to the management of cultivation tanks, vaccines and vaccination techniques, final product processing, etc. (Aquanoticias, May 2004: 12). It was in the late 1990s that the Chilean salmon farming sector attained many of its present features of a ‘mature’ oligopoly.2 World prices for salmon fell significantly in the second half of the decade, getting closer to the industry’s long-term unit production costs. Unit gross margins contracted, as competition increased and the markets for salmon became more ‘contested’. The technological and competitive regime of the industry became more demanding as a result of M&As which, on the one hand, made the average size of firm considerably


Table 6.3

The evolution of salmon farming in Chile, 1960–2000

Exports (tons) Main products and markets

1960–73

1974–85

1986–9

1990–5

1996–2002

Negligible

1000 Fresh and frozen coho salmon; trout Brokers buy from producers

11,000 Coho salmon for Japan

100,000 Coho salmon for Japan; Atlantic salmon for US. Collective export activities

500,000 Diversification of markets: US, Asia, Latin America Large foreign retailers buy directly

Development of forward (egg and smolt) and backward linkages (food, vaccines)

Environmental control systems; salmon food; production of eggs, vaccines; traceability

Targets for market research, technology for supporting industries; regulation

Targets for environmental management; sources of productivity growth

Key event in marketing

Issues to be resolved

Transition from catch and release to aquaculture.

Government policies

Technology transfer under government co-operation; support from CORFO, Ministry of Agriculture

Established know-how for freshwater aquaculture and need to develop saltwater equivalent Regulation and technology from CORFO, Fundación Chile, Sernapesca, JICA, others

Brokers buy from producers and wholesalers Rapid expansion in scale of production

Provision of basic road and port infrastructure

Continued


Table 6.3

Continued 1960–73

Typical type of External firm in industry co-operation; no industry yet Intermediate Very few suppliers

1974–85

1986–9

1990–5

1996–2002

Family-owned; small firms; few foreign companies High degree of vertical integration; few domestic input suppliers

Local SMEs growing very fast

Growing presence of foreign firms

M&A by foreign firms

Hatchery, cultivation, and final processing begin to integrate Supporting industries develop

Outsourcing expands and many new suppliers enter the market

Cluster gets stronger and service industries develop

Clustering forces become stronger

Rapid expansion of number of cultivation sites and scale of plant Private sector cooperative activities expand

Mostly local quality standards

International norms and standards diffuse; good manufacturing practices and traceability Productivity, local and international standards; ISO 9000 and 14,000; traceability

Expected externalities

Sources of competitiveness

Natural comparative advantage

Production

Relations among actors in industry

International co-operation; proactive state participation

Public-private co-operation; CORFO, Fundación Chile

Initial forms of globalization emerge

Full-scale globalization after M&As

Source: Based on Iizuka (2004). 10.1057/9780230276123 - The Rise of Technological Power in the South, Edited by Xiaolan Fu and Luc Soete


139

bigger, much more capital-intensive and technologically more sophisticated, while on the other hand, business concentration increased quite significantly. The number of salmon-farming companies increased until 1996, reached a plateau in the late 1990s, and then began to fall. Although there were fewer firms in the industry at the end of the period, the average salmon-farming company was larger and more capital-intensive and technologically more sophisticated. During the 1990s a major wave of FDI occurred as large, international salmon-farming enterprises arrived in Chile and took over a number of successful local SMEs. Concentration ratios increased and the industry turned into a mature oligopoly, firmly inserted into international food chains. The two case studies so far presented illuminate the fact that the inception of new production activities in the economy is normally associated with the co-evolution of economic, technological and institutional forces that come together during the process of growth and retrofit on each other through time. Far from being a conventional neoclassical equilibrium growth model, what the case studies show us is a complex evolutionary story of intertwined economic and institutional changes that no formal algorithm can be expected to illuminate well. The ultimate lesson seems to be that ‘story-telling’ and ‘appreciative’ theorizing are needed if we are to understand correctly what the relationship between economic growth and structural change is all about. An important underlying theme of both case studies refers to the fact that domestic R&D efforts and technology generation activities remain a somewhat marginal force in the explanation of economic growth in Latin America. Both case studies show that the new export-oriented activities developed strong international competitiveness, but neither of them managed to create a strong sectoral innovation system. Much seems to be lacking on this front if these new activities are to become more knowledge-based in the years ahead, expanding their domestic value-added content and exploring in a more fundamental way the technological frontier of their respective fields of operation.

6.4 The poor technological performance of Latin American firms In spite of the fact that new production activities have emerged in Latin America over the past two decades and that the modern sector of the economy now features new state-of-the-art production facilities, it is still the case that much of the above is based on imported capital goods


140

Jorge Katz

and foreign product licences, and that not a great deal has so far happened in terms of developing a dynamic and proactive national innovation system with local firms assuming a forward-looking stand as far as R&D spending and technology-generating efforts are concerned. Latin American firms continue to be characterized by their low level of commitment to R&D activities and by their low propensity to undertake long-term knowledge-generation efforts in exploring the technological frontier of their field of operation. Most Latin American firms limit their technological activities to those needed to introduce ‘minor’ changes in product designs, process technologies and production organization routines. These efforts normally involve spending in the order of 1 per cent of sales on engineering activities which allow them gradually to obtain productivity improvements and a better utilization of their installed production capacity. But they rarely move further ahead to explore more complex and long-term technological research options. Table 6.4 and Figure 6.1 help us further to elaborate on this point. Table 6.4 R&D expenditure as a percentage of GDP

Japan US Germany France Canada Australia Italy Spain Brazil Portugal Chile Argentina Mexico Panama Bolivia Uruguay Colombia Peru Paraguay Nicaragua

Year

Percentage

2003 2004 2004 2004 2004 2002 2002 2003 2003 2003 2003 2004 2002 2003 2002 2002 2001 2003 2002 2002

3.15 2.68 2.49 2.16 1.93 1.69 1.16 1.05 0.95 0.78 0.60 0.46 o.40 0.34 0.26 0.22 0.17 0.11 0.10 0.07

Source: PADI Programa de Analysis del Desarrollo Industrial. DIvision de Desarrollo Productivo y Empresarial, Economic Commission for Latin America and the Caribbean (ECLAC).



141

Domestic learning and the evolution of local technological capabilities. State of the art State of the art research. Technological gap

Latin Amercian technological capabilities

New production processes and product designs. Major improvements in production processes. ´Adaptive´ product design efforts. Adaptive´ process engineering efforts. Plant operation capabilities.

Figure 6.1 Domestic learning and the evolution of local technological capabilities

Latin American firms can be thought of as undertaking knowledgegenerating efforts on the initial and less complex steps of the technology-generation ladder. These steps are related to mastering and upgrading the technology with which they operate, as well as to gradually improving their capital equipment and production facilities. The incremental know-how they manage to produce in all of the above directions is highly ‘firm-specific’ and appropriable at the individual company level. The situation is very different for a firm that decides to proceed into more risky and time-consuming research strategies, exploring the technological frontier of its field of concern. R&D expenditure increases quite sharply as more expensive pilot plants and experimental equipment are required. Ventures of this sort normally demand networking with universities, public sector laboratories, engineering firms, etc. Public regulatory agencies, government environmental protection bodies and municipal authorities are also frequently involved in cases of this sort, making the whole operation more risky and uncertain. Intellectual property rights (IPRs) become more difficult to enforce. Firms carrying out activities of this sort usually spend 10 per cent or


142

Jorge Katz

more of their sales in R&D activities. Long lead-in periods are involved before significant new results are obtained and the rate of failure is quite high. Only a small number of Latin American firms actually commit themselves to R&D activities and technological targets of this second type. Unlike firms in South Korea, Finland, Israel or Singapore, Latin American companies prefer to innovate through the use of imported machinery and taking out product design licences, instead of performing their own R&D efforts in search of globally novel products and processes. This was the case during the ‘inward-oriented’ industrialization period of the post-war decades, and continues to be the case now, even though the economy has gone though a major restructuring and modernization process, becoming closer to the international state of the art in various areas of economic activity. Recent studies carried out in Brazil, Chile and Argentina show this to be the case. Close to 80 per cent of what Latin American countries spend on R&D activities – in itself only a small fraction of what more developed countries allocate for such purpose – is financed by the public sector and largely carried out in public R&D laboratories.

6.5

Concluding remarks

There are different forms of capitalism around the world and each country has to come up with the local brand that best suits its need, history and social and political institutions. Market-oriented structural reforms and the globalization of economic activities have brought major changes to Latin America, inducing the gradual phasing out of many industries and institutions of the ‘inward-oriented’ period of industrialization. Their production structures now feature many new sectors of economic activity more closely tied to their natural comparative advantages. A modern sector of economic activity has emerged in natural-resource-processing activities, as well as in ‘maquila’-type industries and in service sectors catering for local demand in areas such as telecoms, banking and financial services, water provision and sanitation services, and more. But this structural transformation has not been strong enough as to incorporate the vast majority of the population. Thus, the modernization process has occurred hand in hand with a significant expansion in the gap between the rich and poor segments of society. This has clearly resulted in a dramatic sequel of frustration and despair that makes political governability an increasingly difficult proposition.



143

Lack of ‘initial entitlements’ resulting from low-quality education, poor health services, etc., together with insufficient provision of public goods ‘levelling the playing field’ and different forms of market failure have been instrumental in market-oriented structural reforms inability to deliver a broader pattern of improvement in economic efficiency. Conventional neoclassical growth theory has little to offer us in terms of policy advice under such circumstances. Neoclassical growth models are specified in terms of an institutional free equilibrium algorithm which is not very useful when we come to examine policy questions related as to how to make an economy more productive and internationally competitive, less heterogeneous and more equitable during the course of time. The standard refrain of the need for ‘good institutions’ is something of a mantra and sounds particularly irrelevant under the present circumstances. Good institutions in the neoclassical sense seem to be those that facilitate good market functioning. The lesson here seems to be that we have to go well beyond market institutions if we are to come up with a useful set of public policies addressing issues of the upgrading of domestic capabilities and the need for more social equity. Current research shows that no universal recipe is available for the above and that countries need to exercise a great deal of pragmatism as they move into this complex territory. They have to learn how to regulate markets, build adequate local technological capabilities and integrate well into World Trade Organization’s Trade-related Intellectual Property Rights environment. They have to join world markets ‘on their own terms’ as Southeast Asian countries – and now China – have learnt to do it. This seems to be the ultimate lesson that rapidly emerging economies all around the world are teaching us today.

Notes J. Katz. University of Chile. The present paper elaborates on a first version presented by the author at the Sanjaya Lall Conference on ‘The challenges of Technology for Development’ celebrated in QEH, Oxford, in May 2008. Thanks are due to the participants for the comments and to ONUDI for making possible my participation in the referred conference. I am solely responsible for the outcome. 1. This section is based on a previous paper by the present author written for the World Bank in 2004 (Katz, 2006). 2. What a ‘mature’ industry actually is, and how the notion applies to the case of salmon farming, was made quite clear in a recent public conference by Mr Torben Petersen, CEO of Fjord Seafood Chile, a subsidiary of the Norwegian company of the same name when he stated that ‘The real maturation process begins when we see that company actions are aimed at the markets and not


144

Jorge Katz

at production, in other words, when salmon farming growth is determined by its market and not by its production. See Acuanoticias 79, 18 May 2004.

References Goldstein, A. (2005), ‘The political economy of industrial policy in China. The case of aircraft manufacturing’. Mimeo. Paris: OECD. Harberger, A. (1998) ‘A vision of the growth process’, American Economic Review, 88, March. Katz, J. (2006), ‘Salmon farming in Chile’ in V. Chandra (ed.), Technology, Adaptation and Exports. The World Bank, Washington, 193–223. Iizuka, M. (2004). ‘Organizational capability and export performance: the salmon industry’ in Chile. Paper presented at the DRUID Winter Conference, 22–24 January.


Part III Innovation Systems and Technological Capabilities




7 The Finance of Innovative Investment in Emerging Economies Jörg Mayer

7.1

Introduction

A sustained rise in living standards can be achieved only through continuous productivity growth. This presupposes high rates of investment in physical infrastructure, plant and equipment, as well as in more intangible elements, such as education and R&D. For private investment to take place, firms not only need an incentive in terms of expected future profits, they also need to have access to reliable, adequate and cost-effective sources for financing their investment. This chapter analyses the role of different sources of finance for innovative investment and looks at the experience of the BRICS (Brazil, Russia, India, China and South Africa) in this regard. Section 7.2 discusses the conceptual framework and Section 7.3 draws on enterprise data to provide statistical evidence on the role of different sources of investment finance. Section 7.4 examines the role of different sources of investment finance for innovation, while section 7.5 concludes by discussing measures that governments can take to support innovative investment by private enterprises.

7.2

Investment finance and information asymmetries

In making their decisions on how to finance investment, entrepreneurs have a well grounded microeconomic rationale not to consider different sources of investment financing as perfect substitutes. The so-called ‘pecking-order theory’ of capital structure postulates the relevance of specific forms of investment finance for investment decisions. It suggests that the choice of capital structure depends on financial factors (e.g. the availability of internal finance, access to new debt or equity 147


148

Jörg Mayer

finance and the functioning of particular credit markets) and a firm’s characteristics (e.g. its investment opportunities, profitability and size). According to this view, firms generally follow a hierarchy in financing real investment, with a preference for internal over external finance and for debt over equity. Highly profitable firms might be able to finance their growth by using retained earnings and by maintaining a constant debt ratio. In contrast, firms that are less, or not yet, profitable are forced to resort to external financing. Accordingly, changes in a firm’s debt ratio are driven by its need for external funds, which in turn is determined by the extent to which investment opportunities outrun internally generated funds (Myers and Majluf, 1984).1 According to the pecking-order theory, a firm prefers internal sources (i.e. internal cash flow stemming from depreciation and retained earnings) because they allow it to safeguard the manager’s insider information on the value of the firm’s existing assets and the quality of its investment opportunities. Asymmetric information makes it very costly, or even impossible, for providers of external finance to assess fully the quality of a firm’s assets and its investment opportunities. Moreover, utilizing internal finance avoids agency costs (i.e. costs associated with mitigating a potential conflict of interest between the firm’s management and providers of external finance). Information asymmetry is also the reason why debt financing is preferable to issuing equity. The degree of information asymmetry, and hence the agency cost, is relatively lower for debt than for equity finance. This is because debt financing, such as through bank loans, allows screening and monitoring of investment projects and their execution directly at the level of the firm. Banks can demand collateral, and, in the event of financial distress, debt generally has the prior claim on assets and earnings, while equity is the residual claim. Moreover, capital markets will assume that an enterprise issues equity only when it considers its existing assets to be overvalued and see recourse to equity financing as an indication that the enterprise is unable to obtain other financing because its investment opportunities are extremely risky or as an indication that the enterprise’s debt ratio already is at a level that raises serious concern about upcoming financial distress (i.e. difficulties in meeting principle and interest obligations). As a result, for a firm that is seeking financing for investment, the conditions attached to issuing equity will tend to be worse than those associated with debt financing. A further reason for preferring debt to equity is that equity financing exposes a firm to the risk of takeover, especially when financial markets undervalue the firm’s assets. The pricing process on stock markets


Innovative Investment in Emerging Economies

149

may work well in terms of information arbitrage efficiency, or financial arbitrage, which ensures that all stock market participants have immediate access to all new information concerning a firm’s shares so that no participant can make a profit on such public information. By contrast, this pricing process may not work so well in terms of fundamental valuation efficiency, which would ensure that share prices accurately reflect a firm’s fundamentals (i.e. its long-term expected profitability). Firms in developing and transition economies often face different problems in sourcing finance for their investment projects than those in developed countries. Financing needs may frequently exceed the availability of internal finance, particularly when technological upgrading and new-product development require a fast turnover of capital equipment. According to Singh (1997), this was the case for many firms in East Asia, which had to use substantial amounts of both internal and external resources to finance their investments and expand their world market shares. Industrialization and economic catch-up generally require the application of novel techniques (i.e. novel for the respective economy) for producing new products or using new processes. Traditionally, large firms and business conglomerates were considered to have an advantage in driving industrialization in sectors that require large-scale, heavy capital investment, prior manufacturing experience and the coordination of investment activities across a number of industries (Amsden, 2001). However, over the past few years increasing importance has been given to the use of information and communication technologies (ICT) as an important condition for achieving productivity growth. This has resulted in a growing emphasis on the role of new and often small firms in the application of novel techniques. New and innovative firms are not likely to have the ability to resort to internal finance or be able to generate sufficient cash flow, and their projects may be deemed excessively risky by outsiders. In these cases, information asymmetries are particularly pronounced because there is no track record on either the entrepreneurial skills of the manager or the profitability of the innovative firm; moreover, information of the firm’s previous engagement in non-innovative activities may not be of much help. Thus, outsiders cannot easily distinguish between highand low-value opportunities. New and innovative firms are likely to encounter enormous difficulties in procuring bank credit because the only collateral they may be able to provide will be in the form of intangible assets, which are partly embedded in human capital and generally very specific to the particular firms in which they reside (Hall, 2002).


150 Jörg Mayer

While the innovator could convey all the information about the innovative investment project to potential outside sources, this would involve disclosure of insider information, which would expose the firm to imitation and severely diminish its ability to appropriate the returns on its investment. On the other hand, banks will be reluctant to finance an initial investment that could make productive investment and productivity gains if they are unable to acquire a share of the productivity gains commensurate with their earlier risk taking. This may create a situation where every bank waits for others to move first so that they can reap the benefits of other banks’ revelation of information about the capability of the entrepreneur to undertake profitable investment. In such a situation, informal financing from the entrepreneur’s family or friends can be an important source of risk capital in the early stages of innovative projects when the need for financial resources is limited.2 But when this need strongly increases, informal financing will no longer suffice and the firm may try to access venture capital.3 Venture capital is equity or equity-linked investment finance in young, privately held companies, where the investor is a financial intermediary that collects financing from a group of investors (banks, pension funds, insurance companies, foundations, etc).4 Venture capitalists may be considered specialists in the accumulation of information on balance-sheet positions and on investment projects of firms with a high growth potential. Since venture capitalists often also possess technical knowledge, they suffer less from information asymmetry than providers of traditional bank loans or equity capital. Venture capitalists often lend their expertise to the firms in exchange for part of the value that the firms generate. Their technical knowledge and experience also enable them to perform non-financial advisory or managerial functions, which permit a better assessment of the industrial and commercial viability of an investment project. These non-financial functions may actually prove to be more important than the mere financial contribution, because they help manage the downside risks and maximize the return from a given investment (Lerner, 1995). Since the venture capitalist usually disinvests after some time, venture capital may be best considered as a hybrid form of debt and equity finance (Hall, 2002). This means that an innovative enterprise is likely to follow a slightly different hierarchy in the pecking order of capital structure and, as far as external finance is concerned, resort to bank financing only after obtaining resources from venture capitalists.5 However, the venture capital solution to financing investment has its limitations, particularly in developing countries, because there must be



151

an active stock market to provide an exit strategy for venture capitalists (typically through an initial public offering in which the enterprise issues shares to the public) and allow them to move on to financing other enterprises (ibid.). Moreover, in order to limit the number of partners in a firm, venture capitalists need to invest a certain minimum amount. This amount may exceed the means at the disposal of most potential venture capitalists in developing countries. Developing countries have traditionally used public banks, including national development banks, to cover gaps in access to investment finance. Amsden (2001), for example, provides a detailed account of the role played by national development banks in many late industrializing economies. As a result of a large share of non-performing loans in their liabilities, several public and national development banks were dismantled in many countries as part of the financial reforms in the 1990s. However, more recently, there has been renewed interest in their usefulness as an instrument in development strategies.

7.3

Different sources of investment finance

Empirical evidence based on cross-country averages, which cover more than 32,000 firms from 100 developed, developed and transition economies and the period 2002–6, shows that internal funds and retained earnings are the main source of investment finance (UNCTAD, 2008: 120). Firms worldwide finance about two-thirds of their investments from internal funds and retained earnings and another 16–23 per cent, depending on firm size, from bank loans. Equity financing is of relatively little importance, accounting for only about 3 per cent of investment financing – a share that is even smaller than investment finance from family and friends, trade credit or not specifically identified sources combined in the category ‘other’. Country-specific evidence for the BRICS also underlines the varying importance of different sources for the financing of fixed investment (Table 7.1). Perhaps most importantly, the capital structure of Chinese firms in 2003 differed significantly from that of firms in other countries, in that they appear to have sourced a very low share of investment finance from internal funds and retained earnings, while the category ‘other’ played a significant role. This category includes funds raised by enterprises and, for state-owned enterprises (SOEs), financing by local governments, as well as external sources of funds raised through various channels, including capital markets.6 Given that the category ‘other’ cannot be disaggregated further, it may also largely


Table 7.1

Sources of innovative investment finance, selected countries, 1999–2006

Number of firms

Internal funds and retained earnings %

Local and foreignowned commercial Investment Trade banks % funds % credit %

Equity %

Family and friends % Other %

Developed a major new product line Brazil 2003 All firms 1351 56.3 Small firms 226 58.0 Medium firms 736 58.6 Large firms 384 51.2

14.3 10.8 14.8 15.0

8.5 5.7 6.4 14.1

8.7 13.0 8.2 7.4

4.3 3.5 3.8 5.7

1.2 2.2 1.4 0.3

6.7 6.7 6.9 6.2

China (2003) All firms Small firms Medium firms Large firms

1342 169 478 686

15.2 13.7 14.6 16.2

20.4 8.6 15.2 26.8

0.5 0.9 0.6 0.4

1.0 0.0 1.1 1.2

12.4 16.7 12.4 11.4

5.9 11.0 8.6 2.7

44.5 49.0 47.5 41.1

China (1999) All firms Small firms Medium firms Large firms

94 42 27 24

59.6 64.9 61.6 48.4

9.7 6.8 8.0 16.3

6.4 5.0 10.1 4.6

2.9 1.0 3.9 5.0

2.8 0.3 3.9 5.6

6.2 9.0 3.9 4.1

12.5 13.0 8.6 15.9

263 831

11.5 15.9

25.3 18.4

1.0 0.3

0.0 1.1

4.7 14.1

1.2 8.7

56.3 41.6

1476

52.0

32.2

0.0

4.5

1.1

6.9

3.3

China (2003) State-owned firms Private domestic firms India (2005) All firms


Small firms Medium firms Large firms

612 497 284

51.2 54.5 51.4

25.9 33.2 41.6

1.0 0.0 0.0

6.4 4.1 2.0

1.1 0.8 1.8

10.9 4.6 2.1

4.6 2.7 1.2

Russia (2005) All firms Small firms Medium firms Large firms

431 183 132 116

85.0 90.9 82.2 78.8

6.5 3.5 7.3 10.3

1.2 0.0 1.5 2.8

2.4 1.5 3.6 2.6

0.2 0.0 0.0 0.7

1.1 1.5 1.6 0.1

3.6 2.6 3.9 4.7

South Africa (2003) All firms 539 Small firms 49 Medium firms 225 Large firms 260

58.4 57.1 58.1 59.3

16.6 9.0 17.1 17.4

0.7 2.0 0.4 0.7

0.7 2.7 0.5 0.4

0.1 0.0 0.2 0.0

0.9 1.0 0.9 0.8

22.5 28.2 22.7 21.4

Notes: New firms = firms aged two years or less. Small firms = fewer than 20 employees; medium firms = 20–99 employees; large firms = more than 99 employees. For China (1999): small firms = fewer than 50 employees; medium firms = 50–500 employees; large firms = more than 500 employees. The numbers for small, medium and large firms may not add up to the total number given for all firms because some firms gave no indication of their size. Emerging-market economies in Europe: Czech Repuiblic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia and Slovenia. Investment funds includes funds from investment funds, development banks and other state services. Source: Author’s calculations, based on World Bank, Enterprise Survey database.


154

Jörg Mayer

include misclassified retained earnings. Indeed, according to the results from a 1999 survey (reported in the third panel for China in the table), Chinese firms financed about 60 per cent of their fixed investments from internal funds and retained earnings at that time (i.e. roughly as much as in other countries). On the other hand, Allen et al. (2005) argue that informal financing channels – including informal associations, private financial houses and money-lenders that function like banks but charge very high interest rates – have played a significant role in the Chinese economy, particularly for those private entrepreneurs who have no access to the formal banking system. Chinese firms also make relatively extensive use of equity finance. It is likely that a large part of this feature reflects the government’s use of the equity market as a vehicle for privatization, while the number of domestic enterprises enlarging their capital base through new equity issues is still relatively small. Chen (2004: 1346), in contrast, argues that equity financing may be particularly important for Chinese firms because of country-specific factors, such as insufficient enforcement of enterprise law and individual shareholders who lack adequate investment protection, with the result that equity ‘has become somewhat a “free” source of finance’. A major source of investment finance in Russia and South Africa is self-financing and retained earnings, while in India it is bank loans. In Brazil, special development finance – which falls under the investment funds category – plays a relatively important role. Brazil’s national development bank, Banco do Desinvolvimento de Todos os Brasileiros (BNDES), is an example of a financially sound institution that survived the reduction of the state presence in banking activities in the 1990s.7 It focuses on investment projects in infrastructure and industry, which account for about half and one-third of its disbursements, respectively, and more than four-fifths of its operations are in support of small enterprises. Unlike private commercial banks, public and development banks have a development objective: their loan analysis takes account of the economic and social development impact of an investment project in addition to its financial return. Public and development banks provide finance for investment projects that would typically be judged too risky by a private bank, either because full recovery of the cost of investment is a long-term process, such as from infrastructure investment, or because investment is carried out by small and/or innovative enterprises that aim to produce new products or apply new production processes. The developmental role of public banks implies that their activities tend to be concentrated in areas characterized by information asymmetries and



155

intangible assets. Thus, public banks should not be expected to have the same degree of profitability as private commercial banks. Indeed, disproportionate pressure for profitability would cause managers of public banks to deviate from their developmental mandate (Levy et al., 2007).8 Some of the projects financed will necessarily be a commercial failure for the very reason that it is only by undertaking such projects that their profitability – or lack of it – will be discovered. Thus, in order to act as a source of public risk capital, an optimal strategy of a development bank would be to minimize the costs of mistakes when they occur, rather than minimizing the risks of making such mistakes. Another part of the development objective of state-owned and development banks regards the coordination of investment projects. Investment can fail to be profitable unless there is simultaneous investment in upstream or downstream activities, particularly if such activities are not tradable or require geographic proximity. Physical infrastructure is a prime example. But a similar argument applies to the availability of appropriate production inputs (i.e. appropriately skilled labour as well as physical inputs that match a country’s level of technology) or the presence of a buyer for a firm’s products. In this sense, a major problem for entrepreneurs, who act as independent agents and only in their self-interest, is how to coordinate investment in a way that enables them to benefit mutually from upstream and downstream linkages. Where such mutual benefits occur, the economy-wide impact of an investment project exceeds its private profitability. So, it is likely that a bank acting in the interests of national economic development as a whole (i.e. a public or development bank) will have an advantage in financing investments, the profitability of which depends strongly on complementary investment. National development banks often suffer from underfunding, in particular when they lack access to resources through client and government deposits. This is one of the reasons why their loan disbursements are often made in association with private banks. For example, over the past few years, BNDES has made about half of its loans in association with private commercial banks.9 This kind of syndicated loan allows the development bank to invest in more projects and diversify its project-specific risk. At the same time, involving another bank gives a second opinion on the viability of the investment opportunity, thereby reducing the risk of funding bad projects. The experience of China’s state-owned banks has been less successful, lending support to the argument that in the absence of a complementary institutional set-up, state-owned banks might not allocate credit


156

Jörg Mayer

optimally. Lending decisions based on political and other non-economic reasons caused non-performing loans of the four largest stateowned banks to become a serious problem for China’s banking system during the 1990s. According to official statistics, non-performing loans have fallen in recent years both in value and as a percentage of total loans despite the emergence of new non-performing loans (Allen et al., 2008). The continued accrual of non-performing loans may have biased the capital structure of Chinese firms depending on ownership types. Cull et al. (2007, for example, argue that lending from state-owned banks has become increasingly inefficient because the government led stateowned banks to bail out poorly performing state-owned enterprises, as these enterprises provide a lot of employment. In order to maintain sales, the state-owned enterprises themselves may have extended trade credit to private domestic enterprises that have little or no access to loans from state-owned banks. The results in the third panel for China in Table 7.1 indeed show that bank loans are a more important source of investment finance for state-owned than for private firms, while the reverse holds for trade credit. However, it would appear that the magnitude of this bias is too small to have significant economic effects.

7.4 Financing of innovative investment As already mentioned, innovative firms face particularly severe constraints for financing their investment because they are not likely to have the possibility of resorting to internal finance or to be able to generate sufficient cash flows, and their projects may be deemed excessively risky by providers of external finance. Moreover, if the innovation is rapidly imitated by other firms, innovative investors cannot fully achieve the envisaged returns from their investment. A low propensity to undertake innovation may be a third reason for low innovative activity. This will be the case when the expected return on innovation is low because of a relative shortage of complementary factors of production (skilled workers, complementary intermediate production inputs, etc.) and poor infrastructure that act together to depress the overall productivity of innovation, or because of an uncertain macroeconomic environment that induces short planning horizons and makes innovation particularly risky. This section addresses the first of these three constraints – the financing of innovative investment. It examines enterprise survey data on the relative importance of different sources of investment in 1) developing



157

a new product line, 2) upgrading an existing product line and 3) introducing new technology that substantially changes the way the firm’s main product is produced. The examination is based on the World Bank’s Enterprise Survey database and relates to the period 2002–6.10 Only about one-third of the enterprises which indicated the relative importance of different financial sources for their investment responded that they had undertaken innovative investment (first column of Table 7.2); about another third responded that they had not undertaken innovation and the remaining enterprises did not respond. The majority of large firms indicated that they had undertaken innovative investment, while the majority of small firms indicated that they had not done so and the replies by medium-sized firms provided a mixed picture. The results for the entire sample (Table 7.2) give no indication to suggest that a specific financial source is particularly conducive to innovative investment. The results are similar to those discussed above for total investment: innovative enterprises source the bulk of their finance for innovative investment from internal funds and retained earnings, followed by bank loans. Within this general pattern there are substantial differences across firm sizes. In particular, bank financing is generally more prevalent among larger innovative firms, whereas small innovative firms rely to a greater extent on retained earnings and finance from family and friends. Within groups of equally sized innovative firms, there is wide variation in the importance of specific financial sources across different country groups. The results of a probit estimation regarding the financing of investment by large firms (i.e. the most innovative group of firms) in order to develop a major new product line are reported in the first five columns of Table 7.3. They indicate that innovation is more likely to occur when investment is financed from internal funds and retained earnings or from bank credit, although this probability is much higher in Latin America than in Africa or Asia. Moreover, in Latin America, financing from equity markets, trade credit and investment funds strongly increases the probability of innovative investment, while this is not the case in the other developing countries. Overall, these results suggest that whether or not firms undertake innovative investment is determined more by size- and region-specific factors than by specific financial sources. The results of a similar estimation for investment by new firms in the upgrading of an existing product line are reported in column 6 of Table 7.3.11 Financing sourced from investment funds has by far the


Table 7.2

Sources of innovative investment finance, 2002–2006

Number of firms


Local and foreignowned commercial banks %

Investment funds %

Trade credit %

Equity %

Family and friends % Other %

Developed a major new product line Yes All firms 11,590 55.0 Small firms 3477 64.0 Medium firms 4296 54.3 Large firms 3750 47.6

22.6 14.4 22.9 29.7

1.7 1.0 1.8 2.3

4.3 36.8 5.1 4.0

5.2 5.0 4.7 6.1

3.3 5.5 3.1 1.5

7.8 6.3 8.2 8.8

No All firms Small firms Medium firms Large firms

18.0 12.2 19.7 27.0

1.3 0.7 1.4 2.5

3.0 3.0 3.4 2.4

6.3 5.1 6.9 7.8

3.8 4.7 3.7 2.0

9.3 7.1 10.2 12.5

Upgraded an existing product line Yes All firms 12,919 52.9 Small firms 3814 61.1 Medium firms 4589 5.37 Large firms 4439 45.3 New Firms 281 46.4

22.0 13.7 21.4 29. 15.3

1.9 0.9 2.0 2.7 3.7

3.3 3.3 3.7 2.9 3.7

7.2 7.8 6.2 7.6 14.3

3.7 6.1 3.5 1.8 8.2

9.0 7.1 9.4 10.2 8.3

No All firms Small firms

16.9 12.0

1.1 0.6

2.5 2.3

5.6 4.3

3.5 4.3

10.1 7.2

13,547 6135 4292 3050

8894 4261

58.4 67.2 54.8 45.9

60.4 69.3


Medium firms Large firms New Firms

2790 1788 166

3.3 1.6 5.3

11.4 14.7 4.1

Introduced new technology that has substantially changed the way that the main product is produced Yes All firms 11,076 54.0 23.2 1.8 4.1 5.6 Small firms 3186 62.3 15.7 1.0 3.7 5.3 Medium firms 4081 54.7 22.7 1.8 5.0 4.7 Large firms 3775 46.3 30.1 2.4 3.6 7.0

3.1 5.1 3.1 1.5

8.0 6.8 7.9 9.1


3.9 4.9 3.7 1.9

9.0 6.9 10.1 11.8

14,018 6313 4567 3043

55.4 47.1 55.9

58.8 67.7 54.5 47.6

19.0 25.3 20.2

18.0 11.9 20.2 26.8

1.1 2.2 0.5

1.2 0.6 1.3 2.3

2.7 2.3 3.0

3.2 3.1 3.5 3.0

7.1 6.7 11.2

5.9 5.0 6.7 6.6

Notes: New firms = firms aged two years or less. Small firms =fewer than 20 employees; medium firms = 20–99 employees; large firms = more than 99 employees. The numbers for small, medium and large firms may not add up to the total number given for all firms because some firms gave no indication of their size. Investment funds includes funds from investment funds, development banks and other state services. Source: Author’s calculations, based on World Bank, Enterprise Survey database.


Table 7.3

Financial determinants of innovative investment Upgrading an existing product line, new firms

Developing a major new product line, large firms

Internal funds and retained earnings Local and foreignowned commercial banks Investment funds Trade credit Equity Family and friends Other

Number of firms Groups

(1)

(2)

(3)

(4)

(5)

(6)

1.52 (6.29)*** 2.09

2.98 (9.76)*** 2.10

1.32 (2.29)*** 2.39

1.18 (2.51)*** 0.94

6.79 (11.7)*** 8.27

1.98 (2.21)*** 0.78

(6.45)*** 0.75 (0.65) 6.34 (5.52)*** -0.83 (-1.24) -1.22 (-0.89) -1.45 (-2.80)***

(5.92)*** 3.12 (2.22)** 7.69 (5.79)*** -1.01 (-1.44) -1.21 (-0.86) -1.21 (-3.61)***

(2.05)*** -4.72 (-0.81)** 1.51 (0.32) -1.37 (-0.39) 5.02 (0.87) 5.02 (2.56)***

(2.29)*** -2.23 (-1.08) 3.94 (1.95)* -1.58 (-2.11)* -2.36 (1.53) -2.36 (-6.08)***

(2.68)*** 8.46 (3.83)*** 10.4 (5.37)*** 12.8 (3.53)*** 10.6 (1.90)* 10.6 (2.51)***

(0.46)*** 20.3 (3.49)*** 5.39 (1.25) 5.08 (2.61)*** 6.61 (2.38)** 6.61 (2.89)***

6800 All countries

5002 All developing countries

803 Africa

3037 Asia

1162 Latin America

445 All developing countries

Notes: Large firms=more than 99 employes; new firms=firms aged two years or less. Binary probit estimates with quadratic hill climbing. The coefficients are multiplied by 1000. Robust z-statistics in paratheses. *** significant at 1 per cent, ** significant at 5 per cent, * significant at 10 per cent. Source: Author’s calculations, based on World Bank, Enterprise Survey database. 10.1057/9780230276123 - The Rise of Technological Power in the South, Edited by Xiaolan Fu and Luc Soete


161

greatest probability of leading to innovative investment by new firms, followed by financing from family and friends and equity. Bank finance appears to play virtually no role, while internal finance and retained profits make innovative investment more probable but the coefficient is of small size and barely statistically significant. Country-specific evidence for two of the BRICS countries, namely Brazil and China, shows that about two-thirds of the Brazilian enterprises that responded to the survey indicated that they had undertaken innovative investment (particularly for upgrading an existing product line), as shown in Table 7.4; the opposite holds for the Chinese enterprises (Table 7.5). In both countries, as reported above also for the total sample, the share of innovators is greatest in the group of large firms and least in the group of small firms. The results also indicate that in Brazil (Table 7.4) financing from investment funds, family and friends and the category ‘other’ is more likely to be used for innovative investment. This gives further support to the earlier finding that development banks can play an important role in the financing of innovation. In China (Table 7.5), it is in particular financing from internal funding and retained earnings, as well as bank credit, that makes innovative investment more likely to occur, while the opposite holds for the category ‘other’. To the extent that this category reflects informal financing from the curb market, the unsure legal status of this source and the associated high interest rates may explain why it is not conducive to innovative investment. The table also shows that the relatively large share of equity in investment financing by Chinese enterprises does not make innovation more likely.12

7.5

Conclusions

The pattern of how firms finance their productive investment displays a number of characteristics that apply to all countries, such as the relatively greater importance of internal finance in comparison to external finance and the relatively lower importance of equity finance. But within this general pattern there are substantial differences both across regional country groups and firms. In particular, bank financing is generally more prevalent among larger firms, whereas small and new firms rely to a greater extent on retained earnings and finance from family and friends. This variation in the relative importance of different sources of investment finance can be traced to information asymmetries between a firm’s managers and potential providers of external finance with


Table 7.4

Sources of innovative investment finance, Brazil, 2003

Number of firms


Local and foreign- owned commercial banks %

Investment funds %

Trade credit %

Equity %

Family and friends %

Other %

Developed a major new product line Yes All firms Small firms Medium firms Large firms

937 136 507

54.5 58.3 56.9

13.9 10.5 13.9

9.2 6.4 6.9

9.3 10.3 9.7

4.4 2.8 3.8

1.4 3.5 1.4

7.3 8.3 7.3

290

48.7

15.3

14.6

8.1

6.4

0.3

6.7

414 90 229

60.4 57.7 62.3

15.3 11.3 16.9

6.7 4.7 5.2

7.5 17.0 4.7

4.0 4.7 3.9

0.9 0.4 1.2

5.2 4.2 5.8

94

58..7

14.2

12.6

5.4

3.7

0.6

4.9

Upgraded an existing product line Yes All firms 1288 56.0 Small firms 205 56.5 Medium 705 58.3 firms Large firms 374 51.7

14.5 11.8 14.8

8.5 6.3 6.4

8.7 12.9 8.3

4.3 3.5 3.8

1.1 2.0 1.3

6.9 7.1 7.0

14.9

13.7

7.1

5.9

0.3

6.4




63 21 31

62.4 72.6 64.7

11.1 1.7 14.7

7.7 0.0 6.6

9.9 14.3 4.7

3.5 4.3 4.2

2.7 4.3 1.9

2.7 2.9 3.2

10

30.0

21.0

28.0

18.0

0.0

2.0

1.0

Introduced new technology that has substantially changed the way that the main product is produced Yes All firms Small firms Medium firms Large firms No All firms Small firms Medium firms Large firms

947 138 522

56.8 57.7 58.8

13.4 9.5 14.0

8.3 5.7 6.6

9.3 13.4 9.1

4.1 4.3 3.4

1.2 1.7 1.4

6.9 7.8 6.5

284

52.8

14.1

12.8

7.7

5.2

0.4

6.9

404 88 214

55.3 58.6 58.0

16.3 12.9 16.7

8.7 5.8 5.8

7.4 12.3 5.8

4.8 2.3 4.8

1.3 3.1 1.2

6.2 4.9 7.7

100

46.7

17.6

17.7

6.5

7.1

0.2

4.2

Notes: Small firms = fewer than 20 employees; medium firms = 20–99 employees; Large firms = more than 99 employees. The numbers for small, medium and large firms may not add up to the total number given for all firms because some firms gave no indication of their size. Investment funds includes funds from investment funds, development banks and other state services. Source: Author’s calculations, based on World Bank, Enterprise Survey database


Table 7.5

Sources of innovative investment finance, China, 2003

Number of firms


Local and foreign- owned commercial banks %

Investment funds %

Trade credit %

Equity %

Family and friends %

Other %

Developed a major new product line Yes All firms Small firms Medium firms Large firms No All firms Small firms Medium firms Large firms

348 32 110

17.2 14.5 18.3

25.1 10.3 17.0

0.3 0.0 0.1

0.9 0.0 0.9

12.1 13.8 11.4

5.5 12.3 9.2

38.9 49.0 43.2

20

17.3

31.8

0.5

1.0

12.5

2.5

34.5

983 133 366

14.7 13.9 13.6

18.4 7.7 14.4

0.6 1.2 0.7

1.1 0.0 1.2

12.4 17.2 12.6

6.0 11.1 8.4

46.7 49.0 49.1

479

15.8

24.6

0.4

1.3

11.1

2.7

44.0

21.7 4.9 15.0

0.5 0.0 1.2

1.5 0.0 1.5

12.9 18.7 10.1

4.7 8.4 8.9

39.5 52.0 44.9

Upgraded an existing product line Yes All firms 592 19.2 Small firms 47 16.0 Medium 181 18.4 firms


Large firms No All firms Small firms Medium firms Large firms

361

20.2

27.2

0.3

1.7

13.6

2.2

35.0

740 118 294

12.3 13.3 12.4

18.9 9.5 14.8

0.6 1.4 0.2

0.7 0.0 0.9

11.9 15.6 13.6

6.8 12.5 8.5

48.9 47.8 49.6

323

12.0

26.2

0.6

0.8

9.1

3.2

48.1

Introduced new technology that has substantially changed the way that the main product is produced Yes All firms Small firms Medium firms Large firms No All firms Small firms Medium firms Large firms

490 36 136

18.9 14.2 17.0

2.3 12.0 16.3

0.4 0.0 0.8

1.2 0.0 0.9

12.3 16.4 7.2

6.6 10.3 14.0

37.1 47.2 43.8

315

20.5

27.8

0.3

1.5

14.2

2.9

32.7

842 130 339

13.1 13.9 13.5

18.4 7.1 14.6

0.6 1.2 0.5

0.9 0.0 1.2

12.3 16.4 14.4

5.4 11.5 6.5

49.2 49.9 49.4

368

12.7

25.8

0.6

1.0

9.2

2.4

48.3

Notes: Small firms = fewer than 20 employees; medium firms = 20–99 employees; large firms = more than 99 employees. The numbers for small, medium and large firms may not add up to the total number given for all firms because some firms gave no indication of their size. Investment funds includes funds from investment funds, development banks and other state services. Source: Author’s calculations, based on World Bank, Enterprise Survey database.


166

Jörg Mayer

respect to the value of its existing assets and the quality of its investment opportunities. The use of retained earnings allows a firm’s manager to protect insider information, the disclosure of which would expose the firm to imitation and severely restrict its ability to achieve returns on its investment. However, small and medium-sized firms and new firms encounter serious obstacles to accessing suitable external financing for their investments. Therefore they resort to internal or informal sources of finance, not out of choice but generally for lack of an alternative. From a policy perspective, what matters is ensuring access of firms to reliable, adequate and cost-effective sources for financing productive investment. At one level, positive demand and profit expectations, as well as secure property rights and related patent and copyright policies, are an important condition for entrepreneurs to envisage undertaking productive investment and for potential lenders to finance such investment. However, to the extent that the availability of funds, and in particular the amount of profits retained by firms, determines investment, measures that increase the liquidity of firms are likely to spur investment. Possible measures include a range of fiscal incentives, such as preferential tax treatment for reinvested or retained profits and special depreciation allowances aimed at accelerating capital accumulation and enhancing productive capacity. They also include measures that allow firms to set-up various reserve funds against risk, so as to defer paying taxes on profits, or grant import duty rebates on capital equipment. The impact of such measures can be amplified if banks are encouraged to make loans more easily available for investment. Restrictions on lending for consumption or for speculative purposes could induce banks to extend longer-term loans for investment purposes. To the extent that high lending rates reflect perceived risks, government guarantees for loans to finance promising investment projects of firms that otherwise may have very limited access to longer-term bank credit (or may be able to obtain such credit only at extremely high cost that would make their investment unviable) may be envisaged. To maximize the developmental impact of investment, it is important to consider not only the microeconomic profitability of an investment project, but also the external benefits that the project generates for the economy as a whole. One way to bring both considerations to bear on credit allocation could be through joint financing of certain investment projects by private and public banks. Whereas the commercial bank would contribute its expertise in assessing the viability of a project from a private-sector perspective, the public financial institutions



167

would make a judgement from the point of view of the project’s overall developmental merits, and through their participation in the financing could reduce the risk of the commercial bank. This kind of arrangement has several precedents in some developed countries in the postwar period, in some successful late industrializers in East Asia and also in the activities of BNDES in Brazil. The debate on the role of public and development banks has often centred on the argument that state ownership and the existence of national development banks may increase the opportunities for corruption and patronage, rather than on the economic merits of such institutions. It is clear that public and development banks can fulfil their developmental role only if they are subject to strict rules on accountability (e.g. by giving them clear mandates and making further activities conditional on reporting and explaining deviations from the targets set in the mandate) and if the projects financed by them need to meet performance standards (e.g. related to productivity performance). It should also not be forgotten that a state that is prone to corruption and patronage will likely be unable to provide contracts and regulation that would allow private banks to step in – and in this situation privatization may involve large amounts of bribes and eventually lead to socially suboptimal outcomes. Taken together, these examples highlight the importance of an appropriate governance structure to provide conditions where the benefits of public bank ownership outweigh the inefficiencies that may be generated by its political nature.

Notes Acknowledgements: Parts of this chapter were prepared for UNCTAD’s Trade and Development Report 2008. The author is grateful to participants of the ‘Confronting the Challenge of Technology for Development: Experiences from the BRICS’ conference in Oxford in May 2008, as well as Alfredo Calcagno and Detlef Kotte for helpful comments on an earlier draft. The opinions expressed are only those of the author and do not necessarily reflect the views of UNCTAD or its Member states. 1. The pecking-order theory contrasts with the Static Trade-off model (STO). The STO assumes that firms try to adhere to a target capital structure, which is determined by equalizing the marginal benefit from tax savings associated with additional debt and the cost of financial distress when the firm finds it has borrowed too much. While it has proven difficult to distinguish between these hypotheses empirically, Shyam-Sunder and Myers (1999) show that the STO theory cannot account for the usually observed correlation between high profits and low debt ratios. For a discussion of the empirical evidence, see also Hogan and Hutson (2005).


