UNITED NATIONS INDUSTRIAL DEVELOPMENT ORGANIZATION

economy environment employment

Industrial Development Report 2002/2003

Competing through Innovation and Learning

United Nations Industrial Development Organization

Copyright 2002 United Nations Industrial Development Organization The designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat concerning the legal status of any country, territory, city or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries. Designations such as "developed", "industrialized" and "developing" are intended for statistical convenience, and do not necessarily express a judgement about the state reached by a particular country or area in the development process. The mention of firm names or commercial products does not imply endorsement by UNIDO. Material in this publication may be freely quoted or reprinted, but acknowledgement is requested, together with a copy of the publication containing the quotation or reprint.

ISBN

92-1-106420-1

UN sales number E.02.II.B.13 UNIDO publication identification number 414

Foreword

HE CURRENT PREDICAMENT OF MOST DEVELOPING COUNTRIES AND

the state of the global economy deflate the optimism in the United Nations Millennium Declaration and the Millennium Development Goals. The problem is not just the apparent wedge between targets and trends. It is the more fundamental differences in what to expect of the future and how to get there. Take equity and efficiency. Proponents of each see themselves as offering the best possible way to maximize both social fairness and economic progress in the long run. But even a partial marriage of the two (often) polar views is not yet in sight, despite the development communitys sorely felt need. Productivity growth, equity, poverty eradication and security can all reinforce one another. But for that to happen requires attentiveness to the widespread access to wealth-creating assets, especially through education, the basis for acquiring skills and grasping opportunities. Countries need to pass thresholds on the route to becoming productivity-driven economies. They also need to put in place strategies and policies to sustain productivity gains over time. UNIDOs research agenda centres on these effortsto advance todays development agenda. With more ability to create wealth, people and countries can achieve sustainable livelihoods, begin to attack poverty and have the rule of law take hold. But this requires a massive mobilization of skills and capabilities, while fostering greater equity. In turn, an economic and social system that offers the right public goods and that rewards the opportunities for equitable and efficient growth is the best guarantee for the rule of law and security for all. Poor countries stay poor when low productivity leaves poverty and inequity untouched, slowing growth and hobbling the innovative forces of society. The challenge for policymakers is to move to a virtuous circle where productivity gains reduce poverty and enhance social equitywhich can then feed economic growth.

A strand of contemporary thought on development holds that reducing poverty and promoting equity need not compromise growth. UNIDO strongly subscribes to this view. Indeed, UNIDO views social progress as an essential part of sustained growth. To break the negative links between equity and efficiency, direct policy action is needed to enhance the complementarities among equity, productivity and growth. Recent cases of countries catching up strongly suggest that improvements in equity and reductions in absolute poverty do much for attaining sustainable productivity increases and for sharply narrowing the income gap with the advanced economies. That productivity growth has the salutary effect of also favouring further improvements in social equity.

The ingredients of growth with equity

Development thought now stresses innovation, technological change and other intangibles at least as much as capital accumulation. Trickle-down effects through cross-border capital and technology flows have benefited only a handful of developing countries. Why? Because those are the only countries that paid attention to increasing productivity and creating a good climate for investment. Too many other countries are still missing the domestic links between productivity growth, equity and poverty. And they are not taking advantage of increasing returns and agglomeration economies. Their real problems: weak commitments to development and poor implementation of policy reforms. So, trickle-down effects, though desirable, now seem far from automatic. Needed instead is deliberate action. Developing countries have to foster the diffusion of technology to them and within them, both in quantity and quality. They also have to break the negative links between low equality and low productivity. In short, they need development policies directed explicitly to reducing poverty. Only with such policies will it be

iii

possible for a substantial part of the developing country population to share in the benefits of globalization. Economic growth, so essential for eradicating poverty, rests on the accumulation of physical and human capital and, very importantly, on the gains in productivity. Catching up rapidly rests, in addition, on a fast rate of structural transformation. Normally associated with the initial phases of catching up quickly are rapid capital accumulation, dramatic increases in labour participation and the large-scale absorption of foreign technology. To maximize productivity, each requires significant learningso that society can assimilate, adapt, master, develop and efficiently use foreign technology. This learning, and the institutional and policy setups that enable it, starts with mobilizing domestic innovative capabilities to fuel sustained structural change. Opening the economy fosters both domestic competition and inflows of embodied and disembodied technology. And given the drive to export, it spurs the domestic diffusion of international productivity and quality standards through exposure to international competition. As countries begin to catch up, innovation-driven productivity gains have to become the engine of growth. The reason is that input-driven growth eventually runs into rapidly diminishing returnsunless supported by the assimilation, adaptation and mastery of rapidly changing technologies. In a world where continuous innovation and international competition drive one another, countries have to devote special effort to keeping up with the advancing technological frontier. Faced with this challenge, the development agenda needs to give pride of place to policies for sustainable productivity growth. Countries have to establish framework conditions for the rich interplay of resources, markets and institutions. They have to expose themselves to the spurs of competition in the international economy. And they have to supply the public goods needed to match gains in efficiency with improvements in equity.

For the more advanced developing countries the development priority is to deepen and upgrade their links with the world economy so that they can apply their innovative capabilities to international competition and domestic development. For the least developed countries the key development priority is to take the first steps towards being able to do this. They need to set policies that allow them to take a greater part in international trade, investment and technology flows. Yes, macroeconomic stability and a sound incentive system are important. But so is attending to basic human needs, strong institutions and the building of social capital.

Shortcomings of todays policy models

Much economic reform in developing countries was based on the assumption that equity and wealth creation would come from market-led structural reforms. But the record shows that countries heeding the prescriptions saw equity deteriorate and efficiency gains turned out to be short-lived and elusive. The reason is that the policy reform packages of the 1990s were unbalancedin the best of cases having only two of the three legs to support a durable social consensus in the fight against poverty and the drive for sustainable development. Macroeconomic and institutional reforms have made some difference. But the third legensuring enduring productivity gains by mobilizing the innovative potential of societyhas been missing. Innovation and technological learning were expected to follow. But for the most part they did not. Hence the attenuated effect of the macroeconomic and institutional reforms. Overcoming this deficit is still a major taskone that UNIDO addresses as a fundamental part of its mandate.

Spreading access to incomegenerating assets

The ability of a social system to achieve high rates of economic growth and at the same time improve equity lies in its capacity to ensure ample access to income-generating assets across the population. That is one of the key lessons from the recent industrialization experience. The most sustainable income-generating assets are not necessarily physical or financialthey are human, embodied in people and institutions. These assets comprise entrepreneurial, management, technical and scientific knowledge as well

The predicament of the least developed countries

The least developed countries, still struggling to meet the basic human needs of their population, have had their health, social and economic standards slip over the last few decades. Indeed, the real per capita income of 30 developing countries is lower today than it was 35 years ago.

iv

Industrial Development Report 2002/2003

as governance systems and codes of conduct. A society that makes universal access to knowledge and skills a key priority is much more likely to marry equity and efficiency than a society that does not. Investing in formal education is just one aspect of this. An innovative and competitive private sector is essential. So are institutions and incentives geared to eradicating corruption, ensuring the rule of law, promoting social capital and easing resource reallocation. At the base, of course, is a competitive business environment and sound macroeconomic management. Efficiency with equity can be achieved only when markets, agents and institutions interact to diffuse the fruits of technical progress and so to improve all sectors of society.

A rule-based international system in which most members cannot play is doomed. Developing countries, especially the least developed, should not be held to standards they cannot meet today. What would be fair is to hold them to relaxed standards they can meetand to give them assistance for increasing their capacities. Only a major concerted effort by the international community to remove obstacles to market access and support developing countries capacity building to conform with the rules can redress intercountry disparities.

Caring about future generations

Natural resources are as much for creating wealth as for improving human welfare. When economic activity is harmful to the environment, resource productivity today is augmented at the expense of resource productivity tomorrow. To the extent that the neglect of natural capital affects global commons, international norms need to be set and complied withjust as when the health, safety and security of the population are at stake. But developing countries have to conform to the emerging international norms as well, and for this they need home-grown scientific and technical capability.

Providing greater symmetry in the costs and benefits of international compacts

The new development round of trade negotiations addresses the perception that the opportunities to draw on the benefits of the Uruguay Round agreements have not been evenly distributed across countries. The substantial protectionism still prevailing in international trade is clearly biased against activities in which developing countries enjoy greater comparative advantages, such as agriculture and labour-intensive manufactures. Moreover, developing countries can take advantage of market access to industrialized countries only by conforming to their stringent sanitary, phytosanitary and other regulations and standards. To do this, they have to overcome their technical and scientific underdevelopment.

***
These are the main messages of this report, the first of a new UNIDO series devoted to the various dimensions of industrial development highlighted in this foreword. This new series is intended to build on development policy experience and to advance the frontiers of current thinking, with a focus on the least developed countries.

Carlos Alfredo Magarios Director-General UNIDO

Foreword

Acknowledgements
The Industrial Development Report 2002/2003 was prepared under the overall guidance of Carlos Alfredo Magarios, DirectorGeneral of UNIDO, and the supervision of Yo Maruno by a team led by Frdric Richard. Sanjaya Lall was the main adviser and architect of the Industrial Scoreboard. Larry Westphal was the adviser on the substantive content and organization of part 2 of the report. The following UNIDO staff and international experts were involved in drafting the report: Manuel Albaladejo, Raphael Kaplinsky, Sanjaya Lall, John Mathews, Olga Memedovic, Torben Roepstorff, Bruce Ross-Larson, Hubert Schmitz, Francisco Sercovich, Jebamalai Vinanchiarachi and Larry Westphal. Bruno Dissmann, Gerhard Margreiter and Tetsuo Yamada provided statistical data. Assistance was provided by Jrg Mahlich and Ganesh Wignaraja, as well as by UNIDO interns, Perihan Khairat, Vesna Petakovic, Karma Tshering and Chi Yiu Wan. Annaliesse M. Hyser and Christen E. Lungren also provided assistance. John Degnbol-Martinussen conducted a comprehensive review of the final draft of the report. E. Agustin Stellatelli contributed to the design and publication of the report. Elizabeth Cruz, Maria Fermie, Heidi Karbiener de Alvarez, Rosemarie Oestreicher and Chantal Pothier provided clerical and secretarial support. The UNIDO Inter-divisional Advisory Group provided useful comments and suggestions during the preparation of the report. The members were Ahmed Ben Brahim, Jos de Caldas Lima, Bernardo Calzadilla-Sarmiento, Jrgen Estrup, Helmut Forstner, Sarwar Hobohm, Diana Hubbard, Claudia Linke, Wilfried Luetkenhorst, Ralph Luken, Paul Makin, Jrgen Reinhardt, Ghislain Robyn, Ricardo Seidl da Fonseca and Francisco Sercovich. Valuable inputs were also provided by the following UNIDO staff members: Aurelia Calabro in Bellamoli, Michele Clara, Edward Paul Clarence-Smith, Giovanna Ceglie, Andr de Crombrugghe de Looringhe, Ferenc Demjen, Mohamed Dhaoui, Richard Kennedy, Gerardo Patacconi, Frank van Rompaey and Fabio Russo. Much insight was gained from two Expert Group Meetings that were held in April and November 2001 respectively. The discussions and comments made by the participants greatly contributed to improving the scope and analytical content of the report. UNIDO staff and the following international experts attended these meetings: Manuel Albaladejo, Jacques de Bandt, Michael Best, John Degnbol-Martinussen, Sanjaya Lall, John Mathews, Jrg Mahlich, Hubert Schmitz, Carlos Scheel, Simon Teitel, Morris Teubal, Larry Westphal and Ganesh Wignaraja. This report draws on 24 background papers prepared by Jacques de Bandt, Andrew Baxter, Michael Best, Ha-Joon Chang, Daniel Chudnovsky, Phil Cooke, Christian Debresson, Talat Diab, Charles Edquist, Gary Gereffi, Karen Hamann, John Humphrey, Alex Inklaar, Raphael Kaplinsky, Rughvir Khemani, Nagesh Kumar, Sanjaya Lall, John Mathews, Stan Metcalfe, Pierre Mohnen, Mike Morris, Martin Mulligan, Miguel A. Prez G., Julian Perkin, Carlo Pietrobelli, Slavo Radosevic, Jeff Readman, Teresa Salazar de Buckle, Carlos Scheel, Xu Shiqing, Morris Teubal, David Wallace and Xie Wei. Bruce Ross-Larson of Communications Development Inc. was the principal editor of the report. Meta de Coquereaumont, Paul Holtz and Alison Strong, all with Communications Development, also edited parts of the report, while Wendy Guyette and Stephanie Rostron were responsible for layout and coordination.

vi

Industrial Development Report 2002/2003

Explanatory notes
References to dollar ($) are to U.S. dollars, unless otherwise specified. Billion means 1,000 million. References to tonnes are to metric tonnes, unless otherwise specified. A slash (1990/1991) indicates a crop year or a financial year. Country classifications by income levels are from World Bank (2001b) World Development Indicators 2001. Economies are divided according to 1999 gross national income (GNI) per capita, calculated using the World Bank Atlas method. The income groups are low income, $755 of less; lower middle income, $756$2,995; upper middle income, $2,996$9,265; and high income, $9,266 or more. The designation of least developed country follows the United Nations definitions, which is based on three criteria: low income (less than $900 estimated GDP per capita, three-year average), weak human resources and economic vulnerability. The following symbols are used in tables: Two dots (..) indicate that data are not available or are not separately reported. A dash () indicates that the amount is nil or negligible. na is not applicable. Totals may not add precisely because of rounding. The UNIDO Scoreboard database on selected indicators of industrial performance and drivers draws on numerous databases, as detailed in the technical annex. The following abbreviations appear in this publication. BIPM BIS CIP CO2 COTEX CPC EDF EPZs EU FDI GATT GATS GDI GDP GNI GNP HKPC ICICI IMF ISO ITMIN ITRI JPC-SED MAC Bureau International des Poids et Mesures Bureau of Indian Standards competitive industrial performance carbon dioxide Consortium of Textile Exporters China Productivity Centre Enterprise Development Fund export processing zones European Union foreign direct investment General Agreement on Tariffs and Trade General Agreement on Trade in Services gross domestic investment gross domestic product gross national income gross national product Hong Kong Productivity Council Industrial Credit and Investment Corporation of India International Monetary Fund International Organization for Standardization Industrial Technology and Market Information Network Industrial Technology Research Institute Japan Productivity Centre for Socio-Economic Development Manufacturing Advisory Centres

Explanatory note

vii

MEP MHT MIDA MFA MVA NAFTA NAMAC NCPC NCS NGOs NIST OBM ODM OECD OEM PSB R&D SAR SDF SERCOTEC TRIMS TRIPS UNCTAD UNEP UNESCO UNIDO USIMINAS WTO

Manufacturing Extension Partnership medium and high tech Malaysian Investment Development Agency Multi-Fibre Arrangement manufacturing value added North American Free Trade Agreement National Manufacturing Advisory Centre National Cleaner Production Centres Network Computer Systems non-governmental organizations National Institute for Standards and Technology own brand manufacturing own design manufacturing Organisation for Economic Co-operation and Development original equipment manufacturing Productivity and Standards Board research and development Special Administrative Region of China (Hong Kong) Skills Development Fund Servicio de Cooperacin Tcnica trade-related investment measures Trade-Related Intellectual Property Rights United Nations Conference on Trade and Development United Nations Environment Programme United Nations Educational, Scientific and Cultural Organization United Nations Industrial Development Organization Usinas Siderrgicos de Minas Gerais SA World Trade Organization

viii

Industrial Development Report 2002/2003

Contents

Foreword Overview

iii 1

PART 1. ASSESSING WHERE COUNTRIES STAND

1. New technologies, new systems, new rules 2. Global industrial activity 27 41 57 9

3. Benchmarking industrial performance

4. Benchmarking the drivers of industrial performance

PART 2. LAYING THE FOUNDATIONS FOR INDUSTRIAL COMPETITIVENESS

5. Innovation and learning to drive industrial development 6. Innovation and learning in global value chains 7. Supporting innovation and learning by firms 8. The way forward 133 105 117 93