168

Jörg Mayer

2. Another solution would be for the government to grant guarantees for bank loans to new and innovative firms. 3. A domestic market for corporate bonds denominated in the domestic currency would also facilitate the provision of external finance for investment. However, such markets are absent in most developing countries. 4. The role of venture capital strongly expanded during the 1970s and early 1980s. This evolution was linked to the ICT revolution and the fact that this revolution was largely propelled by small private enterprises (Gompers and Lerner, 2001). 5. Hogan and Hutson (2005) provide evidence for this hypothesis from Ireland, and cite similar findings from other developed countries, including Finland, the UK and the US. They argue that venture capitalists seem to be better able than banks to overcome information asymmetry problems, but that the key reason for innovative entrepreneurs to favour venture capital over debt is their willingness to forfeit independence and control in order to obtain the finance needed to proceed with their projects. 6. See China Statistical Yearbook, table 6.4, at http://www.stats.gov.cn/tjsj/ ndsj/2007/indexeh.htm. The category ‘other’ also includes leasing, foreignowned banks and credit cards but, as in other developing countries, these sources are of very little importance in China. 7. Given that BNDES had a sound balance sheet, it was not affected by the Programme of Incentives for the Reduction of States’ Participation in Banking Activities (PROES) launched by the Brazilian government in 1995 (Levy Yeyati et al., 2007: 217–18). 8. Much of the literature on the role of state-owned banks focuses on their effect on growth and financial development. Levy Yeyati et al. (2007) demonstrate that findings showing an adverse effect of state ownership on financial development and growth are far less robust than often thought, and that evidence to support a causal adverse impact of state ownership of banks and growth relies on the unrealistic assumption that there is no correlation between the presence of public banks and the level of financial development. Moreover, they show that public banks in developing countries reduce pro-cyclicality in credit allocation. 9. See the BNDES website (http://www.bndes.gov.br). 10. The data include only the most recent results where country-specific surveys are available for various years during the period 2002–6. The 2006 surveys do not enable an identification of sourcing from foreign-owned banks, leasing and credit cards; however, judging from evidence for the other years, these sources are generally of little importance for developing and transition economies. Results from more recent surveys are not included because at the time of writing they were not part of the World Bank Enterprise Survey standardized database. 11. Developing a new product line and upgrading existing technology are hardly pertinent for new firms. Therefore, the examination of new innovative firms refers only to investment in upgrading an existing product line. 12. These findings for Brazil and China are supported by the results of a probit estimation which are available from the author on request.



169

References Allen, F., J. Qian and M. Qian (2005), ‘Law, finance, and economic growth in China’, Journal of Financial Economics, 77 (1), 57–116. Allen, F., J. Qian and M. Qian (2008), ‘China’s financial system: Past, present, and future’, in L. Brandt and T. Rawski (eds), China’s Great Economic Transformation. Cambridge: Cambridge University Press. Amsden, A.H. (2001), The Rise of the Rest: Challenges to the West from LateIndustrializing Economies. Oxford and New York: Oxford University Press. Chen, J.J. (2004), ‘Determinants of capital structure of Chinese listed companies’, Journal of Business Research, 57 (12), 1341–51. Cull, R., L.C. Xu and T. Zhu (2007), ‘Formal finance and trade credit during China’s transition’. Working Paper 4204. Washington, D.C.: World Bank. Gompers, P. and J. Lerner (2001), ‘The venture capital revolution’, Journal of Economic Perspectives, 15 (2), 145–68. Hall, B.H. (2002), ‘The financing of research and development’. NBER Working Paper 8773. Hogan, T. and E. Hutson (2005), ‘Capital structure in new technology-based firms: Evidence from the Irish software sector’, Global Finance Journal, 15 (3), 369–87. Lerner, J. (1995), ‘Venture capitalists and the oversight of private firms’, Journal of Finance, 50 (1), 301–18. Levy Yeyati, E., A. Micco and U. Panizza (2007), ‘A reappraisal of state-owned banks’, Economia, 7 (2), 209–47. Myers, S.C. and N. Majluf (1984), ‘Corporate financing and investment decisions when firms have information that investors do not have’, Journal of Financial Economics, 13 (2), 187–221. Shyam-Sunder, L. and S.C. Myers (1999), ‘Testing static tradeoff against pecking order models of capital structure’, Journal of Financial Economics, 51 (2), 219–44. Singh, A. (1997), ‘Savings, investment and the corporation in the East Asian miracle’, Journal of Development Studies, 34 (6), 112–37. UNCTAD (2008), ‘Trade and Development Report 2008’. New York and Geneva: United Nations.


8 A Comprehensive Model of Technological Learning: Empirical Research on the Chinese Manufacturing Sector Jin Chen, Xiaoyu Pu and Haihua Shen

8.1

Introduction

Following the rapid pace of industrial development in emerging economies, many enterprises are experiencing a growing need for innovation. However, the absence of strong technological capabilities, which contribute to continuous competitiveness, often impedes developing countries from improving their innovation performance. Technological capability has always been a key component of economic growth and welfare (Archibugi and Coco, 2004; 2005). There is a cycle which links technological learning, technology capability, technical change and production capability (Albu, 1997). Technological learning thus plays a fundamental and inevitable role in acquiring technology capability. This essential link can be investigated for the case of China whose old-style policy of ‘sacrificing the domestic market for technology transfer’ is now severely challenged. To thrive amid fierce global business competition, Chinese firms are nowadays impelled to upgrade their technological capabilities. Enhancing the domestic innovation capability is now the government’s solution to making the shift in economic restructuring and the mode of generating economic growth. To increase the domestic innovation capability, firms can partly rely on the mature mass-production market, which is a major advantage of the ‘made in China’ period. However, it is important to realize that countries and enterprises with technology advantages tend to be cautious about the transfer or licensing of high technologies, 170


A Comprehensive Model of Technological Learning

171

in consideration of their own strategic interest and overall capacity. Hence, international enterprises often keep their core technologies hidden from their partners in developing countries when transferring product lines. This practice puts manufacturing corporations in developing countries into a very disadvantageous situation in global competition. As a result, corporations in less developed countries (LDCs) have been urged to put more emphasis on technological learning and technology upgrading as a way out of the current dilemmas of low added-value in their industries (Carayannis, 1998, Carayannis et al., 1994; 1998; Teece et al., 1997). But there are still many questions not answered by existing theories. What factors influence firms’ willingness to learn? What should they pay attention to during the learning process? How does technological learning actually improve the firms’ innovation performance? How can a firm develop its own core technology by conducting technological learning? These are not only academic questions but also important issues that are essential to the development of Chinese firms and to the theory and practice of technology innovation management. This chapter will study a sample of Chinese manufacturing corporations and provide empirical evidence about how technological learning can improve innovation performance by upgrading corporations’ technological capability. Some key factors influencing the technological learning process can be drawn from the results. In Section 8.2, the relevant and related streams of literature are briefly reviewed. Section 8.3 presents several hypotheses to be tested in the following part of the research. The methodology and results are described in Sections 8.4 and Section 8.5. Section 8.6 will suggest the main implications of this study for manufacturing firms in developing countries.

8.2

Literature review

8.2.1 Technological learning Technological learning is defined as a process through which replication and processes can be made better and quicker, with the possibility of identifying new products (Teece et al., 1997). It is a process of organizational transformation, integrating technology development and management throughout individuals, teams or organizations as a whole in order to improve decision-making as well as to enhance the control of uncertainty and complexity (Carayannis et al., 1994). Effective technological learning can bring a company competitive advantage by extending the scale of strategic behaviour, improving managerial abilities and


172

Jin Chen, Xiaoyu Pu and Haihua Shen

allowing the choice of a suitable strategy according to the company’s environment (ibid.). Albu (1997) constructs the theory of the learning cycle based on the ideas above. He claims that technological ability gained through technological learning can generate and manage technology transformation, feed the experience back to technological learning, and thus upgrade technology capability. Carayannis et al. (1994) developed the triple-layered technology learning structure. In this framework, learning takes place at three interrelated levels – the operational learning level, the tactical learning level and the strategic learning level (ibid.; Carayannis and Kassicieh, 1996). People learn to accumulate experience and learning by doing new things at the operational level. The tactical level comprises the learning of new strategies with which to apply the accumulated experience and the learning process. At the strategic level people learn new strategies and new views of the operating universe. Chinese scholars consider technological learning as an organizational behaviour to obtain new technology. Xie (2001) proposes a process model of technological learning, which introduces the sequential linkage among technology, production capability and innovation capability. Chen and QU (2003) combines the situation of developing countries with the imitation-innovation theory of South Korea developed by Kim (1997) and creates another technological learning model. This model extends the triple-layered technology learning structure suggested by Carayannis et al. (1994) by adding multiple learning contents, learning sources, learning agents and learning methods. So, we can draw the dynamics of technological learning as follows. Technological learning: 1) has specific goals, 2) is risky and needs investigation, 3) needs interaction with the external environment, 4) is a multi-layered behaviour and 5) is a long period of processing; the upgrade of capability is divided into phases. 8.2.2 Technology capability Technology capability is one of the sources of company competitive advantage, which is an extremely important research area in corporate management theories. This topic has already been discussed extensively in the literature, from macro to micro aspects. This chapter is focused on the micro aspect. Technology capability is divided into four abilities, which are the ability of learning by searching and acquiring strategy by learning from employees, the ability of learning from practising, the ability of learning from



173

performance feedback, the ability of learning from changes and the ability of learning by training (Bell, 1984). It is the combination of technological searching capability, technological learning capability and technological innovation capability, as well as the linked process of searching-learninginnovation of technology, that constructs the path for technology capability to upgrade (Dore, 1984). Desai’s (1984) definition of technology capability includes the ability of a firm to acquire technology, to operate, to duplicate and extend, and to innovate. It is clearly generalized as the four layers from technology purchasing, to use, imitation and innovation. Panda and Ramanathan (1996) decompose technology capability according to firm’s value chain and divide the system into three parts, which are strategic technology capability, tactical technology capability and assistant technology capability. They also point out that the control capability is key for technology capability to be effectively utilized and upgraded. Barton (1992) develops the concept of core technology capability in the discussion of core capability of firms. She claims that technology capability comprises four parts, which are employee skills and knowledge base, technical systems, managerial systems and the values and norms of firm. Jonker et al. (2006) define technology capability as the combination of production capability and innovation capability – the former refers to the ability to improve the performance of existing equipment, while the latter includes the creation and supporting of new technology, as well as the rebuilding of existing technology. Wei and Xu (1995) define technology capability as the ability of a firm to acquire advanced technology and information from external sources, to combine them with internal knowledge and to create new technology and information so as to realize the innovation and diffusion of technology while accumulating the technology and knowledge. For Zhang and Guo (1997) technology capability is the ability to operate technology resources. They argue that technology capability is expressed through the operating of technology resources and developed through the learning process. To sum up, the technology capability of a firm includes its potential capability and explicit capability: the former is the capability that is not expressed by the consolidation of the communication of existing knowledge, and the latter is the capability that is already consolidated or expressed. Technology capability has the following characteristics: 1) it is the engine of technology development, 2) it has the comprehensive ability needed by technology development, 3) it is sensitive to other firms’ technology strategies, 4) it can derive or indicate the causal relationship between behaviour and performance, 5) it connotes knowledge, 6) it includes


174


cognition skills and behavioural skills and 7) it is a evolutionary process. By observing these characteristics, we can see that technology capability is the determining factor in the creation, choice and development of technology, and that innovation is a consequence of technology capability.

8.3 Theory developments This section discusses the impact of technology capability on innovation performance from the following five aspects: the source, content, agent, layers and environment of technological learning. 8.3.1 Technological learning 8.3.1.1 The source of technological learning The source of technological learning refers to the knowledge, technology and information sources that enable a company to initiate technological learning. It includes both internal and external learning sources. Companies should integrate all technical information to enhance technological learning continuously, and in this way innovate from existing technologies. As for the internal learning sources, Zhang (1998) and Chen (2000) pointed out that the executives, in-house R&D, marketing and manufacturing departments are all information sources (learning sources) for a company. Note that the learning sources in companies are not isolated, as Von Hippel (1986) claims that innovation ideas are derived from the interaction between R&D and marketing departments. So, the technological learning source comes from interactions between departments, and these interactions promote the spread of various types of knowledge, especially tacit knowledge. External learning sources include customers, suppliers, universities and other research institutes, competitors and distributors. In Chinese firms, external technology information is mainly obtained in the following three ways: 1) seeking all chances to obtain the latest international technological information, 2) accessing market technological information through sales personnel and market researchers and 3) R&D staff visiting markets and clients to get the technological information. Based on these discussions about technological learning sources, this chapter has two hypotheses: H1: Technological learning sources have a positive influence on technology capability, i.e. widening the range of technological learning sources can promote an upgrading of technology capability.



175

H2: Technological learning sources have a positive influence on innovation performance, i.e. widening the range of technological learning sources can improve innovation performance. 8.3.1.2 The content of technological learning Knowledge can be categorized as explicit knowledge and tacit knowledge. The former can be demonstrated and shared – no matter if it is written down or stored in someone’s mind, while the latter is hard to demonstrate – it is experience, know-how, inspiration and the like. Because of the difficulty of measuring and imitating tacit knowledge, it bears a much stronger importance than explicit knowledge. Thus, technological learning should include scientific knowledge, technological knowledge, experiential knowledge and know-how. Scientific knowledge refers to systematical-theoretical knowledge, which is the infrastructure. Technological knowledge refers to knowledge that is relative to technology and focuses on the application of knowledge. Experiential knowledge and know-how are tacit forms of knowledge derived from practice. Tacit knowledge is the focus of technological learning, as it is where most of a corporation’s competitive advantages come from. The absorption of tacit knowledge directly affects the outcome of technological learning (Tang et al., 2004). As the accumulation of tacit knowledge requires a large amount of R&D and manufacturing practice as well as continuous improvements in manufacturing processes, it is of high priority for Chinese firms to build up this type of knowledge in order to use their established technology efficiently. Based on the previous discussion about the content of technological learning, this chapter makes the two hypotheses: H3: The content of technological learning has a positive influence on technology capability, i.e. enhancing the absorption of technological learning contents can promote the upgrading of technology capability; H4: The content of technological learning has a positive influence on innovation performance, i.e. enhancing the absorption of technological learning contents can improve the innovation performance. 8.3.1.3 The agent of technological learning The agent of technological learning includes individuals, teams and the organization as a whole. Individual learning is the process by which an individual obtains knowledge and skills, team learning is


176


the way in which a whole team acquires knowledge and know-how through in-team communication and interaction and organization learning is the ‘Kaizen’ process in which various types of information are effectively processed, interpreted and responded to. An organization is an agent that is able to learn, process information, rethink experience and maintain a huge amount of knowledge, skills and specialization. It is often constrained in learning by subsystems such as the structure, procedure and culture of the organization and its own technology. In order to promote the technological learning of a firm, we need to enforce each of these types of agent’s learning, while enhancing the communication and interaction of each, so as to establish an internal environment that is beneficial for organizational learning. Hence, two hypotheses based on the previous discussion about the agent of technological learning are made: H5: The technological learning agent has a positive influence on technology capability, i.e. enabling the interaction between different agents can promote the upgrading of technology capability; H6: The technological learning agent has a positive influence on innovation performance, i.e. enabling the interaction between different agents can improve the innovation performance. 8.3.1.4 The levels of technological learning Carayannis (1998) developed the triple-layered model of technological learning, which includes the operational, tactical and strategic learning levels. The operational level of learning refers to the knowledge, skills and experiences accumulated during practice, and belongs to the short-medium term of learning, focuses on the mastering of technologies and equipment that are newly introduced into the firm. The tactical level of learning is a medium-long term process of learning. It establishes a new decision-making system for the firm by changing the technology decision systems or adding new rules to the established systems. The strategic level of learning falls into the long-term category, and focuses on the reconstruction of some of the firm’s strategic issues. This process may extend a firm’s knowledge of the limitation and potential of its own strategic environment, which usually results in a change of the rules of the game, or in the establishment of a brand new industry. Since the different layers of technological learning are interdependent, ignoring any one of them will jeopardize the others.



177

In this way, this chapter establishes the following hypotheses: H7: Technological learning levels have a positive influence on technology capability, i.e. emphasizing the comprehensive learning that at all levels helps to promote the upgrading of technology capability; H8: Technological learning levels have a positive influence on innovation performance, i.e. emphasizing the comprehensive learning that at all levels helps to improve the innovation performance. 8.3.1.5 The environment of technological learning Carayannis (1998) defined the environment of technological learning as the scale of learning activities and the environmental factors that influence the content and process of learning. It mainly consists of the executive’s attitude to learning, its funding and the corresponding motivation. The attitude of the executive determines a firm’s learning situation to a large extent. A positive attitude at the executive level towards learning may receive a positive response from the employees and hence intensify the latter’s learning initiative and motivation and establish an active learning environment. Funding support includes software and hardware. The former refers to the training of employees, while the latter refers to the construction of network information systems, etc. The motivation mechanism is the key to encouraging employees to learn and can effectively facilitate the firm’s technological learning. Based on the previous discussion about the environment of technological learning, the following hypotheses are made: H9: The technological learning environment has a positive influence on technology capability, i.e. improving the environment of technological learning helps to promote the upgrading of technology capability; H10: The technological learning environment has a positive influence on innovation performance, i.e. improving the environment of technological learning helps to improve the innovation performance. 8.3.2 Technology capability This chapter treats the firm’s technology capability as an intermediate variable and considers both the direct effects of technological learning on innovation performance and the indirect effects caused by technology capability. So, the following hypothesis about the relationship


178 Jin Chen, Xiaoyu Pu and Haihua Shen

H1

Source

H3

H2

Technology capability

Content H11

H4 H5

Agent H6

H8

H7

Innovation performance

Levels H9 H10 Environment

Figure 8.1 The model: Technological learning’s influence on innovation performance

between technology capability and innovation performance is made: H11: Technology capability has a positive influence on innovation performance, i.e. upgrading the technology capability can improve a firm’s innovation performance. 8.3.3

Conceptual model

The conceptual model of this study is shown in Figure 8.1. In this model, the five elements of technological learning are extracted as the causal variables which have indirect influence on innovation performance through the intermediate variable of technology capability, as well as some direct influence on the final outcome.

8.4 Methodology 8.4.1 Sample This study sampled 127 firms with 251 questionnaires, of which 192 were returned. There were 118 valid questionnaires from 92 firms. The valid reclaim rate was 65.5 per cent. The following steps were taken in order to control the data veracity and validity: 1) Control of the sample subjects. The sample subjects were typically large manufacturing corporations in China, involving companies that produce marine containers,



179

tractors, electricity meters, vehicle axles, etc. Such large corporations are those with strong R&D capability and a need for technological learning. The interviewees consisted mainly of engineers in R&D departments, managers and executives from the marketing, sales or engineering departments; 80 per cent of the interviewees have got at least a bachelor’s degree. 2) Control of the delivery process of questionnaires. At the target corporations located in Shanghai, Hangzhou, Shenzhen, Qingdao and Luoyang, face-to-face interviews were conducted, with immediate feedback on site. The other questionnaires were sent by mail to the relevant interviewees. These initiatives proved effective in controlling the veracity and validity of data. 8.4.2

Reliability and validity analysis

Reliability is gauged via a Cronbach α value, which should be above the threshold of 0.70, and the item-to-total correlation over 0.30, to be considered reliable. The results are shown in Table 8.1. Conducting factor analysis to the 25 independent variables, and extracting the principal component with characteristic roots greater than 1 as factors, we get five factors which can explain 60.040 per cent of the total variance. This indicates that the independent variables in this research are of constructive validity. Conducting the previous analysis on the seven dependent variables, two factors are extracted which explain 71.995 per cent of the total variance, indicating that the dependent variables in the research are also of constructive validity. 8.4.3 Measurement model The original model of this research is shown in Figure 8.2. There are 25 independent variables (S1-S11, C1-C4, A1-A3, L1-L3 and E1-E4) and

Table 8.1 Reliability analysis of variables α

Variable class

Variables

Independent variables

Learning source Learning content Learning agent Learning levels Learning environment

0.7604 0.8011 0.7552 0.7220 0.7847

Technology capability Innovation performance

0.7259 0.8165

Dependent variables

Threshold

α ≥ 0.70


180 e


S1 Source

e

S11

e

C1

e

C2

e

C3

e

C4

e

A1

e

A2

e

A3

e

L1

e

L2

e

L3

e

E1

e

E2

e

E3

e

E4

γ si

Content

γ st

e

γ ct Technology capability

γ ci γ at Agent γ lt


γ li γ et

e

P

e

M

e

H

e

IP1

e

IP2

e

IP3

e

β ti

γ ai

Levels

R

γei

e

Environment

Figure 8.2 Original model of the research

seven dependent variables (R, P, M, H and IP1-IP3). Meanwhile, there are 11 paths corresponding to the 11 hypotheses made in Section 8.3. Running Amos 4.0 to test the overall goodness of fit, we got the following results: χ2 = 1228.480; df = 440; p = 0.000 < 0.05; χ2/df = 2.792 < 3. Other indicators are: TLI = 0.903, CFI = 0.915 > 0.90; standardized RMR = 0.076 falls into the scale of 0.05–0.08. But the NFI = 0.874 < 0.90; RMSEA = 0.124 > 0.08, which indicates that the original model cannot fit very well into the sample data. Hence, an amendment of the model is needed. The result is that deleting the path of γei can bring the most significant decrease to χ2, so the path of environment’s direct influence on innovation performance should be deleted from the model. Testing again for the overall model fitness, we obtain the following results: χ2 = 939.771; df = 441; p = 0.000 < 0.05; χ2/df = 2.131 < 3; TLI = 0.938, CFI = 0.945 > 0.90; Standardized RMR = 0.065; NFI = 0.901 > 0.90; RMSEA = 0.098. Note that all other indicators fit into the required values except RMSEA, which is still larger than 0.08. But according to Steiger (1990), models with RMSEA < 0.10 already have an acceptable fitness, so we accept this adjusted model here.



8.5

181

Results

After the adjustment of the model, the final results are generated by Amos 4.0. The hypotheses are tested according to the previous model, and all are approved except hypotheses 1 and 10. Figure 8.3 shows the adjusted relation model of this research. The paths in Figure 8.3 are all direct-effect paths that can be computed from each of the coefficients in the conceptual model. In order to make clear the overall influence on this model, we needed to test the indirect effects on those paths, and the total effect among those variables, which is the summary of direct and indirect influences. The effects are decomposed as shown in Table 8.2. After the decomposition and adding, it is clear that the key factor that influences innovation performance is the contents of technological learning (with a total effect of 0.725), and the second is learning levels (with a total effect of 0.625); although the learning source has no influence on technology capacity (which means no indirect effect on output), it has a relatively significant direct influence on output (with the direct effect of 0.581). To sum up, we can make the following conclusions 1) While the source of technological learning has a positive influence on innovation performance, it does not influence the technology capability. 2) The content of technological learning has a positive influence on both technology capability and innovation performance, and the latter is the

Source .581 Content

.646

.376 .235

Technology capability .540

Agent .094 .392 Levels

.463


.300

Environment

Figure 8.3 The adjusted model of influence factors


182


Table 8.2

Effect decomposition and key factor identification

Variable relationships Innovation performance ← Learning source Innovation performance ← Learning content Innovation performance ← Learning agent Innovation performance ← Learning levels Innovation performance ← Learning environment Innovation performance ← Technology capability

Direct eff.

Indirect eff.

Total eff.

0.581

0.000

0.581

0.376

0.349

0.725

0.094

0.127

0.221

0.463

0.212

0.675

0.000

0.162

0.162

0.540

0.000

0.540

most significant. 3) The agent of technological learning has a positive influence on both technology capability and innovation performance, but neither is significant. 4) The level of technological learning has a relatively distinct positive influence on both technology capability and innovation performance. 5) The environment of technological learning has a positive influence on technology capability, but it is unclear if it influences the innovation performance. (6) Technology capability has positive influence on innovation performance; it is found to be placed as the intermediate variable. (7) The factors with the most significant influence on innovation performance are the learning content, learning levels and learning sources. Thus, the adjusted model shown in Figure 8.3 is a reliable framework for Chinese firms to analyse their own paths towards higher innovation performance through technological learning. However, since this chapter has not taken into account the comparative analysis of different industries when considering the influence of learning. it cannot estimate the possible effects of the industry factor. Moreover, the sample consists mostly of established corporations, and there is little empirical data on the learning situation of emerging firms. So we are not able to determine the possible effects of the different scales of firms. Further studies should focus on comparative research between different industries or different scales of firms, looking at the influence of industry characteristics and scale factors which can generate more specific results for firms of different industries and different stages.



183

8.6 Conclusions and implications Improving the situation of technological learning is one of the key paths to achieving domestic innovation. From the theoretical and empirical study in this chapter, we can draw the following conclusions as suggestions: The source of technological learning: internal communications should be emphasized. Firms should make efforts to extend the sources of learning. Compared with the clear and specific external learning sources, internal learning tends to be underestimated or ignored. As a result, many resources are not fully utilized, which is a common mistake made by many corporations in China. Internal communication within or between functional units is a solution to this situation, and can make learning more effective and more natural. The content of technological learning: focus on the absorption of tacit knowledge. In this era of tremendous explicit knowledge, it is tacit knowledge which calls for the long-term accumulation and improvement that is needed by the corporations in China. The core technology capability is the real chasm between Chinese firms and established international corporations. However, it cannot be acquired within a short period of time. In order to cross this gap, firms need to focus on their own tacit knowledge and seek opportunities for long-term accumulation. The agent of technological learning enforcing the learning throughout the organization. it is implied by the results of this research that communication and interaction between individuals, teams and the organization as a whole can improve the innovation performance of a firm. Chinese firms are still putting too much emphasis on unitary learning, namely in the form of individual learning, and ignore both team and organization learning. In these circumstances, a more multi-agent learning process might be able to improve the learning efficiency. In addition to an internal learning system, firms should also establish a sound external learning system. The latter is based on industrial networks that consist of customers, suppliers, factories, research institutes and other related elements in a value chain. This network may even include sponsors, industrial associations and government agencies. The levels of technological learning: improving the strategic learning level. The results indicate that multi-layered learning can improve


184


innovation performance effectively. In the interviews and investigations, we noted that Chinese firms are comparatively short of strategic learning and some firms are even operating without an explicit longterm vision, not to mention a reliable environment identification or strategic evolution mechanism. Thus, on the one hand, firms need to promote their strategies successfully among their employees; on the other hand, the executives of firms need to absorb experiences from external sources, and to learn from inter-corporation collaborations, so as to keep an insight into their own industry sectors, and so take the appropriate decisions. The environment of technological learning: optimizing the motivation mechanism for technological learning. Compared to executive and funding support, the lack of an appropriate motivation mechanism is the fatal problem that most Chinese firms are confronted with at present. The motivation mechanism is the key to encouraging employees to learn, but it is not functioning as anticipated in most of the enterprises investigated in this study. Firms should be committed to optimizing their motivation mechanisms, inspiring their learning agents’ enthusiasm for technological learning. Other than physical awards, the opportunity for employees to prove themselves in their teams is more important.

References Albu, M. (1997), ‘Technological learning and innovation in industrial clusters in the South’, SPRU Electronic Working Paper No. 7. Brighton: University of Sussex. Archibugi, D. and A. Coco (2004), ‘A new indicator of Technological capabilities for developed and developing countries (ArCo)’, World Development, 32 (4), 629–54. Archibugi, D. and A. Coco (2005), ‘Measuring technological capabilities at the country level: A survey and a menu for choice’, Research Policy, 34, 175–94. Barton, D.L. (1992), ‘Core capability and core rigidities: A paradox in managing new product development’, Strategy Management Journal, 13, 111–25. Bell, M. (1984), ‘Learning and the accumulation of industrial technological capacity in developing countries’, in M. Fransmanand K. King (eds) Technological Capability in the Third World. London: Macmillan, 187–209. Carayannis, E. (1998), ‘Higher order technological learning as determinant of market success in the multimedia arena; a success story, a failure, and a question mark: Agfa/Bayer AG, Enable Software, and Sun Microsystems’, Technovation, 18 (10), 639–53. Carayannis, E. et al., (1994), ‘A multi-national, resource-based view of training and development and the strategic management of technological learning: Keys for social and corporate survival and success’, in 39th International



185

Council of Small Business Annual World Conference, Strasbourg, France, 27–9 June, 154–63. Carayannis, E. and S. Kassicieh (1996), ‘The relationship between market performance and higher order technological learning in high technology industries’, in Fifth International Conference on Management of Technology, Miami, FL, 27 February–1 March, 309–20. Chen, J. (2000), ‘A first glance into national technology developing system’, Science Press, Beijing. Chen, J. and W.G. Qu (2003), ‘A new technological learning in China’, Technovation, 23, 861–7. Desai, A.V. (1984), ‘Achievements and limitations of India’s technological capability in the Third World’, in M. Fransman and K. King (eds) Technological Capability in the Third World. London: Macmillan. Dore, R. (1984), ‘Technological Self-Reliance’, in M. Fransman and K. King (eds) Technological Capability in the Third World. London: Macmillan. Jonker, M., H. Romijn and A. Szirmai (2006), ‘Technological effort, technological capabilities and economic performance: A case study of the paper manufacturing sector in West Java’, Technovation, 26 (1), 121–34. Kim, L. (1997), Imitation to Innovation: The Dynamics of Korea’s Technological Learning. Boston, MA: Harvard Business School Press. Panda, H. and K. Ramanathan (1996), ‘Technological capability assessment of a firm in the electricity sector’, Technovation, 10, 561–88. Steiger, J.H. (1990), ‘Structural model evaluation and modification: An interval estimation approach’, Multivariate Behavioral Research, 25, 173–80. Tang, C., J. Zhou and T. Liu (2004), ‘A study on the framework of how the firm absorb tacit technological knowledge’, Science Research Management, 25 (4), 41–50. Teece, D.J., G.P. Pisano and A. Shuen (1997), ‘Dynamic capabilities and strategic management’, Strategic Management Journal, 18 (7), 509–33. Von Hippel, E. (1986), ‘Lead user: A source of movel product concept’, Management Science, 32 (7), 39–45. Wei, J. and Q. Xu (1995) ‘Corporate technology capability: Concepts, structure and assessment’, Science of Science and Management of S&T, 16 (9), 29–53. Xie, W. (2001), ‘Four models of technological learning process in developing countries’, Science Management Research, 19 (6), 19–25. Zhang, G. (1998), ‘The incentive sources and the information sources of enterprise’s technological innovation’, Science Research Management, 19 (4), 27–31. Zhang, G. and B. Guo (1997), ‘Technology, technology resource and technology capability’, Journal of Dialectics of Nature, 5, 37–43.


9 The Innovation of SMEs and Development of Industrial Clusters in China Jinmin Wang

9.1

Introduction

During the 1980s and 1990s, developing countries obtained more market share for some labour-intensive products than developed countries because of cost advantages. Subsequently, many export-oriented industrial clusters came into being in such countries. Most clusters of small and medium-sized enterprises (SMEs) are characteristic of labour-intensive, low-tech content and low entry barriers. However, most of these industrial clusters in developing countries had comparative advantage only in terms of low prices and failed to get into a position to enter mainstream international markets. In contrast, the industrial clusters in developed countries maintained competitive advantages in terms of quality, design capacity, technological innovation and rapid response to market changes (Zhu, 2003). Since the 1990s, SME industrial clusters have emerged and developed rapidly in tandem with the reforms in ownership and the rapid development of the market economy in the Yangtze and Pearl River deltas in China. In Jiangsu province, some SME industrial clusters formed on the basis of township and village enterprises (TVEs), while in Zhejiang province hundreds of family workshops, agglomerated in neighbouring villages or towns, are engaged in the same industry. Typical examples include in Zhejiang province the shoemaking cluster in Wenzhou city, the clothing cluster in Ningbo city and the socks cluster in Yiwu city; in Jiangsu province the silk and light fabric cluster in Shengze; and in Guangdong province the leisurewear cluster in Shaxi town and the metal-processing cluster in Xiaolan town, both in Zhongshan City. 186


The Innovation of SMEs 187

SMEs have been playing an important role in promoting regional innovation and economic development in China. According to the National Development and Reform Commission, the number of SMEs had risen to more than 42 million by the end of June 2007, accounting for 99.8 per cent of the total enterprises in the country. Moreover, the GDP, tax revenue, import and export volumes of SMEs amounted to 60, 53 and nearly 68 per cent of total national figures respectively. SMEs have provided approximately 75 per cent of employment opportunities in urban areas and have contributed strongly to the nation’s technical innovation, with 66 per cent of invention patents and more than 82 per cent of new products (Yang, 2008). With fiercer global competition, technological innovation has been of great importance to the growth and upgrading of SME industrial clusters in China. This chapter explores how the specialized wholesale market in China has contributed to the technological innovation of SME industrial clusters.

9.2 Literature review It has been widely recognized that innovation is playing a more and more significant role in promoting economic growth all over the world. Since the 1990s, several new approaches to innovation have been developed including the endogenous growth theory (Romer, 1990), the theory of national competitive advantage (Porter, 1990), the theory of national innovation systems (Nelson, 1993) and the social and network theory (Engel, 1997). Meanwhile, both developed and developing countries have published a series of policy reports to offer a crucial context for improving their innovation strategies and promoting economic growth. Since the 1990s, the debate on the definition of innovation concentrates mainly on two aspects: the evolutionary and interactive features of the innovative process and the fact that knowledge and learning play a key role in the process in both developed and developing economies (Arundel et al., 1998). Innovation can be understood as a social competence. It is a process of social interactions between a variety of social actors and emerges as a result of individual and collective inquiry and learning to improve the social predicament, holding a key to its success in creating value to individuals, groups or communities (Engel, 1997). This definition shows that individuals create value within a framework set by social institutions, organizations, formal and informal conventions and standards


188

Jinmin Wang

of professional and social practice. It is important to overcome institutionalized barriers to innovation and change in order to stimulate innovation in the whole society. The Conference Board of Canada defines innovation as ‘A process through which economic or social value is extracted from knowledge, through the generation, development and implementation of ideas to produce new or significantly improved goods, processes and services’ (Conference Board of Canada, 2004). The Advisory Committee on Measuring Innovation in the 21st Century in the US defines innovation as ‘the design, invention, development and/or implementation of new or altered products, services, processes, systems, organizational structures or business models for the purposes of creating new value for customers and financial returns for the firm’ (Advisory Committee on Measuring Innovation in the 21st Century, 2008). An industrial cluster is not only an economic agglomeration with complex internal mechanisms but also a hybrid of social and economic factors. Camagni (1991) argues that innovation is a collective learning process, depending mainly on the network of firms and the flow of skilled labour in a regional market. The process is usually motivated by common culture, psychology and political background. The cooperative rules and the implicit action guidelines accepted by individuals and firms will help to establish trust and become the prerequisite for improving the capability of regional collective learning. According to Scott (1992), competition in the modern production system results from both the adoption of market principles and the institutional framework. Such institutional frameworks will connect firms with familiar and reciprocal routines and this will bring about multiple forms of cooperation, strengthening the comparative advantage of specific industrial locations. Industrial clusters with strong development incentives usually need collective institutional arrangements on the basis of prevailing social and cultural norms so as to overcome market failure. Saxenian (1994) highlights community building and the reproduction of social networks and institutions as prerequisites for making flexible production systems work in practice. She also shows that external economic factors resulting from spatial proximity alone cannot explain the innovation and dynamics of growth in Silicon Valley, emphasizing the relevant role of local institutions and culture to coordinate decentralized production. Institutions vary and this makes a difference in the ability of firms in a cluster to build their capabilities and improve performance. Developing countries usually acquire embodied technology or equipment for both product and process innovation. Minor or incremental



changes can be the most frequent type of innovation activity in some developing countries, together with innovative applications of existing products or processes. Organizational changes are extremely significant in the innovation process. They have direct impacts on a firm’s performance and contribute to its preparedness to absorb new technologies incorporated in machinery and other equipment. Heterogeneity frequently prevails with regard to technological, organizational and managerial patterns (Ernesto et al., 2005). The opportunities for SME innovation in developing countries depend on the type of relations in which they are involved and the possibilities available to them for moving from one level of relations to another. Business upgrading can affect process, product, functional or inter-sectoral capacities or some combination of these; process upgrading results in increased production efficiency arising from the use of new technology or the improved management of existing technology; product upgrading occurs as a result of moving into higher-value products or services than previously supplied; functional upgrading refers to the redistribution of activity in ways which allow the overall skill content of activities to increase and inter-sectoral upgrading occurs where firms use the knowledge acquired in one area of activity to move into a separate activity, such as when skills acquired assembling television sets are applied to the assembly of computer equipment (Humphrey, 2002). Academic circles in China did not embark on research about industrial clusters until the 1990s. Since then, academic research on industrial clusters in China has mainly focused on the innovation of high-tech industrial clusters. Dijk and Wang (2005) studied a city-wide information and communication technologies (ICT) cluster in Nanjing, China, involving various software parks and universities, as well as a huge concentration of computer shops. They found that ICT clustering in China is still in the initial stage of development. Li (2006) conducted one study on the strategy of cluster upgrading against the background of international industrial relocation in the IT sector. He disclosed the different patterns and models of cluster relocation and upgrading after making a comparative study of typical IT clusters around the world. Liu and Song (2006) conducted an empirical study of the domestic appliance cluster in Qingdao and concluded that the development path of this large-firm-dominated cluster is heavily influenced by the collective innovation of entrepreneurs in the cluster, including the introduction of new technology and institutional innovation with firm support from local government. Cultivating domestic R&D, participating in the formulation of global standards and joining in global industrial


190

Jinmin Wang

associations are all of vital importance for industrial clusters in realizing successful path creation. The literature on industrial clusters in developing countries has mainly focused on the role of multinational companies in promoting innovation by SMEs and forging local-global linkages while ignoring some important institutional arrangements such as the large-scale specialized wholesale market in China. Clustered firms in other developing countries mainly concentrate on the production process, leaving the products to be marketed by professional agents. Even if there are physical wholesale markets, the scales are usually small. Production and sales are undertaken by specialized firms. Moreover, SMEs within industrial clusters do not usually participate in the establishment of the marketing network. However, in China, the large-scale specialized wholesale market has not only helped local SMEs to forge an industrial and marketing network, but also enables them to make constant technological innovations and the upgrade along the global value chain. This chapter will examine the role of the large-scale specialized wholesale market in stimulating the innovation of local SME clusters with a case study of the specialized wholesale market in Yiwu, Zhejiang province, in East China.

9.3 Case study: Yiwu China commodities city and the innovation of local SME clusters As a city located in the middle of Zhejiang province, Yiwu occupies an area of 1105 square kilometres and has a population of 1.67 million. About 1 million of these are temporary rather than permanent residents and are there to do business. In 1988, Yiwu county changed into Yiwu city. Since China’s adoption of the open-door policy, Yiwu city has been making unremitting efforts to implement strategy for developing the city through promoting commerce and fostering a market system based on trade in ‘small commodities’, thereby promoting regional economic development and realizing a historic transition from a traditional agricultural county to a prosperous modern commercial city. Yiwu city has become the biggest distribution and exhibition centre for small commodities in China. The turnover of Yiwu China Commodities City Group Co., Ltd reached RMB381.18 million in 2008 (Yiwu Yearbook Leading Group of Editing Office) (Table 9.1). The growth of SME industrial clusters for socks, ornaments and zips has been closely linked to the rapid expansion and internationalization of Yiwu China Commodities City, the world’s largest specialized market. At present, Yiwu is the


The Innovation of SMEs

191

Table 9.1 The turnover of Yiwu China Commodities City Group Co., Ltd (RMB100 million) Year

Turnover

Percentage growth

2001 2002 2003 2004 2005 2006 2007 2008

212.0 230.0 248.3 266.9 288.5 315.0 348.4 381.8

9.8 8.5 8.0 7.5 8.1 9.2 10.6 9.6

Source: Yiwu Statistics Bureau (2009); Yiwu Statistics Yearbook 2008.

biggest sock-production base in China. The output of ornaments and zips has grown to more than 70 and 25 per cent of national output respectively (Jin et al., 2009). Yiwu’s municipal government has played an important part in the internationalization of the specialized wholesale market, leading to the innovation and upgrading of the local SME industrial clusters through its information diffusion and collective learning mechanisms. The specialized wholesale market has gone through three stages of development in the past 30 years, namely an agglomeration market, a wholesale market and an exhibition and trade market. The socks market is an important part of Yiwu China Commodities City. 9.3.1 An agglomeration market (1982–91) Like other wholesale markets in the small and medium-sized cities of China, Yiwu China Commodities City also became an agglomeration market from 1982 to 1991, during which the market was moved three times. It was expanded, the products upgraded and the environment improved each time with the number of booths increasing from 750 to 10,500 units. Before 1982, transactions were mainly conducted in the regular market for people in the city centre and suburbs. The commodities were mostly daily necessities and agricultural products. As in other small and medium-sized Chinese cities, the Regular market was one of the main trading places in Yiwu, with relatively few traders, commodities and sales volumes. However, the reach of the market was gradually strengthened during this stage. Its target consumers were mainly farmers and the residents of the city centre and suburbs, which


192

Jinmin Wang

amounted to about 200,000 people in 1982. In 1983, the market served about 3 million consumers in the rural area and neighbouring counties, 50 million people in Zhejiang province and some towns in Anhui, Fujian, Jiangxi and Jiangshu provinces. After the third move in 1987, the reputation of the market spread to the whole of East China and neighbouring provinces with about 200 million target consumers. After the fourth move in 1991, Yiwu China Commodities City targeted about 900 million farmers in China. In addition, the quality of commodities was gradually improved. The commodities were mainly necessities for farmers including thread, clothes, sewing needles, kitchenwares, etc. In addition, during the 1983–8 period, the specialized market mainly dealt in daily necessities, clothes, socks and stationery for state-owned enterprises (SOEs) throughout China. This boosted the cash flow of the SOEs. In 1988–91, the number of family workshops operating in the marketing booths of the specialized market increased considerably. Meanwhile, a lot of clothing factories were set up at Dachen town, Yiwu, which is known as the home of shirts in China. At that time the factories produced more than 500,000 shirts every day and one in seven shirts in China was made there (Zhejiang Province Zhengxie Historical Data Committee, 1997). Some specialized villages with family workshops also emerged, including, for example, handbag village, thread and belt villages. The family workshops producing socks, ornaments, toys, clothes, cosmetics, thread and belts expanded rapidly, creating some famous trademarks and brands at both the provincial and national levels. At this stage, the SMEs achieved technical innovation mainly by purchasing abandoned equipment from the SOEs and employing the SOEs’ at the weekends. This model of technology spillover was afterward spread to the rest of China. 9.3.2 The wholesale market (1992–2000) Yiwu China Commodities City had the following features during its second stage of development. First, the positioning of small commodities in the market became more diverse. The small commodities available on the specialized market were mainly from SMEs, which pushed the sales of their products through the wholesale market and trade fairs. They usually went to the wholesale market to look for one general agent or to set up a marketing centre on their own. Yiwu China Commodities City became their best choice. The marketing centre at Yiwu China Commodities City allowed penetration into the national market. The sales volume increased dramatically as many more enterprises entered



the specialized wholesale market in Yiwu, and at times the sales volume of one marketing centre in Yiwu would equal that of several in other cities. As a result, SMEs in Yiwu could develop and launch more new products onto the market so as to increase their profits, and many SMEs achieved great success after setting up one marketing centre in the specialized market. Second, the means of transaction improved greatly. Because clients bought more, it was impossible for them to collect all their purchases themselves. Most clients therefore made selections at the specialized market, issued one bill of lading, and then went to the warehouse or the factory of the seller to obtain the commodities, so transactions were mainly concluded through the combination of the specialized market and the producer’s warehouse. As long as clients sent their commodities to the specified shipping agent, they would be sent to the destination. If commodities were, full compensation would be made by the shipping agent. Third, the means of payment changed. With the increasing level of purchases, the clients needed to make bigger payments. It was risky to bring a lot of cash when travelling and therefore payment was mainly made through cheques, debit cards, credit cards or direct payment to the seller’s bank. Fourth, the market space increased constantly. As China Commodities City became famous around China, it aroused the attention of businessmen in the regional wholesale market, that is, the second-tier wholesalers. Small urban and rural shops, some shopping centres in large and medium-sized cities, businessmen in the second and third-tier wholesale markets mainly bought from the biggest wholesale market in China. Yiwu China Commodities City became the first-tier wholesale centre for small commodities in China. The number of businessmen from South-east Asia, the Middle East, Eastern Europe and East Asia also increased greatly year by year. At this stage, foreign businessmen started to play an important role as regards technological innovation among local SMEs. Foreign entrepreneurs would usually establish joint ventures with domestic firms or fully fund their own operations. Although they lack local embeddedness, most foreign enterprises enter Yiwu to take advantage of the complete production network and low-cost skilled labour to be found there. Therefore they usually establish industrial links with local firms. These foreign entrepreneurs have not only introduced modern management concepts and models, but have also improved the technological capability of clustered firms. In addition, the Yiwu municipal government created a fair and competitive environment through a reasonable division of labour, by classifying the industries into specific markets (Hua Hang Gui Shi) and by agglomerating commodities of the same type in


194

Jinmin Wang

one area (Feng Lei Ji Ju). This has contributed to information diffusion and technological learning among local sock firms and the competitive advantages of the sock cluster have been strengthened. 9.3.3 The exhibition and trade market (2002–present) The exhibition and trade market mainly involves the display of commodities with integrated negotiations, signing of contracts and conducting of e-commerce. Yiwu International Trade City, which first operated in 2002, formed an integrated business pattern with exhibition, negotiation, contracts and e-commerce under one roof, and created a new trading pattern for wholesale markets in China. The exhibition and trade market has combined all the advantages of a large-scale shopping centre, a wholesale market and an exhibition hall. When Yiwu International Trade City was planned in the first half of 2000, it was designed with the concepts of ‘modernization, internationalization, information, exhibition and humanization’. Yiwu International Trade City also serves as a unified commercial platform for SMEs and buyers, and the growth of productive enterprises in the industrial cluster is resulting in innovation and new products. There are four main marketing channels for production enterprises in the industrial clusters in Yiwu. 1) Producing components for large-scale enterprises. As the partner firms of one large-scale enterprise, clustered producers rely on the large enterprise for their survival and development. Small firms must mature with the business expansion of large-scale enterprises. If not, they would have to stop production or change their business. 2) Marketing components through the specialized wholesale market as the producer. There is a clear division of labour among these types of enterprises, who have formed the industrial cluster through producing different kinds of components intensively in one region. The final products will be completed through assembling different components. 3) Organizing production in response to orders from different kinds of exhibitions home and abroad. Production varies according to the volume of orders. 4) Creating one general agent or distributor at the wholesale market. Foreign businessmen from more than 200 countries and regions, mainly small and medium-sized buyers, each usually buy dozens of or even a hundred standard containers full of products every year. Some foreign businessmen who used to trade in the border cities of China or buy at the Guangzhou Foreign Trade Fair, asked clustered firms in China to produce commodities with their own brands. When they found that some commodities were from Yiwu, they came to the International Trade City to make their purchases at lower prices, to start



195

with. Since then, some foreign businessmen have invested in setting up overseas-funded enterprises. Because the labour-intensive industry in Yiwu has relatively mature manufacturing skills, industrial innovation mainly lies in the management, marketing and introduction of new technology and manufacturing equipment. At this stage, entrepreneurs in local SMEs have promoted technological progress mainly by introducing advanced technologies from developed countries. For example, in 2006, there were more than 1400 sock-making enterprises in Yiwu with 60,000 sockweaving machines, 40,000 of which were the latest models in the world (Yiwu Commmercial News, 2006). Langsha, MengNa, Fenli, Bonas and Chinehigh are the five largest sock-making enterprises in the world. They have made a substantial investment in introducing several thousand of the latest sock-making machines from developed countries. When it was first established, Langsha group decided to develop its own brand and aimed to introduce first-class equipment, attract first-class talent and become a first-class enterprise. With the introduction of advanced sockmaking equipment from Japan, Italy and the US, its daily production capacity for sock reached 1 million pairs. The MengNa Group has been keeping to the strategy of specialization and innovation. The group has made great investment in technological progress by introducing new technologies to make socks. In May 1994, the company invested US$2.3 million bringing in 78 sets of Lonati fine-gauge knitting machines from Italy and four sets of finishing-off equipment from Japan. A new workshop, which covers more than 5000 square metres, was also established. In 1995, annual production capacity reached 13 million pairs and the sales volume hit about US$5.2 million. In order to keep track of international markets, the company sent 39 engineering technicians to Italy for training in 2003 and 2004. In order to learn the latest developments in the technology and to expand its international business, it has also established sales offices in Italy, Germany, the UK, Japan, Russia, the US, etc. In addition to quota restrictions, sock exports from China face many other trade barriers including technical, environmental and labour standards. The MengNa group obtained Worldwide Responsible Apparel Production (WRAP) certification granted by Cal Safety Compliance Corporation (CSCC) in the US in 2003 and 2004. It passed detailed investigations of its equipment, production capacity, employee welfare, human rights and environmental facilities by leading retailing giants including Knart, Walmart, and a strict audit by the UK’s Marks and Spencer in 2005, which made it possible for MengNa socks to enter EU


196

Jinmin Wang

Traditional firms

Traditional firms

Informal sector agglomeration

Local government support for Specialized agglomeration Support management Location incentives

Clustered SMEs Figure 9.1 clusters.