ANNEXES
Technical annex Statistical annex 145 149

Bibliography

181

ix

TABLES
2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.1 4.2 4.3 4.4 4.5 A4.1 5.1 6.1 6.2 A.1 A.2 A2.1 A2.2 A2.3 A2.4 A2.5 A2.6 A2.7 A2.8 A2.9 A2.10 A2.11 A2.12 A2.13 A2.14 A2.15 A2.16 A2.17 A2.18 Top 25 exporters of high-tech products, 1985 and 1998 31 Top 25 exporters of medium-tech products, 1985 and 1998 32 Top 25 exporters of low-tech products, 1985 and 1998 33 Top 25 exporters of resource-based products, 1985 and 1998 33 Ranking of economies by the competitive industrial performance index, 1985 and 1998 43 Ranking of economies by the competitive industrial performance index, by region or country group, 1985 and 1998 45 Ranking of least developed countries by the competitive industrial performance index, 1985 and 1998 46 Correlation between components of the competitive industrial performance index, 1998 47 Cluster analysis of competitive industrial performance for industrialized and selected transition economies, 1985 and 1998 49 Cluster analysis of competitive industrial performance for developing economies, 1985 and 1998 49 Leading and lagging exporters, 1998 52 Correlation between industrial performance measures and carbon dioxide emissions, 1998 53 Biggest and smallest polluters, 1998 54 Correlation between drivers of industrial performance, 1998 60 Patents taken out internationally, 1998 63 Reliance of major high-tech exporters on domestic R&D and foreign direct investment, 1985 and 1998 67 Developing economies by industrial performance and average capabilities, 1985 and 1998 73 Using the Scoreboardand going beyond it 74 Ranking of economies by the drivers of industrial performance, 1985 and 1998 75 Technological and organizational capabilities within firms 96 Characteristics of producer-driven and buyer-driven global value chains 108 Global furniture tradetop 10 net exporting countries, 1994 and 1998 113 Technological classification of exports according to SITC revision 2 145 Technological classification of manufacturing value added according to ISIC revision 2 146 Manufacturing value added by income level and region, 1985 and 1998 149 Manufactured exports by income level and region, 1985 and 1998 150 Technological structure of industrial activity by income level and region, 1985 and 1998 151 Ranking by concentration of manufacturing value added and exports in selected economies, 1985 and 1998 152 Tertiary enrolments, total and technical, by income level and region, 1987 and 19951998 153 Ranking by concentration of tertiary enrolments, total and technical, in selected economies, 1987 and 19951998 154 R&D financed by enterprises by income level and region, 1985 and 19951998 155 Ranking by concentration of R&D financed by enterprises in selected economies, 1985 and 19951998 156 Foreign direct investment inflows by income level and region, 19811985 and 19931998 157 Technology licence payments abroad by income level and region, 1985 and 1998 158 Ranking by concentration in technology licence payments abroad in selected economies, 1985 and 1998 159 Information and communication technologies infrastructure by income level and region, 1998 and 2001 160 Comparison of main industrial performance and capability indicators by income level and region, 19851998, selected years 161 Ranking by manufacturing value added, 1985 and 1998 162 Ranking by manufactured exports, 1985 and 1998 163 Ranking by technological structure of manufacturing value added, 1985 and 1998 164 Ranking by technological structure of manufactured exports, 1985 and 1998 165 Ranking by Harbison-Myers index of skills 167

Industrial Development Report 2002/2003

A2.19 A2.20 A2.21 A2.22 A2.23 A2.24 A3.1

Ranking by tertiary enrolments in technical subjects, 1985 and 1998 168 Ranking by productive enterprisefinanced research and development, 1985 and 1998 Ranking by foreign direct investment inflows, 19811985 and 19931997 171 Ranking by royalty and licence payments abroad, 1985 and 1998 173 Ranking by modern physical infrastructure, 1985 and 1998 175 Ranking by traditional physical infrastructure, 1985 and 1998 176 Ranking of economies by basic indicators of industrial performance and by composite index of competitive industrial performance, 1998 177 A3.2 Ranking of economies by basic indicators of industrial performance and by composite index of competitive industrial performance, 1985 179 A3.3 Regression results for export structure and growth in manufactured exports 180

169

FIGURES
1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 3.1 3.2 3.3 3.4 3.5 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 Research and development spending by industry, OECD countries, 1994 14 Share of medium- and high-tech products in global dynamic exports, 19801997 15 Share of top five countries in foreign direct investment receipts 19 Shares of foreign affiliates in research and development, 19961998 21 National shares of developing world manufacturing value added, 1998 29 National shares of developing world manufactured exports, 1998 29 Developing country share of world manufacturing value added by technology intensity, 1985 and 1998 31 Developing country share of world manufactured exports by technology intensity, 1985 and 1998 31 Shares of world manufactured exports of top 5 and 10 exporters by technology intensity, 1985 and 1998 31 Distribution of tertiary enrolments in developing regions, total and technical subjects, 1985 and 1998 34 Regional distribution of developing world R&D financed by productive enterprises, 1985 and 1998 35 Leading developing economies in R&D financed by productive enterprises, 1998 36 Regional distribution of foreign direct investment inflows, 19811984 and 19931997 37 Regional distribution of royalty payments, 1985 and 1998 38 Leading developing economies in royalty fees, 1998 38 Regional distribution of information and communication technologies, 19982001 39 Regional distribution of information and communication technologies per 1,000 population, 19982001 39 Changes in ranking by the competitive industrial performance index between 1985 and 1998 44 Winners and losers in competitive industrial performance rankings between 1985 and 1998 48 Cluster analysis of technological evolution of industry in industrialized and transition economies, 19851998 50 Cluster analysis of technological evolution of industry in developing economies, 19851998 51 Regression of competitive industrial performance index values on carbon dioxide emissions (log model), 1998 54 Competitive industrial performance and its drivers by region, 19811985, 1985, 19931997 and 1998 58 Cluster analysis of skills, infrastructure and R&D in developing economies, 1985 and 1998 61 Cluster analysis of skills, infrastructure and R&D in industrialized and transition economies, 1985 and 1998 62 Economies by technological effort and inventiveness index, 1998 64 Cluster analysis of industrial performance, R&D and foreign direct investment, 1985 65 Cluster analysis of industrial performance, R&D and foreign direct investment, 1998 66 Ranking of economies by R&D spending per unit of foreign direct investment, 1985 and 1998 68 Cluster analysis of R&D, foreign direct investment and high-tech exports, 1985 69 Cluster analysis of R&D, foreign direct investment and high-tech exports, 1998 70

Contents

xi

4.10 A4.1 A4.2 A4.3 A4.4 A4.5 A4.6 A4.7 A4.8 A4.9 A4.10 A4.11 A4.12 A4.13 A4.14 A4.15 A4.16 A4.17 A4.18 5.1 6.1 6.2 6.3 6.4 6.5

Competitive industrial performance index and average drivers of industrial performance in selected economies, 1998 72 Technological structure of manufacturing production and exports in selected industrialized countries Changing ranks in industrial performance indicators and drivers for Ireland 79 Changing ranks in industrial performance indicators and drivers for New Zealand 79 Technological structure of manufacturing production and exports in selected countries in Latin America and the Caribbean 80 Changing ranks in industrial performance indicators and drivers for Mexico 80 Changing ranks in industrial performance indicators and drivers for Jamaica 81 Technological structure of manufacturing production and exports in selected economies in East Asia 82 Changing ranks in industrial performance indicators and drivers for China 82 Changing ranks in industrial performance indicators and drivers for the Philippines 83 Technological structure of manufacturing production and exports in selected countries in South Asia 83 Changing ranks in industrial performance indicators and drivers for India 84 Changing ranks in industrial performance indicators and drivers for Bangladesh 85 Technological structure of manufacturing production and exports in Turkey and selected countries in the Middle East and North Africa 85 Changing ranks in industrial performance indicators and drivers for Turkey 86 Changing ranks in industrial performance indicators and drivers for Egypt 87 Technological structure of manufacturing production and exports in selected countries in Sub-Saharan Africa 87 Changing ranks in industrial performance indicators and drivers for Zimbabwe 88 Changing ranks in industrial performance indicators and drivers for the United Republic of Tanzania 88 Enterprise innovation and learning 98 Simple value chain 106 Linking local producers and global buyers 107 Leverage paths within two dimensions 108 Apparel value chain 109 Links in the wood furniture value chain 113

78

BOXES
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 2.1 3.1 4.1 4.2 5.1 6.1 6.2 Technology and innovation 10 Industry as the engine of growth 11 Innovative uses of information and communication technologies in developing countries 13 Internet access in Ghanaimpressive but expensive 14 Cooperative contracting for research and development in Germany 16 New ways of organizing and managing enterprises 17 New international rules and regulations 21 New standards and quality regulations 22 Standards and technical regulations as barriers to developing country exports 23 More stringent environmental norms and conditions 23 Stricter intellectual property rights 24 The case for strong protection of intellectual property rights 24 Manufactured products by technology intensity 30 The competitive industrial performance index 42 Highlights of the Scoreboard analysis 59 The relationship between industrial performance and its drivers: results of statistical analyses 71 Linking up with othersto start the processes of leveraging and learning 99 Jumping into the leadin global value chains 106 Pluses and minuses of being in a global value chain 107

xii

Industrial Development Report 2002/2003

6.3 6.4 6.5 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8.1 8.2 8.3 8.4 8.5 8.6

Races to the bottom 111 Linking to the leaders 111 From trust to triangles to own brand manufacturing 112 Institutional support to technological efforts of firms 118 Reforming poorly performing organizations 119 Activities involved in successful investment promotion 121 Available on the Internet 123 Programmes to help domestic firms achieve standards 124 Technology support from the Hong Kong Productivity Council 126 National cleaner production centres 127 Cluster development in Jaipur, India 129 Leveraging advanced technologies from abroad 130 Framework imperatives for effective industrialization 135 Comparative advantageto be realized 135 Broadening competitive advantage is far from automatic 136 What to promote? 138 Four Tigersfour broad visions 139 Foresight in Hungary 141

Contents

xiii

Overview

HAT IS THE STATE OF INDUSTRIAL DEVELOPMENT AROUND THE

world? Perhaps the clearest impression is that of diversity and divergence. A few developing economies have done very well in recent years in coping with the fast-changing industrial scene. But others, a disturbingly large number, have done badly. This is hardly news. It is now well known that economic performance, particularly industrial performance, is highly variable among developing economies. This is accepted as part of the hard reality of development and globalization. Early models of inevitable convergence, based on simple neoclassical growth models, have given way to more diffuse analyses stressing that structural, institutional and social factors may continue to drive economies apart. UNIDO endeavors to build on the consensus for macroeconomic stability, institutional reform and open trade and investment. It takes it as given that technological change will continue at a rapid paceand that economies will be knitted together by freer flows of information and productive factors and by the international rules of trade and investment. In this setting, the ability to compete internationally will be the basic condition for growth in the industrial sector. Relying on such static endowments as primary resources and cheap unskilled labour may be a good way to start, but it is a bad way to continue.

important, how it reflects structural factors. Those structural factors are difficult to alter in the short to medium termand often cannot be left to reverse themselves. Nor can they be expected to improve simply by exposing economies to rapid liberalization and globalization. They thus raise strong policy concerns. The international community and national governments together have to address the growing structural gaps that drive divergence. If they do not, there is a real risk of serious long-term marginalization of many countries from the dynamics of industrial development. The clear solution is to follow the high road to competitivenessto develop capabilities and increase productivity growth through concerted innovation and learning. The report also shows that successful developing economies have used widely differing strategies to build industrial capabilities and compete in world markets: building capabilities through domestic research and development (R&D), through foreign direct investment or through a combination of the two. Some, but relatively few, have succeeded by drawing in foreign technology largely at arms length while building strong technological and innovative capabilities in local firms. Others, a larger number, have gone some way by plugging into global value chains, becoming suppliers of labourintensive products and components, without having strong domestic capabilities. Of these economies, a few have managed to combine their reliance on foreign direct investment with strong industrial policy, targeting the activities they wish to enter and the functions they wish to upgrade. Others have tapped the potential of foreign direct investment by more passive policies, benefiting from sound economic management, pro-business attitudes, attractive locations and plain good luck. The less successful developing economiesand there are manyhave not managed to follow any of these strategies effectively. At first sight, the best strategy for developing economies without strong technological capabilities is to find their way into the production systems of global value chains and let local capabilities develop slowly. While recent experience of global production systems shows that this works, some cau-

Most of the effort has to come from within economies, providing the right environment for capability building and investing in the necessary factors and institutions. But such local efforts should be helped from outside. Opening markets completely in developed economies will help greatly, but much more is needed to narrow the widening gap between economies and to build industrial capabilities in developing economies. Indeed, this is the mission of UNIDObuilding and enhancing industrial capabilities. UNIDO continues to work to narrow that gap and to ensure the international communitys support with financial and other resources. It is with this purpose that UNIDO launches its first Industrial Development Report. This report shows starkly how wide the dispersion is in levels of industrial development, how much it has grown and, most

tion is called for. Latecomers entering global production systems will find it difficult to sustain growth as wages rise unless they can raise their skill, technological and institutional bases. Plugging into global value chains does not by itself ensure that participants will upgrade their capabilities. Yet such upgrading is essential. Moreover, global production systems are highly concentrated, and the concentration rises with the sophistication of the technology. With globalization and liberalization on the rise, economies must be internationally competitive to prosper and grow. Governments have reduced or are reducing restrictions on trade, international finance and foreign direct investment. Domestic liberalization is being strengthened by new international rules of the game for economic activity. Production across national boundaries is being integrated under common ownership or controloften in the hands of a small number of large private companiesmaking it even more difficult to isolate economies from world market forces. Technical change is underpinning these processes. The result is that enterprises are exposed to global competition with an immediacy and intensity rarely seen before.

economies shows East Asia leading the CIP ranking in 1998, followed by Latin America and the Caribbean, Middle East and North Africa, South Asia and Sub-Saharan Africa. The stability of the CIP ranking over time confirms that industrial performance is path-dependent and difficult to change. But there have been some leaps. Major improvements have been experienced since 1985 by middle-income developing economies (China, Costa Rica, Malaysia, Mexico, the Philippines and Thailand). Enhanced industrial performance in these economies has been triggered, to a great extent, by their insertion into global value chains through transnational corporations. Low-income economies remain at the bottom in the CIP index and the gap between least developed economies and other developing economies widened during the period 19851998. This points to growing industrial divergence within developing economies. Low-income economies have not moved up the technology ladder. Evidence suggests that 42 developing economies had a technology structure in 1998 similar to that in 1985. Only 16 developing economies (of the 58 in our sample) have shown dynamic production and export structures towards technology-intensive products. Industrial production and manufactured exports within developing economies are highly concentrated. The top 5 countries account for 60 percent of developing country industrial production and 61 percent of exports. By contrast, the bottom 30 countries account for only 2 percent and 1 percent. Most worrying, these shares declined during 19851998.

How can countries see where they stand?

The Scoreboard introduced in this report provides information on crucial aspects of industrial development and competitiveness. It has two parts: an index of a countrys ability to produce and export manufactures, and benchmarks of the structural drivers of industrial performance.

Benchmarking the "drivers" of industrial performance

Industrial performance is the outcome of many social, political and economic factors interacting in complex and dynamic ways. The purpose here is to benchmark economies on their key structural variablesreferred to here as driversusing available data. This report focuses on five proxy variables for drivers directly relevant to industry: skills, technological effort, inward foreign direct investment, royalty and technical payments abroad and modern infrastructure. Singapore led developing economies (and the world) in foreign direct investment per capita in 1998, while Hong Kong Special Administrative Region (SAR) of China was in fifth place. Other developing economies among the top 20 recipients of foreign direct investment were Malaysia and Chile; one transition economy, Hungary, also ranked among the top 20. Singapore and Hong Kong SAR ranked among the top 5 in payments for technology per capita, followed in the developing world by Malaysia, Taiwan Province of China and the

Benchmarking industrial performance

The competitive industrial performance (CIP) index measures the ability of countries to produce and export manufactures competitively. It is constructed from four indicators: manufacturing value added per capita, manufactured exports per capita, and the shares of medium- and high-tech products in manufacturing value added and in manufactured exports. The first two indicators tell about industrial capacity. The other two reflect technological complexity and industrial upgrading. A ranking of 87 economies (selected on the basis of data availability for inter-economy comparison) by the CIP index reveals a general and expected pattern: industrialized economies congregate near the top, transition economies and middle-income developing economies around the middle, low-income developing economies and least developed economies at the bottom. Looking at the regional averages for developing

Industrial Development Report 2002/2003

Republic of Korea. Singapore ranked third in physical infrastructure, with Bahrain and Hong Kong SAR also in the top 20. The ranking of economies by each driver of industrial performance shows considerable stability over time (just as the ranking by the CIP index does). Thus the ranking of economies by R&D spending per capita for 1998 is highly correlated with that for 1985, and so on. Even so, some countries changed their relative position significantly between 1985 and 1998, such as Uruguay in the skills index, Ecuador in R&D per capita and Tunisia in foreign direct investment per capita. Indigenous technological effort (proxied by enterprise R&D) appears to be one of the most important factors for improving industrial performance, in industrialized and developing countries alike. Foreign direct investment has become central to competitive performance (especially in fast-moving industries) as global production systems have grown in importance. And skills and infrastructure continue to be key drivers. But indigenous technological capabilities do not always match industrial performance. Some economies with high capabilities have "underperformed" due to a disabling regulatory environment, macroeconomic instability and other fundamental factors. Bahrain, Hong Kong SAR and Panama are among them. Similarly, economies with relatively low capabilities have "overperformed", rapidly upgrading their export structures, led by transnational corporations. They include Malaysia, Mexico, the Philippines and Thailand. Among developing economies, industrial capabilities are highly concentrated, with East Asia leading in all factors. Industrial divergence among developing economies is even more acute when looking at technological capabilities. For instance, the bottom 30 economies account for only 2 percent of developing economy foreign direct investment inflows in 1998, and their R&D expenditure, technology license payments and Internet hosts are almost negligible.

sources of technology and market access remains vital for industrial success. Tapping into global value chains, especially in knowledge-driven sectors, can be a good means to enter global markets and gain access to new technology and knowhow. Enterprises in developing countries generally start the innovation and learning process by importing new technology; they then invest in building their capabilities to master the tacit elements. How much they invest depends on the incentives thrown up by markets, mainly by the competition faced in foreign and domestic markets, as well as on the ability to assess complementary supporting activities. Enterprises draw on internal and external resourcesboth foreign and domesticto build their capabilities. The process starts with capabilities needed to master the technology for production purposes and may deepen over time into improving the technology and creating new technology. Linking, leveraging and learning capture what enterprisesand countrieshave to do to enable their technological development. Linkingconnecting with outsiders to acquire needed technologies and skills. Leveraginggoing beyond arms-length transactions to squeeze as much as possible from the new relationships with those outsiders. Learningmaking the many efforts to master process and product technologies, consciously building the foundation for improving current technologies and creating new ones. Whatever the process, enterprises have to start with their initial complement of resources, technologies, skills and capabilities. It is what they do with these elements that counts. The most important thing an enterprise can do is accelerate its acquisition of capabilities by looking overseas to obtain information, purchase machinery, acquire bits of technology, bring in consultants and so on. An important part of this can be linking up with other enterprises or institutions, locally or overseas, through formal or informal ties. Strategically, it makes a lot of difference what choice is madebut the choices are also heavily constrained by the enterprises competence and the options available to it.

Innovation and learning drive industrial development

The report demonstrates that building technological capabilities is a long, costly and risky process, which many developing economies cannot afford entirely on their own. But the emergent global setting opens up alternatives for developing economies to build up such capabilities. Although external sources can be used by developing economies to stimulate industrialization, building domestic industrial capabilities is a must if industrial growth is to be rewarding and sustainable. In todays interdependent world, connectivity to the external

Latching onto global value chains

The report shows how firms and economies can build a foundation for ongoing innovation and learning by competing in

Overview

global value chains. Spread around the world, enterprises in global value chains perform related activities to bring a product (or service) from design and product development to production, marketing and sales and to consumption, after-sales services and eventual recycling. The advantage of latching onto global value chains is that firms can seek involvement at their level of technological competence. Competing in global value chains can build foundations for industrial innovation and learning. Crucial factors for latching onto a global value chain are not only the hard facts of price, quality and punctuality but also the willingness to learn and to absorb advice from the lead enterprises. Global value chains can thus unleash enterprisesbut they can also constrain them. Particularly in manufacturing, the insertion of local activities in wider networks is an opportunity for developing countries to upgrade their capabilities. But entering global value chains does not provide an assured ride up a capability escalator. It is often a fast track to acquiring production capabilities, but moving further up the chain can lead to conflicts with existing customers. Some firms even have had their capabilities downgraded as a result of their integration in global value chains. So, it makes sense for latecomers to use all the resources they can acquire from the advanced world, in return for providing such services as lowcost manufacturing. But this requires a strategic choice to use the links for domestic development.

What principles, then, should guide the provision of the subsidized services for innovation and learning? The report considers three to be paramount. First, support institutions should be established and managed and subsidized services provided within the framework of the national strategy for industrial development. Second, as a general rule, subsidized provision of industrial services has more justification the more widely shared the specific services rendered. Third, the services and organizations should not be supplied solely by government. As quickly as is feasible, they should be supplied in public-private partnerships or by private firms and associationswith subsidies, if justified, or without, if the market can supply the services.

Formulating strategies
Developing economies can build competitive industrial capabilities in the current setting. Also clear is that building these capabilities needs extensive policy support. The success of developing economies that employed industrial development policies in export-oriented environmentswith complementary policies to build skills, technological capabilities and supporting institutions and to leverage foreign resourcesshows that such strategies can radically transform the industrial landscape in just a few decades. The report argues that the basis of any coherently framed industrial strategy is a national vision of industrial developmenta vision to get on the high road to competitiveness by increasing productivity growth through concerted innovation and learning. Foresight exercises offer a disciplined means for determining targets and the ways to achieve them in formulating industrial development strategies. The focus of these exercises in developing economies differs from that in industrialized economies in that the objective is to catch up with the global technological frontier, not to remain on, or at the forefront of, the changing frontier. Even so, developing economies require foresight in relation to existing industriesnot simply for keeping up but also for catching up to a shifting frontierand in relation to industrial activities for which potential competitive advantage is within grasp. The report makes three major points on the policy process. First, policy needs vary with the level of development. As markets and institutions become more efficient and complex, the need for direct interventions falls and the potential costs rise. Second, industrial development policy must be systemic. No

Helping firms solve problems and grasp opportunities

The report also details howand whethersupport institutions can help firms meet the information, skill, finance and other needs that are difficult to satisfy in open markets. A nurturing environment is required to foster vibrant industrial development. And ensuring access to vital services that support innovation and learning is a critical part of establishing that environment. Many of these services are supplied through the market in industrialized countries, but even these countries find it necessary to augment what is supplied through the market with subsidized services. Various considerations provide ample justification for the provision of subsidized services to support the process of innovation and learningeven more for developing countries. Many types of institutions are essential in supporting innovation and learning by firms. Infrastructure determines the cost of operation and interacting with the outside world. Training and specialized education are very important, as are financial services. The focus here is on the institutions directly supporting the innovation and learning efforts of firms.

Industrial Development Report 2002/2003

strategy can succeed unless it dovetails physical investment in capacity with technology development, skill building, cluster strengthening and so on. Third, policies must correspond to the phase of learning and so must change accordingly: policies in the infant phases of capability building must differ from those in the mature phase, when R&D and frontier innovation become vital.

erate effort to build up the widest possible base of higher order skills, capabilities and institutions, growth will slow down or grind to a halt. Countries have to build the capacities to take on, at competitive levels, more complex activities that use emerging technologies and sustain rising wages. This entails building the institutions and providing the support to create new skills, information and capabilities. These ingredients of success are hardly a secret. What is difficult is to devise and implement practical strategies to suit the specific needs of particular developing economies. The task is broad and challenging. It is also slow, difficult and detailed. It requires understanding and tackling the basics of small, incremental changes on which received theory provides little guidance. It entails constant adaptation and learning on the part of policymakers. And it has to call forth the cooperation of a range of agents, private and public, as well as new forms of governance that are difficult to introduce. In all this, countries need consciously to build their technological capabilities through concerted innovation and learning. To get the productivity gains promised by such efforts they need, in addition, to put in place the institutions to support their proactive integration into the global flows of trade, capital and technology. Again, the international community can do much to support these effortsand again, UNIDO will work to ensure that such support is forthcoming.

Enhancing capacities to sustain productivity growth

Industrial development can drive human development and national development. By attending to the drivers of industrial development countries can greatly improve their potential industrial performance. Needed are policies to ensure sustained productivity growth with an emphasis on equitable distribution. Only when countries achieve this will they turn onto the high road to development, drawing fully on world trade and investment flows and connecting their people to the global economy. Developing economies have to deliberately mobilize the key ingredients of productivity growth and spread them widely. Getting the macroeconomy right and opening up to trade and investment is only a first step. And it is one that needs to be handled carefully: many countries have rushed into it without adequate preparation. If they stop here and make no delib-

Overview

Part 1 Assessing where countries stand

1
T

New technologies, new systems, new rules

HE INDUSTRIAL SCENE IS CHANGING QUICKLYDRIVEN BY

constantly emerging, rapidly spreading new technologies that are altering the relationships between enterprises and other entities and influencing how enterprises are organized and managed. National and international rules and regulations are also changing, improving the functioning of markets. Although many of these changes offer enormous benefits to developing countries that can use them in their economic interest, countries that cannot could be marginalized and excluded. Countries at all levels of development face the same challenge: ensuring that industrial enterprises become and remain internationally competitive. Becoming internationally competitive can be much harder than it sounds. Why? Because industrial competitiveness does not result from merely opening economies to global trade, investment and technology flowsthough, if done carefully, that can be important. Nor does it mean cutting wagesa response that is, at best, a short-term defensive strategy (often termed the low road) incompatible with sustained growth. Instead, industrial competitiveness requires building capabilities in the use of new technologies (the high road). To develop those technological capabilities, countries have to acquire enterprise-specific knowledge, skills and practices through an incremental learning process. This process can be slow and difficult. Depending on the country and the technology, it can involve heavy costs and great risks and uncertainties. But if countries fail to build the capabilities to compete internationally, they can become bystanders at the technological feast, stuck with the crumbsstuck with simple manufacturing activities that do not lead to sustained, diversified growth. Capability development takes place primarily in enterprises. It is, however, strongly conditioned by the environment in which enterprises operate. Responding to market, policy and technological signals, enterprises are sensitive to macroeconomic changes, growth prospects, national security issues

and physical and intellectual property rights. They need a variety of inputs from markets, institutions and other enterprises to build and strengthen capabilities. These inputsincluding finance, skills, machines, information and technical knowledgehave to keep pace with rapid technical change and intense competition. Thus capability building requires complex interaction among actors. The policy challenge for developing countries is to foster dynamic competitiveness. The complexity of the capability building process varies by industry. It also varies by a countrys level of industrial development. With industrialized countries constantly increasing their competitiveness and strengthening technology systems, capability development is crucial in developing countries and requires strong policy support. Policy needs are even greater in the worlds least developed countries. This chapter assesses the opportunities and challenges that the new industrial scene creates for the the process of industrialization in developing countries.

Assessing technical progressits promise for growth

Recent scientific and technical advances provide a wealth of productive knowledge. Such knowledge, applied properly, can raise incomes and employment in developing countries. A rice producer in India, a goods transporter in the United Republic of Tanzania, a small fruit processor in Thailand, an automobile component manufacturer in Brazilall can benefit from new technologies. The rapid pace of technical progress shows no signs of slowing, and its reach is pervasive, affecting nearly every aspect of life. Except if directly taught, productive knowledge can move from innovator to user only if it is codified into transmittable forminto information. New technologies facilitate such codification and transmission, allowing knowledge to spread faster and cheaper than ever before.1 Moreover, it is not only

Box 1.1 Technology and innovation Opportunities

The technology structure of production and exports matters for

long-term development.
With adaptive capabilities, technology can be upgraded to increase

competitiveness in all countries and industries and at all technological levels.

Technological upgrading can be achieved by pursuing a strategy

of innovation based on linking, leveraging and learning.

Technological upgrading is facilitated by entering into high-tech

global value chains, even at the assembly level (for export-oriented operations). Challenges
Stricter intellectual property rights have raised the entry fee for

Statistics on global deprivation and inequality, though well known, are worth reiterating. About half of the worlds peoplearound 3 billionlive on less than $2 a day. Around 1.2 billion people struggle on less than $l a day (the yardstick of extreme poverty). Some 15 percent more people live in poverty in developing countries than 10 years ago; 800 million lack access to health care; and 500 million are not expected to survive to age 40. Women and children suffer the most: 10 million children under five died in 1999, mostly from preventable diseases.3 In 1960 per capita incomes in the richest 5 percent of countries were 30 times those in the poorest 5 percent. By 1997 they were 74 times as high. Inequality has also increased in the manufacturing industry, both between industrialized and developing countries and within the developing world. In 1985 per capita manufacturing value added in the most industrialized 5 percent of countries was 297 times that in the least industrialized 5 percentwhile in 2000 it was 344 times as high. The industrial leaders among developing countries did quite well. But in 1985 per capita manufacturing value added in the five leading developing countries was 276 times that in the five laggardsand in 2000 was 437 times as high. To the extent that manufacturing remains a driving force in sustained developmentand the next section argues that it doesthe growing divergence in manufacturing performance presages a similar divergence in economic performance more generally. But if the international economy is to promote political and social stability, it cannot sustain this pattern for long. The broad acceptance of global integration in a democratic framework requires that the process benefit all participantsand that the benefits be reasonably equitably distributed. This is not the case today, creating hardship and raising resistance to further reform. Unless the divergence is reversed, the promise of growth based on technical progress may remain just thata promise that marginalized people no longer believe in.

technological upgrading.
Low-tech, low-wage, resource-based industrialization is a slow-

growth strategy. Sustainable growth requires rapid increases in wages and productivity.
Large-scale investments are required in information and commu-

nication technologies, infrastructure and capabilities.

Narrowing the economic divide requires bridging the digital

divide.
Replicating East Asias rapid, technology-led growth will be diffi-

cult given the new global setting, new rules, different preconditions and new competition from China and India.

knowledge that moves more easily: so do products, money, skills, machines and other inputs into production. Thus industry remains the focal point of technical change and diffusion. Accordingly, this reports main concern is to determine why many developing countries are unable to use new industrial technologies efficiently. With falling costs of distance, the economic world is shrinking rapidly and irreversibly. Globalization, the term that describes this process, has huge technological potential to changeand improveeconomic life. But it also carries costs and risks (box 1.1). The ability to move information easily does not mean that productive technologies spread easily or that their benefits are distributed equally. On the contrary, resources tend to flow to relatively few countriesthose able to use them efficiently and profitably. Because globalization lacks inherent forces to balance such divergence, it is not always an engine of beneficial and sustainable economic integration.2 Indeed, it can also be a powerful force for deprivation, inequality, marginalization and ecological disruption.

Being competitive in manufacturingthe imperative

Is industry still important for economic development? Most analysts would say yes. Since the industrial revolution, manufacturing has been the main engine for growth and for transforming the economic structure of poor countries. It has been the catalyst for shifting them from simple, low-value activities with poor growth prospects to activities with high productivity, increasing returns and strong growth potential.4 The rapid growth of technology-driven economic activity does not change this, despite the rising share of services in income and

10

Industrial Development Report 2002/2003

the much-hyped growth of the new economy. Indeed, rapid technical progress makes industrialization even more important for developing countries (box 1.2). With globalization and liberalization on the rise, countries must be internationally competitive to survive and grow. That was not the case when industrial development started in todays industrialized countries and most newly industrializing economies. Many governments used import protection, subsidies, procurement and other measures to promote industrial enterprises and catch up with the leaders. The leaders, in turn, tried to protect their positions through measures such as prohibiting the emigration of skilled workers and even (in early nineteenth century
Box 1.2 Industry as the engine of growth Industry has long been the main source, user and diffuser of technical progress and associated skills and attitudes. No other productive activity comes close. Industrys special role can be understood only in a world of dynamic learning and technical change, where large enterprises strive to increase their size and capabilities to realize economies of scale and societies constantly transform their structures and habits. In this world the manufacturing industry is not just an ingredient of developmentit is the essential ingredient. Applying technological progress to production. Manufacturing is the main vehicle for applying technological progress to production. Agriculture also benefits from technical progress, but at a much slower pace than manufacturing. Manufacturing can apply a limitless variety of inputs and equipment. Moreover, many industrial technologies involve increasing returns to scale and offer enormous potential for further learning and incremental improvements. That is why the shift from low- to high-productivity activities always involves a shift from agriculture and traditional services to industry. In recent years information and communication services have also attracted innovative activity. But that innovation was only possible because of technological advances in the hardware of information processing and telecommunications. Driving innovation. Manufacturing is the main source of innovation. Research and development by private industrial enterprises accounts for the bulk of innovation in industrialized countries; these enterprises also finance significant research and development in universities and other laboratories. Moreover, formal research and development is only part of the technology development process. A significant portion occurs in the engineering, production, procurement, quality management and other departments of enterprises. The scope for such innovation is enormous in manufacturing, perhaps more so than in other activities. Diffusing innovation. Manufacturing is often the hub for diffusing innovation to other activities, providing capital goods and transmitting new technical and organizational knowledge. Historically, the capital goods sector served as such a hub; today the electronics industry is the hub. In particular, the use of information technologies by all activities involves the considerable spread of new technologies, accompanied by close interaction between suppliers and users. Developing new skills and attitudes. Manufacturing is a vital source of new skills and attitudes, transforming traditional economic structures. It creates an industrial work ethic, spreading the discipline and organization required in modern societies. It fosters entrepreneurial capabilities, with small enterprises as the springboard, and it develops
Sources: UNIDO; Chenery, Robinson and Syrquin (1986).