Learning Competition

Industrial clusters in neighbouring regions

Clustered SMEs

The specialized wholesale market and the growth of local industrial

Source: Compiled by the author.

market. The leading firms have diffused the advanced technology to the SMEs in the cluster. Meanwhile, the specialized wholesale market has also promoted the rapid growth of neighbouring SME clusters in Zhejiang province (Figure 9.1). One typical example is the Datang socks cluster, which consists of 12 town-like rural settlements. Datang town is located in the centrewest of the Zhuji area, which is 53 kilometres from Yiwu. More than 4.8 billion pairs of socks are made in Datang every year. The annual output value reaches RMB8 billion. The socks industry contributes about 90 per cent of the gross domestic product (GDP) of the township of Datang and the majority of socks are marketed through the specialized wholesale market in Yiwu (Zhu, 2003). In addition, the local government has established industrial parks adjacent to the townships, specializing in products so as to encourage the SMEs to carry out technological innovation and expand their businesses in a better investment environment. Many clustered firms which have accumulated the necessary capital are eager to undertake technological improvement and business expansion. The economic development zone not only solves the issue of land use for clustered firms but also offers a series of preferential policies including favourable land prices, tax breaks, etc. In addition, infrastructure in the state-level and provincial-level economic (technological) development zones and



the export-processing zones is fully constructed and capable of meeting the requirements of a variety of high- and new-tech industrial projects. Fierce market competition has accelerated the transition of traditional industrial clusters to scientific industrial clusters. The emergence and growth of scientific enterprises within the clusters have led to great improvements in the regional innovative capability. As the SMEs have strengthened their innovative capacities in science and technology, the pace of industrialization has accelerated.

9.4 Horizontal cooperation and technological innovation of SMEs According to Schmitz and Nadvi (1999), joint actions within an industrial cluster consist of horizontal cooperation among competitive firms and vertical integration along the industrial chain. Active horizontal cooperation has contributed to technological innovation within local SME industrial clusters. Inadequate innovation incentives among local SMEs can be overcome as long as horizontal cooperation and technological innovation have been integrated closely by the local specialized wholesale market and informal social network. Horizontal cooperation within the local SME cluster not only lies in cooperative research and development, but also covers subcontracting, financing, information flows, leasing of machines, joint marketing, product development and cooperative innovation, joint training programmes, etc. A prerequisite for incentive compatibility in order to achieve horizontal cooperation is the formation of a common belief that an enterprise will be driven out of the horizontal alliance forever if it fails to participate effectively. This common belief plays an important role in coordinating the activities of clustered firms and in helping the clustered firms to form their opinions. If a clustered firm is found not to be cooperating seriously, this information will be transmitted rapidly within the SME industrial clusters through the specialised wholesale market and informal social network. The establishment of horizontal cooperation within the local SME industrial cluster is closely related to the local social network. The SMEs in the local industrial clusters can access the latest developments of their rivals through the entrepreneurs, managers and employees within their social network. Innovation is one dynamic collective learning process for the local SMEs that make up the different kinds of network. Informal social networks help to transmit technological information and knowledge within the cluster and thus to diffuse innovation among


198

Jinmin Wang

its entrepreneurs. Innovation by entrepreneurs can be transmitted from the first-tier subcontractors to the lowest-tier manufacturing firms and thus upgrade the manufacturing level and innovation capacity of the whole cluster. When manufacturing new products, entrepreneurs usually send technicians to offer technical guidance and undertake the monitoring of product quality. Since entrepreneurs make demands on subcontractors for more advanced technology and processing capacity, competition among subcontractors has become fiercer. Those subcontractors with a high innovative capacity usually have more opportunities. As a result, entrepreneurs are able to learn more about the latest innovations from subcontractors and have more innovative resources. The private network of entrepreneurs is another channel through which innovation can be diffused. In general, the entrepreneurs who manufacture homogeneous products and push sales in the same regional market communicate very little because of fierce competition with each other. However, those entrepreneurs who are within the same value chain but target different segments usually keep in close contact with each other. The entrepreneurs within the local SME clusters in Yiwu have woven an enormous network of direct and indirect links. Under most circumstances, innovation by these entrepreneurs has been diffused through social networks with them learning some of the latest information from their competitors. Although network effects might lead to free-riding behaviour, mature and innovative entrepreneurs become motivated to improve the competitive advantages of their firms when they face fierce competition from their counterparts.

9.5 Conclusion The growth and development of SME industrial clusters in Yiwu has shown that competitiveness of industrial clusters is subject to the innovative capability of SMEs. It is a typical example of a spontaneous relationship between market expansion, technological innovation and regional economic development. The positive interaction between market expansion, the growth of SMEs and the development of industrial clusters has led to Yiwu’s unique economic development. Unlike the development paths of other areas in China that began with rural industrialization, Yiwu’s economic growth started with the specialized wholesale market and prompted local economic development through the Chinese commodity trade. The growth in scale of the market, the evolution of its structure, the extension of its operations along the production and sales chain and the promotion of regional economic



199

development through the agglomeration of firms and the spread of its reach have generated a series of institutional changes and technological innovations. Research has shown that innovation among SMEs has mainly been realized by collective learning through the local specialized wholesale market. This institution has enhanced information flow among SMEs and offered an important trading platform for manufacturers, retailers and consumers. The integration of physical and virtual markets has further improved the competitiveness of SMEs in the province. The development path of Yiwu exhibits strong market openness, first regionally then internationally, making it possible for local SMEs to absorb international advanced technology. Active horizontal cooperation is one of the main reasons for technological innovation within SME industrial clusters in Yiwu. The strategic linkage between horizontal cooperation and technological innovation, with the intermediation of the local specialized wholesale market and informal social network, can lead to a special institutional arrangement that can overcome inadequate incentives for innovation among local SMEs and contribute to the upgrading of SME industrial clusters on the global value chain. The clustered firms choose to make innovation and conduct horizontal cooperation through the medium of the specialized wholesale market. The common belief and strategy of maintaining the level playing field are self-fulfilling. Although there is fierce competition and widespread imitation among clustered firms in China, entrepreneurs within the Yiwu sock cluster are still making substantial investments in technological innovation because their enterprises could easily be excluded from the horizontal cooperative alliance within the cluster if they broke the rules of the game, even if they could initially make gains through. The existence of social networks on the basis of blood and regional ties has made the threat of exclusion from horizontal cooperation credible.

References Advisory Committee on Measuring Innovation in the 21st Century (2008), ‘Measuring innovation in the 21st century economy’. http//www.innovation metrics.gov (accessed on 16 August 2008). Arundel, A., K. Smith, P. Patel and G. Sirilli (1998), ‘The future of innovation in Europe: Concepts, problems and practical directions’, IDEA (Indicators and Data for European Analysis) Report 3. Camagni, R. (ed.) (1991), Innovative Networks: Spatial Perspectives. London and New York: Belhaven Press. Conference Board of Canada (2004), Exploring Canada’s Innovation Character: Benchmarking Against Global Best. Ottawa: The Conference Board of Canada.


200 Jinmin Wang Dijk, M. and Q. Wang (2005), ‘Cluster governance in an emerging city-wide ICT cluster in Nanjing, China’, in E. Giuliani, R. Rabellotti and M. Dijk (eds) Clusters Facing Competition: The Importance of External Linkages. Aldershot, Ashgate, 215–31. Engel, P. (1997), The Social Organisation of Innovation: A Focus on Stakeholder Interaction. Amsterdam: KIT Press. Ernesto P., L. Gustavo and P. Fernando (2005), ‘Innovation in developing countries: Characteristics and measurement priorities’, UNU-INTECH Technology Policy Briefs, 4 (1). Humphrey, J. and S. Humphrey (2002), ‘How does insertion in global value chains affect upgarading in industrial clusters?’ Regional Studies, 36(9), 1017–27. Jin, X.R., X.W. Zhu, Z.Y. You and S. Zhang (2009), ‘The industrial clusters in Yiwu and Huangyan Districts’. http://www.ide.go.jp/English/Publish/Download/ Jrp/pdf/144_2.pdf (accessed on 1 August 2009). Li, J. (2006), ‘One study on the strategy of cluster upgrading under the background of international industrial relocation – The case of IT industrial cluster’, in The Fifth International Conference on Industrial Clustering and Regional Development,Beijing, 2006. Beijing: Peking University. Liu, S. and D. Song (2006), ‘From path dependence to path creation: Path transition theory and empirical studies of clusters’, in The Fifth International Conference on Industrial Clustering and Regional Development, Beijing, 2006. Beijing: Peking University. Nelson, R. (1993), National Innovation Systems: A Comparative Analysis. Oxford, Oxford University Press. Porter, M. (1990), The Competitive Advantage of Nations. New York: Free Press. Romer, P. (1990), ‘Endogenous technological change’, Journal of Political Economy, 98, 71–102. Saxenian, A (1994), Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge: Harvard University Press. Schmitz, H. and K. Nadvi (1999), ‘Clustering and industrialisation: introduction’, World Development, 27 (9), 1503–14. Scott, A. (1992), ‘The collective order of flexible production agglomerations: Lessons for local economic development policy and strategic choice’, Economic Geography, 68 (3), 219–33. Yang, W. (2008), ‘The development of Small and Medium-sized Enterprises in China’, China Comment, 5. http://news.xinhuanet.com/banyt/2008-03/26/ content_7863635.htm (Accessed 8 October 2008). Yiwu Commercial News (2006), ‘Yiwu awarded China famous socks city and China famous knitting city. (14 January), 1. Yiwu Statistics Bureau (2009), Yiwu Statistics Yearbook 2008. Yiwu: Yiwu Statistics Bureau. Zhejiang Province Zhengxie Historical Data Committee (ZPZHD Committee) (ed.) (1997), Small Commodities, Big Market: The Memoirs of the ‘Founders of Yiwu China Commodity City. Hangzhou: Zhejiang People’s Press. Zhu, H. (2003), Industrial Clusters in Zhejiang Province: Industrial Network, Growth Path and Development Dynamics. Hangzhou: Zhejiang University Press.


Part IV Foreign Direct Investment and Technology Transfer




10 FDI, R&D and Innovation Output in the Chinese Automobile Industry Chen Fang and Pierre Mohnen

10.1

Introduction

After joining the World Trade Organization (WTO) China experienced a major inflow of Foreign Direct Investment (FDI). Many famous automobile firms from developed countries decided to invest in China in order to cooperate with domestic firms. The question is whether FDI benefited the development of the Chinese automobile industry. On the one hand, foreign investors brought with them new technologies, methods of management and worldwide network linkages, but on the other they might have crowded out local producers of complete goods and components in the automobile industry. The Chinese automobile industry grew rapidly after 2006 and the country became the third biggest producer in the world after the US and Japan. But to ensure that this growth is sustained, it is necessary for China to keep innovating. If China merely completes the process of introducing, assimilating and imitating, it is likely to remain a junior partner to the major world players. To go on innovating China needs to invest in R&D. Foreign Direct Investment (FDI) could play an important role in this regard. But it is a priori unclear whether the effect of FDI on R&D and innovation is stimulating. It is often said that China does not build up its own innovation ability in the declining automobile market and that it simply imports technologies without developing the ability to innovate on its own. Instead, one of the primary motivations for China should be to attract FDI from developed countries so as to obtain advanced technology on which to build its own innovation capability. Given the inconclusiveness of the role of FDI on R&D and 203


204

Chen Fang and Pierre Mohnen

innovation so far, we revisit this issue using firm-level data collected by the State Statistical Bureau of China. The chapter is organized as follows. In Section 10.2 we review the evidence on the effects of FDI in China. In Section 10.3 we lay out the model used in this chapter. In Section 10.4 we describe the data used. In Section 10.5 we analyse the results of the estimation and in Section 10.6 we summarize and conclude.

10.2 Review of the literature The presence of foreign multinational enterprises (MNEs) may exert a significant influence on the host country’s own innovation. However, different theoretical models and empirical studies come up with different conclusions regarding the relationship between FDI and R&D. FDI can increase or decrease technological innovation depending on the specific context. Those in favour of FDI argue that when attracting major FDI, developing countries are ready to give up part of their domestic market n order to improve their domestic firms’ competitive advantage, because FDI can bring in direct or indirect technology transfer. In order to compete with the overseas-funded firms domestic enterprises need to improve or update their production, management and marketing techniques. Firms have to increase R&D inputs in order to raise their technology level. Those against FDI argue that the FDI competes with domestic enterprises, decreases their profits and may even drive some of them out of the market. Furthermore, domestic enterprises may not have sufficient technological capacity to innovate or carry out R&D on their own. Kokko (1994) pointed to a positive effect of FDI if the difference in technology between the MNEs and the host countries is not very large. In recent years, the relationship between FDI and domestic innovation in China has received a lot of attention. The conclusions are mixed. Chen (2001), Huang (2003) and He (2000) find that FDI had a negative effect. Jiang and Xia (2005), Xi and Yan (2005) and Wang et al. (2006) find that FDI helped China to improve its technological innovation capability. In the Chinese electronics industry, Hu and Jefferson (2002) find a significant drop in productivity rather than a positive spillover effect of FDI on domestic firms. Fu (2008) investigates the impact of FDI on the development of regional innovation capabilities using a panel dataset of Chinese firms. She finds that FDI had a significant positive impact on the overall regional innovation capacity and on innovation efficiency in the host region. She concludes that the type and quality of FDI inflows, as well as the strength of both local


FDI, R&D and Innovation Output

205

absorptive capacity and complementary assets in the host regions, are crucial for FDI to serve as a driver of knowledge-based development. Fu and Gong (2008) explore the drivers of technology upgrading in emerging economies using Chinese firm-level panel data from the 2001–5 period. The R&D activities of foreign firms exert a significant depressive effect on technical change by local firms over the sample period. It is domestic R&D activities at the industry level that push Chinese firms to the technology frontier. A similar conclusion is reached by Huang and Sharif (2009), who find that FDI from Hong Kong, Macao and Taiwan did not raise the productivity of firms in Guangdong province and that these foreign-funded firms perform less R&D and are less innovative than domestic firms.

10.3 Econometric model We measure and try to identify the determinants of innovation in the Chinese automobile industry, on the input side by way of R&D expenditures, and on the output side by way of the share in total sales resulting from new or substantially improved products. The input indicator can be seen as a predictor of future innovation and the output indicator as an indicator of past innovation efforts. In the literature on the knowledge-production function, the output measure of knowledge or innovation is generally the number of patents (see Griliches, 1990 for a review). With the advent of the Oslo Manual (OECD, 1992) innovation surveys were launched in many countries containing a new measure of innovation output, the share in total sales resulting in new or subtantially improved products (also sometimes called the share of innovative sales). We shall use this measure in our analysis (for a comparison of the two measures of innovation output, see Crépon et al., 1998). Given the large number of zero observations for both R&D and the share of innovative sales, we estimate a generalized tobit model, also known as tobit type II model. This model estimates both the propensity to carry out R&D, resp. innovate, and, in the case of R&D performers, resp. innovators, the R&D, resp. innovation, intensity. The generalized tobit model consists of two parts. A first equation determines the level of the latent variable (y1i* ) y1i* = x1ib1 + b1i

(1)

that selects the R&D performers (resp. innovators) (y1i = 1) and the nonR&D performers (resp.) non-innovators (y1i = 0) depending on whether


206 Chen Fang and Pierre Mohnen

its value falls above or below a given threshold, that in all generality we can set as equal to zero = 1 y1i =  = 0

if if

y1*i > 0 y1*i ≤ 0 .

A second equation determines the level of a latent variable (y*2i ) that corresponds to the observed intensity of R&D (resp. innovation) (y2i) for R&D performers (resp. innovators) and is equal to zero for non R&D performers (resp. non-innovators) y*2i = x2ib2 + u2i

(2)

with = y * ......if ...y1*i > 0 . y2 i =  2 i *  = 0........if ...y1i ≤ 0 xki (k = 1,2) are the explanatory variables in both equations, bi their respective coefficients and u1i and u2i are the error terms in both equations that are assumed to follow a bivariate normal distribution with correlation coefficient and standard errors 1 (for reasons of identification) and 2 respectively.1 We estimate the model by maximum likelihood (see Mohnen et al., 2007). To say it more concretely, we observe and try to explain whether a firm has R&D expenditure and if so, how much R&D it does. Likewise we look at whether a firm has introduced new products and, if so, what the share in total sales resulting from the new products is. The first explanatory variable that we shall consider is firm ownership. Existing studies are inconclusive as to whether or not the nationality of ownership of a firm has an impact on its R&D. Caves et al. (1980: 193) suggest that foreign activity reduces the rate of R&D activity in Canada. Haddad and Harrison (1993), based on Moroccan companylevel data, prove that FDI involving higher technology will not necessarily raise domestic R&D capacity. Aitken and Harrison (1999), based on firm-level panel data from Venezuela, found that the impact of FDI on R&D by domestic enterprises is negative. We classify firms into three groups: domestic-owned firms, foreign-funded firms and firms from Hong Kong, Macao and Taiwan (HMT). Domestic-funded firms are the



207

reference group and dummy variables are created for the other two groups. The intensity of foreign ownership is measured by the intensity of FDI capital over total capital. Innovation is also postulated to be a function of firm size. According to Schumpeter’s hypothesis we expect large firms to be more innovative than small firms, because large firms have easier access to finance, can spread the fixed costs of innovation over a larger volume of sales and may benefit from economies of scale and complementarities between R&D and other manufacturing activities. Firm size is measured by the total number of employees. In the selection equation, instead of using the continuous variable for size we classify firms into three groups: a firm is a large-scale firm if the number of employees is greater than or equal to 2000, middle-scale if the number of employees is between 300 and 2000, and small-scale otherwise. The small-scale firm is the reference group. There is probably a strong link between R&D, which can be seen as an innovation input, and the share of innovative sales as a measure of innovation output. The endogeneity of R&D should be accounted for (as emphasized in Crépon et al., 1998). The predicted incidence of R&D will be used in the equation on the incidence of innovation output and the predicted intensity of R&D in the equation on the intensity of innovation output. The standard errors of the estimates will be corrected for the fact that the R&D variables are generated regressors. Schumpeter’s hypothesis is also sometimes cast in terms of market power. Firms with a large market share are more innovative than those with a smaller market share because they have more to lose by not innovating. We measure market share by the firm’s sales as a percentage of the total sales in the four-digit industry to which it belongs. Of course, this measure only captures the domestic market share, but in the case of the Chinese automobile industry this seems to be the relevant market share to consider. Since the innovative environment differs across industries and space, we also allowed innovative activity to vary with industries and geographical location. The reference industry is the car maintenance and repair sub-industry, and binary variables corresponding to each of other five categories (cars, vehicle rebuilding, trams, bodies and trailers and parts and accessories) have been constructed. The geographical reference area is the West and binary variables have been constructed for the Middle and the East.


Table 10.1

Explanatory variables introduced in the two generalized tobit models R&D

Variables

Explanation of variables

D-R&D

Dummy equal to 1 if the firm has R&D inputs

R&D D-FDIF

R&D inputs/total capital Dummy equal to 1 if FDI is more than 25%

D-HMTDIF

Dummy equal to 1 if HMTFDI is more than 25%

FDI* HMTFDI* L-SIZE

Foreign capital/total capital HMT capital/total capital Dummy equal to 1 if the number of employees is greater than or equal to 2000 Dummy equal to 1 if the number of employees is between 300 and 1999 Number of employees Sales/total sales in the same (4-digital code) industry

M-SIZE SIZE MSHARE

Propensity – –

New products

Intensity

Propensity

– –

* –

Intensity – *

*

–

*

–

* – – *

– * * –

* – – *

– * * –

*

–

*

–

–

*

–

–

*

*

*

*

D-IND

Dummies for the sub-industries (cars, vehicle rebuilding, trams, bodies and trailers, parts and accessories), the repair and maintenance sub-industry being the reference group

*

*

*

*

D-AREA

Dummies for geographical areas (East and Middle), West being the reference group

*

*

*

*

Note: * Zero values are replaced by 0.00001 when taking logs.



209

10.4 Data and descriptive statistics The chapter uses data of the Chinese automobile industry from 2002 to 2006. The year 2004 is excluded because of missing data. To minimize the influence of outliers, we cleaned the data by excluding firms that at one point in the sample had sales of less than RMB5 million, or non-positive data on the number of employees, assets, sales, costs and salaries (incl. social welfare benefits). Further, we restricted ourselves to firms with positive value-added, more than ten employees and less than a 50 per cent share of R&D in total sales. Finally, we eliminated several firms from Xizang province. After the process of cleaning the data, the sample encompassed 3244 firms in 2002 and up to 6795 in 2006 (see Table 10.2). As Table 10.3 shows, our sample is made up of roughly 80 per cent domestic-owned firms and to 10º15 per cent foreign-funded firms other than firms from HMT. The percentage of firms with new products and the share of new products in total sales were always the highest in foreign-funded firms and the smallest in HMT firms.2 The percentage of firms with R&D activities was always higher in foreign-funded firms than in domestic or HMT firms, but the intensity of R&D was lower in foreign-funded firms. Their R&D/sales ratio was on average lower than 1.5. Thus it already appears from a cursory glance at Table 10.3 that foreign-funded firms tend to be more innovative in new products and rely on R&D conducted abroad. HMT firms, while always less innovative in new products than domestic firms, are sometimes more R&D-intensive than domestic firms.

Table 10.2 Data cleaning: Chinese automobile industry, firm-level data, 2002–3, 2005–6 Process of cleaning data

2002

2003

2005

2006

Number of firms in the original database Less than RMB 5 million in sales Non-positive number of employees, total assets, sales, costs (incl. administration costs), salaries (incl. social welfare benefits) Non-positive value-added, more than ten employees, less than 50 per cent share of R&D in total sales Minus Xizang province (Tibet)

4632 3675 3323

5182 4454 4064

7371 6918 6270

8233 7820 6974

3244

4000

6114

6796

3244

4000

6112

6795


Table 10.3

Innovation indicators by type of ownership, Chinese automobileindustry, firm-level data, 2002–3 and 2005–6

Distribution (%)

Firms with new products (%)

Share of new products in total sales (%)

Share of new products in total sales in firms with new products (%)

Firms with R&D (%)

R&D/ sales in firms R&D/sales with (%) R&D (%)

Year

Funding

Number of firms

2002

Domestic HMT Foreign Total

2674 227 343 3244

82.43 7.00 10.57 100

16.83 9.25 19.83 16.62

23.27 11.29 50.78 30.55

49.08 31.39 68.92 56.20

26.55 22.47 41.40 27.84

1.25 0.76 0.88 1.12

1.64 1.08 1.04 1.43

2003


3278 267 455 4000

81.95 6.68 11.38 100

15.77 9.74 19.34 15.78

24.39 23.45 52.02 36.37

50.64 38.46 77.57 63.80

26.24 24.34 39.78 27.65

0.71 0.83 0.70 0.71

0.94 1.21 0.85 0.91

2005


4877 394 841 6112

79.79 6.45 13.76 100

15.32 14.21 20.69 15.98

27.26 21.28 43.42 34.03

47.12 44.92 71.91 58.15

16.69 18.27 31.39 18.82

1.04 1.10 0.93 0.99

1.60 2.12 1.26 1.45

2006


5352 446 997 6795

78.76 6.56 14.67 100

16.01 13.45 21.97 16.72

26.80 15.92 46.54 36.34

45.57 38.64 68.37 57.85

18.14 18.83 30.39 19.99

1.14 0.72 0.79 0.95

1.71 1.46 1.08 1.36

Note: HMT: Hong-Kong, Macao, Taiwan.


Table 10.4

Innovation indicators by industry, Chinese automobile industry, firm-level data, 2002–3 and 2005–6

Number of firms

Distribution (%)

Share of new Share of new products in Firms products in total sales in with new total sales firms with new Firms with R&D/ products (%) (%) products (%) R&D (%) sales (%)

R&D/ sales in firms with R&D (%)

Year

Funding

2002

Cars Vehicle rebuilding Trams Bodies/trailers Parts/accessories Maintenance/repair

161 230 2 88 2396 367

4.96 7.09 0.06 2.71 73.86 11.31

42.86 27.83 50.00 11.36 16.49 0.00

45.92 23.95 64.87 13.38 10.79 0

63.63 50.33 99.62 68.18 33.02 0

59.63 50.43 50.00 20.45 27.05 6.54

1.67 0.51 0.08 0.10 0.53 0.05

1.73 0.59 0.12 0.28 1.04 0.26

2003


180 291 5 107 3012 405

4.50 7.28 0.13 2.68 75.30 10.13

46.67 27.15 20.00 14.02 14.94 0.49

51.16 30.24 14.30 12.24 10.79 0.39

72.86 45.23 90.39 55.40 34.29 19.90

64.44 45.36 0 21.50 27.12 4.44

0.83 0.31 0 0.22 0.63 0.03

0.88 0.38 0 0.58 1.28 0.18

2005


212 363 10 169 4867 491

3.47 5.94 0.16 2.77 79.63 8.03

48.11 30.03 10.00 12.43 14.92 3.67

52.85 26.72 2.19 9.29 11.69 0.36

66.34 45.44 43.12 35.72 35.86 12.54

59.43 36.09 10.00 11.24 17.77 1.63

1.46 0.70 0.59 0.23 0.48 0.02

1.63 1.08 1.48 1.04 1.06 1.10

2006


224 367 10 188 5517 489

3.30 5.40 0.15 2.77 81.19 7.20

50.00 28.88 10.00 9.57 16.01 3.27

54.58 26.94 0.64 19.10 10.58 0.29

63.60 42.04 15.16 55.12 36.81 6.53

62.05 41.96 10.00 14.89 18.63 1.64

1.22 0.65 3.94 0.16 0.63 0

1.35 0.88 30.51 0.67 1.54 0.19

Note: The six sub-industries correspond to the following codes in the 2002 Chinese industrial classification (GB/T 4754-2002): cars (3721), vehicle rebuilding (3722), trams (3723), trailers (3724), parts and accessories (3725) and repair and maintenance (3726). 10.1057/9780230276123 - The Rise of Technological Power in the South, Edited by Xiaolan Fu and Luc Soete

212


As Table 10.4 shows, most of the firms in our sample produce car accessories (around 75–80 per cent). The percentage of firms with new products and with R&D inputs is the highest in the cars sub-industry. If we exclude the tram sub-industry, which has very few firms in our sample, the percentage of innovators and their innovation intensity, both in terms of new products and in terms of R&D, is substantially smaller in the other subsectors. Many firms in vehicle rebuilding are R&D performers, but with a relatively lower R&D intensity compared to, for example, the firms that produce parts and accessories, which are sometimes even more R&D-intensive than the car manufacturers. We examine the growth of the Chinese automobile industry after China joined the WTO by distinguishing the 2002–3 period, when growth was rapid, and the more stable 2005–6 period. We chose firms that were present in both 2002 and 2003 and 2005 and 2006. There were 2462 firms present in both 2002 and 2003, of which 1991 were domestic-funded and 289 were foreign-funded. Firms present in both 2005 and 2006 numbered 5097, of which 4027 were domestic-funded and 711 were foreign-funded. Table 10.5 shows that the percentage of firms with R&D inputs is higher in the groups of large firms and foreign-funded firms. On the whole, the percentage of firms with R&D decreased between 2002 and 2005. Table 10.6 indicates that the R&D intensity and the shares of foreign capital and of HMT capital in total capital increased between 2002 and 2005 for firms with R&D inputs.

Table 10.5

R&D propensity 2002

2005

Number of Percentage firms with of firms R&D with R&D

Number of firms with R&D

Percentage of firms with R&D

Funding Domestic-funded HMT-funded Foreign-funded

556 42 129

27.93 23.08 44.64

700 62 242

17.38 17.27 34.04

Scale

Large Middle Small

56 313 358

83.58 49.53 20.31

87 407 510

80.56 45.47 12.46

Total

727

29.53

1004

19.70

Note: HMT: Hong-Kong, Macao, Taiwan.


FDI, R&D and Innovation Output Table 10.6

213

R&D intensity for firms with R&D

Year Variables

Mean

Stan. dev.

25%

50%

75%

95%

2002 R&D FDI HMTFDI Log(SIZE) MSHARE

0.009 0.096 0.037 5.837 0.004

0.014 0.231 0.153 1.243 0.016

0.002 0 0 4.898 0.000

0.004 0 0 5.724 0.001

0.011 0 0 6.718 0.002

0.034 0.600 0.342 7.847 0.020

2005 R&D FDI HMTFDI Log(SIZE) MSHARE

0.013 0.155 0.045 5.794 0.003

0.021 0.308 0.177 1.243 0.017

0.002 0 0 4.875 0.000

0.005 0 0 5.677 0.000

0.016 0 0 6.564 0.001

0.053 1.000 0.488 7.919 0.009

Table 10.7 Innovation propensity 2003

2006

Number Percentage Number Percentage of firms of firms of firms of firms with new with new with new with new products products products products Funding

Domestic-funded HMT-funded Foreign-funded

371 19 62

18.63 10.44 21.45

677 54 160

16.81 15.04 22.50

Scale

Large Middle Small

54 236 162

68.35 34.96 9.48

79 370 442

72.48 38.14 11.00

Carrying out R&D

No Yes

144 308

8.55 39.59

369 522

9.31 45.99

452

18.36

891

17.48

Total

Table 10.7 reveals that there is clearly a higher percentage of firms with new products among the foreign-funded and large firms, and those with R&D. Table 10.8 shows that between 2003 and 2006 the share in total sales of new products decreased from 39.2 to 37.1 per cent, the R&D intensity in the preceding year increased from 0.6 to 0.9 per cent, the FDI share in total capital climbed from 7.4 to 10 per cent and the HMT-sourced FDI in total capital went up from 1.5 to 4 per cent. The compared averages in Tables 10.4 to 10.7 do not correspond to the same firms and are therefore only indicative of changes over time. To disentangle the various determinants of the propensity


214


Table 10.8

Innovation intensity for firms with new products

Year Variables

Mean

Stan. dev.

25%

50%

75%

95%

2003 NEWP R&D -1 FDI HMTDI SIZE

0.392 0.006 0.074 0.015 1034

0.286 0.012 0.195 0.082 1967

0.142 0 0 0 196

0.347 0 0 0 447

0.601 0.007 0 0 1037

1 0.029 0.503 0.000 3847

2006 NEWP R&D -1 FDI HMTDI SIZE

0.371 0.009 0.100 0.040 936

0.304 0.019 0.240 0.163 3578

0.104 0 0 0 125

0.279 0.000 0 0 300

0.584 0.009 0 0 749

1 0.047 0.663 0.342 2952

Note: R&D -1 refers to 2002 (resp. 2005).

and the intensity of doing R&D and of innovating in products we now revert to a multivariate analysis. What is also clear from these tables is that the sample means are often above the medians and influenced by some extreme values.

10.5 Results The innovation input and output models that simultaneously carry out the innovation propensity and intensity equations have been estimated separately for the rapid growth period of 2002–3 and the more stable period of 2005–6. Just after China joined the WTO, its automobile industry grew rapidly, driven by a high domestic demand. After 2004, the development slowed down and development problems arose. Thus the industry had a different pace of development in 2002–3 compared to 2005–6. Therefore the two periods are calculated separately, but in each pair of years the data are pooled. Table 10.9 reveals that the propensity to engage in R&D increases with size. Medium-sized firms have a higher propensity than small firms, and large firms have an even higher propensity than medium-sized firms. Firms with a higher market share have also a higher propensity to be R&D performers. These effects are significant for both time periods. FDI firms were not significantly more likely to be R&D performers than domestic-funded firms in 2002–3. In 2005–6, however, their effect is significant. The propensity of FDI firms to do R&D is 1.7 percentage points higher. HMT-funded firms were never significantly more likely to be R&D performers than domestic-funded firms. The intensity of R&D


Table 10.9

Generalized tobit estimation of R&D efforts in the Chinese automobile industry 2002–3 pooled data Propensity

Coefficient P-value L-SIZE M-SIZE SIZE (in logs) MSHARE (in logs) DFDI DHMTFDI FDI (in logs) HMTFDI (in logs) Intercept (Std. err.) 2 (Std. err.) Number of observations

2005–6 pooled data

Intensity log(R&D/sales) Margin. effect Coefficient P-value

Propensity

Margin. effect* Coefficient

P-value

Intensity log(R&D/sales) Margin. Effect

Coefficient P-value

Margin. effect*

0.431 0.316 –

0.001 0 –

0.119 0.087 –

– – –0.431

– – 0

– – –0.431

0.493 0.411 –

0 0 –

0.107 0.089 –

– – –0.222

– – 0

– – –0.222

0.295

0

0.081

0.645

0

0.499

0.316

0

0.069

0.394

0

0.374

0.060 0.015 –

0.294 0.836 –

0.016 0.004 –

– – –0.126

– – 0

– – –0.126

0.080 –0.011 –

0.053 0.857 –

0.017 –0.002 –

– – –0.079

– – 0

– – –0.079

–

–

–

–0.089

0

–0.089

–

–

–

–0.050

0

–0.050

0.424 – – 7244

0.005 – – –

– – – –

–2.048 0.386 1.953 2009

0.001 (0.087) (0.058) –

– – – –

0.125 – – 12907

0.4 – – –

– – – –

–3.157 0.051 1.891 2508

0 (0.117) (0.028) –

– – – –

Note: Sub-industry, region and year dummies are controlled for but not reported. *Marginal effect conditional on doing R&D.


Table 10.10

Generalized tobit estimation of innovation in new products in the Chinese automotive industry 2002–3 pooled data Propensity Coefficient P-value

Predicted R&D probability Predicted log of R&D intensity L-SIZE M-SIZE MSHARE (in logs) DFDI DHMTFDI FDI (in logs) HMTFDI (in logs) Intercept ρ (Std. err.) σ2 (Std. err.) Number of observations

2005–6 pooled data

Intensity: log(share in sales of new-to-firm products) Margin. effect

Coefficient P-value

Margin. effect*

Intensity: log(share in sales of new-to-firm products)

Propensity Coefficient P-value

Margin. Effect Coefficient P-value

Margin. effect*

0.197 – 0.725 0.498 0.148

0.776 – 0.029 0.028 0.468

0.039 – 0.143 0.098 0.029

–0.433 – – –0.105

– 0.001 – – 0.190

– 0.433 – – –0.546

0.665 – 0.602 0.232 –0.041

0.058 – 0.004 0.125 0.710

0.141 – 0.128 0.049 –0.009

– 0.282 – – 0.786

– 0.062 – – 0.034

– 0.282 – – –0.125

-0.161 -0.342 –

0.042 0.001 –

–0.032 –0.068 –

– – 0.060

– – 0.003

– – 0.060

–0.089 –0.054 –

0.091 0.373 –

–0.019 –0.011 –

– – 0.033

– – 0.006

– – 0.031

–

–

–

0.019

0.378

0.019

–

–

–

0.115

0.017

0.021

-1.422 – – 7244

0.001 – – –

– – – –

0.192 0.038 1.317 1170

0.880 (0.131) (0.028) –

– – – –

–0.552 – – 12,907

0 – – –

– – – –

0.601 –0.041 1.370 2113

0.994 (0.135) (0.022) –

– – – –

Note: Sub-industry, region and year dummies are controlled for but not reported. * Marginal effect conditional on being innovative.



217

decreases with firm size, whereas it increases with market share. What is interesting to notice is that the elasticity of R&D/sales with respect to the percentage, that is, if foreign ownership increases by 10 per cent, R&D/ sales decreases by 1 per cent. The elasticity is somewhat lower for HMTfunded capital, and somewhat lower in 2005–6 than in 2002–3. The determinants of innovation output measured by the share in total sales resulting from new products are reported in Table 10.10. We have included the same set of explanatory variables as for the explanation of R&D, except for three differences. First, R&D as an innovation input is naturally introduced as a determinant of innovation output, recognizing the endogeneity of R&D. This is done by introducing the estimated probability of doing R&D into the probability of innovation output equation and the estimated intensity of innovation into the innovation output intensity equation.3 Time dummies were removed in both equations because they were not significant and size was removed from the innovation output intensity equation after a likelihood-ratio test showed that including size did not significantly increase the likelihood. Actually, since time and size are already included as regressors in the R&D equations they enter indirectly as explanatory variables in the innovation output equations. During the two periods that we examine, R&D seems to have had a positive effect on innovation output. Firms performing R&D were 4 per cent more likely to introduce new products in 2002–3, a probability that increased to 14 per cent in 2005–6. The elasticity of the share of innovative sales with respect to R&D intensity was around 0.4 in the first subperiod and 0.3 in the second. Obviously, large- and middle-scale firms are more likely to introduce new products into the market than smallscale ones. The Schumpeterian market-share argument is not confirmed for the Chinese automobile industry. In 2002–3 foreign-controlled firms, especially HMT-controlled ones, were less likely to innovate than domestically controlled firms. The effect of foreign ownership on the likelihood to innovate was no longer significant in 2005–6. But for those that innovated in new products, the share of innovative sales increased with foreign ownership. If foreign ownership doubled, the share of sales resulting from products new to the firm increased by 3 to 6 per cent, a figure that was closer to 2 per cent for HMT-owned firms.4

10.6 Conclusion Since China joined the WTO its vehicle producers have kept innovating by introducing new products onto the market, but the percentage of


218 Chen Fang and Pierre Mohnen

R&D performers, at least in our the sample, has been decreasing. At the same time the proportion of foreign-funded firms has been increasing, especially from sources other than Hong Kong, Macao and Taiwan. The question we have been investigating in this chapter is whether R&D and innovation output differ for domestic-funded and foreign-funded firms. It does not appear from our multivariate analysis that FDI firms are more prone to be R&D performers than the domestic-funded firms. They also seem to be less likely to introduce new products, although the differences in innovation propensities with respect to domestic-funded firms are most of the time insignificant. What is striking, though, is that foreign-funded firms are less R&D-intensive but, when they innovate, they have a higher share of their total sales attributable to new products than domestic-funded firms do. This finding is reminiscent of the often-voiced argument that foreign-owned firms are innovative but generally keep their R&D in their home base. Our results for the Chinese automobile industry confirm those obtained by Fu and Gong (2008) for all industries in China and by Huang and Sharif (2009) for the province of Guangdong. Unless FDI fosters innovation in Chinese-controlled firms in the automobile industry (something we have not investigated in this chapter), the burden of R&D that, as we have shown, stimulates innovation, rests on the shoulders of Chinese-controlled firms. The fact the foreign-funded firms are more innovative when they do innovate suggests that Chinese firms have to do more R&D to compete with foreign-funded firms and/or that they have to increase the productivity of their R&D in transforming research into marketable new products.

Notes 1. The time subscript has been deleted to simplify notation. 2. Foreign-funded enterprises include joint-venture enterprises, cooperative enterprises, enterprises with sole funds and share-holding Ltd corporations. Joint-venture enterprises and cooperative enterprises are responsible for the amount of investment stipulated by the contract. According to Chinese legislative regulations, when a share-holding Ltd corporation registers with agencies of the Administration for Industry and Commerce, it is classified as a foreign-funded firm only if the foreign equity stake is at or above 25 per cent. More detailed discussion of the classification of foreign-funded firms in China can be found in Huang (2003: 4 and 35). 3. Because the predicted value for R&D is used as a regressor, the standard errors in Table 10.10 are somewhat too small. Attempts to bootstrap the standard errors failed, perhaps because of the small size of our sample.



219

4. The marginal effects reported in Tables 10.9 and 10.10 are obtained by taking the first derivative of the expression for the expected conditional intensity E ( y2 | x1 , x2 ) = x2 b2 + rs 2 [w( x1b1 ) / F( x1b1 )] with respect to each element contained in x1 and x2, which are the regressors in the selection equation and in the intensity equation respectively.

References Aitken, B. and A. Harrison (1999), ‘Do domestic firms benefit from direct foreign investment? evidence from Venezuela’, American Economic Review, 89 (3), 605–18. Caves, R.E., M.E. Porter, A.M. Spence and J.T. Scott (1980), Competition in the Open Economy: A Model Applied to Canada. Cambridge, MA: Harvard University Press. Chen, M. (2001), ‘The effectiveness analysis of China’s introducing FDI’, Strategy and Management, 3. Crépon, B., E. Duguet and J. Mairesse (1998), ‘Research, innovation and productivity: An econometric analysis at the firm level’, Economics of Innovation and New Technology, 7, 115–58. Fu, X. (2008), ‘Foreign direct investment, absorptive capacity and regional innovation capabilities: Evidence from China’, Oxford Development Studies, 36 (1), 89–110. Fu, X. and Y. Gong (2008), ‘Indigenous and foreign innovation efforts and drivers of technological upgrading: Evidence from China’, SLPTMD Working Paper Series, 016. Griliches, Z. (1990), ‘Patent statistics as economic indicators: A survey’, Journal of Economic Literature, 28 (4), 1661–707. Haddad, M. and A. Harrison (1993), ‘Are there spillovers from direct foreign investment? evidence from panel data for Morocco’, Journal of Development Economics, 42, 51–74. He J. (2000), ‘The further precise quantification of FDI on China’s industrial sector spillover effects’, Journal of World Economy, (12), 29–36. Hu, A. and G. Jefferson (2002), ‘FDI impact and spillover: Evidence from China’s electronic and textile industries’, World Economy, 38 (4), 1063–76. Huang, C. and N. Sharif (2009), ‘manufacturing dynamics and spillovers: The case of Guangdong Province and Hong Kong, Macao and Taiwan’, Research Policy, forthcoming. Huang, Y. (2003), Selling China, Foreign Direct Investment during the Reform Era. Cambridge: Cambridge University Press. Jiang, D. and L. Xia (2005), ‘The empirical study of the function of fdi on innovation in China’s high-tech industries’, Journal of World Economy, 8, 3–10. Kokko, A. (1994), ‘Technology, market characteristics, and spillovers’, Journal of Development Economics, 43, 279–93. Mohnen, P., J. Mairesse and M. Dagenais (2007), ‘Innovativity: A comparison across seven European countries’, Economics of Innovation and New Technologies, 15 (4/5), 391–413. Organisation for Economic Co-operation and Development (OECD) (1992, 1996), Oslo Manual: Guidelines for Collecting and Interpreting Innovation Data, OECD: Paris.


220


Wang H., T. Li and J. Feng (2006), ‘Does R&D facilitate or dampen indigenous R&D’, Economic Research Journal, 41(2), 44–56. Xi, G., and B. Yan (2005), ‘The spillover effect of FDI on China’s innovation capacity’, Journal of World Economy, 10, 18–25.


11 The Role of FDI in the Development of Innovative Capacity: The Case of Russian Companies Juha Väätänen, Daria Podmetina and Marina Aleksandrova

11.1

Introduction

This chapter studies the role of foreign direct investment (FDI) in the development of innovative capacity and labour productivity of Russian enterprises. FDI has had a very significant role in the development of competitiveness in the transitional economies of Central and Eastern Europe. According to World Bank study, the productivity of Russian R&D is very low in international comparison (Schaffer and Kuznetsov, 2007). Globalization has increased opportunities and pressures for domestic firms in emerging market economies to innovate and improve their competitive position (Gorodnichenko et al., 2008). Exporting allows firms in developing countries to enlarge their markets and benefit from economies of scale. Moreover, export and import operations are proven effective channels of technology transfer between countries (Pack, 1993). Some companies in developing countries establish R&D centres or acquire companies from developed countries in order to obtain skills and knowledge (Bell and Pavitt, 1993). The Russian economy is currently highly dependent on the export of natural resources, such as oil and gas. In the last eight years Russia’s gross domestic product (GDP) has been growing more than 5 per cent annually, thanks to high oil and gas prices on the world markets. FDI and exports are important co-operation channels for developing countries and the rest of the world and in a previous study the authors assessed the effect of FDI on the development of the innovative capacity of Russian companies on a regional level. The potential impact of FDI 221


222

Juha Väätänen et al.

Table 11.1 Companies with foreign participation, share of GRP 2005 (%) Regions

Federal districts

Moscow Omsk region Sverdlovsk region St Petersburg Leningrad region Belgorod region Perm region Tatarstan Russia, total Moscow region Yaroslavl region Primosky region Tyumen region

Central Siberia Ural North-west North-west Central Volga Volga Central Central Far East Ural

Turnover/GRP 157.7 142.5 94.3 93.4 88.8 81.7 76.5 74.4 72.1 72.1 64.4 53.2 49.9

Source: Rosstat, authors’ calculations.

is higher in regions where the relative share of FDI is higher (Väätänen and Podmetina 2007). There are 16,128 companies with foreign participation (foreign companies) according to the statistics of Federal State Statistic Service – Rosstat (2007) (See Appendix 11A.1). The economic impact of foreign companies is assessed by comparing the turnover of foreign companies to gross regional product (GRP). A more accurate indicator would be value-added compared to GRP, which would allow us to classify foreign companies by value-added created and to track whether regions would differ by their FDI profiles. However, the Russian statistical data do not allow for this more sophisticated option. Table 11.1 illustrates that the impact of foreign companies is highest in Moscow, Omsk, Sverdlovsk and St Petersburg. The comparison of the share of foreign investments to total fixed investments and GRP on a regional basis is a valuable indicator for the potential effects of FDI on the development of domestic companies. The regional comparisons are presented in Table 11.2. The share of foreign investment is highest in Moscow, Omsk and St Petersburg. The largest Russian cities (Moscow and St Petersburg) are the most relevant when assessing the impact of FDI on local companies because of the better investment climate. These regions host almost 60 per cent of all foreign companies in Russia. The potential impact of foreign companies is very high especially in Moscow and the Moscow region, where 13.8 and 19.7 per cent of fixed investments are carried out by foreign companies. Foreign companies have also high impact on the regional


The Case of Russian Companies 223

Table 11.2 FDI as a share of gross fixed capital formation (GFCF) and GRP in 2004 (%) Regions

Federal districts

Omsk region Moscow region Moscow Russia Leningrad region Krasnodar region Tyumen region Stavropol region Bryansk region St. Petersburg Pskov region Rostov region Kaliningrad region

Siberia Central Central North-west South Ural South Central North-west North-west South North-west

FDI/GFCF

FDI/GRP

25.9 19.7 13.8 10.7 8.2 7.9 5.1 4.9 4.8 4.7 4.1 3.0 2.2

0.2 4.0 2.0 1.9 2.2 0.6 1.2 0.1 0.4 0.6 0.9 0.3 1.0


level in the above-mentioned regions when comparing turnover to GRP. Thus the selection of Moscow, the Moscow region, St Petersburg and the Leningrad region for closer analysis provides a unique view of the impact of FDI on the potential development of innovative capacity in the most developed regions of Russia. The role of FDI in the development of innovative capacity of domestic companies is assessed based on the combination of Russian statistical data and an innovation survey of 176 companies located mainly in St Petersburg and Moscow. The survey was conducted in 2008 and aimed to study the innovative capabilities of foreign and domestic R&D-oriented companies. The sample consists of companies that are actively innovating or represent an industry with high innovation intensity (OECD, 1993; 2007). The authors hope that the combination of Russian statistical data and survey results will enable the evaluation of the potential impact of FDI on the development of innovative capacity and labour productivity of Russian companies.

11.2 Literature review The theoretical background of this chapter is based on FDI’s effect on the development of innovative capacity of domestic companies. Dyker (2006) studied the process of development and dissemination of technology in Russia through co-operation between Russian organizations and foreign firms. It is important to understand, that FDI in Russia


224 Juha Väätänen et al.

facilitates technology transfer from abroad. The interesting point is that the success of privatization in Russia can be estimated by ‘the diversity of enterprise forms, sizes, and strategies which is essential for knowledge diffusion and generation’. The existence of a strong relationship between internationalization (FDI and exports) and innovation is obvious for technology-oriented companies. However, globalization processes more often push companies to enter foreign markets and acquire specific knowledge that will enable them to implement more technology innovations. Thus: innovation has moved from an international reality dominated by the idea of technology transfer, where agents develop knowledge and transfer it to other countries, to a much more complicated situation where, although, that reality has not disappeared, there are also new ways of developing innovation in which the international ambit also affects the creation of knowledge stage and which multinational companies acquire new protagonism. (Molero, 2008) FDI is an important channel through which to increase the innovative capacity of the companies (Hoekmann and Javorcik, 2006) and it may positively affect the productivity of domestic companies with three effects: the competition effect, the linkage effect and the employment effect (Görg and Strobl, 2001). The competition effect from foreign companies forces local companies to focus on their core competences (Blomström and Sjöholm, 1999), such as innovation. From the other perspective, implementing innovations is a significant driver for increasing the competitiveness of domestic companies on both the home and foreign markets. However, transitional countries should be the greatest beneficiaries of globalization, especially from the transfer of capabilities via FDI (Sutton, 2007), because competition caused by foreign companies should strengthen domestic ones. Entry to the foreign market plays ‘a more creative role’, serving as an instrument for the introduction and diffusion of innovations. ‘Entry often plays a major role early in the life of most products ... in early stages, outsiders are the source of most innovations and use these as a vehicle of entry ... there is a shift from product towards process innovation ...’ (Geroski, 1991). The effects of FDI could be either positive (Blomström and Wolff, 1994; Kokko, 1994) or negative (Djankov and Hoekman, 2000). This is true both for the effects of increased competition and spillover. Aitken and Harrison (1999) state the effect to be positive for small enterprises



with high foreign participation, but for larger enterprises and enterprises with little or no foreign participation the effect is rather negative. For non-exporting firms the effect should be positive, since they are less likely have faced earlier absorption of foreign knowledge. The import of technology was complemented by a huge effort to develop local capabilities in developing countries in East Asia: technological co-operation with foreign partners covers not only ‘the acquisition of competencies for operating and maintaining, but also the acquisition of various combinations of design, engineering and project management skills’. Companies invest in postgraduate education and training in developed countries so their personnel can become members of the informal international networks (Bell and Pavitt, 1993). Yudaeva et al. (2003) found the spillover effect for local companies to be positive in those Russian regions that have more open policies towards foreign investors and a less corrupt administration. Foreign firms located in more reform-oriented regions, especially, tend to be more productive than others. The development of a market economy in Russia has to be based on networks of innovative companies utilizing FDI (Dyker, 2006). Sinani and Meyer (2004) claim that the potential effects of FDI depend on the receiving firm’s size, ownership structure and trade orientation. Ownership structure, whether the companies are private or state-owned, is a very important determinant of spillovers in transitional economies like Russia. Positive effects of FDI could be limited to certain sectors, either the foreign company’s own sector or the ones vertically integrated into it. Kinoshita (2000) studied companies in the transitional economy of the Czech Republic and found positive spillovers from FDI in electrical machinery, which is also active in R&D spending. Spillovers could also be limited by geography (Aitken and Harrison 1999). Perhaps only local companies located in the same industrial district might be able to adapt new knowledge from foreign firms. Whether local companies are able to absorb knowledge depends on the technology gap between them and foreign firms (Chaturvedi and Chataway, 2006; Kokko, 1994; Lane and Lubatkin, 1998). Probably the most important requirements for spillovers are a small technology gap, R&D investment and an educated workforce. Local companies have to have a certain level of technological capability; if the technology gap is too large, the catching-up process becomes too difficult (Damijan et al., 2003). Companies with high R&D investment and an effective R&D department are able to adapt new technologies quickly. In countries with a well-educated workforce, such


226


as in Russia, it is potentially easier to absorb technology than it is in many other emerging economies. In the end, enhancement of the innovative capacity and productivity of local firms depends on their ability to absorb knowledge from outsiders, such as foreign companies. Hargadon and Fanelli (2002) define innovation as ‘... the adaptation of existing knowledge into new activities’. The absorptive capacity is defined by a firm’s ability to value, assimilate and utilize new external knowledge (Cohen and Levinthal 1990). The innovative capacity is measured in this study by the capability to introduce new products. Traditionally innovation capability is measured by patenting activity (Acs and Audretsch, 1989). The authors believe that the measure based on the introduction of new products is more relevant than relying solely on R&D expenditure or patent data. As is well known, in Russia the efficiency of R&D measured by patents is low despite relatively high R&D spending (Schaffer and Kuznetsov, 2007). Thus, the inclusion of new product development provides a wider perspective than traditional R&D and patent approaches, especially when not all innovations are patented (Bottazzi et al., 2002). Labour productivity is applied as a competitiveness indicator in this study. Yudaeva et al. (2003) found that foreign companies are more productive in reform-oriented regions such as St Petersburg and Moscow. This supports the focus of this study, especially when results from earlier studies show that FDI has significant effect on the regional innovation capacity as well as on the productivity of innovation (Fu, 2007). There are also specific factors relating to transitional economies that affect the impact of FDI. Sinani and Meyer (2004) reported that company ownership structure is one of the factors behind positive spillover effects. This is an important factor in transitional economies such as Russia. These factors are included as background factors in the study.

11.3

Survey data

The study is based on a survey of 176 R&D-oriented Russian companies conducted in early 2008. The sample was drawn from companies active in innovating or representing an industry with a high innovation intensity (OECD, 1993; 2007). Thus the sample chosen was based on the expectation for the firms to be innovation-oriented and to emphasize R&D as a source of their long-term competitive advantage. Innovativeness indicators, such as R&D expenditure, new product development and patenting activity are used to evaluate the innovative capacity at the firm level.



To achieve the results a number of industries and regions were included in the sample. The survey was conducted in the regions with the highest impact of FDI and highest innovation sector development, mainly in St Petersburg and Moscow (Väätänen and Podmetina 2007; Torkkeli et al., 2009). The break-down by industry of the sample is as follows: the largest number are service companies (27.8 per cent), followed by machine building (22.7), information and communications technologies (ICT) (14.2), electronics (14.2), energy, oil and gas (7.4) and construction (6.3). The study applies the classification of enterprise backgrounds based on the BEEPS study (Business Environment and Enterprises Performance Survey by the World Bank and the European Bank for Reconstruction and Development) – state-owned, privatized, new enterprise and foreign-owned – in order to be able to analyse whether enterprise history is a significant explanatory factor in innovative capacity. Innovation indicators, such as R&D expenditure, new product development, innovation risks and patenting activity are used to evaluate the innovative capacity on the firm level. Most studies use mainly patent data and R&D expenditure, which is problematic. Patents have several weakness because they measure inventions rather innovations, they are very industry-, country- and process-dependant, and companies often use other methods to protect their inventions. Using R&D expenditure can also be problematic, because not all innovations are generated by R&D expenditure, R&D does not necessarily lead to innovation and formal R&D measures are biased against small firms. Privately owned enterprises are dominant in the sample. Only 6 per cent of companies are state-owned, the greater number of companies were private from their establishment (68 per cent) and only 5 per cent were privatized in the mass-privatization process following the collapse of the Soviet Union. Foreign-owned companies have the highest productivity. In the research data, the average R&D expenditure against sales is 6.5 per cent, the ICT and electronics sectors having the highest shares, at 8.5 and 7.8 per cent respectively. Foreign companies make up 12.5 per cent of those surveyed; most of them are located in Moscow and St Petersburg. Foreign companies are concentrated (Table 11.3) in the ICT (28 per cent), machine-building (12.5) and transportation sectors (12). The average sales of foreign companies are slightly higher (€67.1 million) than domestic companies (€59.2 million); productivity of foreign companies is 10 per cent higher – €37,640 per employee against €33,230 per employee for domestic companies; in addition, the share of exports



Table 11.3 Industry sectors, R&D expenditure and foreign ownership Share of sample (%) Services Machinery Electronics ICT Energy, oil and gas Construction Others Total

27.8 22.7 14.2 14.2 7.4 6.3 7.4 100.0

R&D expenditure/ sales (%) Share of FDI (%) 7.0 5.5 7.8 8.5 4.6 2.4 4.1 6.5

10.2 12.5 8.0 28.0 0.0 9.1 15.4 12.5

Source: Survey data, authors’ calculations.

Table 11.4 Financial indicators

Percentage of total Sales/company, € millions Employees/company Productivity (Sales/employees) Exporting companies (%) Export/company, € millions Export/Sales (%)

Foreign

Domestic

12.5 67.1 5891 37,640 54.5 12.4 25.5

87.5 59.2 3313 33,230 44.2 9.9 19.8


is higher (Table 11.4). There are no significant differences in the education level of employees between domestic and foreign companies.

11.4

Results of the analysis

The authors’ aim is to analyse the effect of FDI on the development of labour productivity and innovative capacity of Russian companies. The performance measures are labour productivity, new product development and patent activity. In addition to ownership (foreign v domestic), the independent variables of the study are industry sector and company size. Company size is classified into four categories: micro (fewer than ten employees), small (10–50 employees), medium (50–250 employees) and large (more than 250 employees). The interaction effects are tested separately for each dependent variable (innovative capacity indicators). The interaction effects are analysed by applying the General linear model (GLM) univariate test, which allows investigating interactions between factors.



Table 11.5 Test results on labour productivity Labour productivity Corrected model

Ownership Industry Size

2006 F R Sq Sig F Sig F Sig F Sig

4.527 0.503 0.000** 1.451 0.230 10.062 0.000** 11.209 0.000**

Note: ** Significant effect on 95% confidence level. Source: Survey data, authors’ calculations.

Table 11.6 R&D operations

R&D expenditure/Sales (%) R&D expenditure/R&D personnel. R&D expenditure TOP4 (%) Internal R&D Machinery & equipment Acquisition of external knowledge Acquisition of external R&D

Foreign

Domestic

6.2 3,637

6.5 3,155

30.0 17.6 20.7 17.1

29.7 20.4 20.3 19.4


The test results on labour productivity are presented in Table 11.5. The results show that industry and company size have a significant effect on labour productivity. The electronics industry has the highest labour productivity. Similarly small companies (10–50 employees) have the highest labour productivity. There is also a significant interaction effect (sig. 0.021) between industry and ownership. Foreign companies have labour productivity three times higher than domestic companies in the electronics industry (€99,330/person v €34,910/person). In other industry sectors there are no significant differences between foreign and domestic companies. When comparing R&D expenditure, there are no significant differences between foreign and domestic companies (Table 11.6). The share of R&D spending of total sales is 6.2 per cent for foreign companies and 6.5 per cent for domestic companies. Similarly there are surprisingly few differences between foreign and domestic companies in the structure of R&D spending. Both for foreign (F) and domestic (D)



companies spend 30 per cent on internal R&D; acquisition of machinery and equipment for R&D purposes – 17.6 (F) and 20.4 per cent (D); acquisition of external knowledge (e.g. licences, patents and know-how) – 20.7 (F) and 20.3 per cent (D) and for acquisition of external R&D – 17.1 (F) and 19.4 per cent (D). The test results on R&D expenditure/sales are presented in Table 11.7. The results show no significant differences in the R&D expenditure/ sales by ownership, industry or company size. The ICT and electronics industries have the highest R&D expenditure/sales, at 8.5 and 7.8 per cent respectively. Similarly, there are no significant effects of ownership, industry or company size on R&D expenditure/R&D personnel (sig. 0.191). The test results indicate relatively small differences between foreign and domestic companies on R&D operations.

Table 11.7 Test results on R&D expenditure/sales R&D expenditure/sales Corrected model



0.777 0.299 0.774 0.355 0.553 0.196 0.997 1.755 0.163


The basic indicators of new product development (NPD) are presented in Table 11.8. More than one-third (36 per cent) of foreign companies introduced new products in the last three years compared with 26 per cent of domestic companies. Foreign companies are more likely to cooperate with external partners in the product-development phase. The largest difference is in the sales mix with 61 per cent of foreign companies’ turnover originating from new products compared with 28 per cent for domestic companies. The average duration of new product development, from idea to the market, is longer for foreign companies – 15.8 months against 13.4 for domestic companies. Companies prefer to develop new products themselves; only 25 per cent of foreign companies and 10 per cent of local companies cooperate


The Case of Russian Companies 231 Table 11.8 New product development (NPD) Foreign New product introduced in the last 3 years (%) New product developed by (%): Own company In co-operation with others Turnover, 2006, distributed (%) New product Significantly improved Unchanged Average duration of NPD from idea to market (months)

Domestic

36.3

26.0

75 25

90 10

61.2 15.8 23.0

28.1 39.8 32.1

15.8

13.4

Table 11.9 Test results on NPD NPD Corrected model



1.191 0.308 0.003** 1.662 0.199 2.928 0.010** 0.353 0.787

Note: ** Significant effect on 95% confidence level.

with external partners. The test results on new product development are presented in Table 11.9. The dependent variable tested is whether a company has introduced new products during the last three years. The test results show that industry sector has a significant main effect (sig. 0.010) on the new-product introduction. The ICT and electronics industries have the highest new product introduction rate – over 60 per cent of companies have introduced new products during the last three years. The results indicate relatively the high innovative capacity of the ICT and electronics industries where the new product introduction rate is double that of the Russian average of 32 per cent. However, there are no significant differences between foreign and domestic companies. Patent activity is the last component of innovative capacity in this study. An important indicator of innovation capability is the number of patents a company owns. Foreign companies hold on average 4.23 patents per company in Russia and 5 internationally, compared with 1.92 patents per company among domestic companies in Russia and 0.23


232 Juha Väätänen et al. Table 11.10 Test results on patents/R&D personnel Patents/R&D personnel Corrected model



246.198 0.994 0.000** 0.063 0.804 12.488 0.000** 236.463 0.000**

Note: ** Significant effect on 95% confidence level.

internationally. The difference in the number of international patents between foreign companies and domestic companies, especially, is large. The dependent variables tested are the number of patents/sales and the number of patents/R&D personnel. In the number of patents/sales, the company size has a significant main effect (sig. 0.000). The large companies have the highest number of patents/sales, 2.5 times more than the average of data sample. This is an interesting result, since the company size is already taken into account in the dependent variable. This could be explained by the fact that a certain critical mass of R&D expenditure is needed to obtain patents. This assumption is supported by the fact that the micro companies do not have any patents. The test results on the number patents/R&D personnel are presented in Table 11.10. When comparing the number of patents to the number of R&D personnel, the industry sector and company size have a significant main effect. The electronics industry has 2.5 times more patents/R&D personnel than the average of the data sample. Small companies (10–50 employees) have three times more patents/ R&D personnel than the surveyed companies on average. The results indicate that small companies have more effective R&D operations than medium-sized and large companies. On the other hand, large companies are more effective when the comparison is made using the number of patents/sales. However, in either case ownership does not have any significant effect.

11.5

Conclusions

The survey results showed smaller differences in labour productivity and innovative capacity between foreign and domestic companies in



Russia than the authors expected. The labour productivity of foreign companies was 10 per cent higher than for domestic companies. In the case of the electronics industry, foreign companies had a labour productivity three times higher. The higher labour productivity of foreign companies could be partly explained by higher share of exports and higher share of new products in sales. The export prices could be assumed to be higher than prices on the Russian domestic markets. Similarly the prices of new products could be assumed higher than prices of older products. The domestic companies had a higher R&D expenditure/sales ratio, at 6.5 per cent against 6.2 for foreign companies. The test results show no significant differences in the R&D expenditure/sales by ownership, industry or company size. The ICT and electronics industries have the highest R&D expenditure/sales, at 8.5 and 7.8 per cent respectively. Similarly, there are no significant effects of ownership, industry or company size on R&D expenditure/R&D personnel (sig. 0.191). The test results indicate relatively small differences between foreign and domestic companies on R&D operations. However, contrary to expectations there were no significant performance differences between foreign and domestic companies in the innovative capacity, when measured by new product development or patent activity. The results show that industry and company size have a significant effect on labour productivity. The electronics industry has the highest labour productivity. Similarly, small companies (10–50 employees) have the highest labour productivity. There is also a significant interaction effect (sig. 0.021) between industry and ownership. Foreign companies have labour productivity three times higher than domestic companies in the electronics industry (€99,330/person v €34,910/person). In other industry sectors there are no significant differences between foreign and domestic companies. The ICT and electronics industries had the highest new-product introduction rates. In both sectors more than 60 per cent of companies had introduced new products in the last three years. Foreign companies has a 28 per cent share in the ICT industry and 8 per cent in the electronics industry, when the average within the data sample was 12.5 per cent. The high new-product introduction rate of the ICT industry could be partly explained by large share of foreign companies, whose presence could have increased the innovative capacity of the whole industry through direct and spillover effects. To test the effect statistically would require time-series on innovative capacity data.



If innovative capacity was measured by the average number of patents a company holds, foreign companies dominated domestic companies both in Russian and international patents. However, statistically there were no differences in the number of patents/sales or the number of patents/R&D personnel between foreign and domestic companies. Both industry and company size had a significant effect on patent activity. For the number of patents/sales, company size has a significant main effect (sig. 0.000). Large companies have the highest number of patents/sales, 2.5 times more than the average of data sample. This is an interesting result, since company size is already taken into account in the dependent variable. This could be explained by the fact that a certain critical mass of R&D expenditure is needed to obtain patents. This assumption is supported by the fact that micro sized companies do not have any patents. When comparing the number of patents to the number of R&D personnel, the industry sector and company size have a significant main effect. The electronics industry has 2.5 times more patents/R&D personnel than the average of the data sample. Small companies (10–50 employees) have three times more patents/R&D personnel than the surveyed companies on average. The results indicate that the small companies have more effective R&D operations than medium-sized and large companies. On the other hand, large companies are more effective when the number of patents/sales is compared. However, in ownership does not have any significant effect in either case. The electronics industry was the most effective in patent activity by all indicators. Similarly, small companies were most effective, when compared by R&D personnel numbers. More than one-third (36per cent) of foreign companies have introduced new products in the last three years compared with 26 per cent of domestic companies. Foreign companies are more likely to cooperate with external partners in the product-development phase. The largest difference is in the sales mix. Foreign companies have 61 per cent of turnover originating from new products compared with 28 per cent for domestic companies. The average duration of new product development, from idea to the market, is longer for foreign companies – 15.8 months against 13.4 for domestic companies. Companies prefer to develop new product themselves; only 25 per cent of foreign companies and 10 per cent of local companies cooperate with external partners. The dependent variable tested is whether company has introduced new products during the last three years. The test results show that industry sector has a significant main effect (sig. 0.010) on new product introduction. The ICT and electronics industries have the highest new-product


The Case of Russian Companies 235 Appendix 11A.1 The number of companies with foreign participation, 2005. Region Russia Moscow region, including Moscow Leningrad region, including St Petersburg Krasnodar region Kaliningrad region Primosky region Rostov region Belgorod region Bryansk region Novosibirsk region Pskov region Sverdlovsk region Samara region Khabarovsk region Omsk region Stavropol region Tatarstan Tyumen region Murmansk region Yaroslavl region Perm region

Federal districts

Number of foreign companies

Central

16,128 7112

North-east

2318

South North-west Far East South Central Central Siberia North-west Ural Volga Far East Siberia South Volga Ural North-west Central Volga

508 434 324 265 253 203 190 174 151 150 149 144 138 137 137 136 135 123


introduction rate; over 60 per cent of companies have introduced new products during the last three years. The results indicate the relatively high innovative capacity of the ICT and electronics industries. The newproduct introduction rate is double that of the Russian average of 32 per cent. However, there are no significant differences between foreign and domestic companies. To summarize, the study results showed that, contrary to preliminary expectations, there were only small differences in the efficiency of R&D operations between foreign and domestic companies, when measured by either new product development or patent activity. This could be explained by the fact that currently Russia is not an attractive R&D location for foreign companies. According to international statistics the productivity of Russian R&D has remained very low (Schaffer and Kuznetsov, 2007). It could be that foreign companies are not able to perform R&D operations more effectively than domestic companies in


236


the Russian market. Thus the potential FDI effect on the development of innovative capacity of Russian companies remains limited.

References Acs, Z. and D. Audretsch (1989), ‘Patents as a measure of innovative activity’, Kyklos, 42 (2), 171–80. Aitken, Brian J. and Ann E. Harrison (1999), ‘Do foreign firms benefit from direct foreign investment? Evidence from Venezuela’, The American Economic Review, 89 (3), 605–18 Bell, R.M. and K. Pavitt (1993), ‘Technological accumulation and industrial growth: Contrasts between developed and developing countries’, Industrial and Corporate Change, 2 (2). Blomström, M. and E. Wolff (1994), Multinational Corporations and Productive Convergence in Mexico. In Convergence of Productivity: Cross National Studies and Historical Evidence. Oxford: Oxford University Press. Blomström, M. and F. Sjöholm (1999), ‘Technology transfer and spillovers: Does local participation with multinationals matter?’ European Economic Review, 43, 915–23. Botazzi, L. and P. Giovanni (2003), ‘Innovation and spillovers in regions: Evidence from European patent data’, European Economic Review, 47, 687–710. Chaturvedi, K. and J. Chataway (2006), ‘Strategic integration of knowledge in Indian pharmaceutical firms: Creating competencies for innovation’, International Journal of Business Innovation and Research 1 (1/2), 27–50. Cohen, W.M. and D.A. Levinthal (1990), ‘Absorptive capacity: A new perspective on learning and innovation’, Administrative Science Quarterly, 35, 128–52 Damijan, Joze, Mark Knell, Boris Majcen and Matija Rojec (2003), ‘Technology transfer through FDI in Top-10 transition countries: How important are direct effects, horizontal and vertical spillovers?’ William Davidson Working Paper, 549, February. University of Michigan. Djankov, S. and B. Hoekman (2000), ‘Foreign direct investment and productivity growth in Czech enterprises’, World Bank Economic Review, 14 (1). Dyker, D (2006), Closing the EU East-West Productivity Gap. London: Imperial College Press. Fu, X. (2007), ‘Foreign direct investment, absorptive capacity and regional innovation capabilities. Evidence from China’. University of Oxford, Department of International Development, SLPTMD Working Paper Series 3. Geroski, P.A. (1991), Market Dynamics and Entry. Cambridge, MA: Blackwell. Gorodnichenko,Y., J. Svejnar and K. Terrell (2008)‚ Globalization and innovation in emerging markets’. IZA Working Paper 3299. Görg, H. and E. Strobl (2001), ‘Multinational companies and productivity spillovers: A meta analysis’, The Economic Journal, 111, November, 723–39. Hargadon, A. B. and A. Fanelli (2002), ‘Action and possibility: Reconciling dual perspectives of knowledge in organizations’, Organization Science, 13(3), 290–302. Hoekman, B. and B. Javorcik (2006), Global Integration and Technology Transfer. Washington, D.C.: Palgrave Macmillan and the World Bank. Kinoshita, Yuko (2000) ‘R&D and technology spillovers via FDI: Innovation and absorptive capacity’. William Davidson Working Paper 349.


The Case of Russian Companies 237 Kokko, Ari (1994), ‘Technology, market characteristics, and spillovers’, Journal of Development Economics, 43: 279–93. Lane, P.L. and M. Lubatkin (1998), ‘Relative absorptive capacity and interorganizational learning’, Strategic Management Journal, 19, 461–77 Molero, J. (2008), ‘The challenges of the internationalization of innovation for science and technology policies’. ICEI Paper, Instituo Compluense de Estidios Internacionales. Organisation for Economic Co-operation and Development (OECD) (1993), The Measurement of Scientific and Technological Activities: Proposed Standard Practice for Surveys of Research and Experimental Development (The Frascati Manual). Paris: OECD. Organisation for Economic Co-operation and Development (OECD) (2007), Oslo Manual: Guidelines for Collecting and Interpreting Innovation Data (3rd Edition). Paris: OECD. Pack, H. (1993), ‘Technology gaps between industrial and developing countries: Are there dividends for latecomers?’ in World Bank (ed), Proceedings of the World Bank Annual Conference on Development Economics 1992. Washington, D.C.: World Bank. Rosstat (2007), Russia in Figures 2006. Moscow: Russian Federal State Statistics Service. Schaffer, M. and B. Kuznetsov (2007), ‘Productivity’, in M. Desai and I. Goldberg (eds), Enhancing Russia’s Competitiveness and Innovative Capacity. Washington, D.C.: The World Bank, chapter 2. Sinani, E. and K. Meyer (2004), ‘Spillovers of technology transfer from FDI: The case of Estonia’, Journal of Comparative Economics, 32, 445–66. Sutton, J. (2007), ‘Quality, trade and the moving window: The globalization process’, The Economic Journal, 117 (524), 469–98. Torkkeli M., D. Podmetina and J. Väätänen (2009), ‘Knowledge absorption in emerging economies – the role of foreign investments and trade flows in Russia’, International Journal of Business Excellence, 2 (3/4). Väätänen, J. and D. Podmetina (2007), ‘International technology transfer in the Russian economy – The effect of foreign direct investment spillovers’, Global Business and finance review, 12 (3), Special issue. Yudaeva, K., K. Kozlov, N. Melentieva and N. Ponomareva (2003), ‘Does foreign ownership matter? The Russian experience’, Economics of Transition, 11 (3), 383–409


12 Human Capital and Technological Spillovers from FDI in the Chinese Regions: A Threshold Approach Miao Fu and Tieli Li

12.1

Introduction

Openness is a well-known factor that facilitates the development of a country. Investment from developed to less developed countries are the major channel for realizing this process. Besides the direct effects of investing, foreign direct investment (FDI) brings technology spillover effects to domestic firms. Kokko et al. (1996) and Blomstrom and Kokko (1997) find empirical evidence for FDI technology spillovers. Some authors also discus absorptive capacity for technology spillovers determined by human capital and technology gaps. Nelson and Phelps (1966) argue that education speeds technological diffusion and that technological progress is an increasing function of educational attainment and proportional to the gap between the theoretical level of technology and the level of technology in practice. Keller (1996) suggests that technological information is distributed freely, but the technologies cannot be utilized unless the accumulated human capital reaches a higher level. Eaton and Kortum (1996) also emphasize the importance of human capital in the absorption of internal and international knowledge transfers. The emphasis of human capital as a determinant of technology spillovers does not give clear direction to policy-makers. An appropriate guide should point out a clear target for the human capital quality for developing countries. When these countries reach this target, significant positive technology spillovers, instead of negative or insignificant effects, will take place. We call this target a threshold. Several works have been devoted to the search for the minimal human capital 238


Human Capital and Technological Spillovers

239

conditions, that is, the threshold. Borensztein et al. (1998) point out that FDI from countries in the Organisation for Economic Co-operation and Development (OECD) to developing countries has positive effects on output growth only if host countries have reached a minimum human capital threshold of 0.52–1.13 years (in terms of average years of male secondary school attainment over the age of 25). They discover that the human capital stocks of most developing countries are above this threshold level. Alternatively, Xu (2000) gives a higher threshold level of 1.4–2.4 years, which indicates that most developing countries do not have sufficient human capital to cross that threshold and that it is developed countries rather than developing countries that benefit from FDI technology spillovers. As foreign advanced technologies are usually absorbed by human capital with higher education, a threshold given in terms of average years of secondary school attainment may mislead policy-makers in developing countries to pursue too strongly a target in secondary education. Moreover, because of the rapid development in some countries, the education structure of older generations is completely different to that of the younger generations who are a major part of the active labour force. Thus human capital thresholds based on the population over the age of 25 may underestimate the real educational level of rapidly developing countries. In addition, Borensztein et al. (1998) and Xu (2000) regard a developing country as a whole in their studies. This may cause problems when applying their conclusions to explain the growth of large developing country such as China, where regional differences are pronounced. Empirical results show that technology spillovers vary across regions. Fu (2007) studies the inequalities of Chinese coastal and inland areas and finds that uneven distribution of FDI and human capital is the major reason for the economic inequalities. Fleisher et al. (2007) also find that the distribution of human capital affects technology growth and thus promotes regional inequalities in China. Therefore, the absorptive capacities of East and Mid-west of China are different and spillover patterns are variable within the country. Although the importance of domestic R&D in developing countries cannot be denied, the bulk of new technologies in the world are created in a handful of the richest countries (Eaton and Kortum, 1996; Keller, 2004). Thus, technology spillovers from abroad are crucial for developing countries and may provide a good explanation for the technological progress of a developing country in which the human capital surpasses the threshold. Yet the question remains of what will happen in a developing country which is below the threshold, and is this the case in


240 Miao Fu and Tieli Li

China? In spite of this pertinent question, most studies of Chinese FDI technology spillover are not concerned with threshold effects. Our research will be different from the existing literature in the following aspects. First, we estimate the threshold endogenously by the threshold regression which is provided by Hansen (1999). Borensztein et al. (1998) and Xu (2000) gain their threshold values by splitting the sample into groups according to exogenously chosen thresholds and estimating them separately or with dummy variables. Girma (2005) uses threshold regression to analyse FDI technology spillovers, but the threshold variable he uses is the absorptive capacity, measured by distance from the technological frontier. Second, our estimated thresholds are determined in terms of the proportion of workers who received a higher education. Our thresholds emphasize the absorptive capacity of higher education and the active workforce which is generated by accumulated and depreciated human capital. Third, we consider regional differentiation and interregional technology transfers within the country in addition to FDI technology spillovers. Because the physical capital, human capital, public infrastructure and geographic environment differ significantly across regions, we should not neglect these regional discrepancies and interregional technology transfers when studying the external causes of technology progress. Fourth, our comparison between thresholds and regionally specific human capital will discover which region’s human resource satisfies the minimum requirement and which does not. This will introduce different development patterns for those two types of regions. Fifth, as infrastructure investment motivated by the government play a key role in Chinese economic development, we include the public infrastructure variable in our model. We organize the remainder of the chapter as follows. Section 12.2 presents the human capital threshold models related to knowledge production and technology spillovers. Section 12.3 estimates the provincial total factor productivity (TFP) series. Section 12.4 provides the empirical findings and the last section concludes.

12.2 Empirical specification based on a threshold model Based on the theories of Paul Romer (1990), Grossman and Helpman (1991) and Aghion and Howitt (1992), David Romer (2001) introduces a knowledge-production function, which is presented in equation (1). Romer separates an economy into two sectors, a goods-producing sector and an R&D sector where knowledge is produced. KR and LR in the model represent R&D capital stock and labour force in the R&D



241

sector respectively. A indicates the stock of knowledge and time is denoted by t. We add a subscript i to denote different regions within a country. Because of externalities of knowledge, knowledge in one place can be used elsewhere. Thus A is not restricted to a specific region by i. dA is the increase of knowledge and it is variable across regions and times. dAi ,t = G( LRi ,t , KRi ,t , At )

(1)

The measurement of technology or knowledge is still a problem for the current literature. Grilliches (1979) argues that current and past R&D expenditure, that is, the stock of R&D capital, is a good measure of the current knowledge stock. Romer’s model comprises R&D capital stock in region i. Following this line of thought, we further segment R&D capital stock into two parts, one of which is the R&D capital stock in region i represented by KRi, and the other of which is the weighted sum of R&D capital stocks of the other regions, except region i. The latter is related to the externalities of knowledge. This is consistent with the perspective of Coe and Helpman (1995) who use the weighted sum of R&D capital stocks of other countries to estimate the international spillover effects of technology. Following their model, we use the logarithm of TFP to measure the variation of technology, as well as to change other variables into logarithm forms. We get equation (2). All explanatory variables lag for one year because knowledge outputs are usually the consequences of previous R&D investment. Ci is the intercept that includes the regional fixed effects. ln(TFPi ,t ) = ci + a ln( LRi ,t −1 ) + b ln( KRi ,t −1 ) + g ln(

∑ z KR j

j ,t −1

)

(2)

j ,j ≠i

j is the weight of region j’s R&D capital stock. According to endogenous growth theory, technologically advanced regions are usually economically advanced and vice versa. Therefore, we let j equal the ratio of region j’s GDP to sum of GDP in all regions. and are the output elasticities of R&D labour force and R&D capital stock respectively. measures the technological spillover effects from other domestic regions. Démurger (2001), Fu et al. (2004) and Fleisher et al. (2007) emphasize the effects of public infrastructure on economic growth or technological progress in China. Following this point of view, we add Ri, the length of the



regional highway network, as the measurement of infrastructure development. Then the model takes the following form: ln(TFPi ,t ) = ci + a ln( LRi ,t −1 ) + b ln( KRi ,t −1 ) + g ln(

∑ z KR j

j ,t −1

) + h ln( Ri ,t −1 ) (3)

j ,j ≠i

In addition to domestic interregional technology spillovers, international technology spillovers induced by inward FDI should be considered in developing countries. As Borensztein et al. (1998) and Xu (2000) point out, developing countries should satisfy a minimum human capital threshold before they can benefit from FDI technology spillovers. To estimate the human capital threshold endogenously by the model, we utilize the threshold regression approaches suggested by Hansen (1996; 1999). The econometric equation for our threshold regression is specified as follows: ln(TFPi ,t ) = ci + a ln( LRi ,t −1 ) + b ln( KRi ,t −1 ) + g ln(

∑ z KR j

j ,t −1

) + h ln( Ri ,t −1 )

j ,j ≠i

(4)

!! + l1 Fi ,t −1 I ( Ei ,t −1 ≤ u) + l2 Fi ,t −1 I ( Ei ,t −1 > u) + «i ,t In the above equation, F represents FDI capital stock as a share of total capital stock. E is the proportion of labour with higher education to total labour. I(.) is an indicator function and is the threshold level. 1 and 2 are the FDI technology spillover coefficients. The first step for the estimation of equation (4) is to test the existence of the threshold. The test is based on a likelihood ratio test shown in equation (5). The null hypothesis of this test is the non-existence of the threshold which can be denoted as H0: 1 = 2. ˆ means the estimate of the threshold. S0 is the sum of squared residuals under the null hypothesis and S(ˆ ) is the sum of squared residuals with threshold ˆ. ˆ 2 represents the residual variance when there is a threshold, and is equivalent to S(ˆ )/[N(T–1)]. LR1 =

S0 − S(θˆ) ,… and …θˆ = arg min S(θˆ ) sˆ 2 θ

(5)

If LR1 significantly rejects the null hypothesis, then the threshold exists and its estimate equals ˆ. The asymptotic distribution of LR1 can be



243

simulated by a bootstrap procedure recommended by Hansen (1996). This produces a P value for LR1 and critical values of different significant levels. According to the approach of Hansen, when there is a threshold effect, the next step is to get the confidence interval for the threshold value. The likelihood ratio test for H0: θ = θ 0 is shown in equation (6). The asymptotic distribution of LR(θ 0) follows p(LR(θ 0) ≤ x) = (1–exp(– x/2))2. Under this asymptotic distribution, we form valid asymptotic confidence intervals for the estimated threshold and get critical values corresponding to different significant levels. The critical value for significant level is calculated by equation (7). LR(θ0 ) =

S(θ0 ) − S(θˆ ) sˆ 2

(6)

c( a) = −2 ln(1 − 1 − a )

(7)

After acquiring the threshold value and its confidence interval, we estimate the coefficients of the model with the threshold regression. There may be more than one threshold in our application. In the case of two thresholds, the model takes the form in equation (8) where θ1 and θ2 indicate these thresholds. The estimation of the model with two thresholds is a two-stage process. First, we estimate θ1 as if there was only one threshold. Then we estimate the threshold θ2 as if the first threshold θ1 was given. The inference also includes two stages. ln(TFPi ,t ) = ci + a ln( LRi ,t −1 ) + b( KRi ,t −1 ) + g ln(

∑ z KR j

j ,t −1

) + h ln( Ri ,t −1 )

j ,j ≠i

+ !! l1 Fi ,t −1 I ( Ei ,t −1 ≤ u1 ) + l2 Fi ,t −1 I (u1 < Ei ,t −1 ≤ u2 )

(8)

+ l3 Fi ,t −1 I ( Ei ,t −1 > u2 ) + «i ,t

Based on existing literature, we have the following hypotheses: H1: α and β are positive according to the theories of David Romer (2001). Both labour force and capital stock in the R&D sector contribute to technology progress.



H2: The interregional technology spillover efficient is supposed to be positive following the discoveries of Branstetter (2001) and Coe and Helpman (1995). The former finds that intranational knowledge spillovers contribute more to technological progress than international knowledge spillovers ones do. This hypothesis is in line with externality of knowledge, that is, the knowledge-production investment of one area may have positive effects on other areas, and these effects strengthen when technological distances, knowledge structures and economic relationships among these areas are close. H3: The sign and magnitude of international spillover efficient i (i = 1, 2, 3) vary according to the thresholds. When human capital is above a certain threshold, the international spillover coefficient is assumed to be positive or greater in magnitude; otherwise it is negative, small in magnitude or insignificant. H4: The coefficient of public infrastructure, that is, , is positive as argued by Démurger (2001) and Fu et al. (2004). This hypothesis is especially reasonable for China as both the central and regional governments are deeply interesting in infrastructure investment, and this can facilitate the transmission of technologies.

12.3

TFP estimation and economic growth stages

Before discussing the threshold regression, we first estimate the dependent variable of TFP. Original provincial data are taken from the China statistical yearbooks 1991–2006 and all financial variables have been deflated to the constant price of 1990. The labour force is represented by the number of employed persons at the year end. Capital stock is calculated by the perpetual inventory method suggested by Coe and Helpman (1995) which is presented by equation (9), where K is the capital stock, I is the gross capital formation, is the depreciation rate and i and t are indices for regions and periods respectively. According to Zhang et al. (2004), takes the value of 9.6 per cent for China. The capital stock in the initial year is also taken from the paper of Zhang et al. (2004) for their paper is widely accepted in China. Ki ,t = I i ,t + (1 − d)Ki ,t −1

(9)

To calculate TFP, we first estimate the output elasticities of labour force and capital stock. The estimation equation is shown in equation (10). Yi,t and Li,t denote gross domestic product (GDP) and labour force



245

for region i in period t. Based on provincial panel data and a fixedeffects panel model chosen by the Hausman test, the estimates of and β are 0.3037 and 0.2021 respectively. As α + β = 1, we normalize the two elasticities: α = 0.3037 / (0.2021 + 0.3037) = 0.6004, and α = 1 – β = 0.3995. This result is very close to the estimates of Zhang and Shi (2003). Based on country-level series data from 1952 to 1998, Zhang and Shi (2003) find that α = 0.609 and β = 0.391. lnYi ,t = Ci + lt + a ln Ki ,t + b ln Li ,t + «i ,t

(10)

When elasticities of the labour force and the capital stock are obtained, TFP is calculated by equation (11) and it is assumed to be variable between periods and regions. TFPi ,t =

12.4

Yi ,t Kia,t Lbi ,t

(11)

Empirical findings

With the TFP series estimated above, we regress models (4) and (8). R&D data are from China’s Statistical Yearbooks on Science and Technology 1991–2006 and other data are from the China Statistical Yearbooks 1991–2006. Labour force in the R&D sector is represented by the number of the personnel engaged in scientific and technological activities. R&D capital stock is calculated using the perpetual inventory method. Internal expenditure for science research and technical development is used as the annual increase of R&D capital stock. The depreciation rate of R&D capital stock is set at 10 per cent according to the studies of Griliches (1990) and Coe and Helpman (1995). FDI capital stock is calculated by the perpetual inventory method mentioned in Section 12.3. Annual investments are adjusted by the deflator of price indices of investment in fixed assets. The labour force that received higher education is an accumulated variable of students enrolled in higher education institutions with a depreciation rate of 2.5 per cent, and lags for two years. The lag arises because it takes two years on average for in-school students to graduate. The human capital depreciation assures that human capital thresholds estimated here are based on the active workforce instead of the population as a whole, which would include the older generation. The summary statistics of variables is displayed in Table 12.1.


246

Miao Fu and Tieli Li

Table 12.1

Summary statistics of variables

Variable

Mean

Median

Max

Min

Std. Dev.

Ln(TFP) Ln(LR)

–0.6475 1.8610

–0.6598 1.9558

0.0954 3.6261

–1.2216 –2.2867

0.2401 0.9827

Ln(KR)

4.2243

4.2037

7.0754

0.2262

1.1927

Ln(R)

10.4981

10.6845

12.0296

8.0229

0.7841

F

0.0587

0.0307

0.3074

0.0000

0.0722

E

7.7634

5.1764

41.9001

1.3911

7.6713

12.4.1 Human capital thresholds for the absorption of FDI technology First of all, we test the existence of human capital thresholds. The likelihood ratio test results for the existence of thresholds are displayed in Table 12.2. For all regions in China, bootstrapped p values demonstrate that the single threshold and double threshold are significant at the 3 and 6.33 per cent levels respectively, whereas the triple threshold is not significant. Therefore, two thresholds exist for our case. Additionally, as regional disparities are significant in China, we test the threshold effects for the Mid-west area of China independently and find that the single threshold is significant at the 13.66 per cent level. The double and triple thresholds are not significant for the Mid-west. The threshold effects for East China cannot be tested because of a lack of observations. Next, we estimate the thresholds and their 95 per cent confidence intervals. The results are presented in Table 12.3 and the number of existing thresholds is according to the result in Table 12.2. The single threshold of 9.64 per cent of the Mid-west area is very close to the second threshold of 10.99 per cent for all regions. Comparing the confidence intervals between all regions and the Mid-west area, we can further confirm the correspondence of these two thresholds. The likelihood ratio diagrams of the thresholds for all regions are displayed in Figure 12.1. The flat lines in the figures are drawn according to equation (7) and the confidence level is 95 per cent. When the likelihood ratios are above these lines, they significantly deny the hypothesis of θ = θ 0 at the 5 per cent level. Thus, the confidence intervals for θ = θ 0 lie between the intersecting points of the flat lines and the likelihood ratio curves.


Human Capital and Technological Spillovers Table 12.2

247

Likelihood ratio test for threshold effects All regions

Single threshold Double threshold Triple threshold

Mid-west China

Likelihood ratio

P value

59.7546 32.7223 9.0587

0.0300 0.0633 0.7633

Likelihood ratio

P value

30.0200 12.3170 10.3209

0.1366 0.4300 0.3567

Note: The observations of East China are insufficient for threshold test.

Table 12.3

Threshold estimates and their 95 per cent confidence intervals All regions

Estimate (%) First threshold Second threshold

4.8451 10.9945

95% confidence Estimate 95% confidence interval (%) interval [3.9797, 5.0363] [8.7429, 12.0706]

35 Likelihood ratio

Likelihood ratio

30 25 20 15 10 5 0 0

5

Figure 12.1 regions

Mid-west China

10 15 20 25 30 35 40 First threshold parameter

9.6436 [8.1669, 12.2608]

45 40 35 30 25 20 15 10 5 0 0

5

10 15 20 25 30 35 40 First threshold parameter

Confidence interval construction in double threshold model: All

After identifying the threshold values, we obtain the regression results, which are shown in Table 12.4. The proximity of normal standard errors and White-corrected standard errors means no significant heteroscedasticity exists in our case. Therefore, the significance levels of the estimated coefficients can be decided by normal standard errors. The first explanatory variable to discus is the interaction term of FDI and the indicated human capital threshold. For all regions, when


248

Miao Fu and Tieli Li

Table 12.4 Threshold regression results for knowledge-production model: All regions and Mid-west China All regions

Coefficient Stan. Err. LR LRDS LRDSO INF F·I(E≤1) F·I(1<E≤2) F·I(E>2) 1 2

–0.0032 0.0385* 0.0535*** 0.0945*** –1.3899*** –0.6715*** 0.7346*** 4.8451 10.9945

0.0172 0.0237 0.0211 0.0288 0.1932 0.1996 0.2019 – –

Mid-west China Wht Stan. Wht Stan. Err. Coefficient Stan.Err. Err. 0.0150 0.0223 0.0203 0.0296 0.1726 0.2053 0.1877 – –

–0.0030 0.0383 0.1061*** 0.0112 –3.0266*** – –0.0359# 9.6436 –

0.0213 0.0333 0.0276 0.0403 0.5358 – 0.7566 – –

0.0152 0.0348 0.0265 0.0427 0.4744 – 0.8331 – –

Notes: * Significant at 10%; ** significant at 5%; *** significant at 1%. # Coefficient of F·I(E>1). Observations of East China are insufficient for threshold regression.

human capital (represented by the percentage of labour with higher education) is smaller than the first threshold of 4.85 per cent, the effects of FDI on technological progress are significantly negative and the FDI spillover coefficient is -1.39. This indicates the strong crowding out effects of FDI when human capital is low. There are several reasons for this phenomenon. First, goods produced by foreign firms occupy the host country’s markets and the domestic firms are unable to imitate relatively advanced foreign products because of the poor quality of human capital. Second, as the wages of the employees in foreign corporations are usually higher than those of domestic firms, brain-drain occurs, which may further worsen the lack of human capital within them. Third, foreign firms do not pay much attention to R&D in host countries. What they are most concerned with are markets, human resources and natural resources. Even in technology-intensive foreign firms, most domestic intellectuals are usually engaged in marketing or technical services. Fourth, foreign companies prevent their core technologies from leaking out to the host countries and the protection of technologies is more effective when the absorptive capacities of domestic human capital are low. When the quality of human capital is between 4.85 and 10.99 per cent, the negative coefficient of the interaction term of FDI and the



249

indicated human capital will be halved in absolute magnitude. Thus the negative crowding out effects of FDI are alleviated when the quality of human capital increases. Finally, when human capital surpasses 10.99 per cent, the spillover effects of FDI toward technology progress become positive (from here we will call this threshold the sign change threshold). The changed sign of the coefficient signals very strong evidence for the existence of a human capital threshold for the absorption of FDI technology spillovers. The ideas of Grilliches (1979) may help us to understand the changed sign of the coefficient. Grilliches argues that purchases of foreign hightech goods are not real knowledge spillovers. He insists that true spillovers are ideas borrowed by research teams. Thus, the change of the coefficient sign is similar to the change from the purchase of goods to imitation of goods or independent innovations. As the quality of human capital increases and surpasses the sign change threshold, domestic firms are able to study the foreign products more thoroughly and get benefits from imitation. This also improves their ability to resist the brain-drain process as their profits increase. So the positive sign shows that spillover effects substitute crowding out effects. For Mid-west China, when the human capital is below the single threshold of 9.64 per cent, the negative crowding out effects are significant. This is similar to all other regions. However, when the human capital surpasses that threshold, no significant spillover effects appear. This discrepancy suggests that technology progress trajectories for Midwest China should be different from those for East China. Furthermore, the difference does not deny the positive effects of human capital because the coefficient of crowding out effects changes from significant to insignificant. 12.4.2

The realities of who satisfies the thresholds

Turning to another side of the threshold studies, we will investigate the human capital of different regions in more detail and judge whether they satisfy the estimated thresholds. Table 12.5 presents the human capital situation for the whole country. It comprises the promotion rates and rates of educated workforce expressed as percentages. 1978 was the starting point for Chinese reformation. The population educated since that date forms the major part of the current Chinese workforce. Thus, Table 12.5 gives a description of the active workforce in China. The numbers in the table are usually higher than those estimated for the whole population because the older generations who grew up before 1978 received little education on average.1 The average percentage of



labour with a higher education shown in Table 12.5 is 6.64 per cent, which is below the threshold of 10.99 per cent. Regressing the model using FDI directly without thresholds, we find that the effects of FDI on technological progress are negative and present crowding out effects, which arise because the level of human capital is below the estimated threshold. Comparing the thresholds to those in Borensztein et al. (1998) and Xu (2000), our threshold is that for active labour with higher education whereas Borensztein et al. and Xu’s thresholds are in terms of average secondary schooling years for male adults including older generations. As regards technology spillovers, the active workforce is more important than the retired population. Including the retired population in human capital estimations in developing countries will underestimate the true level of labour quality, especially in rapidly developing countries whose educational structures for older generations are completely different from those of the younger generations. Further, we do not

Table 12.5 Human capital in China: Average promotion rates and proportion of educated workforce Promotion rate Net enrolment rate of school-age children Primary school graduates entering junior secondary schools Junior secondary graduates entering senior secondary schools Senior secondary graduates entering higher education institutions _

Proportion of Percentage educated workforce 97.21 69.63

45.31

Workforce receiving no education Workforce receiving only primary education Workforce with junior secondary diploma

Percentage 2.79 29.53

37.01

21.65

Workforce with senior secondary diploma

24.03

_

Workforce with higher educational diploma

6.64

Notes: Data sources are the China Statistical Yearbooks and all values averaged from 1978 to 2006. The promotion rate is calculated against graduates instead of entrants to eliminate students who have dropped out. Senior secondary education comprises regular senior secondary education, regular specialized secondary education, technical education and vocational senior secondary education because our study focuses on the workforce with different educational levels. Thus our rate of senior secondary graduates entering into higher education institutions is lower than the admission rate provided in the China Statistical Yearbooks which consider only regular senior secondary graduates.