England) banning the export of machinery.5 In the early days of industrialization, high transport and communication costs also provided natural protection. In addition, different countries adopted different technical standards, and governments rarely bought goods from foreign suppliers. Finally, consumers often knew little about competing foreign products. Things are very different today. Governments have reduced or are reducing restrictions on trade, international finance and foreign direct investment (FDI). Domestic liberalization is being strengthened by new international rules of the game for economic activity. Production across national boundaries is being integrated under common ownership or controloften in the

new managerial and technological capabilities, the core of modernization and competitiveness. Leading institutional development. Manufacturing has led the development of modern institutions and legal structures such as joint stock companies, accounting standards and corporate governance norms. Producing beneficial externalities. The innovation and skills created in manufacturing provide large benefits for other activities. Agriculture gains from richer consumers, better equipment and inputs, and improved storage, transport, distribution and processing facilities. Services gain from better equipment and skills. Stimulating modern services. Manufacturing provides the direct demand that stimulates the growth of many modern services. It is often the largest customer for banking, transport, insurance, communications, advertising and utilities. It creates markets for new services and skills, particularly important for finance, education and logistics. It is also the source of new service enterprises, many of them originally part of manufacturing enterprises and hived off to provide design, logistics, maintenance, training and other services. Generating dynamic comparative advantage. Manufacturing is the main source of dynamic comparative advantage, the shift from primary to more advancedand generally more dynamic and highervaluemanufactured exports. Manufacturing now accounts for about 90 percent of global visible trade, a share that has grown steadily over time. Terms of trade for manufactures have also improved steadily. Although modern service exports are also growing, much of this growth comes from industrialized countries that have built modern skills and capabilities through manufacturing. Few countries are able to sell high-value services (excluding tourism) without first undergoing industrial development. Internationalizing economies. The internationalization of an economy often follows the spread of transnational manufacturing corporations, banks, transport providers, advertisers and so on setting up shop around the world to serve their customers. The current phase of globalization, with integrated facilities across countries, is led by manufacturing enterprises. Modernizing enterprises. The exposure to foreign markets, enterprises, skills and practices that manufacturing brings can be the catalyst for modernizing national industrial enterprises, as in the Tiger economies of East Asia. Without industrial development, such modernization would not have been possible.

New technologies, new systems, new rules

11

hands of a small number of large private companiesmaking it even more difficult to isolate countries from world market forces. Technical change is underpinning these processes. The result is that enterprises are exposed to global competition with an immediacy and intensity rarely seen before. Thus it is essential for enterprises and countries to deal with the increase in international competition. To compete internationally, enterprises not only have to be efficient, they also require a supportive economic and business environment. Governments must provide appropriate conditions: political security, good macroeconomic management, sound and enforceable legal and property rights, transparent and predictable policies, well-functioning institutions and a business environment with low transaction costs. Suppliers of physical and service inputs and infrastructure must meet international standards of cost, quality and delivery. Markets for labour, capital and information, along with their supporting institutions, must work reasonably efficiently. Enterprises must be encouraged to invest in building new capabilities, mounting competitive strategies and developing networks and clusters for achieving efficiency and dynamism. The needs of competitiveness thus stretch well beyond the front-line enterprises that face international rivals, encompassing other enterprises, activities, institutions and policies and applying to developing and industrialized countries alike. For latecomers to industrialization that lack the required capabilities, structures and institutions, globalization can pose considerable challenges. But countries that can address these challenges have enormous opportunities for growth. How well countries cope depends on their ability to link with foreign partners and leverage additional resourcesparticularly technology and knowledgefor development. But success in these areas requires investing in and facilitating learning efforts to adopt, adapt and improve on the resources acquired. Competition is constantly taking new forms. Low costs are importantbut so are innovation, flexibility, reliability, service and quality. In industrialized countries new products, processes and services are the main drivers of competitiveness. Enterprises in developing countries do not innovate in this sense and cannot rely on these mechanisms to achieve competitiveness. They compete by using imported technologies together with lower labour and other costsand, where

relevant, natural resources. Using new technologies efficiently, however, requires considerable technological and managerial effort. Mastering technologies to competitive standards requires new skills, technical information, organizational techniques and marketing and supply chain methods. The hardware of new technologies, along with blueprints and instructions, can be imported. But its efficient deployment necessarily involves local learning. This process is continuous, because technologies change constantly. Industrial development also entails a constant shift from simple to complex technologiesonly then can wages and living standards rise. This means moving both across industries (from low- to medium- and high-tech) and within industries (from low to high value-added activities).6 None of this is easy, even for countries that do not innovate at the frontier. It is not easy because achieving technical and managerial efficiency takes considerable effort. Opening the economy to world markets does not, in most developing countries, ensure that enterprises will secure the right technologies and, more important, use them at best practice levels. A user of new technologiesnew to the user, that is, rather than the worldhas to master their tacit elements to achieve best practice. In this process the user has to build new skills, collect new information, set up new systems and forge new links with other actors. This process, often requiring costly and risky learning, is in many ways similar to real innovation in industrialized countries. The content, risk, cost and duration of the effort vary by technology, industry, actor and context. Becoming competitive requires widespread technological effort, which is a constant process of innovation and learning. The efficiency of this innovation and learning determines the success or failure of industrial development. How this occurs in developing countries is the theme of this report. That the worlds industrial setting is changing is evident, but many changes are not new by historical standards. In some ways the global economy was more open a hundred years ago. There were fewer barriers to trade and investment, and there was greater certainty about security and exchange rates. Technical progress, however, has integrated the world economy much more closely today, and the interaction of several factors has created a qualitatively different setting for industrial activity. Product, service, financial and information markets are better linked, each in a state of constant ferment. The many features of the new setting can be grouped in three clusters: those driven by new technologies, those driven by new innovation, managerial and organizational systems in

12

Industrial Development Report 2002/2003

enterprises and those driven by new international rules and regulations.

Exploring and exploiting new technologies

Todays technical change is unprecedented in pace and scope. Information and communication technologies are at the core of such change, making spectacular advances. In 1930 a one-minute telephone call from New York to London cost $300 in todays prices; today it costs a few cents.7 The cost of 1 megahertz of processing power fell from $7,600 in 1970 to 17 cents in 1999a 99.9 percent decline over the same period. The cost of sending 1 trillion bits fell from $150,000 to 12 cents. The entire contents of the U.S. Library of Congress can now be transmitted across the country for $40; soon it may be stored on one computer chip. All of this is revolutionary, but the pace of innovation continues to accelerate. Thus it is not surprising that there is so much interest in knowledge societies and the weightless economy.8 Information and communication technologies, the most visible face of technical progress, also affect the pace of innovation. Today information can be processed at rates unthinkable just 10 or 20 years ago.9 But these technologies are also important in mundane low-tech activities, often opening unexpected opportunities to entrepreneurs in developing countries (box 1.3). Information and communication technologies can also significantly boost economic performance, though there is much debate about their precise effect on recent economic growth.10 Developing economies with fast growth in consumption of information and communication technologies
Box 1.3 Innovative uses of information and communication technologies in developing countries In rural southern Ghana, petrol stations place orders by telephone a task that entailed travelling to Accra. In Zimbabwe a company generated $15 million in new business by advertising on the Internet. In the mountains of Lao Peoples Democratic Republic and Myanmar, drivers of yak caravans use mobile telephones to call ahead and find the best route to deliver their goods to market. Fishers off the shores of Kerala, India, make phone calls from 7 kilometres out to sea to determine which market is paying the most for their catch, then sail there. Farmers in some remote Indian villages get weather and price information on the Internet. And in Bangalore, India, a non-profit trust is promoting a $200 computing device to provide rural dwellers with the same information.
Source: Baxter, Perkin and Mulligan (2001, background paper).

India, the Republic of Korea, Taiwan Province of China and Thailandalso appear to have the fastest growth in productivity and gross domestic product (GDP). In Canada, the United Kingdom and the United States between 1990 and 1996, about half the growth contributed by fixed investment is estimated to have come from information and communication technologies.11 Accelerating the adoption of these technologies are their falling prices, which are dropping faster than those for other capital goods. Information and communication technologies can change and improveinnovation by integrating diverse production systems and formerly unrelated technologies.12 They can also change the geography of industrial activity, bringing together locations once separated by high communication and transport costs. In addition, they can create new opportunities for learning in developing countries, using electronic links to access global knowledge on an unprecedented scale. Distance learning, if properly organized, can be quite successful.13 It is partly in response to these possibilities that many governments are opening their economies to international flows of products, knowledge and resources. Quite apart from the massive increase in the use of information and communication technologies, the information content of industrial activities is rising rapidly. About half of the value of a new car lies in its information content design, process management, marketing, sales and so on. In industrialized countries the weightless part of economic activity seems set to dominate life,14 but it is also going to play a larger role in industrial activity in developing countries.15 Can information and communication technologies facilitate leapfrogging and catching up by developing countries? Can latecomers without industrial bases jump to the forefront without going through traditional industrialization? The Internet became economically useful to enterprises in industrialized countries only around 1997, and its potential is just beginning to be exploited. Countries with no background in information and communication technologies and without a large traditional industrial base can use the Internet to promote growth and employment. Developing countries lack of old computer systems is an advantage. New technologies that do not require fixed communication networks may enable developing countries to leapfrog stages of technological development. In Africa satellites and new wireless technologies may make it possible to bypass fixed telephone landlines (box 1.4). Moreover, as computing reaches mobile telephony, millions of users in Africa may come online. Still, the evidence so far does not offer great hopes for leapfrogging. As noted, most countries that have succeeded

New technologies, new systems, new rules

13

Box 1.4 Internet access in Ghanaimpressive but expensive Africa boasts many technology success stories and centres of excellence. One of them is Ghana, which aspires to be an Internet and rapid communications hub for West Africa. Although there is no one size fits all solution for the best adoption of these technologies, Ghana suggests basic principles for all developing countries. Ghana was the fourth Sub-Saharan country to go online, after South Africa, Botswana and Zambia. Accras only full-service Internet access provider, Network Computer Systems (NCS), offers a gateway to global communications. NCS pioneered Internet access in Ghana in late 1994, before many users in Europe had even heard of the technology. NCS subscribers are a cosmopolitan blend of embassies, chief executives, non-governmental organizations, companies and ministries. Ghanas government, which began promoting the adoption of the technology in 1995, deserves some of the credit for Accras Internet preeminence. Ghanas Internet structure and capacity are ahead of those in the 14 French-speaking countries of West Africa, where electronic networking consists primarily of email, bulletin boards, database access, news feeds and small file transfers. Ghanas true Internet connectivity offers much more, including instant access to messages, browsing through hypertext links, access to newsgroups on thousands of subjects and even video transfers. Costs continue to be an issue, however. NCS charges an annual registration fee of $100 and a monthly use fee of $100. But with the average Ghanaian journalist earning less than $150 a month, the cost of a laptop computer is equal to a years salary. So, while Internet technology appears promising and tantalizing, it is unaffordable for all but the richest people in Ghana. To broaden access, cyber cafs have mushroomed all over Accra, and in 2001 the number of Internet users doubled to about 100,000. Similar growth is expected over the next two or three years. But Ghana faces hurdles to developing a thriving online economy. Although there is high demand for basic services, headier ambitions have been thwarted by Ghanas economic crisis. In addition, dreams of ecommerce and international online trading have not been realized.
Source: Baxter, Perkin and Mulligan (2001, background paper).

countries the critical mass of information and communication technologies and the necessary skills and organizational and managerial capabilitiesthe main determinants of productivity gainsmay not exist for some time.

Using new systems for innovation, management and organization

New innovation systems are widening differences between enterprises, industries, countries and regions. At the same time, new managerial and organizational systems are changing relationships between and within enterprises.

Main actors in innovation

With rapid technical change, the growing divergence in technological opportunities between activities leads to rising differences in innovative activities between industries. One good indicator of innovation is spending on research and development (R&D).18 In 1994 high-tech industries in the countries that form the Organisation for Economic Co-operation and Development (OECD) spent nearly 25 times more of their value added on R&D than did low-tech industries (figure 1.1). In industrialized countries high-tech and medium-high-tech industries account for nearly three-quarters of business R&D. Innovative activities introduce new products, create new demand and substitute for old products more rapidly than do stable activities. As a result R&D-intensive output and exports grow faster than other industrial activities.19 Between 1980 and 1997 medium- and high-tech exports grew faster than other manufactured exports. These complex products now
Figure 1.1 Research and development spending by industry, OECD countries, 1994 Percentage of value added 30

with information and communication technologies (in both hardware and software) have been relatively industrialized. Effective use of these technologies requires massive investment in infrastructure and, more important, in new skills and capabilitiesinvestment that is beyond the means of most developing countries.16 Moreover, industrialized countries show that considerable time can pass before the benefits of information and communication technologies are realized. A critical mass of technology diffusionin coverage, organizational adaptation and learningis needed for widespread productivity gains.17 Organizational and managerial changes at the industrial level, redesigning processes and developing new business cultures, are also needed. Productivity gains often arise not directly from technologies but from the higher productivity of new systems, procedures, skills and attitudes. In many developing

20

10

High tech

Mediumhigh tech

Mediumlow tech

Low tech

Source: Baxter, Perkin and Mulligan (2001, background paper).

14

Industrial Development Report 2002/2003

account for two-thirds of the worlds dynamic exports, gaining ground on low-tech and resource-based activities (figure 1.2). The most high-tech productsadvanced electronics, aerospace, precision instruments, pharmaceuticalshave grown much faster than all other groups.20 The five fastest-growing products in world trade during 19801997 were high-tech information and communication technologies, driven by a flood of new products and their growing application in other activities. But the high-tech industry is highly cyclical. Like many investment goods, it takes the lead in both downward and upward business cycles. And with the world economy slowing, it is heading downwards. Enterprises have always been the main investors in new technologies, particularly in industrialized countries. In OECD countries enterprises conducted 69 percent of total R&D in 1997, up from 66 percent in 1981. The share of higher education institutes remained constant at 17 percent, while that of government fell from 15 percent to 11 percent. Private non-profit institutions account for the rest.21 Among enterprises, manufacturing remains the main source of R&D. But the share of services, driven by software, is risingaccounting for 15 percent of the OECD total in 1997.22 Distinctions between manufacturing and services are somewhat arbitrary, however, as the lines between them blur and industrial enterprises contract functions to independent enterprises. To cope with global competition and the growing complexity of knowledge, enterprises are specializing in their core competencies. As a result large enterprises no longer develop all their innovation in-house, but increasingly procure it from other enterprises. Several channels, discussed later in the report, provide access to the required knowledge. Innovation
Figure 1.2 Share of medium- and high-tech products in global dynamic exports, 19801997 Percent 70

surveys suggest that inter-enterprise collaboration is the most important.23 Enterprises share innovation in two ways. The first is with enterprises in the same value chain, such as for automobiles. Major manufacturers work with first-tier suppliers in developing new models, expecting them to design and develop new components and sub-assemblies.24 This process facilitates faster, riskier and more expensive innovation. It also raises the technological distance between first-tier suppliers, generally with strong R&D capabilities, and suppliers that lack such capabilities. This can have implications for enterprises in developing countries supplying (or hoping to supply) global value chains. The increasing use of information and communication technologies for business-to-business relations makes it easier for such enterprises to plug into supply chains. But the tightening technological links between lead enterprises and first-tier suppliers threaten to exclude them from the upper echelons of the supply hierarchy. The second way enterprises share innovation is between competitors in and across countries. This trend is driven by the rising costs and risks of innovation (particularly in the basic, pre-commercial stages), which lead to more frequent use of strategic alliances and research consortiums. Some 5,100 strategic alliances were formed between 1990 and 1998mainly by enterprises from the United States, which are part of 80 percent of known agreements (half with a partner from outside the United States). Enterprises in Europe participated in 42 percent of the alliances, and those in Japan in 15 percent, along with some enterprises from elsewhere.25 Governments that would otherwise oppose such collaboration on antitrust grounds now often permit or support it, even when they maintain stringent antitrust measures.26 Faster technical change, growing industrial links with science, multiple nodes of innovation and falling costs of transmitting information raise the significance of innovation networks.27 Such networks are spreading over wide areas. Geographic agglomeration remains important for some technologies and types of interaction that require direct contactSilicon Valley is an excellent examplebut it is becoming less so for others. Some networks spread across cities, others over regions, still others around the globe. Plugging into the relevant network, or concentric series of networks, is critical to competitive technology development.28 Given the risks and economies of scale, R&D tends to be concentrated at the enterprise level. In the United States the top 100 enterprises, in terms of turnover, accounted for nearly two-thirds of R&D in 1995, and in 1997 the top 20 accounted for one-third. Of the 35,000 enterprises doing R&D, just 1

65

60

55 1980

1985

1995

1997

Source: UNIDO calculations based on Comtrade database.

New technologies, new systems, new rules

15

percent performed nearly 70 percent of the total.29 Not surprisingly, R&D is even more concentrated in small industrialized countries. In Switzerland three enterprises accounted for 81 percent of R&D in the 1980s, and in the Netherlands four accounted for nearly 70 percent. But technical change is reducing concentration.30 In addition, the list of leading enterprises for R&D is changing rapidly: in the United States some 40 percent of the top performers in 1994 were not on the list 10 years before. Interactions between industrial innovators and external agentssuch as research laboratories and universitiesare also changing. New technologies need closer links to basic science even for commercial innovation, with biotechnology as a good example. Technology clusters near knowledge centres with significant concentrations of universities and research centres are an important competitive advantage. As a result industrial enterprises are spending more to sponsor R&D in such centres and tap their expertise. In many industries, relations with technology institutions such as standards and metrology bodies and support agencies for small and medium-size enterprisesalso become important in enterprise technological activity. Most industrialized countries have an array of institutions providing specialized technical inputs to industry. Given the public goods that they produce, many depend on government subsidies, as with Germanys Fraunhofer institutes (box 1.5). Industry associations, export agencies and the like can also provide support and technical assistance. Together these institutions create an environment rich in various kinds of informationcrucial for fostering sustained growth in industrial innovation and learning (chapter 8).