251

think that secondary school attainment is a better measurement of the absorptive capacity of FDI technology spillover than higher education. Since China now implements a policy of nine years of compulsory education, which includes primary and junior secondary education, it is easy for the average secondary education year to surpass Borensztein et al. and Xu’s thresholds. Human capital with a higher education, not just a secondary education plays a key role in the absorption of FDI technology spillovers. If we judge the spillover effects by secondary education thresholds, or even by junior secondary education thresholds, it may produce plausible results. In the past, the large percentage of labour which received primary and secondary education led to the prosperity of low-tech manufacturing industries in China. Yet the lack of highly qualified human capital impeded independent domestic innovations in the country. Therefore, we use the proportion of labour with a higher education as a threshold variable in our research. Although China as a whole does not satisfy the sign change threshold, some regions in the country do meet this minimum requirement for human capital. Table 12.6 gives the percentages for provinces in the three regimes separated by the two thresholds. Before 1998, most regions are below the first threshold of 4.85 per cent, and the major effects of FDI on technology progress are crowding out ones. After 1999, the human capital of most provinces surpasses the first threshold. Most importantly, the percentage of provinces above the sign change threshold of 10.99 per cent is increasing with time. In 2005, 37.59 per cent of Chinese provinces have surpassed this threshold and acquire benefits from the positive technology spillovers of FDI. In Mid-west China, the percentage of provinces above the sign change threshold is relatively lower than that of all other regions, but catching up is taking place and the percentage for the Mid-west regions is very close to that for all regions in recent years. In 2005, 33.33 per cent of the Mid-west provinces have surpassed the sign change threshold. 12.4.3 Interregional technology spillovers as the complement of international spillovers As we know that some regions surpass the sign change threshold and others do not, the remaining problem is how the regions below the threshold benefit from foreign technology spillover. The answer can be discovered by studying the estimated coefficients of interregional technology spillover, which is represented by the sum of other regions’ R&D capital stock. In Table 12.4, this coefficient is significant and this suggests that the region of interest adopts technologies transferred


252

Miao Fu and Tieli Li Table 12.6 Percentage of provinces in the three regimes separated by thresholds Year

E ≤4.85%

4.85%<E≤10.99%

10.99%<E

10.99%<E (Mid-west)

1990 1991

75.87 72.42

13.79 17.24

10.34 10.34

0 0

1992

65.52

24.14

10.34

0

1993

65.52

24.14

10.34

0

1994

65.52

24.14

10.34

0

1995

62.07

27.59

10.34

0

1996

58.63

31.03

10.34

0

1997

55.18

34.48

10.34

0

1998

51.73

34.48

13.79

0

1999

41.38

41.38

17.24

5.56

2000

37.93

44.83

17.24

5.56

2001

31.04

51.72

17.24

5.56

2002

17.24

55.17

27.59

22.22

2003

13.79

58.62

27.59

22.22

2004

13.80

51.72

34.48

27.78

2005

3.45

58.62

37.93

33.33

Notes: E is the human capital represented by ratio of workers receiving a higher education. The last column applies to Mid-west China only; other columns are for all regions.

from other regions. We can find that both all regions and the Midwest area benefit significantly from interregional technology spillovers. Moreover, the coefficient of Mid-west China is 0.1061, and thus much bigger than the 0.0535 of all regions. For the Mid-west provinces, it is difficult to absorb foreign technology spillover directly because of the relatively low quality of human capital, but the quality gap of human capital between advanced regions and backward regions in the same county is usually narrower than the international one, and backward regions are capable of getting technology transfers from other domestic regions more easily than from international sources. Interregional technology transfers involve the distribution of FDI technologies already



253

absorbed by the advanced regions. The bigger interregional spillover coefficient of backward areas demonstrates that backward areas benefit more from interregional technology spillovers. 12.4.4 Other explanations for technology progress In Table 12.4, for all regions, the coefficient of local R&D stock is significantly positive, and the coefficient of local R&D stock in the Mid-west provinces almost equals the coefficient for all regions. This result indicates that independent R&D is relevant to local technology progress and roles of local R&D are similar between East and Mid-west areas, in spite of economic disparities between the regions. This is partially because the efficiency of R&D depends on some non-economic factors such as social institutions, intelligence levels and research cultures which are not so variable across regions. The coefficient of R&D labour is negative but insignificant both in all regions and in the Mid-west area, reflecting that the R&D labour force does not affect the technology progress significantly. This result contradicts hypothesis 1. The explanation for this may be broken linkages from research institutes to industrial production and an ineffective labour force in the R&D sector. For this reason, the dependent variable of TFP does not reflect technologies in theoretical fields, but represents the technologies in use. The broken linkage becomes even worse if the R&D labour force works alone without sufficient support in the form of R&D capital. Infrastructure is important for TFP growth because it improves the technology efficiency embodied in TFP. There is significant evidence for this assertion. The coefficient of infrastructure is 0.0945, which is bigger than those of the R&D capital stock and the R&D labour force. This result demonstrates the salient feature of Chinese development. As China is a big country and the cost of transport is one of the major concerns considered by investors, local governments pay increasing attention to the construction of public infrastructure to attract investment. The public transport infrastructure for can mitigate those problems caused by unequal distribution of natural and human resources. Our empirical results support these infrastructure construction policies. Comparing all regions with the Mid-west area, the coefficients of infrastructure are bigger for all regions than in the Mid-west, which means that newly added infrastructure will be more valuable in developed regions that already have a highly connected network. However, it is true that most Chinese governments invest more in infrastructure than in public R&D. As the transport infrastructure will gradually become more concentrated, this implies that policy-makers



should move their focus from construction to innovative technology progress in the future.

12.5 Conclusions Differences exist about the technology spillover effects of FDI. Some researchers argue that it is positive and others insist that it is insignificant or even negative. The human capital threshold analysis of the absorptive capacity of FDI technology spillovers will probably give the conditions needed to take in the two different views. Based on threshold regression, which includes the threshold variable endogenously, we gain the following results. In respect of human capital as absorptive capacity of FDI technology spillovers, two thresholds are found in terms of the percentage of labour with a higher education. There are negative effects when human capital is below the first threshold of 4.85 per cent, and these negative effects abate, nearly halving in the absolute value once this threshold is surpassed. The sign change threshold that transforms the negative effects into positive effects is 19.99 per cent. The sign change threshold has very important meaning for policy-makers. As most developing countries, and even some developed ones, are below the sign change threshold, the first step to change this situation is to know which countries or regions within a country do not reach it. Second, we find significant interregional technology spillovers. In comparison with FDI spillovers, positive interregional spillovers are pure effects and do not depend on other absorptive variables such as human capital. In addition, the coefficient of backward regions is bigger than that of advanced regions for interregional technology spillovers. Considering that some of the advanced areas are above the threshold, it is possible that even though a country as a whole is below the sign change threshold, positive FDI technology spillovers still take place in some advanced regions, and then the foreign technologies assimilated by advanced regions can be transferred to backward regions by interregional spillovers, and China is an instance of this. Thus, our threshold does not deny the possibilities of FDI technology spillovers in developing countries whose qualities of human capital are below the threshold. Furthermore, in China, the number of regions above the sign change threshold keeps growing. In addition to external technology spillovers, internal factors, such as R&D capital stock and public infrastructure also have significant positive effects. Public infrastructure, especially, has a bigger coefficient



255

than other explanatory variables, except positive FDI spillovers. In contrast with the coefficient of interregional spillovers, that of public infrastructure in advanced regions is greater than that in backward regions. This suggests that an add-on to a matured infrastructure network will get better connecting advantages. This paper gives clear human capital targets for developing countries to reach, providing a threshold for mitigating the crowding out effects of FDI and another threshold to overcome them, and turn the effects into positive ones. The thresholds are applicable as they are represented in terms of the proportion of labour with a higher education and a well educated labour force forms the majority of the population engaged in technology progress. This human capital index is easier to calculate because it involves only higher educational levels. The index also decreases the risk of misleading policy. Our conclusions on both positive FDI technology spillovers in technologically advanced regions and significant interregional technology transfers encourage the efforts of countries currently below the sign change threshold. Those developing countries can rely on their more advanced regions to absorb new technologies and then distribute them to backward regions. Higher returns to FDI spillover and public infrastructure in more developed areas and the higher interregional spillover coefficient in relatively backward areas also suggest different paths of technology progress in more and less developed areas.

Note The earlier version of paper was supported by UNIDO and this research is also supported by the Ministry of Education of P.R.C., HSSR Project (Grant No. 08JA790028). We are indebted to Xiaolan Fu for her help and grateful for the comments from Jaime Moll De Alba Cabot. Comments by participants in the conference of Confronting the Challenge of Technology for Development in Oxford University were helpful. 1. For example, there were 79,000 entrants to tertiary schools in 1952, and 5,461,000 in 2006. According to data from the World Bank, the average number of years in education for the Chinese population over the age of 15, including older generations, was 6.36 in 2000.

References Aghion, P. and P. Howitt (1992), ‘A model of growth through creative destruction’, Econometrica, 60 (2), March, 323–51.


256 Miao Fu and Tieli Li Blomstrom, M. and A. Kokko (1997), ‘How foreign investment affects host countries’. World Bank PRD Working Paper, 1745, Washington, D.C.: World Bank. Borensztein, E., J. Gregorio and J. Lee (1998), ‘How does foreign direct investment affect economic growth?’ Journal of International Economics, 45, 115–35. Branstetter, L.G. (2001), ‘Are knowledge spillovers are international or intranational in scope? Microeconometric evidence from the U.S. and Japan’, Journal of International Economics, 53 (1), 53–79. Coe, D.T. and E. Helpman (1995), ‘International R&D spillovers’, European Economic Review, 39, 859–71. Démurger, S. (2001), ‘Infrastructure development and economic growth: An explanation for regional disparities in China’, Journal of Comparative Economics, 29 (1), March, 95–117. Eaton, J. and S. Kortum (1996), ‘Trade in ideas: Patenting and productivity in the OECD’, Journal of International Economics, 40 (3–4), 251–78. Fleisher, B., H. Li and M.Q. Zhao (2007), ‘Human capital, economic growth, and regional inequality in China’. IZA Discussion Papers, IZA DP 2703. Fu, Feng-Cheng, C.C. Vijverberg and W.P.M. Vijverberg (2004), ‘Public infrastructure as a determinant of intertemporal and interregional productive performance in China’. IZA Discussion Paper 1019, February. Fu, Xiaolan (2007), ‘Trade-cum-FDI, human capital inequality and regional disparities in China: The singer perspective’, Economic Change and Restructuring, 40 (1–2): 137–55. Girma, S. (2005), ‘Absorptive capacity and productivity spillovers from FDI: A threshold regression analysis’, Oxford Bulletin of Economics and Statistics, 67 (3), 281–306. Grilliches, Z. (1979), ‘Issues in assessing the contribution of research and development to productivity growth’, Bell Journal of Economics, 10 (1), 92–116. Griliches, Z. (1990), ‘Patent statistics as economic indicators: A survey’, Journal of Economic Literature, 28 (4), December, 1661–707. Grossman, G.M. and E. Helpman (1991), Innovation and Growth in the Global Economy. Cambridge, MA: MIT Press. Hansen, B.E. (1996), ‘Inference when a nuisance parameter is not identified under the null hypothesis’, Econometrica, 64, 413–30. Hansen, B.E. (1999), ‘Threshold effects in non-dynamic panels: Estimation, testing, and inference’, Journal of Econometrics, 93, 345–68. Keller, W. (1996), ‘Absorptive capacity: On the creation and acquisition of technology in development’, Journal of Development Economics, 49, 199–227. Keller, W. (2004), ‘International technology diffusion’, Journal of Economic Literature, 42, September, 752–82. Kokko, A., R. Tansini and M.C. Zejan (1996), ‘Local technological capability and productivity spillovers from FDI in the Uruguayan manufacturing sector’, Journal of Development Studies, 32 (4), 602–11. Nelson, R. R. and E.S. Phelps (1966), ‘Investment in humans, technological diffusion, and economic growth’, American Economic Review, 56 (2), 69–75. Romer, David (1996), Advanced Maroeconomics (2nd edition). Boston, London: McGraw-Hill Companies, Inc. Romer, Paul M. (1990), ‘Endogenous technological change’, Journal of Political Economy, 98 (2), October, 71–102.



257

Xu, Bin (2000), ‘Multinational enterprises, technology diffusion, and host country productivity growth’, Journal of Development Economics, 62 (2), 477–93. Zhang, Jun and S. Shi (2003), ‘Change of total factor productivity in Chinese economy: 1952–1998’, World Economic Forum (China), 2, 17–24. Zhang, Jun, G. Wu and J. Zhang (2004), ‘The estimation of China’s provincial capital stock: 1952–2000’, Economic Research Journal (China), 10, 35–44.


13 Transnational Corporations from Emerging Economies and South-South FDI Torbjörn Fredriksson

13.1

Introduction

The rise of transnational corporations (TNCs) from developing and transition economies has accelerated in recent years. The presence of such companies on the world stage was documented several decades ago, by such distinguished scholars as Louis Wells (1977; 1983) and Sanjaya Lall (1983). The situation in the 1970s and 1980s differs, however, in a number of ways from that of the present time. The scale of the phenomenon is today much larger and the geographical composition of flows and stocks is different as are the drivers and determinants. This chapter reviews some recent developments with respect to the growing importance of TNCs from the South and their overseas expansion. Section 13.2 presents data from various sources to shed light on the nature and scope of the phenomenon. Section 13.3 examines some salient features of the recent surge of foreign direct investment (FDI) by these firms. Section 13.4 examines some policy implications for governments in developing as well as developed countries.

13.2 The rise of Southern TNCs accelerates The overseas expansion by TNCs from the South has accelerated at record speed in the past few years. In the United Nations Conference on Trade and Development’s most recent ranking of the 100 largest TNCs in the world, seven of them were based in developing countries (UNCTAD, 2009). In terms of foreign assets, the largest among the seven, Hutchison Whampoa (Hong Kong), was in 2007 ranked number 258


Transnational Corporations

259

22 in the world, with foreign assets of US$83 billion, just above the Honda Motor Co. (Japan). Among the BRICS countries (Brazil, Russia, India, China and South Africa), only China was represented among the world’s top 100 TNCs in 2007.1 Other corporate rankings give further testimony to the rapid growth of companies in the South. For example, in 1990, only 18 companies from the South were represented among the Fortune Global 500 rankings. By 2005, that figure had risen to 47, and in the subsequent two years, another 30 companies entered the list. As a result, in the 2008 list, 77 (or more than 15 per cent) of the largest 500 companies were based in developing or transition economies (Figure 13.1).2 This rapid corporate growth is reflected in data on foreign direct investment (FDI). According to UNCTAD, the stock of outward FDI from developing and transition economies almost tripled between 2000 and 2008, from US$0.9 trillion to US$2.6 trillion (UNCTAD, 2009). During the same period, annual FDI outflows from these economies shot up from US$102 billion to almost US$350 billion. The trend of internationalization of companies is not confined to only a few countries. In fact, the number of countries with an outward FDI stock of more than US$5 billion jumped from seven in 1990 to 31 in 2008 (Figure 13.2). With regard to the BRICS countries, outward FDI started to grow in earnest only after the turn of the century (Figure 13.3). Since 2002, the most rapid expansion has been witnessed for FDI from China and

100 90 80 70 60 50 40 30 20 10 0 1990

2005

2006

2007

2008

2009

Figure 13.1 Number of entries from developing and transition economies among the list of Fortune Global 500 companies, 1990, 2005–9 Source: Fortune Global 500.


260 Torbjörn Fredriksson

35 30 25 20 15 10 5 0 1980

1990

2000

2006

2008

Figure 13.2 Number of developing and transition economies with a stock of outward FDI valued at more than US$5 billion, 1980, 1990, 2000, 2006 and 2008 Source: UNCTAD.

60

Brazil China India Russia South Africa

50 40 30 20 10 0 −10

1995

Figure 13.3

1997

1999

2001

2003

2005

2007

Outward FDI from the BRICS, 1995–2008 (US$ billions)

Source: UNCTAD FDI/TNC database.

Russia. In 2008, these two economies both experienced outflows of more than US$50 billion (UNCTAD, 2009).3 Brazil and India have also experienced significant increases in their outflows, while those from South Africa have remained below US$10 billion. Geographically, the rise of TNCs from the South is today predominantly an Asian phenomenon. In this respect, it differs considerably from the situation of the 1970s and 1980s, when Latin American firms



261

were the most active outward investors from the South. Indeed, while Latin America in 1970 accounted for two-thirds of the outward FDI stock from the South, by 2008, it was Asia that held a similar share of that stock (UNCTAD, 2006; 2009). Asia’s dominance is confirmed in various other statistics. Out of the seven Southern TNCs featuring among the top 100 TNCs in the world, six are Asian and only one is from Latin America. In UNCTAD’s most recent ranking of the top 100 TNCs from developing countries, Asia accounted for 82 of the entries, while Africa and Latin America had nine entries each (UNCTAD, 2009). The BRICS accounted for 23 of the entries if Hong Kong is excluded and 50 if it is included. And in the 2009 list of Fortune Global 500 companies, 70 of the 89 entries from developing or transition economies were from Asia, 11 from Latin America and eight from Russia (Table 13.1). Moreover, between 2007 and 2009, 15 new entries were noted for Asia (most of which Chinese companies), while Latin American firms increased by only one entry.

Table 13.1 Number of companies from developing and transition economies among the Fortune Global 500 companies, 2007–9 Number of companies Home economy Africa Asia and Ocenia China Republic of Korea India Taiwan Province of China Malaysia Saudi Arabia Singapore Thailand Turkey Latin America Brazil Mexico Venezuela South-east Europe and the CIS Russian Federation Total

2007 0 55 24 14 6 6 1 1 1 1 1 10 5 5 0 4 4 69

2008 0 62 29 15 7 6 1 1 1 1 1 10 5 5 0 5 5 77

2009 0 70 37 14 7 6 1 1 2 1 1 11 6 4 1 8 8 89

Note: China here includes Hong Kong. Source: Fortune Global 500.


262

Torbjörn Fredriksson

In terms of sectoral distribution, TNCs from the South have appeared in a wide range of industries, with large regional variations. As highlighted in the World Investment Report 2006, some of them have taken global positions in the automotive industry, chemicals, electronics, petroleum refining and steel, and in services such as banking, shipping, information technology (IT) services and construction (UNCTAD, 2006). In some specific segments, such as container shipping, port management and petroleum refining TNCs from the South have a particularly strong presence. In all developing regions and in Russia, major TNCs have emerged in the primary sector (oil, gas and mining) and resource-based manufacturing (metals and steel). Some of them are now competing head-on with their developed-country rivals. Examples include Sasol (South Africa); Vale (Brazil); ENAP (Chile); Petrobras (Brazil); Petroleos de Venezuela (Venezuela); Baosteel, CNPC and CNOOC (China); Petronas (Malaysia); Posco (Republic of Korea); PTTEP (Thailand) and Gazprom and Lukoil (Russia). Another important cluster relates to the production of goods and services that are generally difficult to export and therefore require FDI in order to serve foreign markets. Examples include financial services, infrastructure services (electricity, telecommunications and transport) and such goods as cement, food and beverages. Because of their nontradable nature, these economic activities typically require FDI in order to serve foreign markets. With a few exceptions (such as Cemex), however, most of the Southern TNCs in these areas are primarily regional players, with limited (if any) activities in parts of the world farther from home. A third cluster of activities comprises those sectors that are the most exposed to global competition, such as the automotive industry, electronics (including semiconductors and telecommunications equipment), garments and IT services. Almost all major TNCs from the South in these industries are of Asian origin. Electronics companies such as Acer (Taiwan), Huawei (China) and Samsung Electronics (South Korea); automobile firms such as Hyundai Motors and Kia Motors (South Korea) or smaller TNCs in the IT services industry, such as Infosys and Wipro Technologies (India), are already leaders in their respective areas.

13.3 Drivers and modes The internationalization of TNCs from the South has been driven by a diverse set of factors, ranging from the traditional quest for access to



263

natural resources, production factors or foreign markets to the desire to acquire technological and other strategic assets. Compared with earlier waves of FDI from developing countries, it appears that the current wave involves a greater element of ‘asset-augmenting’ as opposed to ‘asset-exploiting’ motives for going abroad (UNCTAD, 2006). In a globalizing world economy, with high levels of competition and rapid technological progress, tapping into foreign centres of excellence or gaining control of assets controlled by foreign firms can be essential for sustaining or enhancing a company’s competitiveness. Outward FDI is an important mode of tapping into other countries’ national innovation systems. TNCs from South Korea and Taiwan have long located R&D centres in the US, Europe and Asia in order to gain access to new technologies. Such technologies have been applied in the home country to develop new products and processes for global markets. More recently, companies from China and India have set up R&D units in the US and Europe (see below). The search for technology and other strategic assets is also reflected in the surge of cross-border mergers and acquisitions (M&As) with TNCs from the South on the acquiring side. The value of foreign takeovers by such companies shot up from less than US$30 billion in 2002 to almost US$130 billion in 2006 (Figure 13.4). Some of the largest recent such deals include Vale’s (Brazil) acquisition of Inco (Canada), Cemex’s

140,000 120,000

South-east Europe and the CIS Developing economies

US$ millions

100,000 80,000 60,000 40,000 20,000 -

2000

2001

2002

2003

2004

2005

2006

Figure 13.4 Cross-border M&A purchases by companies based in developing or transition economies, 2000–6 (US$ millions) Source: UNCTAD.



(Mexico) takeover of Rinker (Australia) and Tata Steel’s (India) purchase of Corus (UK/Netherlands). Another good illustration is Lenovo’s (China) takeover of IBM’s PC division, which served to help the bidder to establish itself as a global brand while at the same time acquiring vital technology expertise. Three trends, which have been specifically addressed in UNCTAD’s research in the past few years, deserve particular attention from the perspective of the forces driving TNCs from the South to expand internationally. The first relates to the quest for natural resources, such as metal ores, oil and gas (UNCTAD, 2007), he second is the growing importance of Southern TNCs in the area of infrastructure (UNCTAD, 2008) andnd the third is the trend toward internationalization of research and development (R&D) by firms in developing countries (UNCTAD, 2005). The three trends provide an illustration of the diversity and complexity of the drivers behind the recent wave of FDI from the South. 13.3.1

Natural-resource seeking TNCs

As highlighted in the World Investment Report 2007, the search for natural resources has accelerated in very recent years against the backdrop of rising commodity prices, which in turn have been influenced by the high demand for metals and energy in emerging economies, and especially in China. Although national oil companies control most of the global production of oil and gas, the degree to which they are internationalized is typically low in contrast to that of the top oil majors. However, some companies from the South are fast becoming global players (Figure 13.5). The combined overseas production of CNOOC, CNPC, Sinopec (all China); Lukoil (Russia); ONGC (India); Petrobras (Brazil) and Petronas (Malaysia) exceeded 528 million barrels of oil equivalent in 2005, up from only 22 million barrels ten years earlier (UNCTAD, 2007). Both CNPC and Petronas are now involved in oil and gas production in more than ten foreign countries. One distinctive feature of the ‘emerging’ oil TNCs is that they are state-owned enterprises (SOEs). Their overseas expansion often reflects strategic national interests in addition to purely commercial ones. They often enjoy financial backing from their governments that may allow them to assume greater risks when investing abroad and to pay more for access to mineral resources. For example, some of these state-owned oil TNCs have invested in host countries that developed-country TNCs are less likely to operate in, for a variety of reasons, including trade sanctions. Moreover, in 2006, a new record in signature bonuses was reached when Sinopec outbid its competitors by paying US$2.2 billion



1427

ExxonMobil BP Royal Dutch Shell Total ENI Chevron ConocoPhillips Repsol-YPF CNPC/Petro China Inpex BG Petronas Petrobras Statoil Sinopec CNOOC Lukoil Norsk Hydro ONGC

1291 1045 749 584 550 512 366 188 129 114 98 66 53 49 46 46 35 35 0

Figure 13.5 try, 2005

265

200

400

600

800

1000

1200

1400

Oil and gas production of selected TNCs outside their home coun-

Source: UNCTAD (2007), figure IV.10.

in return for the right to explore for oil in two Angolan exploration blocks.4 13.3.2 Infrastructure TNCs The second trend relates to the growing role of infrastructure TNCs from the South. When UNCTAD ranked the top 100 infrastructure TNCs in the world in 2006, as many as 22 companies were based in a developing country or in Russia (UNCTAD, 2008). As noted above, TNCs from the South claimed only seven of the top 100 slots in UNCTAD’s global ranking of TNCs. Within the field of infrastructure, TNCs from the South are particularly strong in the areas of transport and telecommunications. In some cases, massive infrastructure commitments have been made in return for access to natural resources.5 For example, in September 2007, China agreed to invest in big infrastructure projects in the Democratic Republic of Congo (DRC), to be paid for with the host country’s reserves of copper and cobalt. The Chinese companies would start work on infrastructure projects in 2008 in such areas as water, electricity, education, health and transport.6 These works will cost an estimated US$9 billion. The idea is for Congolese and Chinese SOEs to set up a joint venture, Socomin, that is incorporated under Congolese law. This mining company will invest US$3 billion in mainly new mining areas and another correcting loan of US$100 million, with which the back pay of foreign


266


and Congolese ex-employees, among others, can be paid. The profits of Socomin will be used to repay the investments in mining and big infrastructure projects. The infrastructure will be developed by two Chinese SOEs (Sinohydro and the China Railways Engineering Company).7 Under the deal, the Chinese investors will receive loans directly from the Chinese Exim Bank, thereby circumventing the bureaucracy of the DRC. In this way, the DRC’s mineral resources will be used as collateral for infrastructure investment. The agreement between the DRC and China illustrates how the rise of Southern TNCs in a number of infrastructure industries offers new pools of capital and expertise into which low-income can countries. At the same time, it is important to engage these new players in ongoing processes aimed at securing sustainable development gains from international investments. TNCs from the South already make significant contributions to infrastructure development in developing countries. In Asia, almost half of the foreign investment commitments originates from developing countries. In Africa, developing-country TNCs dominate foreign investment in telecommunications and account for a considerable share of all foreign infrastructure investment. In least developed countries (LDCs), their role is most pronounced in transport, in which such companies as DP World and Hutchison Whampoa have pushed up the share of South-South flows to as much as 65 per cent of all foreign investment commitments. 13.3.3 Internationalization of R&D R&D is one of the business functions that have traditionally been the least internationalized. However, this pattern has shown signs of rapid change (UNCTAD, 2005). For example, for Swedish TNCs, the share of R&D conducted abroad rose from 22 to 43 per cent between 1995 and 2003; German firms established more overseas R&D centres in the 1990s than in the preceding 50 years together (Ambos, 2005) and in a cross-country survey, the proportion of foreign to total R&D rose from 15 per cent in 1995 to 22 in 2001 (Roberts, 2001), a trend that has also been confirmed in other studies (Edler et al., 2002; von Zedtwitz and Gassmann, 2002). Moreover, this trend has not been confined to firms in developed countries. As part of their search for foreign technology, a growing number of TNCs from the South are also engaging in R&D internationalization. This involves investments in both developed countries (in which case the aim might be to ‘catch up’ with developed-country



267

competitors) and in other developing countries (in which case the goal may be to support local business activities).8 There are many examples of companies that have targeted the knowledge base of developed countries (UNCTAD, 2005): Samsung (South Korea) has set up R&D laboratories in Europe; Acer (Taiwan) has laboratories in the US; Ingenuity Solutions (Malaysia) has invested in R&D in the US; Bionova (Mexico) has acquired DNA Plant Technology (US); and Cordlife (Singapore) has acquired Cytomatrix (US). There are similarly many examples of South-South R&D-related FDI: Acer has set up R&D laboratories in China; Huawei (China) has an R&D centre in Bangalore, India; A number of firms from South Korea, Singapore, Malaysia and Thailand have set up R&D activities in India, particularly related to software development (Reddy, 2000). Samsung Electronics (South Korea) has opened R&D centres in China and India (as well as in Russia); Bogasari International (Indonesia, food processing) chose Singapore partly because of the country’s favourable R&D incentive schemes for foreign investors. Chinese TNCs are becoming international players in R&D. A recent study of large Chinese companies found that they had at least 37 R&D units abroad at the end of 2004 (von Zedtwitz, 2005). Twenty-six of these foreign R&D units were located in developed countries, predominantly in the US (11) and Europe (11), mostly serving as observation posts or in product design roles. The remaining 11 units are located in developing countries. Huawei9 and Haier,10 are illustrative of the trend, with R&D units located mostly in developed countries. Other Chinese companies from the electronics industry, such as ZTE and UTStarcom, have established R&D centres in India, aiming essentially at offshore (embedded) software development. Indian TNCs are also globalizing their R&D, focusing mainly on serving their customers in specific regional markets. The leading software firms have all invested abroad, mostly in developed countries. For example, Infosys, Wipro, Birlasoft (part of Aditya Birla Group) and HCL Technologies have set up operations in the US. They are also moving into China and other emerging markets.11 Some Indian



software R&D affiliates are located in other developing regions – Tata, for example, invested in Latin America and the Caribbean (Uruguay) – and in new EU member countries (Hungary). There are also examples of FDI in R&D in other industries, such as pharmaceuticals and chemicals.12

13.4 Most of their outward expansion targets developing countries While there has been increased interest among TNCs in the South in acquiring assets in developed countries, most of their investment to date has flowed into other developing countries. While data in this area are admittedly patchy, UNCTAD has found that South-South FDI has expanded particularly rapidly over the past two decades. Total outflows from developing and transition economies (excluding offshore financial centres) increased from about US$4 billion in 1985 to US$61 billion in 2004, primarily as a result of inflows into developing economies. In fact, the South-South part of these FDI flows increased from US$2 billion in 1985 to US$60 billion in 2004 (UNCTAD, 2006). The emergence of these new sources of FDI may be of particular relevance to low-income host countries. In all three trends highlighted in Section 13.3 – natural resources, infrastructure and internationalization of R&D – South-South investment features prominently. TNCs from the South have emerged as significant investors in many LDCs. Developing countries with the highest dependence on FDI from developing and transition economies include China, Kyrgyzstan, Paraguay and Thailand, and LDCs such as Bangladesh, Ethiopia, Laos, Myanmar and Tanzania. Indeed, South-South FDI accounts for well over 40 per cent of the total inward FDI of a number of LDCs (Table 13.2). In addition, the level of FDI from developing and transition economies to LDCs may well be understated in official FDI data, as a significant proportion of such investment goes into the informal sector. The bulk of South-South FDI (excluding offshore financial centres) is intraregional in nature. In fact, during the 2002–4 period, average annual intra-Asian flows amounted to an estimated US$48 billion. The next largest stream of FDI within the group of developing countries was within Latin America, mainly driven by investors in Argentina, Brazil and Mexico. Intraregional flows within Africa amounted to an estimated $2 billion reflecting, in particular, South African FDI to the rest of the continent. For example, South Africa accounts for more than 50 per cent of all FDI inflows into Botswana, the DRC, Lesotho, Malawi



269

Table 13.2 FDI from developing and transition economies in selected LDCs, various years (%) Flows

Recipient economy Bangladesh Cambodia Ethiopia Laos Madagascar Mozambique Myanmar Nepal Solomon Islands Tanzania Uganda Vanuatu Zambia

Stock

Period/year

Share in total FDI (%)

Year

Share in total FDI (%)

1995–7 2002–4 1995–7 2002–4 1992–4 2002–4 1995–7 2002–4 2003 2004 2003 2004 1995–7 2002–4 1990–2 1996–8 1994–6 – 1999–2001 – – 1999 2000–2 – –

9 39 63 64* 100 51 93* 45* 29 54 103** 47** 39* 56* 46** 65** 56** – 41 – – 7 19 – –

1996 2001 1994 2002 1994 – 1990 1999 2003 2004 – – 1990 2004 1990 1999 – 1996 2001 1999 2003 – – 2000 2001

17 13 81 73 77 – 47** 70** 27 36 – – 33** 61** 58** 63* – 36 44 46 36 – – 21 20

Notes: * Based on information provided by the Secretariat of the Association of Southeast Asian Nations (ASEAN). ** Data are on an provisional basis. Source: UNCTAD (2006).

and Swaziland. Interregional South-South FDI has gone primarily from Asia to Africa, while the second largest has been from Latin America to Asia. Perhaps somewhat surprisingly, total flows from Asia to the Latin American region were modest during the 2002–4 period, while those between Latin America and Africa have until recently been negligible (UNCTAD, 2006). However, rising interest Africa’s natural resources has provoked new Latin American investments into Africa, for example in the area of biofuel production.


270


More recently, we have seen the emergence of new forms of SouthNorth FDI, as illustrated by investment by sovereign wealth funds (SWFs). Fuelled by rising current account surpluses, as a result of oil revenues or revenues from exports of manufactures, a number of SWFs in developing countries have become more willing also to invest in direct investment assets. Some have welcomed the capital injections of SWFs into troubled banks (including Citibank and UBS) in certain developed countries, while others have voiced concerns related to these acquisitions. Similar concerns have been raised vis-à-vis takeovers, or attempted takeovers, by SOEs from developing countries. State ownership has been perceived as increasing the risk that a transaction is undertaken for other than purely economic motives. It appears that concerns have been particularly pronounced when acquisitions are related to energy, infrastructure services or other industries with a potential ‘national security dimension’ (UNCTAD, 2006; 2007).

13.5 Policy implications for developing and developed countries 13.5.1

Developing home countries

The rise in FDI from the South reflects considerable policy changes in those countries that have emerged as important sources for it. More and more developing and transition economies are dismantling barriers to outward FDI. While some form of capital control is often still in place to mitigate the risk of capital flight or financial instability, remaining restrictions mostly aim at limiting international capital flows other than FDI. Just a handful of countries retain outright bans on outward FDI. In fact, a number of governments, especially in developing Asia, are even encouraging their firms to invest abroad. Active promotion of outward FDI is most common in South, East and South-east Asia.13 In several countries of this region, governments discharge their promotional policies via trade promotion organizations, investment promotion agencies (IPAs), export credit agencies and export-import (EXIM) banks. A variety of supportive measures are used, including information provision, matching-up services and financial or fiscal incentives, as well as insurance coverage for overseas investment. But there is no one-size-fits-all approach that can be recommended for dealing with outward FDI. Every home country has to adopt and implement policies that fit its specific situation. Whether it will benefit by moving from ‘passive liberalization’ to ‘active promotion’ of outward FDI depends on many factors, such as the capabilities of the its



271

enterprises, and the links between the investing companies and the rest of the economy. As reflected in the various factors driving firms to internationalize, outward FDI can bring improved access to foreign markets, resources and strategic assets. Arguably, the most important potential gain for home countries from outward FDI is the improved competitiveness and performance of the firms and industries involved. Such gains may contribute to industrial transformation and upgrading of valueadded activities, improved export performance, higher national income and better employment opportunities. The improved competitiveness of outward investing TNCs can be transmitted through linkages and spillovers to local businesses and other institutions such as universities and research centres. The more embedded the outward investing TNCs are, the greater will be the expected benefits for the home economy. At the same time, for large technological and competitive benefits to materialize in a home economy, local firms and institutions need to have the necessary absorptive capacity. This makes it important for policy-makers to nurture the development of technological and productive capabilities in the home country. In many low-income countries, it may indeed be more fruitful to focus on creating an attractive business environment and enhancing domestic firm capabilities at home rather than encouraging existing firms to go abroad. The growth of FDI from the South also has implications for the role of international investment agreements (IIAs). Increased FDI from developing economies is likely to generate greater demand from the business community in these economies for protection of their overseas investments. As a consequence, the focus of developing-country governments may swing from a singular emphasis on using IIAs to facilitate inward investment toward protection of outward FDI. This may influence the substantive content of future treaties as well as generate calls for the renegotiation of existing treaties (UNCTAD, 2006). The political and social aspects of their activities may give rise to controversy, partly because of the size of TNC operations. In developing host economies, such problems have sometimes been exacerbated by the absence of an adequate regulatory framework and disparity in the allocation of economic benefits from inward FDI. It is essential that TNCs that have just begun their overseas expansion – as well as their home country governments – quickly learn to manage their foreign operations in accordance with internationally acceptable social and economic standards. They will have to become used to addressing various issues related to corporate responsibility. A recent review of the top mining and oil and gas TNCs, for example, found that very few of those



based in developing and transition economies were explicitly committed to relevant international initiatives in the area of corporate social responsibility (UNCTAD, 2007). Although this may entail costs for the companies concerned, it should pay off in the longer term for both the investing firms and their home countries. Investors from developing or transition economies that are anxious to tap the markets and resources of developed countries may face growing pressure to address more fully issues related to corporate governance and transparency. 13.5.2 Developing host countries There are also implications for host countries in the developing world. As highlighted above, most of the FDI from developing countries flows to other developing economies. A key question is what the host countries can do to leverage this expansion fully. In terms of enhancing the positive impact of such FDI, they need to consider the entire range of policies that can influence the behaviour of foreign affiliates and their interaction with the local business environment. This requires taking the specific characteristics of different industries and activities into account in designing a strategy to attract desirable kinds of FDI. In addition, it is important to promote the number and quality of linkages between foreign affiliates and domestic firms. Host-country governments can use various measures to encourage linkages between domestic suppliers and foreign affiliates and strengthen the likelihood of spillovers in the areas of knowledge, technology and training. In terms of addressing the potential concerns and negative effects associated with inward FDI, there is no principal difference between the policies to be applied for FDI from developed countries and that from developing and transition economies. The scope for ‘South-South’ FDI has led many developing-country host to adopt specific strategies to attract such investment. In a 2006 UNCTAD survey of IPAs, more than 90 per cent of all African respondents stated that they currently target FDI from other developing countries, especially from within their region. For African IPAs, South Africa topped the list of developing countries targeted, while in Latin America and the Caribbean, Brazil was the most targeted source. The new TNCs undoubtedly represent a new source of development finance. As shown above, many low-income countries already rely primarily on other developing countries for inward FDI. For mineral-rich low-income economies, the rapid growth of China and other emerging economies has also led to increased demand and higher prices for their commodities. This expansion in the diversity of potential FDI sources



273

implies greater bargaining power for recipient countries to the extent that they are able to attract a greater number of investors to compete for existing investment opportunities. To the extent that the motivations and competitive advantages of developing-country TNCs differ from those from developed countries, their impact on host developing economies may also vary. For example, the technology and business model of developing-country TNCs may be closer to those used by firms in a host economy, suggesting a potentially greater likelihood of beneficial linkages and technology absorption. 13.5.3

Developed host countries

For developed countries, the emergence of TNCs in the South has provoked mixed feelings among policy-makers. On the one hand, developed countries also see opportunities to attract new capital from these emerging TNCs, which are actively courted by IPAs in the North. Among developed-country IPAs, China was the most commonly mentioned target source, followed by other Asian economies such as India, South Korea, Singapore and Taiwan. A significant number of developed-country IPAs have also established local offices to attract investment in places like Brazil, China, India, South Korea, Singapore and South Africa (UNCTAD, 2006). On the other hand, the speed at which some of these firms have climbed to prominent positions in their respective industries, sometimes via hostile takeovers, appears to have taken competitors and governments in developed countries by surprise. As a result, many observers have responded in defensive ways, some of which are akin to protectionism. This has been visible in the US, with cases such as CNOOC’s bid for Unocal and the responses to Dubai Port World’s takeover of a number of American ports as a result of its acquisition of P&O. In other cases, there have been concerns that takeovers by developing-country firms may lead to jobs shifting to the home country of the investor (UNCTAD, 2006). Finally, there may be concerns that the establishment of R&D activities by firms from the South may lead to outflows of strategic knowledge from a host economy’s national innovation system. Responses to foreign investments involving companies from the South need to be carefully balanced. What is to be regarded as a threat to the national security of a country is not well defined (GAO, 2007). Countries must therefore use this tool very carefully so as not to fuel a trend of



increasing protectionism and retaliation. In today’s multifaceted global economy, it is becoming equally important for developed-country firms to secure access to developing-country markets as for developing-country firms to access the markets in the North. Developed countries and their firms will face new competition for various resources, markets and strategic assets, but they will also find new opportunities for collaboration with newcomer TNCs.

13.6

Concluding remarks

FDI from developing and transition economies has only just started to expand. It is a phenomenon that is likely to evolve rapidly in the next few years, as more and more companies in these economies gain the necessary skills and experience to venture abroad. This trend will be accentuated by firms based in countries characterized by rising current account surpluses. The changes that we are witnessing today will undoubtedly have far-reaching implications, and their full effects are difficult to forecast with any precision. The overseas expansion of latecomer TNCs undoubtedly opens up new possibilities for them to access knowledge and technology in foreign locations. It represents an important complement to other channels of technology transfer – such as licensing, imports and inward FDI. In combination with the observed trend toward more internationalization of R&D – by TNCs from both developed and developing countries – this is contributing to strengthening the interlinkages between national innovation systems. As all international business is associated with significant risks, there are bound to be many failures. However, failures are also a way of gaining experience. Among the main challenges for these emerging TNCs that will allow them to manage their international operations better is finding the right human resources. Many studies have reported the shortage of talent as one of the key obstacles facing managers of these companies. Management training needs to be placed high up on the agenda as policy-makers in the new home countries define their education priorities for the coming years. In fact, policy-makers in countries at all levels of development need to pay greater attention to the emergence of new sources of FDI with a view to maximizing the developmental impact of this recent phenomenon. There is scope for policy-makers from developing and transition economies to share experiences in this area – within and across



275

regions. South-South cooperation between host and home countries may enhance opportunities for cross-border investments and contribute to their mutual development. There is a similar need for an enhanced South-North dialogue with a view to raising awareness and improving the understanding of the factors that determine FDI from the South and of their potential impacts. UNCTAD, other international organizations and the academic community can play an important role in this context by providing analysis, technical assistance and, not least, forums for the exchange of views and experiences. For the world as a whole, it is important to make the process more inclusive. Too many people around the world do not yet see many benefits from globalization. It is furthermore essential to explore ways to avoid the focus on productive efficiency clouding other essential objectives such as alleviating poverty, protecting human rights and fighting climate change.

Notes Head of Policy Issues Section, Investment Analysis Branch, Division on Investment and Enterprise, United Nations Conference on Trade and Development (UNCTAD). The paper draws primarily on research conducted at UNCTAD. However, views expressed are those of the author and do not necessarily represent those of the United Nations. 1. Hutchison Whampoa (Hong Kong, China) and CITIC Group (China). The other companies were Cemex (Mexico); LG Corporation, Samsung Electronics and Hyundai Motor Company (South Korea) and Petronas (Malaysia). 2. See http://money.cnn.com/magazines/fortune/global500/2008/full_list/ 3. Outflows from Hong Kong amounted to almost US$60 billion in the same year (UNCTAD, 2009). 4. See http://www.globalinsight.com/SDA/SDADetail5873.htm 5. As of September 2006, China Exim Bank entertained relations with 36 African countries, and had 259 African projects in its portfolio. Almost fourfifths of its commitments to Africa was in infrastructure (Bossard, 2007). 6. See for example, ‘Congo: Huge development project solidifies China’s clout’, Global Information Network (11 February 2008). 7. These companies will, among other things, build high-voltage power lines and power plants; repair and expand water supplies; construct potable water distribution centres; build 31 hospitals, 145 health centres and four large universities; renovate two important railways and build 250 km of roads. 8. Maximilian von Zedtwitz (2005) denotes such R&D investments as ‘catching up’ and ‘expansionary’ strategies. 9. In addition to Bangalore, Huawei has invested in Stockholm (Sweden), Moscow (Russia) and Dallas (US). 10. Haier operates ten small-scale research units abroad, which focus on technology monitoring and other R&D activities.


276 Torbjörn Fredriksson 11. Jointly with General Electric, for instance, TCS has established an R&D centre in Hangzhou, the capital city of Zhejiang province in China. Similarly, other top Indian IT services players such as Infosys, Satyam and Wipro have invested in China. 12. For example, in 2003, the pharmaceutical firm Ranbaxy set up a new plant in Abu Dhabi that will also do R&D. 13. Some governments in Africa and Latin America have also publicly stressed the importance of outward FDI, but these statements have rarely been followed by concrete promotional measures.

References Ambos, Björn (2005), ‘Foreign direct investment in industrial research and development: A Study of German MNCs’, Research Policy, 34(4), 395–410. Bossard, P. (2007), ‘China’s role in financing African infrastructure’. Berkeley: International Rivers Network, May, 1–19. Edler, J., F. Meyer-Krahmer and G. Reger (2002), ‘Changes in the strategic management of technology: Results of a global benchmark survey’, R&D Management, 32, 149–64. Government Accountability Office (GAO) (2007). Foreign Investment: Laws and Policies Regulating Foreign Investment in 10 Countries, Report to the Honorable Richard Shelby, Ranking Member, Committee on Banking, Housing and Urban Affairs, U.S. Senate. Washington, D.C: GAO. Lall, S. (1983), The New Multinationals: The Spread of Third World Enterprises. New York: John Wiley & Sons. Reddy, P. (2000), The Globalization of Corporate R&D. London: Routledge. Roberts, E.B. (2001), ‘Benchmarking global strategic management of technology’, Research Technology Management, 44 (2), 25–36. United Nations Conference on Trade and Development (UNCTAD) (2005), World Investment Report 2005: Transnational Corporations and the Internationalization of R&D. New York and Geneva: United Nations. United Nations Conference on Trade and Development (UNCTAD) (2006), World Investment Report 2006. FDI from Developing and Transition Economies: Implications for Development. New York and Geneva: United Nations. United Nations Conference on Trade and Development (UNCTAD) (2007), World Investment Report 2007. Transnational Corporation, Extractive Industries and Development. New York and Geneva: United Nations. United Nations Conference on Trade and Development (UNCTAD) (2008), World Investment Report 2008: Transnational Corporations and the Infrastructure Challenge. New York and Geneva: United Nations. United Nations Conference on Trade and Development (UNCTAD) (2009), World Investment Report 2009: Transnational Corporations, Agricultural Production and Development. New York and Geneva: United Nations. von Zedtwitz, M. (2005), ‘International R&D strategies in companies from developing countries – The case of China’. Presentation at the UNCTAD Expert Meeting on the Impact of FDI on Development: Globalization of R&D by Transnational Corporations and implication for Developing Countries, Geneva, 24–6 January.



277

von Zedtwitz, M. and O. Gassmann (2002), ‘Market versus technology drive in R&D internationalization: Four different patterns of managing research and development’, Research Policy, 31 (4), 569–88. Wells, L.T. (1977), ‘The internationalisation of firms from developing countries’, in Tamir Agmon and Charles P. Kindleberger (eds), Multinationals from Small Countries. Cambridge, MA: MIT Press. Wells, L.T. (1983), Third World Multinationals: The Rise of Foreign Investment from Developing Countries. Cambridge, MA: MIT Press.




Part V Technology and Sustainable Development




14 Technological Competences in Sustainability Technologies in the BRICS Countries Rainer Walz

14.1

Introduction

The challenge posed by sustainable development is becoming increasingly urgent from a global perspective. The question raised is how economic growth in transforming and newly industrializing countries can be designed in such a way that it does not undermine the achievement of ecological sustainability goals. At the same time, sustainable innovations can also play an important role for the economic and technological development of these countries. In addition, the prospect of establishing lead markets for sustainability technologies adds an additional incentive for emerging economies to move toward sustainability technologies. Section 14.2 deals with conceptual issues. First, the importance of innovation and technology co-operation are discussed within the traditional view of environmental economics on global environmental challenges. In Section 14.3 the prerequisites for successful technology co-operation and establishing lead markets for sustainable development are presented. In Section 14.4, the empirical research concept and its integration into a system of sustainability innovations are explained. The remainder of the chapter analyses the BRICS countries (Brazil, Russia, India, China and South Africa). The data are based on an updated and enlarged version of Walz et al. (2008). The countries’ technological capabilities and R&D systems are analysed with regard to five fields of sustainability technologies: 1) energy efficiency, both in buildings and in industry; 2) environmentally friendly energy supply technologies, including renewable energy, cogeneration and carbon capture and 281


282

Rainer Walz

storage (CCS); 3) material efficiency, including renewable resources, ecodesign of products and recycling; 4) transport technologies and 5) water technologies. The empirical results include the general framework condition for sustainability research and R&D profiles of the countries with regard to the sustainability fields. Furthermore, the technological fields are analysed with innovation indicators such as patents and specialization in trade. In order to point out the specific strengths and weaknesses of each country, disaggregated results are presented. Based on these results, preliminary conclusions are drawn on the role of sustainability innovations for the economic development process in the BRICS countries. The starting points for the analysis of the demand side and regulation will be presented and conclusions for the future research agenda developed.

14.2

Conceptual issues

14.2.1 Technological innovations as a key to tackling environmental challenges There is a general consensus that environmental sustainability requires an integration of environmentally friendly technologies in the economic catching-up process of the emerging economies. This challenge is discussed within the concept of the Environmental Kuznets Curve (EKC). According to the EKC hypothesis, environmental pressure grows faster than income in the first stage of economic development. This is followed by a second stage, in which environmental pressure still increases, but at a slower rate than gross domestic product (GDP). After a particular income level has been reached, environmental pressure declines despite continued income growth. Graphically, this leads to an inverted U-curve similar to the relationship Kuznets suggested for income inequality and economic per capita income. Within the global environmental debate, it is argued that emerging economies do not necessarily have to follow the path of the industrialized countries. An alternative development path can be labelled leapfrogging or ‘tunnelling through the EKC’ (Munasinghe, 1999). It is argued that countries catching-up economically can draw on the experience of industrialized countries, allowing them to use clean technology and to build a ‘strategic tunnel’ through the EKC. In this concept, technological co-operation and knowledge transfer play a key role. However, there are two critical questions regarding this concept. First, is the interest of the emerging economies strong enough to push in that direction, and


Technological Competences in Sustainability Technologies

283

second, are the countries – given their stage of development – able to absorb the sustainability technologies? 14.2.2

Prerequisites for sustainability technologies innovation

Based on the pollution haven hypothesis and the environmental dumping mechanism, it can be argued that there might be a disincentive for strong environmental policies in the emerging economies in order to attract pollution-intensive industries (see Copeland and Taylor, 2004). However, there are also different incentives for emerging economies to push for sustainability technologies. First, the sustainability technologies influence many of the most discussed environmental problems (see Figure 14.1). Thus, their diffusion would help to improve the environment in the home countries. However, this argument holds less for global environmental challenges such as global warming, which lead to impacts across the world and which require global action. Second, the sustainability technologies analysed improve the infrastructure, for example, in the energy, water or transportation sectors, or address the growing demand for raw materials in the emerging economies. Thus,

Global warming Energy efficiency Acidification Energy supply

Trophospheric ozone

Transport

Environment in megacities

Water technologies

Toxic contamination

Material use and efficiency

Eutrophication

Energy and natural resources Figure 14.1 The influence environmental themes

of

selected

sustainability

technologies

on


284

Rainer Walz

they are part of a necessary modernization strategy. Third, moving toward environmental sustainability will create huge international markets for technologies. It is estimated that the sustainability technologies will be a major market in the future, surpassing other key sectors such as the automotive or machine industries. Forecasts for the world markets up to 2020 show that the average annual growth rates for technology demand in the fields of energy supply, energy efficiency, transport, water and material efficiency amount to 5–8 per cent per year (Berger, 2007). Thus, another incentive is that the emerging economies engage in the development and production of these technologies and compete with the North for lead roles in supplying the world market with sustainability technologies. The potential for technological co-operation focuses on the knowledge base required by the technologies and on enabling competences in the countries. Since the end of the 1980s, the concepts of social or absorptive capacity (Abramovitz, 1986; Cohen and Levinthal, 1990) have been widely known. The results of knowledge management research (e.g. Nonaka and Takeuchi, 1995) and of research on catching up of the last years (e.g. Fagerberg and Godinho, 2005; Malerba and Nelson, 2008) have underlined the importance of absorptive capacity. Archibugi and Pietrobelli (2003) stress the point that the import of technology has little impact on learning, and call for policies to upgrade co-operation strategies toward technological partnering. Nelson (2007) highlights the changing legal environment and the fact that the scientific and technical communities have been moving much closer together. All these factors lead to the conclusion that indigenous competences in sustainability-related science and technology fields are a prerequisite for the successful absorption of sustainability technologies in emerging economies. Above all, for technology-intensive goods, success in foreign trade depends on the innovation ability and the achieved learning effects of a national economy and its early market presence. If there is a forced national strategy to increase the use of sustainability technologies, these countries tend to specialize early in the supply of the necessary technologies. If there is a subsequent expansion in the international demand for these technologies, as indicated by the high market forecasts quoted above, then these countries are in a good position to dominate international competition because of their early specialization in this field (see Blümle, 1994; Porter and van der Linde, 1995; Walz, 2006). Thus, emerging economies could develop an economic interest in pushing for



285

the diffusion and development of sustainability technologies in their countries, in order to reap the benefits of this first-mover advantage. In order to profit from first-mover advantages, however, the domestic suppliers of the technologies have to be competitive internationally so that they – and not foreign suppliers – meet the growing demand at home and on the world market. Taking the globalization of markets into account, this requires the establishment of lead markets with competence clusters that are difficult to transfer to other countries. These competence clusters must consist of high-tech capabilities linked to a demand which is open to new innovations, and horizontally and vertically integrated production structures. Altogether, it is increasingly acknowledged that the absorption of specialized technologies and the development of abilities to advance them and their international marketing are closely interwoven (Nelson, 2007). For both strategies – the transfer of knowledge from traditional industrialized countries and the establishment of an export-oriented lead market position – it is necessary to develop a functioning system of sustainability innovations. 14.2.3 Research concept In the 1990s, the heuristic approach of systems of innovation gained wide acceptance (for an overview see Edquist, 2005). The framework of systems of innovation has been applied traditionally to national innovation systems. More recently, however, it has been also applied to analysing technological or sectoral systems (e.g. Carlsson and Stankiewicz, 1995; Malerba, 2005). These approaches share the starting point that innovations can be best explained by characterizing the components of an innovation system, such as actors, networks and institutions, and their interaction with each other. In the tradition of empirical studies on systems of innovation, the analysis is always context specific and the effects of the various factors depend on the systems’ conditions. Clearly both technology push and demand pull aspects play an important role, leading to numerous interrelated system components (Figure 14.2). However, sustainability technologies differ from ‘normal’ innovations in manufacturing. The formation of demand depends strongly on the specific role of regulations which are especially important for innovations in the energy, water and transportation fields. Environmental regulation acts as an important driver for the demand of technologies. Economic sector regulation dealing with monopolistic bottlenecks in network-based industries also influences the incentives of the actors in technology


Rainer Walz

Market

Regulation

International trade

Demand for sustainability technologies

Industrial system

Environmental problems

Research system

General framework conditions for innovation Figure 14.2

resource availability

Policy coordi nation

Context factors for policy design and impacts

Environmental regulation Public utility regulation

Industrial policy

International policies

286

R&D policy

Scheme of a system of sustainability innovations

decisions. Thus, sustainability technologies face a triple regulatory challenge (Walz, 2007). Furthermore, policy coordination between the different regulatory regimes becomes a major challenge for policymaking. The concept of a system of sustainability innovation (Figure 14.2) can also be used to explain the manifold aspects which must be addressed in empirical research. Below, empirical results for the following aspects are presented: 1. The framework conditions for innovations are analysed by using science and innovation indicators, the level of FDI and survey data from World Economic Forum, WEF (2006). Thus, the results depend on the analytical framework of these approaches, and must be interpreted cautiously. 2. The research infrastructure in the BRICS countries is analysed by means of a questionnaire. The main aim was to find out the research priorities in the countries in the field of sustainability technologies. 3. The technological capabilities of the countries are analysed with innovation indicators (see Grupp, 1998; Smith, 2005). Both patent indicators and trade indicators are used: ●

The patent searches primarily draw on patent applications at the World Intellectual Property Organization (WIPO) and thus international patents. In this way, a method of mapping international



●

●

●

287

patents is employed which does not target individual markets such as Europe but is much more transnational in character. The BRICS countries’ patents identified in this way reveal those segments in which patent applicants are already taking a broader international perspective. The years 2000–4 were chosen as the period of study so that a statistically more reliable population is achieved in which chance fluctuations in individual years are evened out. The UN-COMTRADE database is referred to for foreign trade figures. This is not limited to trade with countries in the Organisation for Economic Co-operation and Development (OECD), but also covers South-South trade relations. In addition, the classification of the technologies uses the Harmonised System (HS) 2002. This foreign trade classification allows more disaggregation and therefore a better targeting of the sustainability technologies compared with the older classifications common in international comparisons (Standard International Trade Classification (SITC)). For both types of indicators, the share of the BRICS countries in the world total was calculated (patent share and world export share). Furthermore, relative indicators (relative patent share (RPA), relative export share (RXS) and revealed comparative advantage (RCA)) were calculated, in order to analyse whether or not the BRICS countries specialize in sustainability technologies.1 All specialization indicators are normalized between +100 and -100 (see Grupp, 1998). Positive values indicate an above-average specialization in the analysed technologies, a negative value shows that the country specializes more in other technologies. Sustainability technologies neither are a patent class nor do they have a classification in the HS-2002 classification of the trade data from the UN-COMTRAD databank which can be easily detected. Thus, for each technology, it was necessary to identify the key technological concepts and segments. These were transformed into specific search concepts for the patent and trade data. This required an enormous amount of work and substantial engineering skills. Furthermore, there is a dual-use problem of the identified segments, and some segments – especially in the trade data – do not necessarily indicate that the technology contributes to sustainability as such. In order to reflect that ambiguity, the term ‘sustainability technology’ has to be interpreted as sustainability-related technology.


288

Rainer Walz

14.3 Empirical results 14.3.1

General framework conditions for innovation

The data on quantitative innovation capacity give a first indication of the general conditions for innovation. The volume of national R&D expenditure, the sectoral share of the R&D expenditure of industry and the number of scientists is much higher in China than in the other BRICS countries. As far as the specific values are concerned, the BRICS countries are on a similar level, except for India, which lags behind. Based on the main indicator of R&D intensity, however, there is still a clear gap between BRICS and OECD countries. A second approach for the analysis of the general framework conditions follows the survey data of the WEF (2006). The indicators are divided into human resources, technological absorption, innovationfriendliness and importance of environmental protection. According to these survey results, among the BRICS countries, India and South Africa can be classified as those countries with the best framework conditions. However, environmental protection is judged to be rather weak in India. In South Africa, the greatest problem lies in the availability of human resources. Brazil and China form the lower midfield. For Brazil, the survey identifies innovation-friendliness (especially regarding regulations on founding new businesses) as the weakest aspect. It is not possible to

100

Human capital Innovation-friendliness

Technological absorption Environmental protection

50

0 Brazil

Russia

India

China

South Africa

Figure 14.3 Results according to survey data from the WEF (2006) on general innovation conditions in the BRICS countries



289

restrict China’s problems to one single factor. The general framework conditions are evaluated as most unfavourable for Russia. FDI can also be interpreted as both a direct source of technological absorption and an indirect proxy for the economic attractiveness of a country. Data about worldwide FDI inflows to the BRICS countries show that China (excluding Hong Kong) lies far ahead of Brazil and Russia. India and South Africa by contrast have received the lowest inflows. China was also ahead of Brazil and Russia with inward stocks in 2005. The accumulated investments for South Africa are currently still clearly higher than for India. 14.3.2 R&D policies The application of sustainability technologies and possibilities for their further development depend greatly on the performance of the R&D system. Whereas literature and science and technology (S&T) statistics contain background information on national innovation policies, these sources provide little systematic information about sustainability research up to now. For this reason, a survey was carried out on selected S&T experts in each BRICS country. In order to collect structured information, an open questionnaire was formulated and was mailed to the experts. In total about 100 experts participated in the survey. It is not possible to compare the figures of R&D funding for the sustainability areas. For this comparison, no figures are available. However, based on the qualitative opinions of the experts questioned, the national strengths and priority areas of sustainability-related R&D policies could be identified. The analysis of the expert’s responses shows that in none of the five countries is the research and innovation policy especially aimed at decoupling environment and resource consumption from economic development. In all BRICS countries, the general support for innovation activity in the business sector has priority. By and large, sustainability research is not institutionally differentiated within the S&T policies of the BRICS countries. Only few R&D programmes in the BRICS countries were specially developed for problems of sustainability research. The South African Water Research Commission and the sectoral R&D funds in the energy, water and transport sector in Brazil, for instance, are among the few exceptions. Mostly sustainability topics are integrated into general, technology-independent funding instruments. Thus, sustainability issues do not represent a separate field of research support. However, there are numerous links between the established research priorities and some sustainability issues. This is especially valid for research priorities in the energy, water and transport sector.


290 Rainer Walz

In all BRICS countries, apart from China, a lack of young scientists represents a decisive barrier for the development of capacity in both sustainability research and public research in general. Against the background of strong economic growth and related career opportunities in the private sector, a scientific career seems to be less attractive in these countries compared to others. 14.3.3 Technological capacity in the area of sustainability-related technologies 14.3.3.1

Overview of the results

The shares of the BRICS countries in the identified transnational patents for sustainability-related technologies are between a few per million to 1 per cent for China. China shows the highest exports of sustainability-related technologies among the BRICS countries, followed by Brazil. In both countries, the world trade shares are considerably higher than the patent shares. That shows that both countries are quite active in exporting sustainability-related technologies, but based on a rather below-average patent base. When looking at the absolute

8

Patent share

World trade share

7 6

(%)

5 4 3 2 1 0 Brazil

China

India

Russia

South Africa

Figure 14.4 Share of BRICS countries in patents and world exports for sustainability-related technologies, 2000–4



291

values – as a proportion of the number of inhabitants – it is noticeable that India has a relatively low international importance in this respect compared to the other BRICS countries. Here, a considerable gap in the development of India compared to the other BRICS countries is shown. The importance of the sustainability-related technologies within the individual countries is also reflected in the specialization profile. Specialization indicators such as the RPA, RXA and RCA show the knowledge and technological competence in sustainability technologies for each country compared to the average of all technologies. Positive values show an above average level of activity of sustainability-relevant technologies, while negative values show a belowaverage level. The results show considerable differences between the countries: In Brazil the sustainability technologies show an above average level for technological capacity; Russia shows knowledge competences, but considerable weaknesses in their transference into internationally competitive technologies; Sustainability-related technologies play a below-average role in India; Sustainability-related technologies play a near-average role in China; In South Africa, sustainability-related technologies form an aboveaverage block in patents, but overall the transference into internationally competitive products is only average.

Specialization exports

100

BR CN IN RU SA

0

−100 −100

0 Specialization patents

100

Figure 14.5 Specialization pattern of BRICS countries for sustainability technologies Source: Fraunhofer ISI.


292

Rainer Walz

14.3.3.2

Disaggregated country profiles

The aggregated numbers disguise the fact that there are strong differences within each country. Thus, a more technology-specific analysis is necessary for differentiating between the technologies. The technological capacity in the selected fields shows a very uneven picture for Brazil. There is considerable specialization in material efficiency, especially in raw materials. Brazil has definitely specialized in this field, which also fits in well with its research competences in the agricultural area. Foreign trade is also above-average in sustainability-related mobility technologies, in particular for ethanol. Energy-efficiency technologies, as well as those of the water sector – which is promoted by a special sectoral fund – display an above-average generation of knowledge. However, against the backdrop of the general innovation problems outlined above, this knowledge can not yet be transformed into internationally marketable products. The electricity sector has been dominated by the use of hydro-electric power until now, so alternative technologies – not only the use of coal, but also other, renewable sources – have not been an issue before now. In this respect, the low capacity in efficient electricity supply and renewable energy technologies is not surprising. Finally, it must be noted that activities addressing innovation and diffusion processes have been only below-average up to now. In Russia, the selected sustainability fields do disproportionately well in technological research. This can be explained by the integration of the fields in the chemicals and mechanical engineering sectors, in which Russia has traditionally been strong. However, what all areas have in common is that they play a very below-average role in Russia’s export portfolio, which is dominated by the export of energy. Only on disaggregated level of single technologies, can positive export specializations be found, for example, for technologies related to coal-fired power stations or for biomass-to-biofuel technologies. On the whole, it appears that the sustainability fields can build on an above-average knowledge base, but considerable deficits seem to exist in their transformation into corresponding goods. The evaluation of the specialization indicators in the selected fields presents an unambiguous picture for India: none of the selected sustainability fields belongs to the above-average segments of Indian foreign trade. Only at the disaggregated level of single technologies can some examples for positive export specialization be found, notably wind turbines, and technologies related to biopolymers and sea-water desalination. Starting points for an above-average knowledge base are



293

found only in material efficiency (e.g. for renewable raw materials) and in water technologies (e.g. sea-water desalination); the other areas are all distinctly below average. On the whole, it emerges that the fields in which India shows particularly high capabilities offer fewer possibilities for locking in to sustainability fields than in the other BRICS countries. In China, the patent applications indicate a rather average activity rate; in parts they are even below average, for example transport. On the other hand, the selected sustainability fields belong to those in which the principle of the ‘extended workbench’ is reflected in above-average foreign trade successes, such as in transport technologies, and on a more disaggregated level in the construction and in photovoltaic cells. The foci of the R&D strategies can be explained in tandem with trade performance: on the one hand, the areas that are being promoted are those that do well (above-average) in foreign trade (e.g. transport technology and construction within the energy-efficiency field). This corresponds to the country’s economic development strategy as a whole. On the other hand, foreseeable bottlenecks in the supply of energy and material resources, in which the country’s technology base has demonstrated weaknesses up to now, are being addressed. Nevertheless, it must be stated that China’s technological capabilities have so far not been directed toward sustainability. This also applies to socio-economic sustainability research. With the exception of the construction field, in South Africa the sustainability fields investigated are characterized by above-average activity for accumulating knowledge. The field of material efficiency stands out with regard to foreign trade, and this can be traced back primarily to regenerative raw materials and sub-sectors in recycling. A similarly strong foreign trade position can also be ascertained for sub-sectors of water treatment, even if this is outweighed in the whole picture of water technologies by weaker positions in the other sub-sectors. 14.3.3.3 Disaggregated sustainability field profile This disaggregation also yields interesting results with regard to comparing the different sustainability fields. Despite the crucial importance of meeting increasing demand for energy, the BRICS countries’ specialization in energy efficiency and supply technologies is, on average, rather weak. This is rather surprising. On the one hand, China and India will increase their use of energy tremendously, acting as the driving forces for the future increase in world energy use. Thus, there will be also an extraordinary increase for technologies related to energy supply and use. On the other hand, Russia, Brazil and South Africa are


Energy efficiency 100

0

−100 −100

0

100


Rainer Walz


294

Specialization patents

Energy supply 100

0

−100 −100 0 100 Specialization patents Transportation



Material efficiency 100

0

100

0

−100 −100 −100 0 100 −100 0 100 Specialization patents Specialization patents Fraunhofer ISI


Water 100

BR CN IN RU SA

0

−100 −100

0

100

Specialization patents Fraunhofer ISI Figure 14.6 Specializations of the BRICS countries within the analysed sustainability fields

also major energy-exporting countries. With the exception of a few knowledge capacities in Russia in some areas, however, this has not led to major developments to augment the advantages of resource availability with technological competences in the next step of the product chain, namely supply and use of energy. The situation is clearly different with regard to material efficiency. Important supply countries such as Brazil and South Africa have also moved to specialize in technologies related to processing the materials.



295

Furthermore, the profile for China and India indicates that these major importing countries are also in the process of building up their knowledge base for these technologies. The situation is mixed for transport policies. China and Brazil, in particular, have specialized in exporting goods here, but without building up their own knowledge base to a similar extent. Altogether this reveals that companies still depend a lot on foreign knowledge. Water-related technologies are equally not among the strengths of the BRICS countries. However, in contrast to the energy sector, there seems to be a stronger emphasis on building a domestic knowledge base. Perhaps this reflects the insight about the pressing problems of the water sector.

14.4

Conclusions and outlook

Global environmental sustainability requires that environmentally friendly technologies are put in place all over the world. However, this requires an economic interest on the part of the emerging economies. One perspective is that these technologies help to reduce national environmental problems and to modernize the infrastructure. Another incentive is to gain enough competences in sustainability-related technologies in order to compete in this growing segment of the world market. Both perspectives require that the countries build up (technological and social) competences in the field of these technologies and their diffusion. In this chapter, a preliminary picture of the existing (technological) competences of the BRICS countries in the field of sustainability relevant technologies has been presented. Various indicators are used, although these are not without caveats. Thus, the results must be interpreted with caution. To sum up, it can be seen that the sustainability-related technologies play an above-average role in Brazil and South Africa. In Brazil, this can be traced back mainly to the strengths in biomass-related activities, including biofuels. In South Africa, the competences are more equally distributed among the five sustainability-related technology areas analysed. In China, the sustainability-related technologies are around average. In Russia, there is a considerable knowledge base, measured through patent specialization, but a trade specialization that is very much below average. In India, which is highly rated among surveys with regard to future innovation activities, specialization in sustainability-related technologies is, by and large, below average.


296 Rainer Walz

However, the analysis also reveals that there are quite considerable differences between the five technological sustainability areas within the BRICS countries, which are reflected in the specialization patterns. The following strengths can be identified for the BRICS countries: ●

●

●

●

●

Brazil is highly specialized in renewable raw materials, including biofuels. At the same time, this field is also a clear research priority. Clearly this area seems to be the best suited for Brazil to aim at a lead market position. South Africa is particularly strong in the areas of material efficiency (including technologies related to recycling and biofuels) and mobility technologies. Furthermore, there are important links to the knowledge base in coal gasification relevant for CCS. China has an above-average competitiveness in transport technologies and in technologies related to the use of energy (especially cooling). However, in both cases, there is a need to upgrade the technologies as regards environmental aspects. Furthermore, China has recently been pushing its exports of photovoltaic cells. The patent base is especially strong in Russia in the areas of material efficiency and renewable energy and CCS. The competences in coal gasification, especially, can play an important role for CCS strategies. India is above average in the international trade in wind energy technology, and has an above-average-patent specialisation in material efficiency and some water technologies (e.g. desalination).

On the other hand the BRICS countries partly also show similar patterns in some sustainability fields. There are high levels of patent activity in the area of water technologies but a below-average level of exports and still a considerable dependence on imports to cover domestic demand. A similar pattern, however, with a smaller level of knowledge-generating activities, also emerges for most energy supply technologies. Even though there is still a high dependency on technology imports in a lot of the areas analysed, the overall picture does not reflect a ‘classical’ dependency on the traditional industrial countries as technology providers. In addition to the countries with an above average-position in some sustainability-related technologies, international patent activities show – particularly in China and to a lesser extent in India – an enormous upward trend. In this respect it might seem realistic that the BRICS countries will also become more and more successful in these technology areas in future. Furthermore, the results differ with regard to the disaggregated technology areas. This also opens up the potential



297

for increased South-South co-operation in future technology development. Examples are the use of renewable resources or perhaps coal gasification, which are both important for almost all BRICS countries. However, moving toward a greater role for sustainability-technologies also requires additional efforts by the BRICS countries. First of all, more attention must be paid to sustainability technologies within the national research priorities. Second, supply-oriented R&D policies have to be augmented with a demand-oriented innovation policy. The demand for the sustainability technologies is strongly influenced by the (environmental) regulatory system. Thus, the latter must be tailored to enhance further innovations. The importance of regulation for shaping the demand for sustainability technologies was pointed out above. However, the empirical analysis presented in this chapter could not picture how the regulatory systems in the BRICS countries support or hinder the development of such technologies. This is not only a question of additional empirical work to be performed. First, a lot of additional research is necessary to develop a clear methodology on how to operationalize the innovationfriendliness of regulation. One promising approach is a heterodox one using the sectoral systems of innovation approach as guiding heuristics and using the functions of an innovation system in order to evaluate the innovation-friendliness of the regulatory system (Walz, 2007; Walz et al., 2008a). It will be necessary to develop these approaches further in the future and to apply them empirically to both traditional industrialized and emerging economies.

Note Head of Department, Institute Systems and Innovation Research, Karlsruhe, Germany; Email of contact: [email protected] 1. For every country i and every technology field j the Relative Patent Activity (RPA) is calculated according to: RPA ij = 100* tanh ln [(pij /

∑p

ij

)/(

i

∑ p /∑ p )] ij

ij

j

ij

Similarly, for exports x the Relative Export Activity (RXA) is calculated according to: RXA ij = 100* tanh ln [(x ij /

∑x

ij

)/(

i

∑ x /∑ x )] ij

ij

ij

j

In addition to exports, the Revealed Comparative Advantage (RCA) also takes the imports m into account and is calculated according to: RCA ij = 100* tanh ln [ (x ij /mij ) / (

∑ x / ∑ m )] ij

j

ij

j


298

Rainer Walz

References Abramovitz, M. (1986), ‘Catching up, forging ahead, and falling behind’, Journal of Economic History, 46, 386–406. Archibugi, D. and C. Pietrobelli (2003), ‘The globalisation of technology and its implications for developing countries – Windows of opportunity or further burden?’ Technological Forecasting and Social Change, 70, 861–83. Berger, Roland (2007), ‘Umweltpolitische Innovations- und Wachstumsmärkte aus Sicht der Unternehmen’, Forschungsprojekt im Auftrag des Umweltbundesamtes, Förderkennzeichen (UFOPLAN) 206 14 132/04, Reihe ‘Umwelt, Innovation, Beschäftigung’, Band 2/07 Berlin. Blümle, G. (1994), ‘The importance of environmental policy for international competitiveness’, in T. Matsugi and A. Oberhauser (eds), Interactions Between Economy and Ecology. Berlin: Duncker & Humblot, 35–57. Carlsson, B. and R. Stankiewicz (1995), ‘On the nature, function and composition of technological systems’, in B. Carlsson (ed.), Technological Systems and Economic Performance: The Case of Factory Automation. Dordrecht: Kluwer Academic. Cohen, W. and D. Levinthal (1990), ‘Absorptive capacity: A new perspective on learning and innovation’, Administrative Science Quarterly, 35, 123–33. Copeland, B.R. and M.S. Taylor (2004), ‘Trade, growth and the environment’, Journal of Economic Literature, 42 (1), 7–71. Edquist, C. (2005), ‘Systems of innovation: Perspectives and challenges’, in J. Fagerberg, D.C. Mowery and R.R. Nelson (eds), The Oxford Handbook of Innovation. Oxford: Oxford University Press, 181–208. Fagerberg, J. and M. Godinho (2005), ‘Innovation and catching-up’, in J. Fagerberg D.C. Mowery and R.R. Nelson (eds), The Oxford Handbook of Innovation. Oxford: Oxford University Press, 514–42. Grupp, H. (1998), Foundations of the Economics of Innovation: Theory, Measurement, and Practice. Cheltenham: Edward Elgar. Malerba, F. (2005), ‘Sectoral systems: How and why innovation differ across sectors’, in J. Fagerberg D.C. Mowery and R.R. Nelson (eds), The Oxford Handbook of Innovation Oxford: Oxford University Press, 308–406. Malerba, F. and R. Nelson (2008), ‘Catching up: In different sectoral systems’, Globelics Working Paper Series 08-01. Munasinghe, M. (1999), ‘Growth-oriented economic policies and their environmental impacts’, in C.J.M. van den Bergh (ed.), Handbook of Environmental Economics, Cheltenham: Edward Elgar, 678–708. Nelson, R.R. (2007), ‘The changing institutional requirements for technological and institutional catch up’, International Journal of Technological Learning, Innovation, and Development, 1, 4–12. Nonaka, I. and H. Takeuchi (1995), The Knowledge Creating Company. Oxford: Oxford University Press. Porter, M.E. and C. van derLinde (1995), ‘Toward a new conception of the environment competitiveness relationship’, Journal of Economic Perspectives, 9 (4), 97–118. Smith, K. (2005), ‘Measuring innovation’, in J. Fagerberg D.C. Mowery and R.R.Nelson (eds), The Oxford Handbook of Innovation. Oxford: Oxford University Press, 148–77.



299

Walz, R. (2006) ‘Increasing renewable energy in Europe – Impacts on competitiveness and lead markets’, Energy & Environment, 17 (6), 951–75. Walz, R. (2007), ‘The role of regulation for sustainable infrastructure innovations: The case of wind energy’, International Journal of Public Policy, 2 (1/2), 57–88. Walz, R., K. Ostertag, W. Eichhammer, N. Glienke, A. Jappe-Heinze, W. Mannsbart and J. Peuckert (2008) Research and Technology Competence for a Sustainable Development in the BRICS Countries. Stuttgart: IRB. Walz, R., M. Ragwitz and J. Schleich (2008a), ‘Regulation and innovation: The case of renewable energy technologies’. DIME Working Papers on Environmental Innovation, (2), http://www.dime-eu.org/working-papers/wp25. WEF (2006), Global Competitiveness Report 2006. Davos: World Economic Forum.


15 Coordination, Convergence or Contradiction: Information and Communication Technologies for Integration and Development in Southern Africa and the Southern Cone Patience I. Akpan-Obong and Mary Jane C. Parmentier

15.1

Introduction

Various factors compel different countries and regions to engage in bilateral or multilateral relations and cooperation. These interactions often require a forfeiture of both sovereignty and national autonomy in varying degrees. For some countries, such relationships have not been equal or bilateral in any real sense of the word. Historically, countries in the developing world were structurally connected with Western countries in a colonial or neo/post-colonial relationship of exploitation and dependence. However, in recent decades, there has been considerable push by contiguous countries to interact more systematically with each other through formal and informal structures of cooperation and horizontal integration. A significant motivation for interactions among many regions in the developing world is the need to reduce their dependency on primary commodity exports to global markets. There is an additional imperative for such regions to reduce their technological dependency on former colonial powers. Thus, information and communication technologies (ICTs) for integration and internal development should, logically, be mutually reinforcing. However, do policies in these areas share the same goals? Development can have multiple meanings, and it is possible that policy goals are not always the same. For instance, the adoption of ICTs to enhance national competitiveness in the global 300


Coordination, Convergence or Contradiction

301

market might assume economic growth rates as the prime indicator of development. Also, national policies for ICT might be more focused on socio-economic disparities and poverty alleviation. The utilization of ICTs for regional integration further complicates this dynamic as countries within regions seek closer economic cooperation and integration as means of achieving national economic growth. These complexities raise the need to develop a framework that integrates the multiple meanings, goals and strategies of development and regional integration. Such a framework can facilitate an understanding of the process and interconnections in ways fundamentally useful for research and policy. Development discourse has spanned several decades, stages and strategies. Post-World War Two discourse and European reconstruction and development coincided with decolonization of non-Western countries. Classical development discourse concentrated on the need and strategies for propelling newly independent countries from their state of underdevelopment to industrialization (Martinussen, 1997: 74–5). Some of these strategies included the development of ICTs, such as the radio, to enhance the flow of information from the West to the emerging independent countries. The assumption was that through the mass media, people in underdeveloped countries would become familiar with, and adopt the culture and values, of the Western world. More direct strategies of development included export-led growth (in the 1950s), import substitution industrialization (in the 1960s and 1970s) and neo-liberal economics (in the 1980s) that manifested in free markets and structural adjustment programmes (Gwynne and Kay, 2004; Kaufman and Haggard, 1992). In the past two decades, advances in ICTs have shifted to the usefulness of these technologies as tools for socio-economic development (Crede and Mansell, 1998). The emphasis on ICTs as tools for development has been accompanied by a simultaneous trending toward integration and cooperation among contiguous countries. While regional cooperation is not a novel phenomenon, there is insufficient theoretical analysis specific to developing countries. Theories of international integration such as functionalism and neo-functionalism were mostly based on European integration after the end of World War Two. Given its the historical context, the ultimate goal of European regionalism was necessarily political (Pentland, 1973). In contrast, the basic objective of regional integration in the developing world is framed in economic terms, with political cooperation occurring only as an auxiliary goal. The economic nature of regionalization in developing countries partly explains the prominence associated with the use of ICTs as a strategy for achieving


302

Patience I. Akpan-Obong and Mary Jane C. Parmentier

the twin goals of regional integration and economic development. This underscores the similarities in the goals of regional integration, economic development and exploitation of ICTs, even though the connections between these are not always clearly articulated. We argue that an analysis of these interconnections can facilitate an understanding of the processes and policies in regions and countries seeking to promote economic development through regional integration and ICT. Already, many governments in the developing world have formulated and implemented ICT policies deliberately to help in achieving socio-economic goals, both nationally and regionally. Moreover, there is a near consensus on the potential of ICTs to accelerate integration processes among contiguous countries. Given this existing practice, it is therefore counterintuitive to research or analyze regional integration, ICT for development (ICT4D) and economic development as distinct phenomena. Integration by itself is understood as a multi-layered rather than linear process (Nye, 1968). This provides a further premise for examining the policies and processes of adoptions of ICTs in order to achieve the goals of regional integration on one hand, and national and regional development on the other. Questions concerning the particular policies and processes in each of the three areas (regionalism, development and ICTs) and their relationships can frame the parameters of such research. Accordingly, we developed a framework that facilitates analysis of the relationship between integration, development and ICTs (See AkpanObong and Parmentier, 2007; 2009). 15.1.1 Towards a theoretical framework of integration, development and ICTs A specific theory of ICTs and development is yet to emerge, mostly because much research in the field has overwhelmingly focused on prescriptive assumptions about the utility of the technologies for development. However a recent trend toward more critical research has prompted tentative theorization. To this end, Heeks (2006) suggests a coherent approach grounded in the social sciences. Research driven by such an approach can be carried out in a systematic way, beginning with identification of the issues, a potentially useful theory (or framework), examples of possible applications and evaluative analysis. We agree that the relatively new but growing body of literature on ICTs and development should contribute to theoretical understanding, as well as provide methodologies and frameworks for subsequent research. We have therefore adopted the Heeks approach by first identifying our research focus: ICTs as tools for economic development and regional



303

integration. We then examine this through the lens of a framework created specifically to develop the interconnections of theory and research in the issue areas (Figure 15.1). We evaluate the vigour of our framework and thus its contribution to research through a case study of selected (ICT4D) strategies in Brazil and South Africa. These countries represent two critical regional blocs in the developing world: Southern Africa Development Community (SADC) and Mercado Comun del Sur (MERCOSUR). The roles of Brazil and South Africa as economic leaders in Latin America and southern Africa and the process of harnessing ICTs for development and economic cooperation are particularly significant. We begin with an overview of the regions and highlight how policies and strategies on ICTs for integration enhance or hinder efforts at regional development. This is followed by a discussion of South Africa and Brazil and their respective policies on ICTs and development. This explores the common trajectory of socio-economic development in the seemingly disparate paths of ICT adoption and regional integration. We conclude with an analysis of the effectiveness of our framework in conceptually bridging the disconnections in research and policy in the fields of ICTs for development and ICTs for regional integration.

Development Simultaneous overlaps and intersection of theories and practice of integration, ICTs and development

ICTs and development ICTs as: tools to meet basic needs; address socio–economic disparities

Regional Integration Closer regional cooperation through technological networks; integration enhances economic growth

Figure 15.1 An integrated theoretical model representing the interrelationships between regional integration, ICTs and development and socio-economic development


304 Patience I. Akpan-Obong and Mary Jane C. Parmentier

15.2 Coordination, convergence or contradiction in SADC and MERCOSUR 15.2.1

Common values, common principles: SADC

SADC began in 1980 as a group of democratized southern African countries. The emphasis on majority-ruled democracies pointedly excluded apartheid South Africa from the alliance. The countries coordinated self-sufficient development projects to reduce their dependence on their big but undemocratic neighbour (SADC, 2008). The nine original members were: Angola, Botswana, Lesotho, Malawi, Mozambique, Swaziland, Tanzania, Zambia and Zimbabwe and the alliance had its headquarters in Botswana. However, the organization did not become a legal entity until 1992 when its declaration and treaty were signed at the Summit of Heads of State and Government in Namibia. Its membership later expanded to include the Democratic Republic of Congo (DRC), Madagascar, Mauritius, Namibia and post-apartheid South Africa. While the formation of an economic community by countries in southern Africa was ostensibly driven by the imperative to disassociate from South Africa, SADC was equally forged by visions of a common destiny. The commonality of interests was aimed at guaranteeing the survival of member states while improving the ‘standards of living and quality of life, freedom and social justice and peace and security for the peoples of southern Africa. This shared vision is anchored on common values and principles and the historical and cultural affinities that exist between the peoples of Southern Africa’ (SADC, 2008). The community has gone through various revisions of its goals since it was first established. For instance, it admitted a major actor in the region, South Africa, in 1994 following the collapse of apartheid. Another major modification occurred in 2001 in Namibia when heads of the SADC states approved the restructuring of its institutions through enhanced objectives and a common agenda for the community. This began the path toward ‘a more explicit Common Agenda which takes into account a number of principles such as development orientation; subsidiarity; market integration and development; facilitation and promotion of trade and investment; and variable geometry’ (SADC, 2008). The enhanced objectives also included the ‘promotion of sustainable and equitable economic growth and socioeconomic development that will ensure poverty alleviation with the ultimate objective of its eradication (and the) promotion of common political values, systems and other shared values’ (SADC, 2008).



305

A functional telecommunications infrastructure in the region was fundamental to the attainment of these objectives (Jaddoo, 2005). Different countries in the region therefore began the process of privatization, liberalization and regulation of telecommunications infrastructure. Initially, as elsewhere in Africa, telecommunications was narrowly defined as telephony (land lines and mobile phones) and resources were therefore invested in these areas. Following the separation of its post, telecommunications and broadcasting functions, the South African government deregulated the telecommunications sector. It broke up the monopoly of the public telecommunications company by licensing a second service provider in 2002. Shortly after this, the emphasis shifted to mobile telephony such that within a few years nearly 71 per cent of South Africans were mobile phone users. Actual numbers moved from 10.7 million in 2001 to 39.6 million in 2006 (ITU, 2008). Alongside this growth in mobile phone usage within member countries, SADC was also reviewing ‘existing protocols so as to put more emphasis on information and communication technologies (ICT) rather than just on telecommunication infrastructures’ (Jaddoo, 2005). The Telecommunication Regulators Association of Southern Africa (TRASA) was thus ‘created to harmonize the ICT legislation of the region’ along the guidelines of universal access and service, licensing policy, fair competition and pricing (Jaddoo, 2005). A regional trade group, (the Southern Africa Telecommunications Operators Association) also emerged to ‘harmonize interconnectivity through radio, fiber, and microwave backbone’ (ibid.). Several countries responded by formulating policies based on an expanded definition of ICTs and their instrumentality in order to achieve SADC’s socio-economic objectives. More pivotal to ICT-related developments in the region was the Declaration on ICT signed by the 14 SADC heads of state in 2001; it became a major catalyst for ICT-related activities in the region. Other SADC documents and policy statements articulated by officials in the different member states reflect the essential element of the Declaration on ICT: We shall undertake to work together to remove barriers to electronic commerce in our SADC countries as a means to opening opportunities and benefits such as increased access to markets, opportunities to create economic value from cultural assets, reduced administrative costs, and improvement of public services. There is a need to adopt and adapt technologies that enable e-commerce capability to avoid increasing exclusion from the global economy. (SADC, 2008)



The Declaration on ICT forges the three elements that underpin our research. The specific goal of integration and cooperation in the various regions of the world is economic growth by itself, as well as to facilitate competition in the global economy. ICTs potentially enhance integration as a key to economic growth as well as acting as tools for economic growth directly. These functions are universal but even more so in developing parts of the world where regionalism might very well be a major coping mechanism in an increasingly globalized and ‘flattened’ world where only the fast, biggest, strongest and fittest survive (Friedman, 2005). In line with the Declaration on ICT, SADC has implemented projects aimed at achieving its goals of economic development through regional integration and ICTs. First, it set up the Regional Information Infrastructure (SRII) with the possibility of linking it to the EASSy cable. The $248 million EASSy cable, developed by a consortium of telecommunication companies with partial funding from the World Bank and African Development Bank, ‘is the most advanced of all the sea cables proposed for the East and Southern African region. It is expected to “bring faster and affordable internet to about 250 million people in the region’ (Biryabarema, 2008). Despite projects such as this, effective utilization of ICTS for economic development at the regional level encounters key challenges, including lack of financial capacity and human resources, difficulties in implementing universal access and service by member states and the non-participation of telecommunications operators in regional agenda (Unuth, 2007). The challenges confronted by SADC indicate that success in the utilization of ICTs for integration and economic development is likely to depend on the capacity and resources of member states, as well as their willingness to translate regional goals into national practice. Interestingly, the presence of large economies with the most developed ICT sectors in these regions could result in the tendency to ignore regional imperatives and focus on national interests. However, the ICT processes of the larger economies in each region can dictate the tone for others and provide a model of they and ICT policies can facilitate or hinder regional integration, and economic development, at both the regional and national levels. 15.2.2 Inward and outward connections: MERCOSUR The Mercado Comun del Sur (Southern Common Market) was begun more recently than SADC, with the Treaty of Asunsion in 1991. This treaty drew Argentina, Brazil, Paraguay and Uruguay into a regional integration plan. The primary goal was the expansion and integration



307

of national markets in order to attract foreign direct investment (FDI) to the region. The Treaty of Asunsion, therefore, centres on a regional integration that emphasizes the coordination of national policies and infrastructure as they relate to trade and investment. It also stresses the development of a common external trade policy toward forging a customs union (MERCOSUR, 1991). In 1994, the agreement was strengthened and the structure of MERCOSUR was established with the Protocol of Ouro Preto to further harmonize trade relations in the region. While the neo-liberal economic development strategy clearly relies on the market, there is also an emphasis on fighting poverty and socio-economic inequities and on the pursuit of ‘economic development with social justice’ (MERCOSUR, Portal Official). More recently, there have been policy recommendations in MERCOSUR to focus on a ‘new economic thinking’ which underscores ‘integration for development’ (MERCOSUR, 2005: 8). The emphasis is still strongly on the enhancement of trade for regional competitiveness, but there are explicit suggestions that development should prioritize regional inequities, and that these inequities are socio-economic, as well as technological and infrastructural. Further, it is recognized that ‘asymmetries’ in public policy are barriers that are as significant as the lack of an integrated and coherent physical infrastructure for integration and development (MERCOSUR, 2005). The enlargement and enhancement of the regional market is seen as the number one tool for development, although national policies could impede this process (ibid.). The coordination of national policies is also a priority for continental initiatives by the Interamerican Development Bank and the United Nations. As regards ICTs for integration and for development, the multinational agencies and MERCOSUR share the objective of coordinating national policies, addressing uneven regional development and creating universal access to technology. Information and communication technologies for development can be classified into groups with different foci. We distinguish between the development strategies that focus on the use of ICTs as enablers of development and strategies that focus on innovation and production in the global ICT industry. Within MERCOSUR, the emphasis is ostensibly on the implementation of ICTs as enablers of integration in pursuit of the goal of expanding the markets of the southern cone region of South America. A working group was created specifically to address the role of ICTs in the integration processes in the region. The group was committed to the coordination of public services for communication, including postal services, radio and television media and



telecommunication (MERCOSUR, 2007). It emphasized the interfacing of the telecommunications systems in the regions with global systems, which would align with the goal of connecting to the global economy for trade and capital. The creation of the group was predicated on the fact that as recently as 2007, people in the region were still facing infrastructural challenges with the older technologies. For instance, much of the region was not connected to the telephone network, thus hindering integration between peoples and communities. While there is no basic telephone system in most of the region, Internet usage through regional networks is widespread. Various networks have been created both officially by MERCOSUR, and unofficially by groups in the regional community. An official network created by MERCOSUR in 1998, the Mercosur Economic Research Network is designed to contribute directly to economic integration and regional growth by networking key institutions of higher education and research in economics. The network, funded by international organizations such as the Organization of American States and the World Bank as well as private foundations, currently links 12 institutions from all four member countries, and organizes collaborative research and conferences (Red Mercosur). Another project, sponsored by the Swiss Agency for Development and Cooperation, was completed in 2007 and targets the region’s software industry. It also promotes R&D in each country while encouraging collaboration and harmonization in order to create synergy for global competitiveness. Yet another initiative culminated in a joint European Union (EU) – MERCOSUR project, Digital Mercosur) in 2007. This seeks to correct the relative lack of regional policies for ICTs and development (MERCOSUR Information Society, 2007). With €9 billion of funding, the agreement highlights the need to address the vast socio-economic inequities in the region by redirecting the integrative efforts of MERCOSUR to development with an emphasis on the role of civil society organizations such as those concerned with marginalized groups and areas. The organization’s official website was launched in February 2009 with a call for general participation in its blog (Projeto Mercosul Digital).