Box 1.5 Cooperative contracting for research and development in Germany Many countries recognize the need for research centres that can conduct technological work for industry and combine it with publicly funded long-term research. One of the best models, Germanys Fraunhofer-Gesellschaft, was created in 1949 and has become Europes leading organization for institutes of applied research. It performs research for industry, services enterprises and the government, providing rapid, economical and practical solutions to technical and organizational problems. In addition, within the framework of technology programmes in the European Union, it participates in industrial consortiums to make industry in Europe more competitive. Fraunhofer-Gesellschafts 56 specialized institutesfunded by Germanys central government, regional governments and private industryhelp develop new technologies for industry and other uses. It has nearly 11,000 staff members (mainly scientists and engineers) and an annual budget of about 900 million euros (just under $1 billion), more than 80 percent of which comes from contract research. About two-thirds of contract revenue comes from industrial and publicly financed research projects and one-third from the federal and Lnder governments. Small and medium-size enterprises account for a large portion of research contracts: in 2000, enterprises with fewer than 100 employees provided nearly 25 percent of Fraunhofers budget and those with fewer than 500 employees about 45 percent. Research for the government is aimed at longer-term social and economic problems, such as the environment. Fraunhofer scientists specialize in a broad range of complex research. When needed, several institutes pool their expertise to develop system solutions. Researchers move easily between science and industry, and more than half of the institutes are headed by academics. Fraunhofer institutes can handle clusters of technologies that universities cannot, and their practical orientation makes them valuable to clients. Companies of all sizes and types use the institutes as high-tech laboratories for development work, special services and organizational and strategic issues. In addition, the Fraunhofer institutes increasingly collaborate with affiliate institutes in Asia, Europe and the United States.
Source: http://www.fraunhofer.de/english/index.html.

New ways of organizing and managing enterprises

New technologies affect relationships between enterprisesthat is, industrial structure and organization. They also affect relationships within enterprisesthe way they are managed. Both developments influence the organization and management of global production systems and encourage global enterprises to cut inventory costs and ensure reliable procurement and delivery (box 1.6). The justin-time system is the best-known manifestation of this, but there are others.

These new organizational systems are not easy to set up and manage, particularly in developing countries. The systems require advanced infrastructure, new contracting mechanisms, greater trust and openness, and new skills and management techniques.32 To manage supply chains effectively, many large enterprises in OECD countries have had to broaden their managerial and technological competence. Information flows, logistics and networking are the new weapons in the competitive armoury, with large potential benefits from lower costs and increased flexibility. In many developing countries, however, policy and business cultures are not conducive to these changes. For two reasons, electronic commerce technologies offer faster, more efficient and potentially more cost-effective ways of connecting enterprises.33 First, these technologies are cheaper and easier to automate in ubiquitous processes such as distribution, sales, after-sales service and inventory management. Electronic data interchange is especially

CHANGES BETWEEN ENTERPRISES

Pressures for specialization, increasing the reliance on suppliers, mean that manufacturers have to manage supply and value chains.31

16

Industrial Development Report 2002/2003

Box 1.6 New ways of organizing and managing enterprises Opportunities

Clustering, networking and specialization increase efficiency and

productivity.
New managerial methods and production techniques also enhance

Smaller inventories. In the United States the average value of inventories is 2.3 percent of annual (non-farm) sales and 4.2 percent of final goods sales. Each stage of the value chain holds significant inventories: 37 percent by manufacturers, 25 percent by wholesalers and 27 percent by retailers. E-commerce can also significantly reduce costs on inventories held. Better forecasts of consumer demand. E-commerce enables more accurate forecasts of consumer demand and increased customization of orders. Collaborative forecasting is expected to cut inventory levels in the United States by 2530 percent, or $250300 billion (OECD 2000a, p. 48). Another change is the growing importance in several industries of the geographic clustering of enterprises, particularly small and medium-size enterprises.34 The benefits of agglomeration arise from external economies such as the availability of information or proximity to pools of suppliers, customers and skilled workers. Clusters are more advanced than passive agglomerations, where enterprises realize external economies just by being there. Combining networking, specialization and joint action,35 clusters could overcome many of the disadvantages associated with small size. Many high-tech clusters have emerged in industrialized countries, inspiring much analysis and policy.36 Many active, competitive clusters also exist in developing countries.37 But their technological dynamism is often limited,38 posing severe challenges in the emerging competitive setting. Such clusters need to shift from realizing largely static external economies to building dynamic capabilities based on new technologies, skills and networks.

efficiency and productivity.

Information and communication technologies provide access to

new knowledge on management methods, production techniques, marketing and export opportunities (e-commerce). Challenges
Increased competition at all levels in both export and domestic

markets due to trade liberalization.

New skills and capabilities required to master information

technologyespecially for new design, production and marketing systems.

suited to supply chain management but may be replaced by the Internet for small suppliers. Second, e-commerce technologies can be applied all along the value chain in an integrated mannersomething not possible with earlier technologies. Efficiency gains through e-commerce applications include: Lower sales costs. In the past, errors forced large enterprises to rework about a quarter of their orders. E-commerce now allows enterprises to check that orders are internally consistent and that orders, receipts and invoices match. General Electrics Trading Post Network, for example, significantly reduced ordering errors. It also cut material costs by 520 percent because competition increased among suppliers and the length of the procurement cycle was cut in half. Cheaper customer support. Cisco Systems, the worlds largest supplier of routers for Internet traffic, has moved 70 percent of its customer support onlineeliminating 250,000 telephone calls a month and saving more than $500 million, about 17 percent of its operating costs. Cheaper, faster procurement. Typical procurement orders cost $80125 to process for low-value requisitions and much more for complex orders (in some cases exceeding the value of the purchase). The use of electronic data interchange can cut these costs by 1050 percent. MCI has cut its personal computer purchasing cycle from 46 weeks to 24 hours, while Bell South has shortened the time required to approve an expense account from 3 weeks to 2 days.

CHANGES WITHIN ENTERPRISES

Enterprises are experiencing important changes in internal management and organization. The need to facilitate information flows is causing enterprises not only to introduce information and communication technologies but also to cut management hierarchies and build new tools to handle informationcalling for new skills throughout.39 On the shop floor the use of new technologies requires new skillsand more continuous training, multiple skills, work teams and the close involvement of workers in quality and productivity improvements.40 Information technology is now pervasive in work methods, plant layouts and process control, quality management, continuous improvement, lean production and just-in-time inventory systems. Other information technology applications include computer-aided design, manufacturing and engineering, manufacturing and enterprise resource planning, product data management, automation, robotics

New technologies, new systems, new rules

17

and flexible manufacturing systems. Information and communication technologies are also being used to automate design, manufacturing and coordinationchanging and improving the innovation process.41 None of this is easy, even in industrialized countries with sophisticated enterprises, ample skills and strong support institutions. The need for new systems and increased interaction with external agents is disruptive to the internal organization of enterprises.42 But enterprises that master the new culture and technologies find it easier to manage operations over long distances. Information and communication technologies also make it feasible for enterprises to separate functions and processeslocating them, almost regardless of distance, where cost, efficiency and market needs dictate. By better managing global networks and spreading activities around the world, enterprises can minimize costs and optimize flexibility and logistics. These possibilities also apply to other activities in the value chainservices, marketing, R&Dthat are also relocating within tightly coordinated international systems. Of all the activities in the value chain, R&D is the slowest to shift, but here too there are signs of change. For several reasons, these activities are not relocated evenly across countries. For example, some activities have to be concentrated in a few sites to reap the benefits of scale economies, agglomeration economies, skill and supplier availability and logistics possibilities. Others can be spread more widely because there are fewer scale or cluster economies or because of the need to be near material inputs or final customers. Other reasons for choosing certain locations may be strategic, including the locations of competitors, need to spread risk, access to innovative work and benefits of firstmover advantages. Countries that insert themselves into the global value chain early can develop skill, technological, supply and infrastructure advantages that build up over time. Moreover, the success of a few sourcing activities can attract other transnational corporations, as direct suppliers or as followers, looking for locations with good images and reputations. In addition, several traditional factors make certain locations more attractive for foreign direct investmentpolitical and macroeconomic stability, welcoming policies and so on. Low wages for unskilled workers increasingly count for less in all but the simplest low-tech activities.

industrial activityfor final products and for inputs such as raw materials, intermediate goods, machinery, finance, technology and, in many cases, high-level skills. It has many manifestations: increased trade, investment, licensing, joint ventures, alliances, networks and subcontracting activities. In most the lead players are transnational corporations from industrialized countries, the main drivers of technical change and the most important agents for transferring technologies and production across the world. But enterprises from newly industrializing economies are also enthusiastic participants. The international role of transnational corporations has been rising steadily, with growing shares of global production, trade, technology transfer and investment. In manufacturing perhaps the most visible manifestation of their activity is the rise of global industrial value chains, linking the entire sequence of activitiesraw material extraction, production, design, R&D, marketing and delivery. Of course, many industrial value chains have long been global in the sense that their materials, components or products have been traded across national boundaries. But some distinct organizational features of emerging global value chains are worth noting: Value chains are organized internationally under the common governance of private enterprises. These enterprises may hold an equity stake in activities in different countries, thus becoming transnational. Or they may have other market or non-market links with local enterprises (through subcontracting, joint ventures, strategic alliances or buying arrangements). Where economies of scale in innovation, production, logistics and marketing are important, the number of key players tends to fall over time. With policy liberalization, the key players rationalize production facilities across countries, often reinforcing their central role. The organization of the global value chain and the strategies of the leading players can affect the entry, upgrading and dynamism of the constituent units. The role of transnational corporations in global value chains (of ownership stakes in activities overseas) is rising, though there are significant differences by industry. In low-tech activities, where it is relatively easy for local enterprises to achieve best practice, arrangements tend to be loose and diverse. Some transnational corporations set up affiliates; others contract local enterprises. Independent buyers often control significant segments of the market, contracting local enterprises and providing specifications, technical assistance and inputs. In high-tech activities, by contrast, links tend to be much tighter because of the need for close coordination, rigorous quality and training needs and the desire to keep valuable technologies within the enterprise. In some of these industries, market leaders are taking specialization to its logical conclusion by

GLOBALIZED PRODUCTION AND NEW GLOBAL ENTERPRISES

Globalization means different things to different people. In this report it signifies the tighter links between all markets affecting

18

Industrial Development Report 2002/2003

renouncing manufacturing altogether. They confine themselves to R&D, design, marketing and after-sales service, letting contract manufacturers handle the entire production process.43 Even in high-tech industries the production systems of transnational corporations are not closed: there is a growing tendency to outsource functions and inputs to capable suppliers. Thus transnational value chains can encompass local enterprises in host countries, with the spread and nature of the links depending on the technologies used, the capabilities of local enterprises and the strategy of the competitors. Industrial activities are being disintegrated across countries by function and stage of production, while remaining tightly linked to ensure the efficiency of the process. Thus an enterprise may design a semiconductor in the United States with an affiliate in India, buy the wafer from a foundry in Taiwan Province of China, assemble and test the chip in the Philippines and use an independent logistics company to ship it to Germany and market it all over Europe. Accounting may be in one country and backoffice functions in another. These divisions, taking advantage of small differences in cost, logistics, skills and efficiency, are made feasible by new communication and management techniques. Different stages of the chain have different levels of value added and technology and so impose different capability needs for participants. Those at the bottom of the chain, with the simplest requirements, are most vulnerable to the erosion of competitive advantages (if the location offers primarily cheap semiskilled labour, it will tend to lose as wages rise). Thus there is constant pressure to upgrade products and processes within value chains, whether facilities are foreign owned or not. That capital and technology are mobile does not mean that local capabilities cease to matter. If anything, they matter more because other factors are so mobile and require strong immobile factors to attract them to particular sites.44 The factors that matter to investors using new technologies and looking for competitive locations are specialized skills, modern infrastructure, strong institutions, low transaction costs, efficient local suppliers, clusters of enterprises and providers of business support services. Thus the spread of transnational corporations promises much to developing countries in investment, technology, skills and market access. But flows of FDI are highly uneven and concentrated. The share of the top five recipients of FDI has declined for the world but risen for developing countries (figure 1.3). Part of this unevenness is due to political, social and policy factors that may deter investment. Part, however, is due to structural economic

factors that lead transnational corporations to concentrate in countries. These structural factors affect FDI location as political and other framework factors converge. Clusters again emerge as an important factor in attracting transnational corporations in activities where complementary factors and capabilities agglomerate. They are particularly important in knowledgeand skill-intensive activities, where the proximity of specialized suppliers, consultants and research and teaching institutes can be critical to competitive dynamism. The evidence suggests that FDI location is increasingly based on such localized factors rather than on general factors of the host country.45 Governments seeking to tap FDI for industrial development have to pay attention to this new reality. Transnational corporations look for efficient complementary factors in making their location decisions, but they also invest in raising the quality of local factors once they have invested. They train employees in new skills, help develop local suppliers, interact with and improve local institutions and so on. In 1989 HewlettPackard, one of the worlds leading electronics companies, started operations in Bangalore, India, with about 10 people, basically to sell hardware. Still growing, it now employs more than 1,000 engineers. Apart from its sales arm, it has two large software development and R&D operations, one in Bangalore and another in Chennai. The second centre collaborates intensively with the locally owned Tata Consultancy Services. Hewlett-Packard has forged strong links with other local enterprisesincluding 25 small and medium-size enterprisesand local research institutions. Its Bangalore affiliate interacts closely with the Indian Institute of Science and funds research in universities around the city. It also helps colleges in the locality develop courses and train teachers. Its
Figure 1.3 Share of top five countries in foreign direct investment receipts 19811985 19931997 Percent 70 60 50 40 30 20 10 0 World Developing

Sources: World Bank (2000); UNCTAD (1995, 1999) and national statistics. Note: Annual averages calculated for available data for 19811985 and 19931997.

New technologies, new systems, new rules

19

engineering employees, who receive six months of rigorous inhouse training, are encouraged to take out patents on their research (some 60 have been granted). Many travel regularly to Israel and the United States, where the enterprise also has R&D centres. There is a minimum of capabilities in host countries below which it is not economical for transnational corporations to locate facilities or invest in further upgrading. The more advanced the technologies and functions being deployed, the higher the local capabilities required. It is up to the host country to ensure that it reaches the critical level. Moreover, it has to ensure that as wages and other costs rise, the quality of local factors improves to attract more complex technologies and functionssuch as design and R&D. In other words, successful participation in the systems of transnational corporations requires constant efforts to build and improve local capabilities. The spread of global chains intensifies this need as more countries compete for high-value FDI. Several features of recent FDI are of direct concern to industrial development: Fast growth. FDI flows are growing faster than other economic aggregates such as GDP, world exports and national gross fixed capital formation. As a result the share of international productionthat under the control of transnational corporations and their affiliatesin global production is steadily increasing. If production by independent enterprises linked to transnational corporations is added, the share is rising even faster. World trade dominance. Transnational corporations dominate the worlds visible trade, handling about two-thirds. This share is growing rapidly in activities with significant scale economies in innovation, production and marketing. These are the high-value end of the manufacturing spectrum, and countries that want to enter these dynamic segments increasingly have to rely on transnational corporations. Global production systems. Of the visible trade in the hands of transnational corporations, about one-third is within corporate systemsbetween different parts of the same enterprises. Important parts of such internalized trade are integrated international production systems, where transnational corporations allocate different functions or stages of production to different countries. In several high-tech activities (semiconductors, hard-disk drives) the bulk of world trade is within such systems. Beyond production. Transnational corporations are also placing accounting, engineering and marketing in

affiliatesoften high-value activities that boost local competitiveness and capabilities. Even research and development. Though one of the least mobile functions internationally, R&D is also being transferred overseas. Many transnational corporations, particularly those from small countries, have long conducted R&D abroad. For instance, more than half the patents filed by transnational corporations from Belgium, the Netherlands, Switzerland and the United Kingdom originate in their affiliates.46 In many host countries foreign affiliates account for large parts of enterprise R&D. More than half of industrial R&D in Ireland, Malaysia and Singapore occurs in affiliates of transnational corporations (figure 1.4).47 Even so, developing countries still account for a small share of overseas R&D by transnational corporations. Developing countries account for less than 10 percent of R&D for transnational corporations in the United States (UNCTAD 1999). The pattern is probably similar for other industrialized countries. This is not surprising: R&D is highly skill-, scale- and linkage-intensive, and most developing countries lack the necessary capabilities.48 Innovation dominance. Innovation is dominated by large transnational corporations. Many are unwilling to part with valuable technologies without a substantial equity stakemaking FDI the most important, and often the only, source of advanced technologies. Exports. Transnational corporations are often central to local exports of technology-intensive products. Many such products are difficult to export independently because of the advanced technologies involved and the need for expensive branding, distribution and after-sales servicing. About two-thirds of consumer electronic exports from the Republic of Korea and Taiwan Province of China are original equipment manufacture.49 Transnational corporations are also active in exports of low-tech products, where market information, branding, distribution and design are important. Preferences for entry by mergers and acquisitions. Crossborder mergers and acquisitions are the preferred mode of entry for transnational corporations, particularly in industrialized countries.50 In 2001 the recession and falling share prices slowed mergers and acquisitions, cutting FDI in industrialized countries by about 40 percent (UNCTAD estimate). The decline is less marked in developing countries but still likely to cause some fall in FDI. Even services. FDI in services is rising rapidly as formerly homebound providers (as in utilities) privatize and globalize.