15.3 Connections and contradictions in South Africa and Brazil 15.3.1 South Africa South Africa, with a population of nearly 48 million in 2007 (Lehohla, 2007) is a middle-income country in Sub-Saharan Africa. It was one of



309

the earliest countries in the region to get on the ICT4D bandwagon, even though it was a late entrant in many ways. For years, the country was economically and politically excluded from global society because of its institutionalization of apartheid, a system of governance that oppressed and suppressed the rights of black people. There were many casualties of the apartheid regime but the most internationally notable was Nelson Mandela, who was sentenced to life imprisonment for his anti-apartheid activism. He was released in 1990 after 27 years in prison. His release and reversal of the ban on the African National Congress (ANC) began the process of democratization in the country that culminated in 1994 with the first multiracial and multiparty elections in the country. Mandela was elected the first lack president. The political events in South Africa coincided with the rising discourse that linked ICTs with socio-economic growth in developing countries. Indeed, Internet usage gathered momentum, becoming a widely used tool for communication and access to information in the 1990s. While South Africa responded to this discourse in the immediate postapartheid years, the government clearly prioritized political, economic and social restructuring with the latter manifesting itself through the activities of the Truth and Reconciliation Commission (TRC), set up in 1995. The leaders also seemed to understand the role of ICTs as tools in the process of reconstruction and reconciliation among the different groups. Thus, the South African Information Technology Strategy (SAITIS) was developed in 1995 and its importance was underscored by it sponsorship by the country’s deputy president, Thabo Mbeki (SAITIS, 2000). As the TRC was fundamental to the socio-political rebuilding of the new democracy, so was SAITIS pivotal to building bridges and ushering in economic development nationally and globally. ‘An important goal in conceptualizing the project was to bridge the global development gap and develop a robust, growing and sustainable South African ICT Sector that directly supports, and contributes to the ... challenge of sustainable economic growth, social upliftment and empowerment’ (ibid.: 2). The role of ICTs in promoting economic goals internally and in enhancing South Africa’s competitive advantage in the international market was particularly emphasized. At the level of practice, various sectors in South Africa embarked on projects to integrate ICT development strategies as means (enablers) of achieving socio-economic goals. Pursuit of ICTs was also framed in terms of how availability and usage were themselves a goal of development. There were so many ICT-related projects in the country that in 2002 a company was hired to compile ‘an initial database of all the


310


information and communication technology and information society projects’, particularly those initiated by the government and its agencies (Bridges, 2003). A year later, an inventory showed 58 active private sector and civil-society initiatives in the country (ibid.). It is safe to assume that the number has doubled (perhaps tripled) now in 2009. While it is impossible to examine all the ICT-related projects in South Africa, here we provide a brief overview of some projects that demonstrate the interconnections between ICTs, integration and national as well as regional socio-economic development in the SADC region. These projects are: the Acacia Initiative, African Information Society Initiative (AISI), Bellanet, Community Information Network for Southern Africa (CINSA) and Southern Africa Non-Governmental Organization Network (SANGONeT). The Acacia Initiative, funded by the Canadian International Development Research Centre (IDRC), was one of the first ICT projects aimed at empowering ‘sub-Saharan African communities with the ability to apply ICT to their own social and economic development’ (ibid.). Emerging in 1997 in an era when the connections between development and ICTs were less nuanced, the Acacia Initiative focused on creating tele-centres, school networks and agriculture-related ICT applications. It emphasized access to information, underscoring an assumption that in the emerging network society, information was by itself a valuable commodity (Castells, 1997). An earlier but similar project, also sponsored by IDRC, was Bellanet, a service organization set up to facilitate the flow of information and related services in the development community. The AISI was a continent-wide project, unlike the Acacia Initiative. Formed in 1996 under the auspices of the United Nations Economic Commission for Africa (UNECA), its primary mandate was to coordinate efforts by various partners collaborating to ‘promote connectivity and information technology in Africa’ and stressed capacity building – “the capacity of Africa to influence ICT investments, to extend their impact, and to encourage the development of African solutions, applications and content’ (Bridges, 2003). South Africa was a major player in each of these projects. Along with its national policies and projects, the country performed more than others SADC members in putting policy statements on ICTs into practice, with many of the projects having regional and continental significance. For instance, the early deregulation of the South African telecommunications sector paved the way for the private sector to get on the ICT bandwagon. This resulted in the fact that at least two South African countries are major providers of mobile telephone services in several



311

countries in sub-Saharan Africa. Ironically, the South African telecommunications industry is one of the most regulated, constantly evoking criticism from neighbouring countries (Business Day, 2008). The country’s frontline status in ICT diffusion and usage then raises an important question. How do these ICT activities connect with the country’s position in SADC, regional economic development and integration? It is a major economic and political actor in the region, a middle-income country, but surrounded by some of the poorest countries in the world – with the notable exception of Botswana. Its GDP (at US$300.4 billion) in 2008 (The World Factbook, 2009) was more than three times that of all the other 13 SADC countries combined. Indeed, many of the countries are so landlocked that they depend on South Africa for access to the sea ports. (Lesotho, for one, is fully surrounded by South Africa.) The significant difference between the economic size of South Africa and its neighbours may serve as a metaphor for the income disparity within South Africa itself. Several years after constitutional democracy began, the structure of apartheid was still evident in the income disparities in the country in 2001 when the ‘the poorest 20 per cent of households (equivalent to 27 per cent of the population), accounted for less than 3 per cent of total income levels. ... The richest 20 per cent of households, (equivalent to less than 3 per cent of the population) account for 65 per cent of total income production’ (ITU, 2001). More recent data are not available, although one can assume that the gap has not closed even if absolute incomes have increased. The structure of inequality also manifests in the distribution of telecommunications infrastructure in the country. About ‘18 per cent of black households and 82 per cent of white households’ have access to the telephone, according to data from the ITU. Moreover, disparities exist between urban and rural households: 64 per cent of urban households have access to a telephone, against and only 9 per cent of rural ones (ITU, 2008). It is remarkable that South Africa’s policies on ICTs as tools for socioeconomic development address two specific areas: growth through the technologies utilized internally in the country, and their capacity to enhance the country’s competition in the international arena. This is demonstrated by the fact that the SAITIS framework conceives of the ‘information age’ as a process driven by global competition a high level of cross-border information infrastructure (SATIS, 2000). Yet, there is insufficient articulation of strategies connecting a conception of ICTs as development tools with that of regional integration as path to socioeconomic growth in SADC. Countries clearly recognize the usefulness of these technologies in regional development either as egoals in


312


themselves or as enablers of other integration efforts. However, many in the region formulate policies with tendencies toward ICT adoption and usage as national, rather than, regional processes. This disconnectivity between ICTs and regional integration as twin tools for development can also be observed in MERCOSUR. However, our theoretical framework highlights the intrinsic connections between these processes. The separation of ICTs from regional integration as tools or strategies for socio-economic development is therefore rather impracticable. The failure to approach these processes as different aspects of the same socioeconomic development trajectory may partly account for some of the problems of underdevelopment that pervade SADC and MERCOSUR countries. An important key therefore is an articulation and implementation of policies that deliberately utilize ICTs to promote regional integration on one hand, and national/regional development on the other. 15.3.2 Brazil In Brazil, the current administration of President Lula da Silva has stressed the problem of extreme poverty and socio-economic inequities, much in line with the United Nations Millennium Goals, and similar to the development emphasis of South Africa. With a Gini coefficient of 0.57, Brazil has one of the highest income gaps in the world, with up to 17 million people living in extreme poverty, the majority of whom are black (Presidency of the Republic, 2004: 15–17). Brazilian and multinational efforts at harnessing ICTs for socio-economic development apparently focus on addressing inequality (Sorj, 2003).Thus, there are policies to extend the Internet and education to marginalized elements of Brazilian society, as well as to small business enterprises unable to afford new technology (Botelho and Tigre, 2005). In 2008, the World Bank’s infoDev programme implemented a joint project with the Brazilian government to aid small businesses in ICT access and use (InfoDev Incubator Support Center, 2008). In general, Brazilian policies utilizing ICTs for development and expanding the ICT industries in the country are much more prolific and coherent than the regional policies of MERCOSUR. This is perhaps not surprising but it is significant in its impact on the harmonization of national policies, universal access and unequal development in the region. In Brazil, there are several policies and initiatives in different sectors aimed at directly linking ICTs to national development, particularly in order to achieve the UN Millennium goals. One of these is the Bolsa Familia programme initiated by President Lula and targeting the country’s poorest and most marginalized populations. An expansive social welfare program, Bolsa



313

Familia also has an ICT component aimed at providing technical support for schools to integrate computer usage and education into the general curriculum (Presidency of the Republic, 2004). In this context, the technologies are instrumental to closing the socio-economic and digital divide. ‘Digital inclusion’ is therefore a ‘priority for the Brazilian government (which also) promotes social inclusion and plays a key role in the fight against poverty, allowing the access of citizens to information and knowledge’ (ibid.: 91–2). Another initiative, Casa Brasil, has built 7000 tele-centres in less-developed regions giving free access to the Internet access and training programmes (ibid.). Many of these programmes are coordinated by the Brazilian Ministry of Science and Technology, which is dedicated to ICT promotion and implementation. It also developed the Information Society Programme, which has a two-pronged approach aimed at both strengthening and expanding the national market via ICTs and promoting the ICT industry itself (Vicentini Jorente, 2008). The ministry currently has committees dealing with ICT services, development and application (Ministerio da Ciencia e Tecnologia). It has developed an Internet community, the Rede Brasil de Tecnologia, which connects universities, research centres, corporations and financial institutions, in order to conduct applied research in developing technological capacity for the private sectors (Rede Brasil de Tecnologia). The community shares similar goals with the informal ‘Mercosur’ grassroots network created by local groups in the three other countries with the conspicuous absence of Brazil. As the scope of ICT activity expands in Brazil, there are concerns that in implementing ICTs for integration and regional growth, participation in ICT development and production might be overlooked to the detriment of knowledge acquisition (Cassiolato and Lastres, 2000). The government has accordingly addressed this concern through the national policy on information technology. The policy facilitates the establishment of a national ICT industry in order to reduce the country’s dependence on the export of natural resources for export, thereby enhancing its competition in the global market (Ministerio da Ciencia e Tecnologia). Several specific initiatives have been implemented, such as the National Research Network (RNP) which connects educational and research institutions on the Internet. This is similar to the Mercosur Economic Research Network, although the latter focuses on integration and economics. The other initiatives include coordinated research among university computer science programmes and the creation of SOFTEX, a programme to increase the export of Brazilian software products. The educational policy of Brazil also addresses the national market,


314


targeting human capacity building in ICTs through a network dedicated to linking schools with technologies studies (Ministerio da Educacao). This ostensibly complements MERCOSUR educational policy objectives to target workforce development, especially in new technologies, in order to enhance regional competitiveness. There does not, however, seem to be any linking of the MERCOSUR and Brazilian networks. Also, at the national level, universal access is a goal of the Brazilian Ministry of Education, which plans to provide computers in all schools, as well as to create distance education programmes with a national spread. But it is a private initiative sponsored by the Development Gateway foundation and several private corporations that centres on bridging competition in the global market with developing national socio-economic equity in Brazil through ICTs. To this end, the e-Brasil project maintains a website ‘dedicated to information and discussions on the use of information and communications technology (ICT) for building a more equitable and competitive Brazil (e-Brasil, 2008).’ The Brazilian ICT market is significant in many areas, and it is certainly a leader in the MERCOSUR region, much like South Africa is in SADC. In 2006, Brazil led Latin America in software revenues; the hardware sector was also growing. It is also the region’s largest market in telecommunications, with 35 per cent of Latin America’s revenue in the sector in 2006 (U.S. Commercial Service, 2007). Policies of liberalization over the past decade have built partnerships between the private sector and academic institutions and governmental support in areas of patents and intellectual property have grown, resulting in incubator R&D projects. These developments have created a ‘meta-innovation system’ in Brazil (Etzkowitz et al., 2005), although the impact is not evenly distributed throughout the country. In 2003, more than twothirds of the incubators were located in the southern and south-eastern part of Brazil where the cities of Rio de Janeiro and Sao Paulo are located (ibid.). Industry continues to be concentrated in the region of Sao Paulo, although there is proposed government funding for the creation of technology parks in other parts of the country (Ministerio da Ciencia e Tecnologia). There are also plans to support university research in lessdeveloped areas of the country. This survey of ICT policies and projects in Brazil and in MERCOSUR suggests a low level of coordination and harmonization between national and regional policies. This supports evidence from an earlier study that examined the coordination of such policies for development. Finquelievich (2003) noted ‘(t)he fact that MERCOSUR countries ... are working on ICT national strategies, does not mean that the MERCOSUR,



315

as an entity, is producing plans to achieve integration to Information Society, nor for ICT use in poverty reduction nor sustainable development.’ Some degree of convergence has been observed, particularly in the promotion of education, research and innovation to encourage regional ICT industries and exports. Both policies have supported programmes to connect research institutions involved in economic research and technology through the Internet. However, there is insufficient synergy between programmes at national and regional levels. For instance, MERCOSUR policies prioritize the coordinating of trade and harmonizing infrastructure for communications, while Brazil’s policies concentrate on the building of national ‘system of innovation’ and increasing the country’s lead in the ICT sector in Latin America. There is regional and national concern about the digital divide and socio-economic disparities, and universal access to ICTs is a recurrent theme in national and regional policies. Yet, national projects are more likely to receive funding from multinational organizations and corporations for extending access to marginalized areas locally rather than regionally. There is therefore a disconnection between the practice of national ICT policies and national/regional goals of integration and development through the instrumentality of ICTs.

15.4 Conclusion This chapter argues that regional integration, socio-economic development and ICTs intersect and tate, therefore, research and policy should reflect these junctures. We have developed a framework that conceptually represents this interrelationship and demonstrated its application by examining policies on ICT for development in South Africa and Brazil, and their respective regional economic integration projects. The overview and analysis show both convergences and inconsistencies in national and regional policies aimed at utilizing ICTs for integration and regional development. We have also presented some similarities in the dynamics and interactions between SADC and MERCOSUR. We argue that national and regional policies regarding ICTs for development are not well coordinated, and that there are similarities between the two regions. For one, they both have policies that contain statements about their commitments to utilize ICTs as tools to address regional underdevelopment as well as to alleviate poverty in member countries. Both regions also lay strong emphasis on developing ICT-enabled industries in order to gain an edge in the global information economy and reduce their dependency on the export of primary products. However, as our


316


analyses show, regional policies are less likely to influence practice in Brazil and South Africa significantly than they are in the less- developed countries in their respective regions. Moreover, the regional preeminence of Brazil and South Africa in the ICT industry and the size of their economies could lead to concentrations of production and exacerbate regional disparities. The regional policies can therefore be seen as being weaker than the national ones, and policies focused on national competitiveness could hinder more equitable regional growth. The framework used in this research facilitates an understanding of the process through which ICT activities converge at national and regional levels, as witnessed by the informal civil-society network, Red Mercosur, which focuses on poorer groups in the region. Further application of the framework can include non-governmental efforts likely to represent connections that are at least inspired by the integration process. Future research should also expand the framework to allow for a more detailed, ongoing assessment of policy articulation and implementation (processes) and the resulting impact on development (outcomes). It is clear that while ICTs for national development and for regional integration and development overlap theoretically, in practice they are more often pursued in separate venues, converging by chance rather than calculated coordination. It may also be possible, with further study, to determine how aligned the regional and national goals for ICTs and development really are. For, despite regional imperatives to integrate for mutual growth and development, the sovereignty and national interests of the member states continue to be significant and represent potential and real barriers to integration and development. We conclude that ICTs are tools that can enhance development, exacerbate disparities or have very little effect at all, depending on their implementation and coordination.

References Akpan-Obong, P. and M.J.C. Parmentier (2007), ‘Information communications technologies and regional integration: The cases of Africa and South America’, Presented at the International Federation for Information Processing Conference, May, Sao Paulo, Brazil. Akpan-Obong, P. and M.J.C. Parmentier (2009), ‘Linkages and connections: A framework for research in information and communication technologies, Regional Integration and Development’, Review of Policy Research, 26 (3). Biryabarema, Elias (2008), ‘East Africa: EASSy project kicks off on friday’, The Monitor (of Kampala), 12 March. Accessed 12 November 2008. Botelho, A.J.J. and P.B. Tigre (2005)’, ‘Information and Communication Technology (ICT) for development of small and medium sized exporters in



317

Latin America: Brazil’. http://www.cepal.org/cgi-bin/getProd.asp?xml=/publicaciones/xml/1/26931/P26931.xml&xsl=/comercio/tpl-i/p9f.xsl&base=/ celade/tpl/top-bottom.xslt. Accessed 30November 2007. Bridges (2003), ‘Inventory and analysis of South Africa ICT initiatives’. www. bridges.org. Accessed 6 March 2008. Business Day (2008)’, ‘Ugandan minister knocks South Africa’s telecoms policy and high SAT3 prices’. http://www.balancingact-africa.com/news/back/balancing-act_400.html. Accessed 30 June 2008. Cassiolato, J.E. and H.M.M. Lastres (2000), ‘Knowledge, learning and development: Policy lessons from the Mercosur experience’. Prepared for the DRUID’s Summer Conference on the Learning Economy – Firms, Regions and Nation Specific Institutions, 15–17 June, Rebild. Castells, M. (1997), The Rise of the Network Society. Oxford: Blackwell. Crede, Andreas and Robin Mansell (1998), Knowledge Societies in a Nutshell: Information Technology for Sustainable Development. Ottawa: IDRC Books. e-Brasil. ‘Portal do e-desenvolvimento’. http://www.e-brasil.org.br. Accessed 21 February 2008. Etzkowitz, H., J.M. Carvalho de Mello and M. Almeida (2005), ‘Towards “metainnovation” in Brazil: The evolution of the incubator and the emergence of a triple helix’, Research Policy, 34, 411–24. Finquelievich, S. (2003), ‘ICT and economic development in Latin America and the Caribbean’. Proceedings of the World Summit of Cities and Local Authorities on the Information Society, 4–5 December 2003, http://www.cities.lyon.fr/uploadfiles/14.pdf. Accessed 8 March 2008. Friedman, T. (2005), The World is Flat. New York: Farrar, Straus and Giroux. Gwynne, R. and C. Kay (2004), Latin America Transformed: Globalization and Modernity. London: Oxford University Press. Heeks, R. (2006), ‘Theorizing ICT4D Research’, Information Technologies and International Development, 3 (3), 1–4. InfoDev Incubator Support Center, (2008), ‘Infodev establishes partnership with Brazil on ICT-enabled innovation and entrepreneurship’ http://www.idisc. net/en/Article.38788.html. Accessed 20 May 2009. International Telecommunications Union (ITU) (2001), ‘Case Study: broadband and the case of South Africa’, http://www.itu.int/osg/spu/ni/broadband/workshop/southafricafinal.pdf. Accessed 10 March 2008. International Telecommunications Union (ITU) (2008), ‘ICT Eye’. http://www.itu. int/ITU-D/icteye/DisplayCountry.aspx?countryId=7. Accessed 10 March 2008. Jaddoo, P.N. (2005), ‘The SADC perspective access – Africa’s key to an inclusive Information Society’. A paper presented at the African Regional Preparatory Conference for the WSISAccra, Ghana 29–30 January. http://www.uneca.org/ disd/events/accra/e-strategy/WSIS%20ACCRA%20presentation%20Jan%20 2005.ppt. Accessed 10 March 2008. Kaufman, S. and R. Haggard, (1992), The Politics of Economic Adjustment. Princeton, NJ: Princeton University Press. Lehohla, P.J. (2007), ‘Statistical release: Mid-year population estimates 2007’. http://www.statssa.gov.za/publications/P0302/P03022007.pdf. Accessed 10 May 2008. Martinussen, J. (1997), Society, State & Market: A Guide to Competing Theories of Development. New York: Zed Books.


318 Patience I. Akpan-Obong and Mary Jane C. Parmentier MERCOSUR (1991) Tratado de Asunsion (Treaty of Asunsion) http://www.mercosur. int/msweb/portal%20intermediario/es/index.htm. Accessed 4 March 2008. MERCOSUR Educacional. ‘Educacion Tecnología’. http://www.sic.inep.gov.br/ index.php?option=com_content&task=blogcategory&id=20&Itemid=38&lan g=es. Accessed 5 March 2008. MERCOSUR, Portal Oficial. ‘Quienes Somos’. http://www.mercosur.int/msweb/ portal%20intermediario/es/index.htm. Accessed 4 March 2008. MERCOSUR, Secretaria, Sector de Asesoria Tecnica Consultoria Economica (2005), ‘Las Asimetrías y las Politicas de Convergencia Estructural en la Integración Sudamericana’. http://www.mercosur.int/msweb/SM/Documento%20 Conjunto/MERCOSUR%20-%20Foro%20de%20La%20Paz.pdf. Accessed 30 November 2007. MERCOSUR, Sub Grupo de Trabajo No. 1 (2007), ‘Comunicaciones’. XXXIII Reunion Ordinaria, 11 May. http://www.ursec.gub.uy/R_internacionales/r_ internacionales_i.htm Accessed 4 March 2008. MERCOSUR Information Society (2007), ‘The Commision of the European Communities’. http://mercosuldigital.blogspot.com/2009/02/o-projeto-mercosul-digital-mercosur.html. Accessed 20 May 2009. Ministerio da Ciencia e Tecnologia, ‘Politica Nacional. Tecnología da Informacao’. http://www.mct.gov.br/. Accessed 6 March 2008. Ministerio da Educacao. ‘Secretaria da Educacao Profissional e Tecnologica’. http://portal.mec.gov.br/setec/. Accessed 7 March 2008. Nye, J. (1968), ‘Comparative regional integration: Concept and measurement’, Internacional Organization, 22 (4), 855–80. Organizacion de los Estados Americanos. Consejo Interamericano para el Desarrollo Integral. Sector Educativo del Mercosur (2006), Perfiles TécnicoProfesionales, Nivel Medio, Orientaciones Didacticas. Proyecto Multilateral ‘Education y Trabajo en el Mercosur’ – Fase II. Pentland, C. (1973), International Theory and European Integration. New York: The Free Press. Presidency of the Republic. Government of the Federative Republic of Brazil (2004), ‘Brazilian Monitoring Report on the Millennium Development Goals’. http://www.undg.org/archive_docs/5338-Brazil_MDG_Report_-_English_ Version.pdf. Accessed 1 November 2007. Programa del Desarrollo Regional. Red Mercosur, ‘Red de Organizaciones Comunitarias y Sociales del Mercosur’. http://www.gpdr.org/redmercosur/ html/FOTOS.htm. Accessed 5 March 2008. Projeto Mercosul Digital (2009), http://mercosuldigital.blogspot.com/2009/02/ o-projeto-mercosul-digital-mercosur.html. Accessed 20 May 2009. Red Mercosur, ‘Red de Investigacionaes Economicas del Mercosur’. http://www. redmercosur.org.uy/softis/cv/3. Accessed 5 March 2008. Rede Brasil de Tecnologia. Ministerio da Ciencia e Tecnologia. Brasil. http:// www.mct.gov.br/index.php/content/view/9427.html. Accessed 5 March 2008. SAITIS (2000), ‘South African ICT Sector Development Framework’. http://www. thedti.gov.za/saitis/about/index.html. Accessed 10 November 2007. Sorj, B. (UNESCO Brazil) (2003), ‘Confronting inequality in the information society’. www.centroedelsteing.org/. Accessed November 30, 2007.



319

South African Development Community (SADC) (2008), ‘SADC Review 2007/2008’. http://www.sadcreview.com/sadc/frsadc.htm. Accessed 10 May 2008. Unuth, R.T. (2007), Harnessing Information and Communication Technologies (ICT) for Development within the SADC Region. http://www.epolafrica.org/events/2007/ sadc-johannesburg/content/Har nessing %20Infor mation%20and%20 Communication%20 Technolog ies(IC T )%20for %20Development %20 within%20the%20SADC%20Region%20-%20Robin%20Unuth%20-%20EN.ppt. Accessed 7 March 2008. Gaborone, Botswana: SADC Secretariat. Unuth, R.T. (2007a), ICT Challenges to Enhance Socio-economic Development and Regional Integration Within SADC. Gaborone: SADC Secretariat. U.S. Commercial Service (2007) Brazil-ICT Newsletter (May/June). http://www. buyusa.gov.brazil. Accessed 23 February 2008. Vicentini Jorente, M.J. (2008), ‘Digital inclusion initiatives in Brazil: Improving education and information seeking behavior through government-academic partnerships’, American Society for Information Science and Technology Bulletin, February/March. http://www.asis.org/Bulletin/Feb-08/FebMar08_ Jorente. html. Accessed 7 March 2008. World Bank (2008), ‘ICT regulation toolkit’. http://www.ictregulationtoolkit. org/en/Section.1652.html. Accessed 4 March 2008. The World Factbook (2009), ‘South Africa’. https://www.cia.gov/library/ publications/the-world-factbook/geos/sf.html. Accessed 15 July 2009.


16 Sustainability of Technology-intensive Social Innovation in India: The Role of Absorptive Capacity and Complementary Assets Xiaolan Fu and Christine Polzin

16.1

Introduction

Over the past 30 years, technology-intensive social innovations have gained an ever-growing role in the development of many low-income countries and can be expected to remain prominent drivers of social and economic change. Information and communication technologies (ICTs), such as internet and mobile phone tools, may help to create links to commercial and social networks, cut transaction costs and foster innovative ways of delivering social services such as education and healthcare to remote areas. One of the most innovative countries in terms of using ICTs for development projects (known as ICT4D) is India. While no inventory exists of past and present ICT4D usage in rural India, numerous studies have shown that it, especially, has become a testing ground for it. For example, Patra et al. (2009) provide an excellent overview of the current state and future challenges of ICT4D usage in various domains (healthcare, agriculture, education, communications and infrastructure, governance, user interface design, etc.), based on extensive literature reviews and expert interviews. Singh (2002) provides an overview of the telecommunications sector and infrastructure in the country and identifies different methods of ICT deployment for broad-based development and economic growth in rural India. The 320


Technology-intensive Social Innovation in India

321

author suggests that ICTs have a catalytic role in spurring complementary innovations. The studies by Cecchini and Prennushi (2002), Kumar (2004) and Rao (2007) focus on the implementation and performance of ICT projects in agriculture and policy design for poverty reduction in rural India. While most studies so far have looked at the supply side of ICT4D in the rural sector, Tiwari’s study (2008) addresses the barely explored demand-side factors and identifies issues critical to enhancing the accessibility of ICT services to the poorest rural households. In India and developing countries sustainability appears to be one of the most challenging aspects of technology-intensive social innovations, given the institutional weaknesses, multiple dimensions of poverty and widespread illiteracy in many (especially rural) communities (Bell, 2006; Pal, 2003; Torero and Von Braun, 2006).1 Evaluating 76 World Bank projects with IT components in Africa, Moussa and Schware (1992) identify a number of core constraints which typically threaten implementation and long-term sustainability in these contexts which may still be relevant today, including institutional weaknesses, lack of human resources, funding issues and local environmental problems, as well as technology and information changes occurring during the projects. In this chapter, we attempt to explore the relevance of the success factors identified in technology innovation for the sustainability of technology-intensive social innovation. We focus on absorptive capacity and complementary assets in ICT4D projects. Drawing on three case studies from India (Drishtee, n-Logue and e-Choupal), this chapter aims to identify and examine factors which determine the sustainability of technology-intensive social innovations in rural settings. The chapter is organized as follows. Section 16.2 analyses the theoretical framework. Section 16.3 discusses the research methodology. Section 16.4 presents the evidence from India and Section 16.5 concludes.

16.2

Theoretical framework

16.2.1 Social innovation – delineating the field Since the term ‘social innovation’ has been used to describe a great variety of different ideas, any discussion about it is meaningless without a clear definition of the concept. In the literature, it first emerged in the early 1970s (Taylor, 1970; Gabor, 1970). For the behavioural scientist James B. Taylor social innovation could result from applied interdisciplinary research responding to social needs by introducing ‘new ways of doing things’ such as new ways of ‘dealing with poverty’


322 Xiaolan Fu and Christine Polzin

(1970: 70). Gabor thought of social innovations as instruments with which to fight for new social arrangements, for example in the form of new laws or technologies. The plethora of other definitions which has been offered by scholars from different fields to date almost seems to deprive the concept of any meaning (e.g. Brooks, 1982; Gershuny, 1983; Henderson, 1993; Lallemand, 2001; Mulgan and Steinberg, 2005; Phills et al., 2008). Following Phills et al. (2008: 36) we define social innovation as a ‘novel solution to a social problem that is more effective, efficient, sustainable, or just than existing solutions and for which the value created accrues primarily to society as a whole rather than private individuals’. In order to have some boundary around the term innovation it may be distinguished from ‘improvement’ (diffusion innovation), which is more incremental in nature ‘creativity’ and ‘invention’ (novel innovation), which are vital components of innovation. These ‘new ideas that work in meeting social goals’ (Mulgan and Steinberg, 2005: 8) could take some form of replicable programmes or organizations. They may arise from the public, private and not-for-profit sectors. In order to approach the definitional problem further, social innovation can be contrasted with ‘business innovation’ (Mulgan et al, 2007: 44; Table 16.1). While the critical resource for the latter is R&D expenditure and manpower, social innovations may also depend on other resources, including political recognition and support, voluntary labour and philanthropic commitment. The motives for social innovation usually go beyond the material dimension and include wider incentives such as recognition, compassion and care. Last but not least, different indicators of success distinguish social innovation from business innovation. Scale or market share, for example, may not be important indicators of success if a social innovation aims to address certain contained needs. Measuring and evaluating social innovation may thus require altogether different metrics. 16.2.2 Sustainability of social innovations Since social innovations are not intended to provide temporary charitable relief but should ideally be scalable and replicable, the question is how to address and theorize the concept of sustainability. How can sustainability be defined and measured? What are the different dimensions of sustainability in social innovations? While different dimensions of sustainability are part of any evaluation of the long-term success of a social innovation, it is hard to find any systematic treatment of this question in the social innovation literature. Many of the so-called ‘success stories’


Technology-intensive Social Innovation in India Table 16.1

323

Main differences between business and social innovations Business innovation

Social innovation

Critical resource

R&D investment and manpower

Political recognition and support Voluntary labour Philanthropic commitment (Financial)

Motives

Profitability Rents Market expansion Quality improvement Strategic purposes

Recognition Compassion Care Social responsibility

Indicators of success

Return on investment Scale Market share Profit

Social goals (e.g. poverty reduction, inclusion of discriminated groups) Sustainability

Major players

Private sector Research institutions and universities

NGOs Governments Social entrepreneurs

Process

R&D project, and commercialization (production and marketing)

New ideas, often originating from the grassroots, seeking financial support and implementation

Sustainability

Sustainable if there is market demand. Market exit when it is no longer profitable.

Many social innovations fail to be sustained when initial funding runs out.

of social innovation, particularly those using ICT to achieve developmental goals, are based on documentations about pilot projects. At this stage, there is a lack of serious engagement with projects which have evolved beyond the pilot stage and proved sustainable in financial and social terms.2 At the same time, there is little or no analysis being done on failure stories.3 For the purpose of this analysis sustainability will comprise financial, social and political components. 16.2.3 Technology-intensive social innovations and sustainability Innovative technology solutions can help to overcome the often characteristic lack of physical infrastructure in rural areas and to reach out into remote regions. The emergence of Internet and mobile phone technologies has sparked a great leap forward in terms of innovative projects


324

Xiaolan Fu and Christine Polzin

for the delivery of social services in rural areas over the last decade. Examples range from e-health (telemedicine) to e-governance, IT education, e-commerce, online matrimonial services, online horoscopes and many other services. This study will look more closely at providers of such services4 and compare experiences in the third and commercial sectors. To examine the sustainability of these technology-intensive social innovations we will first look at how the success of traditional technology innovation has been analysed in the literature. Bearing in mind the differences between business and social innovation described above, we will, in a second step, identify the limitations of drawing such parallels and suggest additional parameters which are necessary for evaluating the sustainability of social innovations based on ICT. In our case studies of innovative e-services, we looked at the importance of absorptive capacity and complementary assets as technologyspecific determinants of long-term sustainability. Before analysing their role in social innovation, it is useful to understand their meaning and importance in the traditional technology-innovation literature. 16.2.3.1 The role of absorptive capacity Traditional (industrial) innovation studies have found that successful development and commercialization of innovations depend crucially on the ability of an organization to acquire, assimilate and exploit knowledge from the environment (Cohen and Levinthal, 1989; 1990). This organizational learning ability, the ‘absorptive capacity’ of a firm, is a key capability when innovation depends on external knowledge. According to Cohen and Levinthal, the level of absorptive capacity of an organization depends on the absorptive capacities of its individual members: ‘To this extent, the development of an organization’s absorptive capacity will build on prior investment in the development of its constituent, individual absorptive capacities’ (Cohen and Levinthal, 1990: 131). Since innovative RD draws mainly on a firm’s existing knowledge base, absorptive capacity may increase the speed, frequency and magnitude of innovation.5 The knowledge which is thus produced through the process and results of innovation becomes part of the firm’s absorptive capacity (Lane et al, 2006: 849). This development of absorptive capacity tends to be a cumulative and path-dependent process, depending on routines and processes within the organization which enable individual-level learning to be shared, communicated and transferred to the organizational level. Absorptive capacity may thus contribute significantly to the firm’s sustainability and long-term ability to innovate.



325

New knowledge may be easily available, but Cohen and Levinthal stress the importance of investment in absorptive capacity. Public technological knowledge, for example, is of little benefit without sufficient capacity to acquire, assimilate and exploit it for commercial ends. Similarly, acquiring new knowledge through membership of interorganizational networks depends on the firm’s ability to exchange, assimilate and exploit this knowledge with the network. This, in turn, depends on the structure of the links and the processes within the latter. While the traditional innovation literature offers a number of valuable lessons for social innovation6 there is a lack of explicit and practical recommendations on how to build and exploit absorptive capacity for sustainability and scaling. Those that have been suggested tend to be so general that they are hard to implement. Lane and Lubatkin (1998: 474), for example, suggest that ‘a firm must develop a thorough understanding of its own knowledge, the processes by which it converts knowledge into capabilities, and the capacity of those capabilities to meet the demands of its environment’. With external knowledge from target communities assuming primary importance in social development projects, absorptive capacity is an equally important strategic resource for technology-intensive social innovation. While technological innovations in business usually evolve from separate R&D departments, social innovations often get their inspiration directly from the needs of communities, are designed with participatory involvement of the grassroots level and aim for social sustainability. Empirical evidence further suggests that the implementation of ICT-enabled social innovations requires absorptive capacity in form of generic and specific skills. Enos (1996, cited in Mansell and Wehn, 1998: 111) distinguishes three types of specific skills, namely participatory, facilitating and control skills. Participatory skills are needed for involvement in networked communication and information sharing. The include computer literacy and fluency in English (unless software and content is available in the local language). Facilitating skills are useful for the design, implementation and maintenance of networks. They include technical skills for implementation, maintenance and the training of users. Last but not least, control skills, based on the authority of governments, are needed for the allocation of funds to acquire appropriate ICT equipment in order to manage access to networks. Our fieldwork interviews confirmed that the lack of IT skills was one of the most frequently encountered challenges to implementing and sustaining technology-intensive innovations in rural India.7 Part of the reason for this shortage is that IT skills, in contrast to most



traditional professional skills, cannot be passed on within families over generations through hereditary training. They must be acquired through the market, which is usually coupled with high fees for training. Acquiring IT skills through such courses thus excludes poor or otherwise disadvantaged sections of society and constrains the growth of absorptive capacity for the implementation of technology-intensive social innovations. Having such skills often draws people towards the urban labour market. As a result of this and other challenges, which shall be discussed later, only a very few social entrepreneurs manage to create and sustain ICT-enabled social innovations in rural India that can be scaled up. 16.2.3.2

The role of complementary assets

Apart from its (internal) absorptive capacity, the degree to which a company can gain from innovations and changes in IT also depends on its ability to exploit investment in (internal and external) complementary assets – human and organizational processes as well as physical capital (Hughes and Morton, 2006: 50). The effective combination and utilization of technological and complementary assets may be a key for successful and sustainable innovation. Broadly defined, complementary factors to IT range from organization and management practices to industry organization and regulation, economic structure, government policy and human capital investment. These assets may influence capital deepening, technical progress and labour quality and thus improve the production process (Dedrick et al., 2003: 3). Teece (1986) differentiates between generic, specialized and co-specialized complementary assets: ‘Generic assets are general purpose assets which do not need to be tailored to the innovation in question. Specialized assets are those where there is unilateral dependence between the innovation and the complementary asset. Co-specialized assets are those for which there is a bilateral dependence.’ The successful commercialization of an innovation usually requires that ‘the know-how in question be utilized in conjunction with other capabilities or assets’ (ibid.: 288) and services such as marketing and after-sales support. These services are often obtained from specialized complementary assets. To commercialize a new product or service via the Internet, for example, information needs to be disseminated over specialized information channels. In the context of IT innovations, the traditional literature has mainly focused on the role of complementary assets as a determinant of the return on IT investments in terms of organizational payoff (Barua et al, 1996; Davern and Kaufman, 2000).8



327

In order to reap the elusive productivity rewards of IT (extracting the greatest value) an organization may not only need to exploit investment in physical capital but also change the way it does business by rethinking its corporate strategy and remaking its basic structure, processes and culture (Hughes and Morton, 2006). The complementary assets and capabilities identified in the traditional innovation literature are similar to those needed for technology-intensive social innovation. In the rural parts of developing countries, physical capital in the form of reliable ICT equipment, power supplies and Internet connections are still a huge operational challenge (Annamalai and Rao, 2003: 19; Bell, 2006: 3–6; Dossani et al, 2005: 31). For many e-services projects, however, the timely access to information is vital for the success of the business model. Human capital in form of committed and well trained staff may be equally important (Dossani et al, 2005: 28). In the traditional literature less attention is paid to ‘social factors’ such as trust, political influence and social networks (Heeks, 2000). It is easy to underestimate the importance of these factors. In our case studies, however, they had significant weight.

16.3 Research methodology To investigate the sustainability of technology-intensive social innovations, we compared different case studies in India. Case investigations provide detailed information that may present unexpected patterns which might not otherwise be revealed. As Bradshaw and Wallace (1991: 166) point out, case studies are particularly appropriate for studies which are adversely impacted by data availability, as it is the case with technology-intensive social innovations in developing countries. A comparative approach is useful in understanding the complex phenomena embedded in long-term dynamics (Miles and Huberman, 1994; Yin, 2003). We have, therefore, chosen three e-services case studies, of which two (Drishtee and e-Choupal) have been widely recognized as successful while the third (n-Logue) has experienced more difficulties in scaling up. Of the two successful cases, Drishtee represents innovative third-sector projects and e-Choupal is an example of a commercial ICT project initiated by the private sector. The sample selection process involved three steps. First, a range of critical criteria was defined which would allow an investigation of technology-intensive social innovation. Given the interest of this study in the sustainability of these innovations, case studies with sufficient experience in order to evaluate the determinants of their long-term


328


performance were sought. Case studies to be selected had to fulfil the following criteria: existence for at least three years and having reached beyond the pilot project stage; selection reflects a range of different business models, allowing for focused comparisons and identification of key variables in the business model components that determine the sustainability and potential of scalability of the projects; reputation as social innovations with the ambition to provide models which could be replicated and scaled up; location in central or south India in order to be easily and cost effectively accessible for the research team. In the second step, an extensive review of the literature on ICT4D (including academic literature, reports, Internet resources, journals and other media sources) and interviews were conducted with the research project’s partners in India, in order to construct a list of innovative ICTenabled service delivery projects in rural India. Seven projects met the criteria.9 The third step was the selection and detailed investigation of a sample of case studies from the list, which reflected both successful and less successful projects in terms of sustainability. For the investigation, semi-structured interviews were conducted with project managers of nine case studies in India (six out of the seven identified in step two (ASA; Drishtee; the information villages project of the M.S. Swaminathan Research Foundation; the i-community project in Kuppam, run by the NGO BASIX; n-Logue and the Indian Tobacco Company’s e-Choupal)10, plus three which were younger than three years but which offered additional insights into new business models and challenges in establishing innovative social innovations in rural India. Out of the six long-term projects, three were chosen for further investigation in this chapter (Drishtee, n-Logue and e-Choupal). One of the main drawbacks to the use of case studies is the uncertainty about whether or not the cases are representative and whether we can draw generalizations from them (Hamel et al, 1993: 20). This research does not attempt to build a theory about technology-intensive social innovations on a single case by assuming that other cases are similar (Bradshaw and Wallace, 1991: 157). Instead, using a comparative method, it combines the specificity and holistic approach of single case studies with the generalizing capability of comparative approaches, keeping in mind the historical and social specificities of our country of investigation, India.



329

16.4 Technology-intensive social innovation – evidence from India Over the past decade, rural e-services have become increasingly popular in India, mainly driven by private social entrepreneurs, corporations and non-governmental organizations (NGOs). In 2005, the Indian Department of Information and Technology announced a plan to establish 100,000 Common Service Centres (CSCs) across the country, financed and implemented by the India government and the private sector.11 The scheme, as approved by the government in 2006, envisions CSCs as delivery points for government, private and social-sector services to citizens in rural areas. 16.4.1

Case studies from rural India – background

Table 16.2 provides a brief overview of the three cases. Drishtee and n-Logue are both commercially focused rural kiosk initiatives, offering e-services through entrepreneurial village computing centres. Drishtee initially developed and implemented software for an e-governance initiative (‘Gyandoot’). Later, it started providing e-governance services to rural villages and computerized a number of government services (e.g. birth certificate applications) in its region of operation and added commercial services in partnership with banks, insurance companies and other private companies. Drishtee sells the start-up equipment for kiosks and provides training. Most operators take loans to finance the initial investment. Kiosk services are based on Drishtee’s website and support for the kiosk owners is provided by Drishtee representatives. Revenues for Drishtee are generated through three different sources: a one-time licence fee, monthly franchise fees and revenues from the services providers (banks, insurances, education and health) for every transaction on the network. Our second case study, n-Logue was founded by the Telecommunications and Computer Networks (TeNet) group at the Indian Institute of Technology in Madras (Jhunjhunwala et al., 2004). The TeNet group developed corDect wireless local-loop technology to provide connectivity in rural areas. Entrepreneurs who act as ‘local service providers’ (LSPs) run a regional kiosk based on a corDect tower that provides phone services and Internet connectivity. Each LSP needs to set up as many kiosks as are required to sustain itself. The company sells kiosk start-up kits, training and support to the LSPs and shares profits and losses with them. Revenue is generated through the provision of phone services and connectivity (based on hourly or fixed monthly fees for unlimited access). After a rapid expansion of its services across several states, n-Logue realized that most


330


Table 16.2

Three cases of social innovation

Case

Essential information

Social innovation

Drishtee

For-profit Ltd. company, set up in 2000 in Dhar (Madhya Pradesh) Microfranchising model Other firms may use Drishtee’s platform to market their products and services Kiosk provider finances equipment (usually through bank loan) In 2008, approximately 1700 Drishtee entrepreneurs operate in ten states of India.

Internet kiosks for villages offering wide variety of e- services, sometimes including e-governance Direct delivery network with potential cost and time savings

n-Logue

For-profit company, set up in 2001 Microfranchising (3 tiers) Targeting very small villages (1000–1500 people) Kiosk providers pay for equipment, maintenance and training (usually through bank loans) and keep all marginal revenues N-Logue derives revenue from kiosk providers who pay for internet access N-Logue gradually set up 3500 Internet kiosks, at least 50% are now closed and the remaining are only half functioning

Wide variety of e-services for very small villages Shared-use model

e-Choupal

Established by the Indian Tobacco Company (ITC Ltd) Started in 2000 with a pilot of 6 e-choupals in Madhya Pradesh In 2008, more than 6400 e-Choupal kiosks are operational in six states of India ITC provides all equipment and pays for telephone bills

Social value for small and medium farmers derives from transparent and informed transactions and often higher prices when selling to ITC. Value from additional services marketed through e-Choupal.

of its new kiosks did not generate sufficient returns on investment. More than half of all the kiosks had to close down. With regard to village computing projects like Drishtee and n-Logue, Bell (2006) notes three important commonalities: 1) the existence of and the need for an umbrella organization to support the network of village computing centers, 2) the need to focus



331

applications and services on specific and clearly defined community needs or desires with the understanding that these may and will likely evolve over time and 3) the need to establish an appropriate funding mix through the creation of ‘win-win’ relationships between all stakeholders to ensure the financial sustainability both of the village computing center and the network that supports it. (Bell 2006: 16) The financial sustainability of such innovative projects depends on sufficient revenues to cover the costs for complementary assets, notably organizational infrastructure and back-end operations such as experienced management resources, software development, client support services and marketing. E-Choupal, a commercially focused e-services initiative following a corporate-based model was established by the Indian Tobacco Company (ITC Ltd), a major Indian conglomerate with various business lines (including agribusiness, fast-moving consumer goods, hotels, packaging, paper and cardboard). At the time of analysis, it is one of very few comparable efforts that have been judged as both effective and sustainable. E-Choupals are operated by a local farmer and provide farmers with information on agricultural market prices, the weather and good farming practices. The computer also allows farmers to place orders for agricultural inputs such as seeds and fertilizers. By providing real-time price information on agricultural products through its e-Choupals, ITC encourages the farmers to sell their agricultural produce directly to its hubs instead of involving numerous intermediaries. The vertical supply chain integration has enabled ITC to have direct access to producers, while the latter may benefit from higher prices for their agricultural produce. The model has also enabled ITC to improve its business conduct through increased transparency of information and operations. 16.4.2

Challenges to sustainable social innovation

Challenges to sustainability may differ between the private and the third sector. Taking Drishtee and n-Logue as examples of innovative third sector projects and e-Choupal as a commercial case study we can identify and compare a number of challenges to sustain technologyintensive social innovations. Despite promising initial success both Drishtee and n-Logue faced problems related to the lack of absorptive capacity and complementary assets in their target market. ITC managed to overcome these challenges through its ability to cross-subsidize and invest substantial resources into R&D. Less obvious perhaps, but still important, is the common underestimation of the degree of freedom


332


that rural customers have in choosing new products and services. Our evidence suggests that it is more difficult for third-sector companies to overcome this challenge. This difficulty is linked to their limited ability to provide comprehensive end-to-end solutions. 16.4.2.1 Absorptive capacity The case studies illustrate the importance of the three specific skills suggested by Enos (1996), that is participatory, facilitating and control skills, in enabling the success and sustainability of rural e-services. A number of more general skills are also mentioned. First, the participatory skills of kiosk operators are a very important strategic resource for social innovation and key for sustainable operations. Both n-Logue and Drishtee demand a certain minimum level of education and computer literacy from their kiosk operators and invest in their training in order to ensure that they can adequately explain and promote their services. With the rapid expansion of their services, however, both Drishtee and n-Logue experienced difficulties in finding sufficiently qualified staff and kiosk operators who could easily acquire, assimilate and exploit knowledge from the environment to foster further innovation and sustainability. Second, social innovations which are highly technology intensive require facilitating skills for the design, implementation and maintenance of networks. For the design and implementation of new services, however, technical skills are not a sufficient guarantee for success when they are not complemented by other assets. The experience of n-Logue shows that innovative e-services often take time to generate acceptance and sustainable demand and clients will not immediately be willing to pay for them. The TeNeT group developed video-conferencing software for n-Logue which works on low bandwidths and enables face-toface interaction with service and information providers in areas such as health, government and agriculture. Despite widespread interest in the villages, however, n-Logue faced ‘difficulties in promoting these videoconferencing based services and developing revenue models around them’ (Gurumurthy et al, 2005: 166). Small private operators in the third sector may not have the capacity to test such services without sufficient revenues. Financial reserves or grants from external partners may be needed to provide such new services for free until they are deemed commercially viable. Even then, social entrepreneurs may need such capacity or other solutions (such as cross-subsidizing) if they wish to enable equal access for poorer sections of the community who cannot afford the fees for such services. With fewer financial resources, n-Logue



333

could not implement as many e-services for its kiosks as Drishtee which received support from various donors and philanthropic organizations including the Acumen Fund, Nike Foundation and the World Bank’s International Finance Corporation (IFC). The control skills of governments were also important determinants of success. In the case of n-Logue, for example, lack of influence on important government regulation hindered the company from implementing its original business model based on cheaper access to rural telephony. It was not able to obtain a licence for telephony, and its prospects for tie-up with the licence-holder were ‘rendered more difficult since new [2003] regulation removed obligations on them for providing certain minimum rural telephony coverage’ (Gurumurthy et al, 2005: 166). General skills and absorptive capacity may be assessed through feasibility surveys indicating the capacity of the local environment to absorb and sustain innovative e-services. Drishtee follows stringent rules in choosing the places for its kiosks. To be eligible, villages must have a minimum population of 1200, be accessible from three to five surrounding villages and have favourable conditions for telecommunication. The literacy level of the targeted age group and the paying capacity of the villagers also play an important role for sustainability. 16.4.2.2

Complementary assets

The development of Drishtee, n-Logue and ITC’s e-Choupal network illustrates how different kinds of complementary assets can contribute to the sustainability of technology-intensive social innovations. Drishtee managed to develop diverse supply chains for the delivery of various e-services including computer courses, matrimonial services, mobile phone top-ups and e-governance. The same network was leveraged to diversify into the distribution of physical products such as mobile phones, low-cost spectacles, seeds and fertilizers.12 The technical progress reflected in its range of increasingly innovative products and services was enabled through networks and partnerships, both of which constitute crucial generic assets for technology-intensive social innovations, especially in third-sector institutions where resources for R&D are often limited. In the commercial sector, ITC may have been less dependent on networks and external partners. However, it also combined its technologies with various complementary assets. Investment in physical and digital infrastructure and human and organizational processes, as well as improved business conduct, were used to eliminate market inefficiencies and to sustain its innovation. Setting up e-Choupal kiosks enabled


334


farmers to unbundle finding out about prices and physical transactions at the mandi (foodgrain trader’s shop). At the same time, the inefficiencies inherent in the physical flow through the mandi were eliminated by establishing procurement hubs for clusters of e-Choupals for scientific inspection of quality and procurement and for the preservation of the produce (storage and testing equipment). Human and organizational processes along the supply chain were changed to manage the digital and physical infrastructure and to fill critical gaps in the chain. ITC’s former commission agents at the mandi have been incorporated into the new supply chain as collaborators (now known as Samjojaks) who provide logistical support and manage the physical infrastructure. One farmer is appointed to handle the digital infrastructure in the village and to inspect the farmers’ produce (the so-called ‘Sanchalak’). According to S. Sivakumar, CEO of ITC’s International Business Division, this human infrastructure is the ‘social capital of the village’ which makes up for the shortage of physical and institutional infrastructure (interview, December 2007). ITC may also use e-Choupals to compete on produce variety in order to increase its own returns as well as that of the farmers. Sanchalaks and Samjojaks constitute vital complementary assets in this process and have also been trained to implement segregation and identity preservation of agricultural produce. ITC’s procurement of wheat through e-Choupals illustrates how the range of complementary assets enables ITC to trace the produce delivered and to respond to consumer concerns. In India’s different agro-climatic zones, wheat grains vary widely in grades and satisfy different consumer preferences. By selling wheat to the mandis, however, farmers had gained no benefit from producing higher-quality grains because all varieties were aggregated as one average quality. ITC’s e-Choupal innovation enables farmers to compare prices for their quality at the village level. Providing specialized knowledge though the internet, e-Choupals may help farmers to customize their produce to the right consumer segments. In combination with ITC’s storage and handling logistics, this configuration of physical infrastructure at the hubs and digital infrastructure in the spokes allows ITC to preserve the identity of different varieties of wheat (and other grains) throughout the supply chain. It also encourages farmers to improve the quality of their produce and attract higher prices. 16.4.2.3 Customer freedom of choice E-services providers usually conduct surveys to analyse the needs of the villagers and to estimate the demand for their products and services.