20

Industrial Development Report 2002/2003

Figure 1.4 Shares of foreign affiliates in research and development, 19961998 Percent 80 70 60 50 40 30 20 10 0
na d Un a Ki ite ng d do m Cz Re ec pu h bl ic nd en an Un y i St te at d es or e nd an nd sia ce Tu rk e nt in Ar ge ai Ita ed Fr an Ja p Sp er m Ire ap Sw Fin Ca ng al M Th ai Ch
21

la

la

Source: OECD (1999b) and national sources.

Telecommunications, power and water enterprises are good examples.

Heeding new international rules and regulations

Rules for international economic activity are changing, allowing it to respond as much as possible to market signals. Opaque rules and differences in trade and investment barriers impede the flow of products, capital, technology, information and skills across countries. New rules are designed to minimize costs and barriersand to lead to more uniform national policies (box 1.7). The best-known rules are those negotiated multilaterally under the General Agreement on Tariffs and Trade (GATT), now administered by the World Trade Organization (WTO). The WTO administers and embodies three main agreements:
Box 1.7 New international rules and regulations Opportunities
By aiming to level the playing field, new rules encourage enter-

the updated version of the 1994 GATT, the General Agreement on Trade in Services (GATS) and the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS). (The WTO also administers four plurilateral agreementson government procurement and on trade in civil aircraft, dairy products and bovine meatnot conditional on WTO membership.) The main agreements are complemented by multilateral agreements on safeguards, antidumping, subsidies, state trading enterprises and balance of payments measures. WTO agreements also include rules on the treatment of goods when they enter importing countries, including customs valuation, technical barriers to trade and import licensing. These agreements are intended to prevent the use of these measures for protectionist purposes. GATT was a provisional agreement among contracting parties and was not a legal institution. In contrast, WTO agreements are ratified by member countries and are permanent, with a sound legal basis. The three main agreementsthe 1994 GATT, GATS and TRIPSform the WTOs institutional structure and are subject to a single set of rules and a single system for resolving disputes. Unlike with GATT, WTO members automatically commit to all WTO agreements, with only a few minor exceptions. Other, less formal rules on trade, FDI and financial liberalization have been issued by the World Bank, International Monetary Fund (IMF) and aid donors. There are also international conventions on minimum labour standards. Several rules result from standards set internationally (for example, the International Organization for Standardization, or ISO) or by dominant regions or countries (such as the European Union or the United States). In addition, some rules are negotiated in regional trade agreements or bilaterally. (Most FDI rules are bilateral.)

prises to spread their operations across the globe and domestic competitors to improve their capabilities.
More uniform rules and regulations facilitate the globalization of

industry. Challenge
Eliminating policies that foster learning by infant industries hinders

the development of new technological capabilities.

New technologies, new systems, new rules

Si

ay

la

ile

ly

The new rules offer benefits but also impose costs. They reduce the scope for intervention in trade and investment important because in many developing countries such interventions have been costly and inefficient. Increased reliance on markets can improve resource allocation and stimulate efficiency and dynamism. By reducing the risk, uncertainty and transaction costs associated with international transactions, the new rules may also raise the quantity and quality of FDI in developing countries. In addition, by strengthening intellectual property rights, the new rules may stimulate innovation and facilitate technology transfer. The costs result from liberalizing when markets and supporting institutions are deficient, as they often are in developing countries.51 The judicious use of infant industry protection, local content rules, FDI restrictions and lax intellectual property rights has yielded spectacular benefits in East Asia.52 Strong intellectual property rights can raise the cost of products and technologies and restrict a valuable avenue for local learning without promoting innovation. Rapid liberalization can impose additional costs, giving an economy too little time to prepare for full market competition. Without the capabilities to attract and use technologies and resources productively, and facing the full forces of competition, poor countries may not draw enough of either. Instead they may lose part of the productive structure they have built up. By renouncing tools that foster learning, they may retard the development of new capabilities. The net balance of benefits and costs remains unclearparticularly because it can vary by country and period. The underlying issue is whether the costs of market failure exceed those of government failure, and if the balance can be changed (an issue not explored here).

Box 1.8 New standards and quality regulations Opportunities

Standards can facilitate technology transfer based on codified best

international practices.
Standards can facilitate international market access because they

are becoming increasingly important for global buyers and as criteria for awarding contracts. Challenges
Standards can substantially increase the costs of entering interna-

tional markets.
Skills and capabilities must be substantially upgraded to meet the

new standards, master new technologies and establish the required institutional information.

reduce technical transaction costs, information asymmetries and uncertainties between sellers and buyers, possibly enabling them to foster innovation. But standards can also impose costs on developing countries, forcing them to upgrade skills and capabilities, master new techniques and establish an institutional infrastructure (accreditation, metrology, standardization and technical support and information). If these costs are very high for a country (relative to its economy and exports), standards can pose a barrier to exporting (box 1.9).

More stringent environmental norms and regulations

The growing emphasis on environmental and social norms such as child labourcan affect industrial and export development in developing countries. Pressures from consumer groups, non-governmental organizations (NGOs) and other bodies have led buyers to impose higher environmental standards on suppliers from developing countries, imposing compliance costs on enterprises. But compliance can also benefit society and the competitive positions of enterprises in developing countries (box 1.10).53 Moreover, the private costs of compliance have not been very high, at least in industrialized countries. In the United States, for example, they are estimated at 0.6 percent of production value. Although there are no data on the costs of environmental compliance for industry in developing countries, case studies suggest that the costs would be similar or even less. The key to using environmental pressures for competitive benefit lies in building the capabilities to transform a potential cost into an opportunity. Countries with weak capabilities may find that compliance costs damage competitiveness. It

New standards and quality regulations

Although most countries are cutting tariffs and quantitative restrictions on trade, standards and various forms of certification have emerged as new entry barriers. Most of the new barriers relate to processes (not, as with tariffs, to products) and include quality standards (ISO 9000), environmental standards (ISO 14000) and labour standards (SA 8000). In addition, many countries have technical regulations, industrial standards and testing and certification procedures designed to protect public safety and health. Standards offer many potential benefits for developing countries. They can be a source of technology transfer and a means of monitoring markets and obtaining information on competitors (box 1.8). They can also be a means of rationalizing costs based on codified best international practices, and can

22

Industrial Development Report 2002/2003

Box 1.9 Standards and technical regulations as barriers to developing country exports Standards and technical regulations provide many benefits to producers and consumers, not least of which is their information value. But they can also create trade barriers and segment marketsas when, for example, countries impose standards for colour televisions that differ from international norms, or protect domestic producers by issuing tailor-made standards such as requiring imported cars to have rain wipers on their headlights. Different countries have different incentives to use standards and technical regulations for protection purposes. For some countries with low tariff protection, liberalizing standards and technical barriers can provide greater economic benefits than further tariff reductions. Whether standards help or hinder exports of developing countries depends on the products being exported and on a countrys level of established standards. Exports can also be affected by higher costs resulting from duplicative testing performed by importers to assess conformity with standards. These duplicative tests are sometimes a response to the perceived weaknesses of standards organizations in developing countries. A recent example shows the effect that standards can have on exports from developing countries. The European Union (EU) banned imports of fish caught in Kenyas Lake Victoria because salmonella was detected in a shipment and, later, because cases of cholera emerged in Kenya. Because of the ban, EU fish imports from Kenya dropped 2537 percenta serious blow because the EU market accounted for 95 percent of Kenyas fish exports. Improving hygienic conditions to reduce the risks of similar actions was estimated to cost $5.8 million. Many country-specific technical regulations and industrial standards, created to protect public safety and health, instead become barriers to trade. To avoid such barriers, the WTO developed the Agreement on Sanitary and Phytosanitary Measures and the Agreement on Technical Barriers to Trade. The Agreement on Sanitary and Phytosanitary Measures recognizes the right of member countries to introduce regulations that protect human and animal health from food-borne risks, human health from animal- and plant-carried diseases, and animals and plants from pests and diseases. These regulations should be based on scientific principles, should not be maintained without sufficient scientific evidence and should not be applied in a way that constitutes a disguised restriction on international trade. The agreement also states that when determining sanitary and phytosanitary protection, members should minimize the negative effects on trade. But are developing countries in a position to identify when this occurs? And when they identify such instances; are they in a position to challenge decisions by industrialized countries based on scientific principles?
Source: UNIDO.

The Agreement on Technical Barriers to Trade states that product standards adopted to protect public health and safety, preserve the environment and serve other consumer interests should not pose unnecessary obstacles to international trade. The agreement encourages member countries to use international standards but does not require them to change their levels of protection. It also sets out a code of good practice to guide central government bodies in preparing, adopting and applying standards and describes how local government and non-governmental bodies should apply their own regulations. The agreements overarching principle is non-discrimination. Fair and equitable procedures must be used when deciding whether a product conforms with national standards, and methods that would give domestically produced goods an unfair advantage are discouraged. To avoid duplicative testing, the agreement also encourages countries to recognize each others testing procedures. Manufacturers and exporters need to know the latest standards in their prospective markets. Developing standards and technical barriers requires a powerful scientific and technical base, which can take decades to build and which industrialized countries have established. Thus it is not surprising that these countries have the highest number of new standards notifications to the WTO. In countries with low levels of protection, standards and technical regulations may provide more protection than traditional trade barriers. Nevertheless, many standards and technical regulations appear to be applied to heavily protected goodsparticularly in industrialized countries. These include agricultural and agroindustrial products as well as textiles, clothing and footwear. Thus any industry- or trade-related technical assistance from international organizations should be complemented by careful analysis of the traditional trade barriers facing the main exports of the countries receiving such assistance. Developing countries face serious difficulties in implementing the Agreement on Sanitary and Phytosanitary Measures and the Agreement on Technical Barriers to Trade. A September 2001 WTO document includes several proposals to facilitate and reduce the costs associated with the implementation and administration of several WTO agreements, including these two. But the proposals do not go far enough, offering at best marginal improvements to an issue that appears to require a complete rethinking. Indeed, a huge gap exists between these reforms and the needs of developing countries.

Box 1.10 More stringent environmental norms and conditions Opportunity

Environmental compliance has positive effects on society and on

the competitiveness of complying enterprises through increased innovation, lower costs, better resource use and first-mover advantages. Challenge
Developing countries lack expertise with accrediting and auditing

then becomes attractive for them to engage in a race to the bottom, lowering environmental standards to attract or retain industrial activity. Many developing countriesparticularly the least developedlack the capabilities to use environmental technologies. They even lack the basic capabilities to run the institutional (accreditation and auditing) framework for environmental compliance.

Stricter intellectual property rights

Industrial and technological development will be influenced by the TRIPS agreement. The agreement could affect invest-

systems for environmental compliance.

New technologies, new systems, new rules

23

Box 1.11 Stricter intellectual property rights Opportunity

Stricter intellectual property rights should stimulate innovation,

learning and risk taking in industrialized and newly industrializing economies. Challenge
Stricter intellectual property rights can raise the cost of technology

nology importsthrough higher licensing fees and product prices, more advanced skill needs to manage the new regime and greater scope for monopolistic practices by holders of intellectual property rights. Finally, stricter intellectual property rights can constrain technology development through copying and reverse engineeringactivities used to great effect by newly industrialized economies and, earlier, by many industrialized countries (box 1.12). At the same time, the TRIPS agreement offers benefits.55 Stricter intellectual property rights stimulate innovationin industrialized countries and newly industrialized economies and even in least developed countries with nascent technological activity. Stricter rights can also boost FDI and sales of advanced technologies (by innovators who need to protect proprietary knowledge). Still, the net benefits will depend on a countrys level of industrial and technological development. In the least developed countries the benefits may take a long time to materialize, and in present value terms (future values discounted at an appropriate interest rate) the costs may outweigh the benefits. Many developing countries are understandably concerned about this important topic, which requires further investigation.

imports for developing countries and limit their ability to reverse engineer and learn from foreign technologies.

ment, technology transfer and innovationand thus the accumulation of technological capabilities (box 1.11). Predicting its net result is difficult. Empirical evidence is scanty, the processes are complex and the effects are highly context-specific. It is widely accepted that stricter intellectual property rights can have negative effects on developing countries, particularly those low on the industrial and technological ladder.54 Such rights may not stimulate local innovation and may not promote overseas innovation relevant to these countries needs. They are also likely to raise the cost of tech-

Box 1.12 The case for strong protection of intellectual property rights Protection of intellectual property rights has played an ambiguous role in technological and industrial development. Many of todays industrialized economies relied on slack intellectual property rights to promote the technological development of their enterprises, shifting to stricter rules only when they had achieved technological parity with the leaders. The most technologically dynamic East Asian Tigersthe Republic of Korea and Taiwan Province of Chinaused copying and reverse engineering for long periods to promote local enterprises, only recently adopting stricter intellectual property rights. Protection of intellectual property rights is based on the premise that innovative activity is seriously constrained if innovators cannot reap the fruits of innovation. Thus copyrights protect the rights of authors (books, music, software), trademark registration protects unique trade logos and symbols, and patents protect the rights of inventions with industrial applicability (products as well as processes). For technology development, patents are most relevant. Patents are supposed to spur innovation. They grant exclusive rights of use, sale and manufacture to owners of intellectual property, compensating them for undertaking expensive and risky innovative activities. But in exchange, owners must disclose the invention on the patent document for anyone skilled in the art to be able to replicate. Thus patents are a trade-off: a market distortion is created in exchange for disclosing information on the technology. This disclosure is intended to benefit society by disseminating new technologies and encouraging competitors to invent around it, encouraging a second round of innovation. Advocacy of strong intellectual property rights presumes that the benefits of appropriation for innovators and disclosure for competitors outweigh the drawbacks of market distortions, making intellectual property rights beneficial to society. This presumption, almost impossible to test empirically, remains the subject of debate. Most develSources: Based on Chang (2001, background paper) and Luthria (2000).

oping countries, seeing themselves as users of existing technologies rather than makers of new ones, consider it premature to adopt Western models of intellectual property right protection. Indeed, technological catch-up could be constrained if developing countries enforced stronger intellectual property rights. Stricter rights could raise the cost of technology imports and restrict the ability to learn through reverse engineering. This argument has some merit. In the absence of a domestic industry lobby, low-income countries have strong intellectual property rights. And for obvious reasons, high-income countries also protect intellectual property rights very strongly. Middle-income countries offer the least protection for intellectual property rights. Two developments may change the shape of things to come. First, investment flows are seeking global destinations, and enterprises ability to protect their knowledge assets is a critical determinant in choosing destinations. Second, all WTO members that are signatories to the TRIPS agreement have agreed to reform their intellectual property rights regimes by 2004. Though the eventual benefits of this universal protection remain to be seen, for now such reform is a bitter pill for domestic industry and consumers to swallow. The challenge will be to help developing countries design policies and instruments that are in line with their technology-follower positions and that balance proprietary motives with access, efficiency and distributional considerations. Doing so would direct attention to drafting competition policies, price regulations and targeted subsidies and other transfer mechanisms that mitigate the potential negative effects of stronger intellectual property rights. Finally, alternative methods of encouraging local innovation may have to be devised to fit particular needs, such as protection and compensation for uses of indigenous knowledge in some societies.

24

Industrial Development Report 2002/2003

The case for stronger intellectual property rights is easier to make for economies such as Brazil, India, the Republic of Korea, Singapore and Taiwan Province of China, with their strong technological bases. In these economies weak intellectual property rights can deter transfers of valuable technologies and investments in risky R&D by domestic enterprises. But a case can be made for less stringent application of the TRIPS agreement in the least developed countries, with more exclusions and longer grace periods,56 so that they can participate meaningfully in global industrial activity.

14. Quah (1999). 15. Dicken (1998). 16. Pigato (2001). 17. Freeman and Perez (1988). 18. OECD (1999b) defines high-tech industries as manufacturers of aircraft, office and computing equipment, pharmaceuticals and communications equipment. Medium-high-tech industries are professional goods, chemicals (excluding drugs), electrical machinery, non-electrical machinery, motor vehicles and other transport equipment. Low-tech industries are paper, textiles and apparel, leather, food, beverages, tobacco and wood products. The remaining activities fall in the medium-low-tech category. 19. In 78 countries accounting for more than 95 percent of global production, high-tech production grew 5.9 percent a year in 19801997, compared with 2.7 percent for other manufacturing activity, and high-tech exports grew 10.8 percent, compared with 7.3 percent for other manufactured exports (NSF 2000). 20. Lall (2000). 21. Data are from OECD (1999b). Business accounted for a larger share of R&D in Japan (73 percent), Sweden and Switzerland (68 percent each) and the United States (64 percent). But its share was lower in Canada, France, Italy, the Netherlands and Spain, where government accounted for more than half of R&D. 22. In the United States, for example, the share of service enterprises in R&D rose from 4 percent in 1980 to 20 percent in 1997. In some countries the share of services is far higherhitting 37 percent in Canada and 32 percent in Denmark and Norway. But among other industrial leaders, such as Germany and Japan, manufacturing still accounts for a large share of innovation, and services account for only about 4 percent of R&D funding. 23. OECD (2000a, p. 32).

Notes
1. Dicken (1998); Freeman and Perez (1988). 2. Streeten (2001) argues that globalization is not international integration: it is partial international integration that, for various reasons, leads to national disintegration. Because industrialized countries have less demand for low-skilled workers, income gaps are widening in these countries. Meanwhile, developing countries seeking to prevent a brain drain are forced to pay higher wages to skilled workers, worsening income distributions in these countries as well. In addition, developing countries have less tax revenue available to pay for social services, though the need for them is rising. Moreover, elites in developing countries are adopting the values of their counterparts in industrialized countries and so neglecting essential social services such as education and health care. And minorities are trying to break away to share directly in the benefits of globalization. In sum, globalization has led to polarization (p. 54). 3. World Bank (2001b). 4. Chenery, Robinson and Syrquin (1986). 5. Reinert (1995). 6. Chenery, Robinson and Syrquin (1986). 7. The Economist (2000b), p. 10.

24. Humphrey (2000). 8. Quah (1999). 9. Dodgson, Gann and Salter (2001). 10. Gordon (2000); Pohjola (1998). 11. OECD (2000a, p. 39). 12. Cantwell and Santangelo (2000). 13. Financial Times, 18 June 2001. 25. In Europe the most active participants were enterprises from the United Kingdom (1,036 alliances), Germany (994), France (715) and the Netherlands (680). Other participants included enterprises from the Republic of Korea (119), former Soviet Union (90), China (86), Australia (63), Israel (51) and Taiwan Province of China (48). Data are from NSF (2000, pp. 257). 26. OECD (2000a, p. 33). 27. Mansell and Wehn (1998); OECD (2000a).

New technologies, new systems, new rules

25

28. Dicken (1998); Radosevic (1999). 29. NSF (1998). 30. Concentration of innovation between countries is also very high. See chapter 3 for data on developed and developing countries separately. 31. The supply chain approachwhich came firstfocuses on activities that include getting raw materials and assemblies into a manufacturing operation smoothly and economically. The value chain approach has a different focus and a larger scope. Value chain analysis looks at every step from raw materials to the end user, down to disposal of the packaging or product after use. The goal is to deliver maximum value to the end user for the least possible cost. Thus supply chain management is a subset of value chain analysis. 32. Mansell and Wehn (1998). 33. OECD (2000a). 34. Best (1990); Humphrey and Schmitz (1998); Nadvi (2001); Schmitz (1999a). 35. Pyke and others (1990). 36. Swann and others (1998). Recent research on geographic agglomeration suggests that agglomeration economies develop cumulatively through an accretion of learning, skills and networks (Krugman 1991; Venables 1996). 37. Schmitz and Nadvi (1999). 38. Bell and Albu (1999). 39. ILO (2001). 40. ILO (1998). 41. Dodgson and others (2001). 42. Pavitt (2001). 43. The trend is most marked in electronics enterprises in the United States, but it is spreading to other industries and countries. By early 2000 contracted manufacturers accounted for about 11 percent of

the market for electronics hardware. The largest, Solectron, was set to sell $20 billion in products by the end of 2000. See The Economist (2000a) and Sturgeon (1997). 44. Narula and Dunning (2000). 45. UNCTAD (2001). 46. Cantwell and Janne (1998). 47. The data come from OECD (1999a), Wong (2000), Rasiah (2000) and private communications from Peter Brimble on Thailand and Daniel Chudnovsky on Argentina and Chile. 48. R&D in developing countries tends to be concentrated in a few economies. For transnational corporations from the United States those economies are Argentina, Brazil, Hong Kong Special Administrative Region (SAR) of China, Malaysia, Mexico and Taiwan Province of China. 49. Hobday (1995). In 1985 more than 40 percent of the Republic of Koreas exports were original equipment manufacture, and in 1990 about three-quarters of its electronics exports were original equipment manufacture (Cyhn 2001). 50. Though it is difficult to compare data for FDI and mergers and acquisitions, about 80 percent of recent FDI in OECD countries has been in mergers and acquisitions (UNCTAD estimate). Mergers and acquisitions have also been important in Latin America and, since its financial crisis, in East Asia. 51. Stiglitz (1996). 52. Lall (1996). 53. Environmental standards can trigger innovation not just in greener production but also in new ways of cutting costs and material and energy waste. Such standards can also create first-mover advantages for regulated enterprises. 54. UNDP (2001); UNCTAD (1996); Maskus (2000); World Bank (2001a). 55. Maskus (2000). 56. World Bank (2001a).

26

Industrial Development Report 2002/2003

2
T

Global industrial activity

HE NEW INDUSTRIAL SETTING IS CHANGING THE NATURE AND

pattern of global industrial activity. Production, trade and innovation are shifting between activities and countries. Rapid technical change is boosting some activities and shrinking others. Economies are exploiting to different degrees the opportunities from new technologies, freer trade and more mobile productive resources. As a result there have been big variations in industrial performancein, say, the growth of output and exports or the upgrading of manufacturings technological structure. These variations are particularly marked in developing countries, where industrial capabilities vary considerably. But they are also evident in industrialized countries. (Detailed tables on selected indicators of industrial performance and drivers at regional and income levels are presented in the statistical annex.) Todays industrial performance and its drivers are as exciting as they are worrisome. As a group, developing countries are doing fairly well on almost all measures of performance. They are increasing their shares of global production and exports. They are moving up the technological ladder, enlarging their bases of human capital, deepening their technological activity and attracting larger portions of mobile resources. Yet the picture is worrying because industrial performance and its drivers are diverging rather than converging, with success confined to a few developing countries. Much of the divergence appears to be a long-term phenomenon, responding to structural factors that develop cumulatively. The implication is that globalization and liberalization may not reverse the divergence. To achieve long-term, sustainable industrial development, countries and firms need a concerted strategy for industrial restructuring and upgradingfor moving from simple to more advanced technologies. Todays map of global industrial activity shows the following features (table 3.2 shows country coverage by region): Manufacturing activity remains heavily concentrated in industrialized countries, though developing countries are

increasing their share. But in intensity of industrialization (measured by manufacturing value added per capita), developing countries still lag far behind. Among developing regions, East Asia (including China) is the best industrial performer in most respects, though it lags slightly in manufacturing value added per capita. It has the highest growth rates in manufacturing production and exports. It is far more export oriented than other developing regions. It has a more technologically advanced structure and is rapidly improving all the main drivers of industrial performance. And (excluding China) it has a commanding lead in skill creation, research and development (R&D) and technology licensing. Latin America and the Caribbean leads developing regions in manufacturing value added per capita and foreign direct investment. It has strong skills, an established export base and good infrastructure for information and communication technologies, and it leverages foreign technology effectively. But its manufacturing production and exports are based on a weak technological structure, particularly if Mexico is excluded. The region lags well behind East Asia in R&D and licensing. Even Mexico, the outlier in technology upgrading because of the North American Free Trade Agreement (NAFTA), suffers from a weak R&D base. South Asia has attained decent manufacturing growth but performs poorly in production per capita and exports. Its export structure is weak and stagnant. It lags in skills creation, technological effort and physical infrastructureand is relatively isolated from inflows of technology. The regions two largest economies, India and Pakistan, have not attracted much foreign direct investment to export activity between 1993 and 1997. The Middle East and North Africa has achieved fair manufacturing value added per capita, a reasonable base of skills and infrastructure, and good access to foreign technology.

27

But its industrial and export structures are not geared to technology upgrading, and its technological effort is weak. Sub-Saharan Africa, excluding South Africa, lags behind all other regions in almost all respects. The technological structure of its industrial production and exports is regressing. Industrial activity and capabilities are highly concentrated in a few leading economies both in industrialized and developing countries. Although this concentration is declining in industrialized countries, it is rising in developing countries in production, exports and technology imports. The 30 poorest countries are losing ground in most measures of industrial performance and its drivers, except for a slightly rising share of tertiary technical enrolments. Technologically, these countries are extremely weak and vulnerable. The situation is similar in the 12 least developed countries, which are seeing diminutions in their already minuscule shares of world industrial production and exports. This reinforces a general impression of highly uneven industrial development, combining spectacular and sustained successes with dismal and prolonged failures. These disparities are not temporary and will not correct themselves over time. Structural drivers of industrial development are slow, difficult and expensive to change, and the new global setting only raises their importance. Some of the drivers can improve only through greater reliance on market forces. But most need strong policy support.

industrial activity in the least developed countries relative to industrialized countries. During the same period MVA per capita in developing countries relative to the least developed countries rose from nearly 5:1 to about 9:1. Divergence in MVA also widened by region. Between 1985 and 1998 East Asia increased its share of the developing worlds total from 43 percent to 53 percent, lowering the shares of all other developing regions except the Middle East and North Africa. China accounted for 56 percent of East Asias growth in MVA. Latin America and the Caribbean remained the most industrialized developing region in per capita terms,1 but its growth failed to keep pace with East Asias, and it suffered the largest fall in regional shares of MVAabout 7 percentage points. South Asia had reasonable growth but remained the least industrialized region after SubSaharan Africa (excluding South Africa). Sub-Saharan Africa (excluding South Africa) accounted for only 1 percent of the developing worlds MVA in 1998, down from nearly 3 percent in 1985. Per capita, it was the least industrialized region, and unlike other regions, its per capita values declined. The least developed countries experienced good growth, but from a small initial base. Bangladesh accounted for 31 percent of this groups MVA in 1985 and 53 percent in 1998. Without Bangladesh, the least developed countries accounted for 0.5 percent of the developing worlds in 1985 and for nearly zero in 1998.

Manufactured exports

Reviewing trends in industrial performance

To assess industrial performance, this report examines manufacturing value added and manufactured exports and their technological composition.

Manufactured exports have grown faster than MVA in every region, reflecting the internationalization of industry. Developing countries again performed better than industrialized countries in both manufacturing growth and exports. By 1998 they had raised their share of world manufactured exports by 8 percentage points (compared with about 2 points for MVA). Per capita exports from industrialized countries were 15 times those from developing countries in 1998, down from 22 times in 1985. The gap in manufactured exports per capita between industrialized countries and the least developed countries widened from 192:1 to 212:1, as did that between developing countries and the least developed countries, from around 9:1 to 14:1. Thus the least developed countries fared much worse in exporting than in MVA. Industrial performance improved for other developing countriesbut weakened for the least developed countries. East Asia dominated developing country exports of manufactured goods even more than it did MVA, accounting for nearly

Manufacturing value added

Global manufacturing value added (MVA) grew about 7 percent a year during 19851998. Although industrialized countries still dominate the total, their share is fallingwith growth rates in developing countries nearly 2 percentage points higher. Still, MVA per capita in industrialized countries was about 17 times that in developing countries compared with about 18 times in 1985. Moreover, in 1985 MVA per capita in industrialized countries was 83 times that in the worlds least developed countriesand by 1998 had jumped to 144 times, revealing a dramatic plunge in the intensity of

28

Industrial Development Report 2002/2003

two-thirds of the total in 1998.2 Its rapid growth is reflected in reduced shares for all other developing regions. The largest fall in share, 4 percentage points, occurred in Latin America and the Caribbeanand excluding Mexico the fall was a massive 10 percentage points. The share of the Middle East and North Africa fell to 3 percent of the developing country total in 1998. Sub-Saharan Africa saw its small share of manufactured exports drop by half. Moreover, in 1998 Mauritius accounted for about a third of the regions $5 billion in manufactured exports (excluding South Africa). In per capita terms, exports from East Asia (excluding China) were 100 times those from Sub-Saharan Africa (excluding South Africa), 13 times those from South Asia and 6 times those from Latin America and the Caribbean. The least developed countries accounted for 1 percent of manufactured exports from developing countries in both 1985 and 1998. Because of its success with apparel, Bangladesh dominated this groups exports more than its MVA, moving from 58 percent of the total in 1985 to 78 percent in 1998. Data on MVA and manufactured exports indicate that: MVA and exports are highly concentrated in industrialized countries, though the share of developing countries is increasing for both. Manufacturing production is more concentrated than exports. The main producing and exporting countries are generally similar, but there are exceptions. Brazil is a leading producer but not exporter. Belgium and the Netherlands are leading exporters but not producers. Country concentrations in world shares of MVA and manufactured exports are declining, mainly due to the declining share of the United States in MVA and the declining shares of most industrialized countries in exports. Country concentrations are rising among developing economies, particularly in exports. Developing economy leaders in MVA are geographically dispersed, with five from East Asia (China, Indonesia, Republic of Korea, Taiwan Province of China, Thailand), three from Latin America and the Caribbean (Argentina, Brazil, Mexico) and one from South Asia (India; figure 2.1). This is not the case for exporterseight are from East Asia (adding Malaysia, the Philippines and Singapore to the leading producers of MVA) and two are from Latin America and the Caribbean (Brazil, Mexico; figure 2.2). China led both groups in 1998, raising its share by 7 percentage points in MVA and 14 points in exports since 1985.

Figure 2.1 National shares of developing world manufacturing value added, 1998 Others 20% China 29%

Indonesia 2% Thailand 3% Turkey 4% Argentina 4% India 5% Taiwan Province of China 6% Brazil 12% Republic of Korea 8% Mexico 7%

Source: UNIDO Scoreboard database (see technical annex).

Figure 2.2 National shares of developing world manufactured exports, 1998

Others 17%

China 16%

Indonesia 3% Philippines 3% Brazil 4% Thailand 5% Malaysia 7% Singapore 11% Taiwan Province of China 11% Mexico 11% Republic of Korea 12%

Source: UNIDO Scoreboard database (see technical annex).

The 30 poorest developing countries now account for 0.5 percent of world MVA and 0.3 percent of exports.

Technological structure of manufacturing value added and exports

MVA and manufactured exports are becoming more technology-intensive, moving from low-tech and resourcebased products to medium- and high-tech products. Classifications of manufactured products by technology intensity are provided in box 2.1. Medium- and high-tech products now account for more than 60 percent of global manufactured exports, mainly because of rapid growth of high-tech exports.3 In international trade, however, another factor is at play: the rise in integrated global production

Global industrial activity

29

Box 2.1 Manufactured products by technology intensity There are many ways to classify manufactured products by technology intensity. The classifications used in this report are based on those of OECD countries but exclude unprocessed primary commodities. Resource-based manufactures: mainly processed foods and tobacco, simple wood products, refined petroleum products, dyes, leather (not leather products), precious stones and organic chemicals. The products can be simple and labour-intensive (simple food or leather processing) or intensive in capital, scale and skills (petroleum refining or modern processed foods). Competitive advantage in these products generally but not alwaysarises from the local availability of natural resources. Low-tech manufactures: mainly textiles, garments, footwear, other leather products, toys, simple metal and plastic products, furniture and glassware. These products tend to have stable, well-diffused technologies largely embodied in capital equipment, with low R&D expenditures and skill requirements and low economies of scale. Labour costs tend to be a major element of cost, and the products tend to be undifferentiated, at least at the mass-produced (nonfashion) end of the scale. Barriers to entry are relatively low; competitive advantages in these productsof interest to developing countriescome from price rather than quality or brand names. Medium-tech manufactures: heavy industry products such as automobiles, industrial chemicals, machinery and relatively standard electrical and electronic products. The products tend to have complex but not fast-changing technologies, with moderate levels of R&D expenditures but advanced engineering and design skills and large scales of production. In engineering products there is emphasis on product design and development capabilities as well as extensive supplier and subcontractor networks. Barriers to entry tend to be high because of capital requirements and strong learning effects in operation, design and (for some products) product differentiation. Innovation and learning in the engineering segment increasingly involves cooperation in the value chain between manufacturers, suppliers and sometimes customers (for large items of equipment). High-tech manufactures: complex electrical and electronic (including telecommunications) products, aerospace, precision instruments, fine chemicals and pharmaceuticals. These products, with advanced and fast-changing technologies and complex skill needs, have the highest entry barriers. The most innovative ones call for large R&D investment, advanced technology infrastructure and close interaction between firms, universities and research institutions. But many activities, particularly in electronics, have final processes with simple technologies, where low wages can be an important competitive factor. The high value-to-weight ratio (for example, electronics products have a higher unit value relative to their weight than automotive products) of these products allows segments of the value chain to be broken up and located across long distances. The data do not allow the same classifications for MVA as for exports: MVA data have more gaps and the categories are often broader. Thus MVA is divided into three categoriesresource-based, low-tech and medium-plus-high-techwhile exports are divided into four. Note that MVA and export data do not distinguish countries by their genuine domestic capabilities in technology-intensive activities. The normal presumption is that production and exports reveal domestic technological capabilities, but the spread of high-tech assembly activity to low-wage countries belies this. Countries with low technological capabilities can appear technologically advanced, giving a misleading picture of industrial performance. This problem is not possible to solve by refining available data on MVA and exports. Thus other evidence on the spread of global integrated production systems dominated by transnational corporations and on local technological effort is used to arrive at a fairly realistic picture of national technological capabilities. The distinction between genuine technological capabilities and high-tech assembly is important, and the implications are discussed more fully in later chapters.

systems governed by transnational corporations. The relocation of different stages of production in different countries results in considerable intrafirm trade. Integrated production systems are most prominent in information and communication technology industries, where the high value-to-weight ratio of the product makes it economical to ship products and components around the world in search of fine differences in cost. Developing countries have less technology-intensive production and exports than do industrialized countries, but the gap is narrowing. The technological upgrading of developing country exports is faster than that of MVA (figures 2.3 and 2.4). High-tech products are experiencing the fastest growth in export shares and, if current growth rates continue, will soon overtake low-tech exports. Indeed, in 1998 developing country exports of electronics ($265 billion) were much larger than exports of textiles, clothing and footwear ($170 billion). But high-tech exports, far more than manufactured exports generally, are highly concentrated in a few countries. Low-income countries (excluding China and India) have far less technology-intensive production and exports than other developing countries, and their upgrading is much slower.4 Indeed,

between 1985 and 1998 their technological structure of production and exports actually regressed. The least developed countries have the lowest technology composition for both production and exports, with that for exports deteriorating. East Asia has the developing worlds most complex industrial production and exports and the fastest technological upgrading. Latin America and the Caribbean has achieved strong technological upgrading, but mainly because of Mexico (driven by NAFTA and concentrated in the maquiladoras on the U.S. border). South Asian exports are low-tech (out of line with its production structure, which reflects Indias heavy industry strategy), and upgrading is slow. Sub-Saharan Africa has experienced some technological downgrading because of rising shares of resource-based industries.