335

With an increased range of products available villagers are supposed to take more informed decisions. What is easily forgotten is that many potential rural customers, however well-informed they may be, are not completely free to take their own decisions. Interlinked contracts involving loans, for example, have been found in all parts of India under various terms and conditions (Bardhan, 1980; Bell and Srinivasan, 1989; Harriss, 1991; Harriss-White, 1996; Sarap, 1991). A common form of contract involves the sale or purchase of goods such as grains or household items. The strength of interlinked contracts may differ from region to region and among different groups of people. In Western Orissa, for example, Sarap (1991) found that interlinked transactions were highest among (collateral poor) tenants and landless labourers, followed by small farmers. Furthermore, they seemed to be confined to households with little land, tenant households and/or households with low educational and caste status. Medium or large farmers did not enter into interlinked credit transactions. This is not a universal pattern, however, and other studies have observed widely differing relations in other parts of India. They also illustrate the variety of terms and conditions of interlinked contracts. Depending on the conditions of these contracts with moneylenders or other middlemen, potential customers may not be able to purchase certain goods and services from e-services providers like Drishtee or n-Logue even if they indicated a need. For the ITC, however, the interlocked contracts of many farmers were a market opportunity. Without modern communication tools, farmers often know the price for their produce only once they are at the mandi. To avoid the uncertainty and high transaction costs for trading through mandis small and marginal farmers, in particular, often enter into interlocked contracts which offer end-to-end solutions, for example buying inputs and taking loans from a village trader who will purchase, transport and sell the produce as well as lend money. These apparently convenient end-to-end solutions may easily drive farmers into cycles of dependence and exploitation (Harriss, 1982; Narayanamoorthy and Kalamkar, 2005; Sarap, 1991). The ITC decided to offer farmers both an unbundled option and a (voluntary) end-to-end solution around its technology-intensive social innovation.

16.5 Conclusions Research on technology-intensive social innovation can benefit from parallels in established research on traditional technology innovation and emerging concepts of social entrepreneurship. This chapter has shown that absorptive capacity and complementary assets – key



concepts in innovation research – can be critical factors for the success and sustainability of innovative ICT-enabled projects in developing countries. Embedding these concepts into social entrepreneurship may form a useful framework for analysis and generate new insights. Analyses of the factors contributing to sustainability of technologyintensive social innovations in developing countries are still rare. Our case studies of ICT-enabled services in rural India help to fill this gap in the literature and suggest that absorptive capacity, complementary assets and consumer choice are important factors for the sustainability of such social innovations. In technology-intensive social innovation, absorptive capacity, such as capacity in learning and basic skills in computer and internet use, is of crucial importance for these social innovations to be able to deliver their promise and therefore hold up the demand pull for its sustainability. Complementary assets are of similar importance, as demonstrated by the different performance of n-Logue and Drishtee and by the way in which the digital and physical infrastructure of the ITC’s e-Choupal system helped the company change its agri-business. Last but not least, a factor which has not featured in the traditional innovation literature but which assumed primary importance in the sustainability of e-services in rural India is the degree of freedom that customers have in choosing and using the services provided through an established service provider. Without this customer freedom of choice, new innovations have limited space for entry, survival and development. Findings from this research have important policy implications. First, the operators of the ICT4D projects or the providers of e-services, no matter whether they are government, NGOs or the private sector, should all provide sufficient IT skills training to the targeted community to ensure that the expected benefits of these social innovations are realized. Second, investment shall not simply focus on computers and software systems. Investment in complementary physical and organizational infrastructure, such as supply chain of physical production material and information, is also crucial to deliver the promise to the community, and maintain the vitality of the social innovations. Third, complementary institutional and regulatory changes are needed to ensure the social innovations have a healthy environment to survive and develop.

Notes The authors are grateful to the Engineering and Physical Science Research Council (EPSRC) (Grant No. EP/E026052/1) financial support, and to John Toye,



337

Andy Dearden, Haider Rizvi, Subodh Gupta, Paul Matthews and conference participants at the Development Studies Association annual conference (2008) for helpful discussions. 1. Some critics hence question the usefulness and priority given to ICTs in development efforts to relieve extreme poverty (Keniston, 2003; Upadhaya, 2002). 2. One of the most cited and documented exceptions is Grameen Bank (see, e.g. Bayes, 2001). 3. For some recent exploratory attempts at documenting and categorizing failures, see, e.g. Dossani et al, 2005; Heeks, 2000; 2002; Klimaro and Nhampossa, 2005; Kumar and Best, 2006. Again, however, there is no systematic treatment of the question of sustainability. 4. Excluded from our study of technology-intensive social innovations are social software and open-source methods (for a discussion of these innovations, see Mulgan and Steinberg, 2005). 5. Therefore, absorptive capacity is commonly measured in terms of the responsiveness of R&D, assuming that R&D generates new knowledge and thus contributes to a firm’s absorptive capacity. R&D intensity reflects the responsiveness of R&D activity (in a static model as R&D/sales) to learning incentives (relevance, ease and appropriability). 6. See Cohen and Levinthal (1990); Zahra and George (2002) for an excellent discussion and review. 7. A review of the literature on IT projects in developing countries suggests that this is not an exceptional result. Moussa and Schware (1992: 1745), for example, report that in 20 per cent of the World Bank projects with IT components implemented between 1983–90 ‘little or no attention was paid to training people who would use the system, service support, absorption capacity of the institution, and data availability and validity’. 8. For extensive literature reviews on the impact of ICT on business performance in general see for example Dedrick et al (2003), Draca et al (2006); Melville et al (2004). 9. The projects thus selected included ASA, the Computer Munchi project by the NGO PRADAN, Drishtee, e-Choupal by the Indian Tobacco Company (ITC), the MSSRF information village project, n-Logue and Warana i-community. 10. The Warana Wired Village Project had to be excluded on the basis of cost and time constraints in the field. 11. http://www.mit.gov.in/default.aspx?id=825. 12. Building these partnerships and seeking funding may have been helped by the large social network of Drishtee’s founder, Satyan Mishra, an Ashoka Fellow.

References Annamalai, K. and S. Rao (2003), ‘ITC’s E-Choupal and profitable rural transformation: Web-based information and procurement tools for Indian farmers’, What Works, August. Bardhan, P.K. (1980), ‘Interlocking factor markets and agrarian development: A review of issues’, Oxford Economic Papers, 32, 82–98.


338


Barua, A., C.H.S. Lee and A.B. Whinston (1996), ‘The calculus of reengineering’, Information Systems Research, 7, 409–28. Bayes, A. (2001), ‘Infrastructure and rural development: Insights from a Grameen Bank village phone initiative in Bangladesh’, Agricultural Economics, 25, 261–72. Bell, C. and T.N. Srinivasan (1989), ‘Interlinked transactions in rural markets: An empirical study of Andhra Pradesh, Bihar and Punjab’, Oxford Bulletin of Economics & Statistics, 51, 73–83. Bell, T. (2006), Village Computing: A State of the Field. Reflections on the Village Computing Consultation. Washington, DC: Grameen Foundation. Bradshaw, Y. and M. Wallace (1991), ‘Informing generality and explaining uniqueness: The place of case studies in comparative research’, International Journal of Comparative Sociology, 32, 154–71. Brooks, H. (1982), ‘Social and technological innovation’, in S.B. Lundstedt and E.W. Colglazier, Jr. (eds), Managing Innovation: The Social Dimensions of Creativity, Invention and Technology. New York: Pergamon Press. Cecchini, S. and G. Prennushi (2002), ‘Using information and communications technology to reduce poverty in rural India’. World Bank PREM Notes No. 70. Available at SSRN: http://ssrn.com/abstract=671645. Cohen, W. M. and D.A. Levinthal (1989), ‘Innovation and learning: The two faces of RandD’, Economic Journal, 99, 569–96. Cohen, W.M. and D.A. Levinthal (1990), ‘Absorptive capacity: A new perspective on learning and innovation’, Administrative Science Quarterly, 35, 128–52. Davern, M.J. and R.J. Kauffman (2000), ‘Discovering potential and realizing value from information technology investments’, Journal of Management Information Systems, 16, 121–43. Dedrick, J., V. Gurbaxani and K.L. Kraemer (2003), ‘Information technology and economic performance: A critical review of the empirical evidence’, ACM Computing Surveys, 35, 1–28. Dossani, R., D.D. Misra and R. Jhaveri (2005), ‘Enabling ICT for rural India’, Project Report, November. Draca, M., R. Sadun and J. Van Reenen (2006), ‘Productivity and ICT: A review of the evidence’, Discussion Paper 749. London: Centre for Economic Performance, London School of Economics. Enos, J. (1996), ‘Human skills and institutions for information-sharing: The challenge to least developed countries’, commissioned by UNU/INTECH and COLCIENCIAS as a background paper for the UNCSTD Working Group on IT and Development. Gabor, D. (1970), Innovations: Scientific, Technological, and Social. New York: Oxford University Press, vi, 113. Gershuny, J. (1983), Social Innovation and the Division of Labour. Oxford: Oxford University Press (Library of political economy). Gurumurthy, A., P.J. Singh and G. Kasinathan (2005), ‘TeNeT, n-Logue and the DHAN foundation: Exploring appropriate ownership models for rural propoor ICTD initiatives’, in S.Ó. Siochrú and B. Girard (eds), Community-Based Networks and Innovative Technologies. New York: UNDP. Hamel, J., S. Dufour and D. Fortin (1993), Case Study Methods. Newbury Park, CA/London: Sage.



339

Harriss, J. (1991), ‘The green revolution in North Arcot: Economic trends, household mobility, and the politics of an “Awkward Class” ’, in P. Hazell and C. Ramasamy (eds), The Green Revolution Reconsidered: The Impact of HighYielding Rice Varieties in South India. London: Johns Hopkins University Press. Harriss-White, B. (1996), A Political Economy of Agricultural Markets in South India: Masters of the Countryside. New Delhi/London: Sage. Heeks, R. (2000), ‘The approach of senior public officials to information technology-related reform: Lessons from India’, Public Administration and Development, 20 (3), 197–205. Heeks, R. (2002), ‘Information systems and developing countries: Failure, success, and local improvisations’, The Information Society, 18, 101–12. Henderson, H. (1993), ‘Social innovation and citizen movements’, Futures, 25 (3), 322–38. Hughes, A. and M.S.S. Morton (2006), ‘The transforming power of complementary assets’, MIT Sloan Management Review, 47, 50–8. Jhunjhunwala, A., A. Ramachandran and A. Bandyopadhyay (2004), ‘n-Logue: The story of a rural service provider in India’, The Journal of Community Informatics, 1 (1), 30–8. Keniston, K. (2003), ‘IT for the common man: Lessons from India’, i4donline (May–June), 4–13. Kimaro, H.C. and J.L. Nhampossa (2005) Analyzing the problem of Unsustainable health information systems in less-developed economies: Case studies from Tanzania and Mozambique. Kumar, R. and M.L. Best (2006), ‘Impact and sustainability of E-government services in developing countries: Lessons learned from Tamil Nadu, India’, The Information Society, 22, 1–12. Kumar, R. (2004), ‘eChoupals: A Study on the financial sustainability of village internet centers in rural Madhya Pradesh’, Information Technologies & International Development, 2, 45–73. Lallemand, D. (2001), Les défis de l’innovation sociale. Issy-les-Moulineaux: ESF (Collection Actions sociales Société). Lane, P.J. and M. Lubatkin (1998), ‘Relative absorptive capacity and interorganizational learning’, Strategic Management Journal, 19, 461–77. Lane, P.J., B.R. Koka and S. Pathak (2006), ‘The reification of absorptive capacity: A critical review and rejuvenation of the construct’, Academy of Management Review, 31, 833–63. Mansell, R. and U. Wehn (eds) (1998), Knowledge Societies: Information Technology for Sustainable Development. Oxford: Oxford University Press. Melville, N., K. Kraemer and V. Gurbaxani (2004), ‘Review: Information Technology and Organizational Performance: An Integrative Model of IT Business Value’, MIS Quarterly, 28 (2), 283–322. Miles, M.B. and A.M. Huberman (1994), Qualitative Data Analysis: An Expanded Sourcebook. Thousand Oaks, C.A/London: Sage. Moussa, A. and R. Schware (1992), ‘Informatics in Africa: Lessons from World Bank experience’, World Development, 20, 1737–52. Mulgan, G. and T. Steinberg (2005), Wide Open: Open Source Methods and their Future Potential. London: Demos.


340 Xiaolan Fu and Christine Polzin Mulgan, G., S. Tucker, R. Ali and B. Sanders (2007), ‘Social innovation: What it is, why it matters and how it can be accelerated’, Skoll Centre for Social Entrepreneurship Working Paper. Narayanamoorthy, A. and S.S. Kalamkar (2005), ‘Indebtedness of farmer households across states: Recent trends, status and determinants’, Indian Journal of Agricultural Economics, 60, 289–301. Pal, J. (2003), ‘The developmental promise of information and communications technology in India’, Contemporary South Asia, 12, 103–19. Patra, R., J. Pal and S. Nedevschi (2009), ‘ICTD State of the Union: Where have we reached and where are we headed’ ICTD 2009, ACM IEEE Conference Proceedings, Doha, April 17. Phills, J.A., K. Deiglmeier and D.T. Miller (2008), ‘Rediscovering social innovation’, Stanford Social Innovation Review, 34–43. Rao, N.H. (2007), ‘A framework for implementing information and communication technologies in agricultural development in India’, Technological Forecasting and Social Change, 74, 491–518. Sarap, K. (1991), Interlinked Agrarian Markets in Rural India. London: Sage. Singh, N. (2002), ‘Information technology as an engine of broad-based growth in India’, in P. Banerjee and F.-J. Richter (eds), The Information Economy in India. London: Palgrave Macmillan, 24–57. Taylor, J.B. (1970), ‘Introducing social innovation’, The Journal of Applied Behavioral Science, 6 (1), 69–77. Teece, D.J. (1986), ‘Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy’, Research Policy, 15 (6), 285–305. Tiwari, M. (2008), ‘ICTs and poverty reduction: User perspective study of rural Madhya Pradesh, India’, European Journal of Development Research, 20, 448–61. Torero, M. and J. Von Braun (eds) (2006), Information and Communication Technologies for Development and Poverty Reduction: The Potential of Telecommunications. Baltimore, MD.: Johns Hopkins University Press. Upadhaya, G.R. (2002), ‘Digital delusions in the South’, HIMAL, August. Available at http://www.himalmag.com/2002/august/essay.htm (Accessed 4 July 2008). Yin, R.K. (2003), Case Study Research: Design and Methods (3rd edition). Thousand Oaks, CA/London: Sage (Applied social research methods series; v. 5). Zahra, S.A. and G. George (2002), ‘Absorptive capacity: A review, reconceptualization, and extension’, Academy of Management Review, 27 (2), 185–203.


Conclusions: Science, Technology and Development – Emerging Concepts and Visions Xiaolan Fu, Luc Soete and Lina Sönne

C.1 Introduction The rate at which the BRICS countries have grown in the last couple of decades has been impressive. As the chapters of this book have shown, a number of factors have contributed to this, including, for example, foreign direct investment (FDI), improved absorptive capacity and growth of domestic high-tech industries as well as the support of small and medium enterprises (SMEs). To this end each of the BRICS has used different strategies and policy mixes. However, this growth and the countries’ policy focus have not necessarily been altogether inclusive. This concluding chapter discusses some of the emerging concepts of science and technology (S&T) for development, related to both the high-tech and high-growth sectors, as well as those that can be deemed more inclusive, targeting the ‘bottom’ of the pyramid. S&T for development has often been talked of in terms of the radical nature, or paradigm shift, of new scientific breakthroughs or technological inventions which appear to offer new windows of opportunity for economic development and might eradicate at once world poverty, diseases and decades of lack of development in many less developed countries (LDCs). Despite the caution of Sanjaya Lall in warning that such ‘miracle’ growth opportunities warrant quite explicit industrial policies,1 there has been a tendency, certainly within the new global era of digital communications and world market transparency, to take those new catching-up opportunities for granted as ‘technology transfer manna’ from the North to be implemented in the South. However, the analytical shift from S&T to innovation which has occurred over the last ten to 20 years brings in a new vision of development: one which now acknowledges the fully ‘endogenous’ nature of 341


342

Xiaolan Fu, Luc Soete and Lina Sönne

innovation, rather than the old, neo-classical external view of technological change and technology transfer. That process of innovation is actually much more complex and challenging in a developing-country context than in a developed-country one. It has only recently become recognized in the endogenous growth literature,2 vindicating many of Lall’s earlier writings on technological capabilities, that the appropriate innovation policy challenge for a country will be closely associated with its level of development. In a high-income country context, the innovation policy challenge will be more and more directed toward questions about the non-sustainability of processes of ‘creative destruction’ within environments that increasingly give premiums to insiders, to security and to risk aversion, and ultimately to the maintenance of income and wealth. However, as already argued by Lall some 30 years ago (see Lall, 1978; 1992), in the emerging, developing-country context of the BRICS, the innovation policy challenge appears more directed toward ‘backing winners’ industrial S&T policies. The challenge is how to broaden further an emerging domestic technological expertise in the direction of international competitiveness and specialization. Such broadening will have to involve a strong recognition on the part of policy-makers of the importance of engineering and design skills and of accumulating ‘experience’ rather than just R&D investment.

C.2 Technology and the emergence of formalized industrial research activities The strong focus on industrial R&D as the factor behind economic growth and development is of relatively recent origin. Up to the late 1950s, it was barely recognized by economists, despite the recognition that ‘something’ (a residual, a measure of our ignorance) was behind most of the economic growth in the twentieth century and the postwar period in particular. But, of course, long before the twentieth century, experimental development work on new or improved products and processes was carried out in ordinary workshops.3 However, what became distinctive about modern, industrial R&D were its scale, its scientific content and the extent of its professional specialization. Suddenly a much greater part of technological progress appeared attributable to R&D work performed in specialized laboratories or pilot plants by fulltime, qualified staff. It was also this sort of work which got recorded officially in R&D statistics, if only because it was totally impracticable to measure the part-time and amateur inventive work typical of the nineteenth century (Freeman and Soete, 2006).


Conclusion 343

As historians have argued, the industrial research ‘revolution’ was not just a question of a change in scale. It also involved a fundamental change in the relationship between society on the one hand and technology and science on the other. The term ‘technology’, with its connotation of a more formal and systematic body of learning, came into general use only when the techniques of production reached a stage of complexity where traditional methods no longer sufficed. The older, more primitive arts and crafts technologies continued to exist next to the new ‘technology’. But the way in which more scientific techniques would be used in producing, distributing and transporting goods led to a gradual shift in the ordering of industries alongside their ‘technology’ intensity. Thus, typically for most developed industrial societies of the twentieth century, there were now both high-tech intensive industries which received heavy, sector-internal R&D investment, and low-tech intensive, more craft-techniques-based industries, with only a very little of their own R&D effort. In many policy debates therefore, industrial dynamism became, somewhat naively, associated with the dominance in a country’s industrial structure of the high-tech intensive sectors. However, the more sophisticated sectoral studies on the particular features of inter-sectoral technology flows, from Pavitt (1984) to Malerba (2004), brought back to the forefront of the analysis many of the unmeasured, indirect sources of technical progress. Unfortunately, many of those insights have not been translated into attempts to broaden the policyrelevant concept of R&D.

C.3 From industrial R&D to innovation: A paradigm shift? As acknowledged by many recent innovation studies scholars ranging from economists such as Paul David and Dominique Foray to S&T studies scholars such as Mike Gibbons and Helga Novotny, a major shift in the understanding of the relationships between research, innovation and socio-economic development has occurred. Here, it is interesting to note that both the more economically embedded innovation research community and the chiefly STS-embedded research community more or less converged on this issue: in each case the perception of the nature of the innovation process has changed significantly. Innovation capability is today seen less in terms of the ability to discover new technological principles, and more as an capacity to exploit the effects produced by new combinations – one is reminded of Schumpeter’s already old notion of ‘neue Kombinationen’ – and the


344


use of items from the existing stock of knowledge (David and Foray, 2002). This alternative model, closely associated with the emergence of numerous knowledge ‘service’ activities, implies in other words a more routine use of an existing technological base allowing for innovation without the need for particular leaps in science and technology, sometimes referred to as ‘innovation without research’. This shift in the nature of the innovation process implies a much more complex structure of knowledge-production activities, involving a greater diversity of organizations having knowledge production as an explicit goal. The previous industrial system was based on a relatively simple dichotomy between knowledge generation and deliberate learning taking place in R&D laboratories, including engineering and design activities, and activities of production and consumption where the motivation was not to acquire new knowledge but rather to produce or use effective outputs. As argued elsewhere (Ghosh and Soete, 2006), the collapse (or partial collapse) of this dichotomy has led to a proliferation of new processes that have as an explicit goal the production and use of new knowledge which may not be readily observable from national R&D statistics but which appears nevertheless essential to sustain innovative activities in a global environment. In short, traditional R&D-based technological progress, still dominant in many of the BRICS’ industrial sectors, ranging from the chemical and pharmaceutical industries to motor vehicles, semiconductors and electronic consumer goods, has been characterized by the S&T system’s ability to organize technological improvements along clear agreed-upon criteria and a continuous ability to evaluate progress. At the same time a crucial part of engineering research has consisted, as Richard Nelson put it, ‘of the ability to hold in place’, to replicate at a larger industrial scale and to imitate experiments carried out in the research laboratory environment. As a result it involved first and foremost a cumulative process of technological progress – a continuous learning from natural and deliberate experiments. The more recent mode of technological progress is associated with the knowledge paradigm and the service economy, for example, information and communications-based (ICT) efficiency improvements in the financial and insurance sectors, the wholesale and retail sectors, health, education, government services and business management and administration. Such progress is much based much more on flexibility which leads it to be confronted with intrinsic difficulties as regards replication. Therefore, learning from previous experiences or from other sectors is difficult and sometimes even misleading. Furthermore, evaluation is


Conclusion 345

problematic as a result of changing external environments in several dimensions: over time, among sectors and across locations. It will often be impossible to separate context-specific variables from real causes and effects. Technological progress therefore is to a greater extent based on trial and error, but (unlike the life sciences) without providing ‘hard’ data that can be scientifically analysed and interpreted. The result is that technological progress will be less predictable, more uncertain and ultimately more closely associated with entrepreneurial risk taking. Attempts at reducing such risks might involve, as Von Hippel (2004) has argued, a much greater importance given to users in the research process itself.

C.4 Implications of innovation for development The implications of this new mode of technological progress for development are straightforward: they bring to the forefront the importance of endogenous innovation processes in the BRICS and other developing countries. In the old industrial S&T model, the focus within the context of development was quite naturally on technology transfer and imitation: the latter to some extent being the opposite of innovation. In the new model, innovation is anything but imitation. Rather, every innovation now appears unique with respect to its application. Reuse and recombination of sometimes routine, sometimes novel, pieces of knowledge are likely to be of significant importance, but their successful application may ultimately involve engineering expertise, design capabilities or even research. Thus, today we see innovation in developing countries emerging from both the ‘tip’ and the ‘base’ of the pyramid. C.4.1 Innovation from the ‘tip’ to the ‘bottom’ of the income pyramid One feature of industrial R&D, and the underlying model of technological progress, which has not received much attention in the development literature is the focus of industrial research on the continuous quality improvement of existing and new consumer goods. The implication of this is that demand for improved and/or new goods is constantly enlarging. A growth model to that effect emerged over the post-war period in the US, Europe and Japan, and appeared to generate its own infinite demand for more material consumer goods – a continuous growth path of rising income with increasing consumer goods’ production and consumption (Pasinetti, 1981). It appeared to assume that


346


consumer goods – contrary to food – would remain completely unaffected by Engel’s law. The continuously rising industrial R&D efforts in high-income countries appeared in other words to match perfectly the continuously rising incomes of the citizens of those countries, leading to a constant enlargement of their shopping basket of new and better-designed, or better-performing, products. The early actual demand for such improvements often arose from niche professional, sometimes military, use, far from the mainstream consumer. However, as a result in large part of the media and its tendency to emphasize the prestige in the use of such niche goods, the average, mainstream consumer would easily become convinced that they required the technologically sophisticated professional-quality characteristics even though those characteristics might ultimately add only marginally to the consumer’s utility. Often these professional niche groups, generally within the highest income bracket, in society that constitute the ‘tip’ of the income pyramid would act as the early adopters and the first try-out group, contributing to the innovation monopoly rents of the innovating firm. In this way, a continuous circle of research was set in motion which centred on the search for new qualitative features4 to be added to existing goods. This ‘professional-use driven’ innovation circle has been the main source for extracting innovation rents out of consumer goods – ranging from consumer electronics to sport goods, footwear, household equipment, computers, mobile phones, medical diagnostics, sleeping comfort and so on – with a physical lifespan that was considered ‘too long’. However, the worldwide risks of this relatively straightforward professional-use driven innovation strategy for the existing global multinational corporations have increased significantly, not least because of globalization. While the world market for new, innovative goods appears, at first sight, enormous and without a doubt sufficient to recoup investments relatively quickly, the huge research, development, prototype and global marketing costs, coupled with ever-increasing numbers of competing international players means that the amount of time during which a firm can enjoy its innovation rents is diminishing very rapidly. Therefore, despite the growth of the high-income classes in the large emerging BRICS economies, the new generation of goods being targeted at and sold to them will not generate enough earnings to fund both the shift toward mass production and the development of the next-generation technology of consumer goods. Thus, having developed incredibly sophisticated technological goods, many firms are now encountering major global sales problems over a much contracted


Conclusion 347

product life cycle with increased competition and rapidly over-saturating markets. Instead, we have recently begun to see a shift in focus from innovation emerging from the tip of the pyramid to innovation emerging from the base. C.4.2 Innovation at the base of the income pyramid: A new form of ‘appropriate innovation’? The need for a shift in research on innovation in private businesses was popularized by CK Prahalad in his by now famous book The Fortune at the Bottom of the Pyramid (2004), which suggests that poverty can be alleviated through profit-making investments in goods and services aimed at the base, or bottom, of the pyramid (BoP). Drawing on his famous BoP innovation example of the multiple-fuel stove developed for the rural poor, in which cow dung and biomass (sticks and grass) can be used as cooking fuels, Prahalad observes that: the process of designing these breakthrough innovations started with the identification of the following four conditions: ... 1. The innovation must result in a product or service of world-class quality. 2. The innovation must achieve a significant price reduction – at least 90 percent off the cost of a comparable product or service in the West. 3. The innovation must be scalable: It must be able to be produced, marketed, and used in many locales and circumstances. 4. The innovation must be affordable at the bottom of the economic pyramid, reaching people with the lowest levels of income in any given society. (CK Prahalad, The Innovation Sandbox) In recent years there has been a flood of similar examples of BoP innovations being introduced by foreign, large multinational corporations (MNCs) from developed countries in developing countries, sometimes in poor rural villages, sometimes in urban slums.5 At first sight the MNC BoP examples seem to contradict Lall’s earlier observations about the limited effectiveness of technology transfer through foreign direct investment (FDI). As Lall noted, back in 1992: ‘With few exceptions, the developing country affiliate receives the result of innovation, not the innovative process itself: it is not efficient for the enterprise concerned to invest in the skill and linkage creation in a new location.’ (Lall, 1992: 179). This is where BoP innovation takes on a completely new meaning, as we explain in the following three broad points.


348


First of all the likely and most successful location of the innovative process activities, the BoP learning lab, will have to be close to BoP users’ contexts. Given the crucial role of users in the innovation process as argued above, this will imply that BoP laboratories will have to be embedded in users’ environments rather than the traditional hightech R&D centres and enclaves, whether in the developed or developing country. To be successful though, such a version will have to pay particular attention to all the elements and features emphasized by Lall back in the early 1990s: the local context, the vertical linkages and the avoidance of innovation ‘truncation’ (Lall, 1980; 1992), meaning the isolation of the innovation process from the host country’s technological and production infrastructure. All this brings to the forefront for successful BoP innovation the need for a local business model that also fully embodies local behavioural responses to innovation. Thus we get the increasingly recognized need in BoP innovation for strategic alliances between large MNCs and local non-governmental organizations (NGOs). Second, the feedback from BoP users and from design developers upstream to more applied research assistance, even fundamental research in some of the core laboratories of Western firms, might well become one of the most interesting examples of reverse transfer of technology (from the South to the North), reinvigorating and motivating the research community in the highly developed world increasingly ‘in search of relevance’ (see Soete, 2008).6 Not surprisingly, the main focus within the developed world at the moment is on BoP innovations in the health area, a sector where applied medical research is increasingly dominated by access to new technologically sophisticated equipment to combat, among other things, antibiotic resistance, infectious diseases or drug-resistant tuberculosis. In fact, health is the sector most in need for what could be called a BoP research reprioritization. Third, the innovation process itself is now likely to be reversed, starting with the design phase, which will, to a large extent, be concerned with finding functional solutions for some of the particular BoP users’ framework conditions. This will involve not just the need to bring the product onto the market at a substantially lower price than existing goods, as Prahalad emphasized, but also, and more in line with Sanjaya Lall’s observations, a clear adaptation to the often poor local infrastructure facilities with respect to energy delivery systems, water access, transport infrastructure and ICT access. Autonomy is the key word here. It is no surprise that the most rapidly spreading technology in developing countries has been in mobile communications with currently


Conclusion 349

more than 3 billion users worldwide. Apart from high-quality energy and water infrastructures, broadband network availability is undoubtedly one of the most pervasive drivers for BoP innovation. Another one might well be ‘cradle-to-cradle’ sustainable innovation. The lack of high-quality logistic infrastructure facilities in rural development settings might well imply that once goods are sold, the repair and/or central recollection of obsolete goods or their parts will be expensive. By contrast local reuse along the principles of cradle-to-cradle might well be a new form of sustainable grassroots innovation. One may therefore talk about ‘appropriate innovation’ and that there seems to be some analytical similarity with the old notion of ‘appropriate technology’.7 Appropriate innovation, in fact, brings us neatly to Schumacher’s emphasis on appropriate small-scale solutions in the local area (Schumacher, 1973). Here there is a clear link with a different kind of innovation taking place at the base of the pyramid that is gaining in importance within the context of development and poverty alleviation. Whilst Prahalad’s BoP innovation is MNC-based, this is the grassroots, social or inclusive innovation that emerges from developing countries’ own entrepreneurs, small firms and NGOs. In a sense the notion of ‘grassroots innovation’ developed by Anil Gupta can be considered as the endogenous, intrinsic version of Prahalad’s external, top-down version of BoP innovation (Gupta, 1997). This entrepreneur-based innovation can essentially be divided into two types. One is where entrepreneurs see business opportunities on a larger scale including the famous M-PESA mobile phone based money transfer service in Kenya, as well as the three Indian examples documented in Chapter 16 of this volume: Drishtee, n-Logue and e-Choupal. The other type includes local entrepreneurs acting as small-scale service providers where the state or major utilities companies have failed. For instance, in India the Sulabh Sanitation Movement provides for-profit sanitation services in slums. Other services provided by small, local entrepreneurs include energy (Ehrhardt, 2000) and water (McIntosh, 2003). The point to note here is that these entrepreneurs are in a good position to offer innovative solutions for their local community since they have a deep knowledge of the market and are small and informal enough to have the flexibility to try out new solutions. In fact, this makes ‘small-scale providers an important part of the private sector which has enormous potential in the provision of basic services’ (Baker and McClain, 2009: 19). The importance of social innovation and the role local, domestic entrepreneurs are playing in the process, is now being recognized at a policy


350


level. For example, the latest OECD report on SMEs, Entrepreneurship and Innovation (OECD, 2009), notes that although further research is required, policy-makers and investors agree that inclusive, or social, innovation is clearly an effective tool for dealing with both social needs and sustainable development.

C.5

Conclusions

Underlying the changes described in this chapter is, of course, the dramatic shift in the globalization of S&T as it has accelerated over the last ten to 15 years. For most countries in the world, including the BRICS, the contribution of domestic S&T to the global stock of knowledge is today relatively small and the contribution to domestic productivity growth equally so. There is little doubt that the largest part of worldwide productivity growth over the last ten years has been associated with an acceleration in the diffusion of technological change and with global access to codified knowledge. The role of ICT has been instrumental here, as has that of increased capital and organizationally embedded forms of technology transfer, as noted by Sanjaya Lall. While a huge worldwide concentration of research investment in a relatively small number of rich countries/regions remains, it is important to realize that such activity, whether privately or publicly funded is increasingly becoming global in focus. The shifts in global demand underlying the process of globalization taking place today also increasingly affect the allocation of private resources and, by extension, the sort of research, knowledge creation and diffusion and innovation being carried out in research laboratories, no matter where they are located. From this perspective it is important to realize that the new, much more globally focused, international business community is becoming concerned, about the sustainability of its long-term growth, because the demand of high-income groups is, in absolute terms, increasing at a much slower rate than that of lower-income groups. This is especially true for the BRICS. Up to a point this trend is similar to what happened in the US at the beginning of the twentieth century, also a period of rapid growth and rising income inequality, when Henry Ford introduced the Model T Ford. His ‘putting America on wheels’ strategy was centred on assemblyline production and on paying workers higher wages so as to create a lasting market for the car. How to create a similar global mass market for consumer goods in the context of the twenty-first century represents a much more complex, global challenge, but the similarity and


Conclusion 351

the timing of such business concerns is striking. It is, in a sense, the ultimate paradox of inequality; the business community itself is becoming concerned with the fact that too much inequality is actually limiting its own long-term output growth potential. Therefore, the vision of innovation for development outlined here appears new, but also very familiar, especially to all those who know Sanjaya Lall’s research and will undoubtedly recognize many of his views and visions in some of the concepts and notions discussed here on how to develop successful innovation-for-development strategies.

Notes 1. As in his most explicit criticism of the World Bank’s 1993 study on the East Asian Miracle, Lall (1994). 2. This view of the philosophy and aims of innovation policies differing amongst countries according to their level of development, reminiscent of many of the arguments of the old infant industry type arguments, has now become very popular in the endogenous growth literature. See Aghion and Howitt (2006). 3. As we noted elsewhere: ‘The early classical economists were well aware of the critical role of technology in economic progress even though they used a different terminology. Adam Smith (1776) observed that improvements in machinery came both from the manufacturers of machines and from “philosophers or men of specialisation, whose trade is not to do anything but to observe everything”. But although be had already noted the importance of “natural philosophers” (the expression “scientist” only came into use in the nineteenth century), in his day the advance of technology was largely due to the inventiveness of people working directly in the production process or immediately associated with it: “... a great part of the machines made use of in those manufactures in which labour is most subdivided, were originally the inventions of common workmen” (Smith, 1776, p. 8. Technical progress was rapid but the techniques were such that experience and mechanical ingenuity enabled many improvements to be made as a result of direct observation and small-scale experiment. Most of the patents in this period were taken out by “mechanics” or “engineers”, who did their own “development” work alongside production or privately. This type of inventive work still continues today and it is essential to remember that is hard to capture it in official R&D statistics.’ (Freeman and Soete, 1997). 4. One may think of audio and sound, vision and clarity, miniaturization and mobility, weight and shock/water resistance, feeling and ergonomiticity, etc. 5. For some of those examples in the sanitation area, see Ramani (2008). For an overview of the BoP literature see Weehuizen (2008). 6. The notion of appropriate technology (Schumacher, 1973) was of course much more formalized in terms of a rational set of economically determined ‘choices of technique’ (Sen, 1968), depending very much on capital-labour substitution possibilities. The term ‘appropriate innovation’ by contrast is much more open.


352 Xiaolan Fu, Luc Soete and Lina Sönne

References Aghion, Phillippe and Peter Howitt (2006), ‘Appropriate growth policy: A unifying framework’, Journal of the European Economic Association, 4 (2–3), 269–314. Baker, J. and K. McClain (2009), ‘Private sector initiatives in slum upgrading’, World Bank Urban Papers, 8. David, P. and D. Foray (2002), ‘An introduction to economy of the knowledge society’, International Social Science Journal, 54 (171), 9–23. Ehrhardt, David (2000), ‘Impact of market structure on service options for the poor’, for the conference infrastructure for development: Private solutions and the poor, 31 May–1 June, London. Freeman, C. and L. Soete (1997), The Economics of Industrial Innovation, 3rd edition. Cambridge, MA: MIT Press. Freeman, C. and L. Soete (2006), ‘Changing STI climate: A sky without horizons’, Blue Sky II Forum, 25–7 September, Ottawa, Canada. Ghosh, R. and L. Soete (2006), ‘Information and Intellectual property: The global challenges’, Industrial and Corporate Change, 15 (6), 919–35. Gupta, A. (1997), ‘The honey bee network: Linking knowledge-rich grassroots innovations’, Development, 40 (4), 36–40. Lall, S. (1978), ‘Transnationals, domestic enterprises and industrial structure in host LDCs’, Oxford Economic Papers, 30 (2), 217–48. Lall, S. (1980), ‘Vertical inter-firm linkages in LDCs’, Oxford Bulletin of Economics and Statistics, 203–28. Lall, S. (1992), ‘Technological capabilities and industrialization’, World Development, 20 (2), 165–86. Lall, S. (1994), ‘The East-Asian miracle’, World Development, 22 (4), 645–54. Malerba, F. (ed) (2004), Sectoral Systems of Innovation. Cambridge, MA: Cambridge University Press. McIntosh, Arthur C. (2003), Asian Water Supplies – Reaching the Poor, Asian Development Bank. London: IWA. OECD (2009), ‘SME’s enterpreneurship and innovation’. Paris: OECD (Draft Report). Pasinetti, L. (1981), Structural Change and Economic Growth: A Theoretical Essay on the Dynamics of the Wealth of Nations. Cambridge: Cambridge University Press. Pavitt, K. (1984), ‘Patterns of technical change: Towards a taxonomy and a theory’, Research Policy, 13 (6), 343–73. Prahalad, C.K. (2004), The Fortune at the Bottom of the Pyramid. Eradicating Poverty Through Profits. Philadelphia, PA: Wharton School. Ramani, S. (2008), ‘Breaking the guardian knot in sanitation – Development of new technology & business models to create a market for toilets for India’s poorest’, UNU-MERIT Research Memorandum 2008–12. Schumacher, E.F. (1973), Small is Beautiful: A Study of Economics as if People Mattered. London: Blond and Briggs. Sen, A. (1968), Choice of Techniques: An Aspect of the Theory of Planned Economic Development. Oxford: Blackwell. Smith, A. (1776), An Enquiry into the Nature and Causes of the Wealth of Nations. Dublin: Whitestone.


Conclusion 353 Soete, Luc (2008), ‘Science, technology and development: Emerging concepts and visions’. UNU-MERIT Working Paper 2008–001. Von Hippel, E. (2004), Democratizing Innovation. Cambridge, MA: MIT Press. Weehuizen, R. (2008), ‘Innovation for the bottom of the pyramid’, March, UNUMERIT, mimeo. Weitz, A. and R. Franceys (2002), ‘Beyond boundaries: Extending services to the urban poor’, Asian Development Bank.




Index Absorptive capacity 10–12, 226, 238, 240, 254, 271, 324–6, 332–3, 335–6 Brazil basic economic indicators 18–19, 20–1, 41, 50–4 impact of Information and Communications Technologies (ICTs) 312–15 BRICS and FDI 101–2, 203–4, 214–17, 226–35, 255–66, 273–4 and investment 147–61 economic and geographic data 2, 51–4, economic growth of 1, 50–6 exports of 39–44, 57–9 industrialisation in 61–5 overview of development and technological performance 341–51 R&D expenditure 2 sustainable technologies in 281, 288–97 technological development in 61–5, 341–51 Catching up experience of China including IT industry 89–104 methods of describing and evaluating 89–90 China basic economic indicators 18–19, 20–1, 41, 50–4 Complementary assets concept of 321, 324–7 relationship to policy in India 329–35 Diasporas role in acquisition of foreign knowledge 25–7

Education role in innovation 16–17, 25, 27–31, 35, 43–4, 62, 65, 143, 240, 242, 245, 249, 251, 254–5, 313–15, 320, 329 Energy technology 34–5, 44, 119, 126, 270, 281, 283–5, 289, 292–6 Environmental technology 281–98 Foreign Direct Investment (FDI) and acquisition of foreign knowledge/technology 8, 10, 23–5 and innovation 23–5, 101–2, 212–7, 221–35 outward FDI from BRICS 255–66, 273–4 receipt by BRICS 101–2, 203–4, 214–17, 226–35 South-South 258–75 spillover effects of 101–2 Human capital and innovation 10, 28–31, 122, 238–40, 245–55, 326 relationship to FDI 238–40, 245–55 Industrial clusters 186–99 case study: China 190–9 Information and Communications Technologies (ICTs) catch up experience of China 95–104, 189 role in development process 27, 30, 114, 189 300–4, 308–12 India basic economic indicators 18–19, 20–1, 41 Industrialisation capacity in BRICS 61–5 case study: China 186–99

355


356

Index

Innovation constraints on 114–17 indicators of 69–70, 139–42 methods of financing 147–61 open innovation 97–9 and patent filings 70–2, 77, 83 process of 127–30, 139–42, 345–50 and public policy 15–17, 41, 89–94, 99–101, 117–21, 166–7, 270–2 social innovation, concept of 11–12, 320–3 Innovation strategy and public policy 15–17, 41, 89–94, 99–101, 117–21, 166–7, 270–2 intellectual property 23, 31, 34, 62, 81, 133, 141, 314 investment sources and financing in BRICS 147–61 inward-oriented development 7, 23, 125, 132, 142 Kaldor-Verdoorn Law 64 Knowledge dissemination of, including indicators 16–17, 37–41 tacit 8, 93, 174–5, 183 Manufactured exports BRICS as exporters of 39–44, 57–9 Mercosur role of ICTs in 303–8, 312–6 Modern Growth Theory relationship to technological capabilities 125–30 Multinational companies (MNCs) relationship to Foreign Direct Investment 26, 31, 33, 133, 190, 224, 347–9 Third World MNCs 258–75 Organizational learning 176, 324

Patents 33–4, 110 as indicators of innovation 69–82 varying country standards 75 Research and Development (R&D) 31–7, 64, 98–9, 140, 266–8, 289–97, 342–5 Reverse engineering 20, 26–7, 35–6, 90–1, 102 SMEs case study: China 190–7 and innovation 8, 93, 96, 125, 129, 136, 139, 186–90, 197–9, 341, 350 Social innovation, concept of 320–3 as distinguished from business innovation 322 case study: India 329–35 measurement of 327–8 South Africa basic economic indicators 50–4 share of hi-tech exports in trade 113–14 Southern African Development Community (SADC) role of ICTs in 304–6 SADC Declaration on ICT 305–6 Spillover effects 10, 69–70, 225, 233 of FDI 10, 101–2, 104, 204, 225, 238–55, 271–2 Sustainable technology 281–98, 323–7 Tacit knowledge 8, 93, 174–5, 183 Technological capabilities of the BRICS impact of government policy on 15–17, 41, 89–94, 99–101, 117–21, 166–7, 270–2 and R&D policies 93–4, 117–21 Technological capabilities 60–3, 132–5, 172–4, 177–8 case study: Argentina 132–5 case study: Chile 135–9 distinction between firm-level v national-level 61–2


Index 357 Technological capabilities – continued estimates and indicators for Brazil, China and India 42 Technological learning 171–2, 174–7, 178–82 and improvements in technology capability 89–92, 181–4

Technology transfer 4, 8–10, 34, 97–104, 204, 221–24, 240, 252, 274, 341–2, 345, 347, 350 Trade Relationship to acquisition of foreign knowledge 20–3


The Rise of Technological Power in the South

Revolution!: South America and the Rise of the New Left

The Indian Slave Trade: The Rise of the English Empire in the American South, 1670-1717

The Impacts of Technological Change

After The Flood 2 - Rise To Power

Anglo American and the Rise of Modern South Africa

Our South: Geographic Fantasy and the Rise of National Literature

Cinderella in the South

Turn in the South

Turn in the South

Summer in the South

In the South Seas

Summer in the South

The Normans In The South

Elsie in the South

Out In The South

Summer in the South

The Rise and Fall of Infrastructures: Dynamics of Evolution and Technological Change in Transport

The Normans In The South

Rise of the TaiGethen

Rise of the Governor

Rise of the Lycans

Rise of the King

Rise of the Nephilim

Rise of the Wolf

Rise of the Carnelians

Rise of the Machines

Rise of the Elgen

Rise of the Wolf

Rise of the Huntress

The Rise of Technological Power in the South

Revolution!: South America and the Rise of the New Left

The Indian Slave Trade: The Rise of the English Empire in the American South, 1670-1717

The Impacts of Technological Change

After The Flood 2 - Rise To Power

Anglo American and the Rise of Modern South Africa

Our South: Geographic Fantasy and the Rise of National Literature

Cinderella in the South

Turn in the South

Turn in the South

Summer in the South

In the South Seas

Summer in the South

The Normans In The South

Elsie in the South

Out In The South

Summer in the South

The Rise and Fall of Infrastructures: Dynamics of Evolution and Technological Change in Transport

The Normans In The South

Rise of the TaiGethen

Rise of the Governor

Rise of the Lycans

Rise of the King

Rise of the Nephilim

Rise of the Wolf

Rise of the Carnelians

Rise of the Machines

Rise of the Elgen

Rise of the Wolf

Rise of the Huntress

Recommend Documents