Assessing global export performance based on technology intensity

The top 5 and 10 exporters in 1985 and 1998 are shown in figure 2.5 according to the technology intensity of their prod-

30

Industrial Development Report 2002/2003

Figure 2.3 Developing country share of world manufacturing value added by technology intensity, 1985 and 1998 1985 1998 Percent 25 20 15 10 5 0 Resource based Low tech Medium and high tech

ucts. Although country concentrations are falling for every product group, they remain high. Complex (medium- and high-tech) products have much higher country concentrations than simple (resource-based and low-tech) products.

High-tech exports
In 1985 and 1998 the top exporters of high-tech manufactured products were the United States, Japan, Germany and the United Kingdom (table 2.1)suggesting that the leading exporters have deep, enduring capabilities. Otherwise, there was more fluidity among the top 25 exporters. Several industrialized economies lost rank significantly (five or more places): Austria, Belgium, Canada, Denmark, Italy and Switzerland. Among developing economies the leading exporters were Singapore, Taiwan Province of China and the Republic of Korea, with Singapore making an impressive leap (from 11th

Source: UNIDO Scoreboard database (see technical annex).

Figure 2.4 Developing country share of world manufactured exports by technology intensity, 1985 and 1998 1985 1998 Percent 35 30 25 20 15 10 5 0 Resource based Low tech Medium and high tech

Table 2.1

Top 25 exporters of high-tech products, 1985 and 1998 (millions of dollars)

1985 High-tech Rank Economy exports 1 United States 41,859 2 Japan 35,731 3 Germany 21,795 4 United Kingdom 13,013 5 France 12,141 6 Italy 7,063 7 Netherlands 5,195 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Taiwan Province of China Canada Switzerland Singapore Sweden Korea, Republic of Belgium Hong Kong SAR Ireland Austria Denmark Malaysia Spain Israel Mexico Finland Poland Brazil

1998 High-tech Economy exports United States 170,513 Japan 109,627 Germany 83,324 United Kingdom 68,276 Singapore 58,678 France 57,025 Taiwan Province of China 36,944 Netherlands 33,930 Korea, Republic of Malaysia China Mexico Italy Ireland Sweden Canada Philippines Switzerland Belgium Thailand Finland Spain Austria Israel Denmark 32,830 30,926 30,518 27,579 23,023 22,801 18,358 18,106 18,081 17,331 14,897 12,667 9,955 8,696 6,519 6,247 5,810 922,661 952,685 97

Source: UNIDO Scoreboard database (see technical annex).

Figure 2.5 Shares of world manufactured exports of top 5 and 10 exporters by technology intensity, 1985 and 1998 10 in 1985 10 in 1998 Percent 100 80 60 40 20 0 Resource based Low tech Medium tech High tech 5 in 1985 5 in 1998

4,480 4,478 4,381 3,879 3,862 3,541 2,827 2,269 2,123 1,464 1,356 1,277 1,255 942 717 716 665 599 177,628 179,380 99

Total for top 25 World total Share of top 25 in world total (percent)

Source: UNIDO Scoreboard database (see technical annex).

Source: Calculated from UN Comtrade database.

Global industrial activity

31

to 5th place). Mexico and Malaysia were also strong performers, with Mexico jumping from 22nd to 12th place and Malaysia from 19th to 10th. Chinas performance was particularly noteworthy: not among the top 25 exporters in 1985, it was in 11th place in 1998. The Philippines and Thailand were also new entrants. High-tech exports are highly concentrated, with the top 25 countries accounting for 9799 percent of world high-tech exports in both years. The top 10 accounted for 84 percent of the total in 1985 and 69 percent in 1998, and the top 5 for 72 percent and 52 percent. The high, stable level of export concentration suggests two trends. The first is that, despite the spread of global production systems, most high-tech exports remain in a few industrialized countries. Second, there are exceptions to this. Some developing countries have become major exporters and have built up strong first-mover advantages. China, which is a new entrant, shows strong competitive capabilities across a range of high-tech exports.

Table 2.2

Top 25 exporters of medium-tech products, 1985 and 1998 (millions of dollars)

1985 Mediumtech Rank Economy exports 1 Japan 101,697 2 Germany 79,256 3 United States 54,514 4 France 28,357 5 Italy 25,500 6 Canada 23,274 7 United Kingdom 20,702 8 Belgium 14,177 9 Sweden 11,184 10 Netherlands 10,543 11 Korea, Republic of 10,362 12 Switzerland 10,308 13 Spain 6,506 14 Austria 5,887 15 Taiwan Province of China 5,818 16 Singapore 3,708 17 Brazil 3,612 18 Finland 3,378 19 Denmark 2,999 20 Hong Kong SAR 2,940 21 Poland 1,953 22 Norway 1,395 23 Mexico 1,375 24 Ireland 1,160 25 Portugal 1,019 Total for top 25 World total Share of top 25 in world total (percent) 431,624 437,990 99

1998 Mediumtech Economy exports Germany 232,429 Japan 190,735 United States 189,215 France 97,154 Italy 93,003 United Kingdom 84,013 Canada 58,724 Belgium 56,975 Korea, Republic of 42,366 Mexico 40,332 Spain 40,301 Netherlands 35,884 China 30,853 Switzerland 29,657 Taiwan Province of China 27,761 Sweden 24,898 Austria 19,719 Singapore 18,214 Malaysia 12,001 Brazil 10,926 Czech Republic 10,675 Finland 10,363 Denmark 8,534 Portugal 7,801 Hungary 7,772 1,380,304 1,444,987 96

Medium-tech exports
Medium-tech exports also show a fair amount of stability, with the same industrialized countries (with slightly shifting ranks) in the top eight places in both 1985 and 1998 (table 2.2). Among developing countries the Republic of Korea was a leading exporter in both years, followed by Mexico (which rose from 23rd to 10th place) and China (absent in 1985, and in 13th place in 1998). Malaysia was another new entrant. The country concentration of medium-tech exports is not very different from that of high-tech exports and fell slightly over time. The top 25 countries accounted for 99 percent of world medium-tech exports in 1985 and 96 percent in 1998. The shares of the top 10 countries were 84 percent and 75 percent, and of the top 5 countries, 66 percent and 56 percent. In some ways medium-tech exports show national technological capabilities better than do high-tech exports. The assembly activities of transnational corporations play a role here too, but less so than in high-tech exports, because strong export performance in medium-tech is often based on deeper local manufacturing. Mobility plays a role as well. Parts and components of high-tech equipment can often be shipped around the world more easily than those of heavy industries. In the Republic of Korea and Taiwan Province of China medium-tech exports are led by domestic firms, while in Brazil, Malaysia, Mexico and Singapore they are led by transnational corporations. (China is a mix.) In both groups medium- and high-tech exports are the outcome of long

Source: Calculated from UN Comtrade database.

processes of domestic technological capability buildingas in the automobile industry in Brazil and Mexico and the electronics in Malaysia and Singapore.

Low-tech exports
Four of the top five low-tech exporters in 1985 and 1998 were industrialized countries, with the United States rising in rank and Japan declining (table 2.3). The number of developing countries in the top 25 low-tech exporters is similar to that in other technology categorieswith seven developing countries in 1985 and eight in 1998. More interesting, the leading exporters were also similar, with most East Asian Tiger economies in the group, along with Mexico. But there are also important differences. China, not among the top 25 exporters in 1985, was the global leader in 1998. Yet its lowtech strengths do not detract from its strong performance in medium- and high-tech products.

32

Industrial Development Report 2002/2003

Table 2.3

Top 25 exporters of low-tech products, 1985 and 1998 (millions of dollars)

Table 2.4

Top 25 exporters of resource-based products, 1985 and 1998 (millions of dollars)

1985 Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Economy Germany Italy Japan Taiwan Province of China France Korea, Republic of Hong Kong SAR United States Belgium United Kingdom Netherlands Spain Austria Switzerland Sweden Canada Brazil Portugal Turkey Finland Denmark India Singapore Greece Israel Low-tech exports 25,263 24,756 21,301 14,604 13,139 11,523 9,683 9,086 8,082 8,059 5,246 4,707 4,535 4,458 4,295 2,965 2,590 2,256 2,235 2,097 2,091 1,950 1,369 1,045 1,031 188,365 197,376 95

1998 Economy China Italy Germany United States France Taiwan Province of China United Kingdom Japan Belgium Korea, Republic of Mexico Netherlands Spain Canada Austria Hong Kong SAR Switzerland Turkey India Thailand Sweden Portugal Poland Denmark Czech Republic Low-tech exports 76,463 70,208 66,756 55,554 36,697 30,716 30,022 29,629 25,647 23,054 17,522 16,755 14,961 14,518 12,932 12,263 11,504 11,259 9,851 9,221 9,216 8,592 7,825 7,008 7,002 615,175 694,138 89

1985 Resourcebased Rank Economy exports 1 United States 22,065 2 Germany 21,795 3 France 17,130 4 Netherlands 17,012 5 Canada 14,759 6 Italy 12,713 7 United Kingdom 12,200 8 Belgium 11,306 9 Japan 9,105 10 Sweden 7,927 11 Singapore 6,883 12 Spain 5,523 13 Finland 5,462 14 Brazil 5,320 15 Switzerland 5,051 16 Denmark 2,962 17 Austria 2,912 18 Taiwan Province of China 2,735 19 Venezuela 2,577 20 Malaysia 2,553 21 Korea, Republic of 2,380 22 Ireland 2,197 23 Israel 2,001 24 Australia 1,776 25 India 1,745 Total for top 25 World total Share of top 25 in world total (percent) 198,089 215,418 92

1998 Resourcebased Economy exports United States 61,055 Germany 54,575 France 41,185 Belgium 34,400 United Kingdom 34,380 Canada 32,624 Netherlands 29,741 Italy 28,266 Japan 23,333 Ireland 16,651 Spain 15,989 China 15,091 Sweden 14,493 Finland 14,280 Singapore 13,764 Switzerland 12,251 Korea, Republic of 11,829 Brazil 11,742 Malaysia Israel Austria Thailand Denmark Mexico India 9,891 7,902 7,802 7,027 6,282 6,117 6,102 516,772 593,812 87

Total for top 25 World total Share of top 25 in world total (percent)

Source: Calculated from UN Comtrade database.

Source: Calculated from UN Comtrade database.

Brazil, revealing competitive weaknesses in low-tech products, was absent from the list in 1998. Hong Kong Special Administrative Region (SAR) of China also weakened, but a significant part of Chinas low-tech exports is based on the operation of Hong Kong SAR firms. Low-tech exports are less concentrated than high- or mediumtech exports. But country concentrations are still high. The top 25 countries accounted for 95 percent of low-tech exports in 1985 and 89 percent in 1998, the top 10 countries for 74 percent and 64 percent, and the top 5 countries for 50 percent and 44 percent. That so many industrialized countries persistently lead in these exports, despite high and rising wages, suggests that cheap unskilled labour is not a dominant competitive advantage. Low wages for productive and skilled workers do matter, as China shows, but strong advantages based on skills, technology and organization persist over time. Thus developing countries that want to establish a long-term lead, outlasting rising wages, have to develop such advantages.

Resource-based exports
Industrialized countries made up the top 10 exporters of resource-based products, with the United States, Germany and France leading in both 1985 and 1998 (table 2.4). Many developing countries rely heavily on primary exports, but competitiveness in processed primary products is firmly in the hands of industrialized countries, many without a large domestic resource base. Again, technologymainly the ability to handle large, capital-intensive and complex processing facilitiesis of great importance. So are complex organization (large integrated production facilities across nations), marketing and branding. In 1985 Singapore was the leading developing economy exporter of resource-based productsreflecting its large petrochemical processing facilities. China led in 1998. The Republic of Korea and Taiwan Province of China were also among the main resource-based exporters, despite their lack of domestic

Global industrial activity

33

resources. Other leading developing economies included resource-rich Brazil, Malaysia, Mexico and Thailand. India appears in the group because of its major business of cutting (imported) gems. Of the four technology groups, resourcebased exports are the least concentrated by country. The top 10 exporters accounted for 68 percent of the total in 1985 and 60 percent in 1998, and the top 5 for 43 percent and 38 percent.

explosive growth of the weightless economy and the high information content of industrial activities. It is difficult to quantify a countrys stock of industrial skills. Few countries publish data on peoples skills by discipline. And even if they did, it would be impossible to estimate levels of relevant, upto-date skills. As a result most comparisons of industrial skills use flows rather than stocks: current education enrolments at the primary, secondary and tertiary levels. Such measures have two main drawbacks. First, they ignore on-the-job learning experience and trainingwhich in many countries is a major source of skill formation. Second, enrolment data do not take into account the significant differences across countries in education quality, completion rates and relevance to industrial needs. Lacking better data, this report relies on formal enrolment figures. Which level of education is most relevant for industrial performance? Primary enrolments may not be as relevant as secondary and tertiary because most countries have achieved universal primary education and much of modern industry requires higher level skills.5 In line with the classic work of Harbison and Myers (1964), this report assumes that secondary and tertiary enrolments are the most relevant skill indicators and that tertiary enrolments should be weighted more heavily than secondary.6 Developing countries account for a large share of the worlds tertiary enrolmentsmore than twice their shares of MVA and exports, and nearly nine times their share of global R&D. Tertiary enrolments grow slowly in all regions, but they grow

Focusing on drivers of industrial performance

Industrial performance reflects the complex interaction of many factors, including institutions, skills, technologies, infrastructure, networking, political and social stability and other factors. It is not possible to quantify all these in a comparable way across countries. It is also not necessary here to include macroeconomic variables analyzed regularly (and in great depth) by other agencies. Mapping the structural influences on industrial performancetermed drivers, for conveniencecalls for selectivity and simplification. This report focuses on five drivers directly relevant to industry and for which comparable data are available: skills, technological effort, inward foreign direct investment, royalty and technical payments abroad and modern infrastructure.

Skills
Skills have always been important for industrial performance. But they have become even more crucial because of the

Figure 2.6 Distribution of tertiary enrolments in developing regions, total and technical subjects, 1985 and 1998 Total 1985 Total 1998 Percent 50 40 30 20 10 0 East Asia East Asia (without China) South Asia Latin America and the Caribbean Latin America and the Caribbean (without Mexico) Sub-Saharan Africa Sub-Saharan Africa (without South Africa) Middle East and North Africa and Turkey Technical 1985 Technical 1998

Source: UNIDO Scoreboard database (see technical annex).

34

Industrial Development Report 2002/2003

faster in developing than in industrialized countries. Still, the intensity of skill creation (measured by enrolments per 1,000 people) is far lower in developing than in industrialized and transition economies. In terms of the number of students, lower-middle-income countries do slightly better than high-income and uppermiddle-income countries in total tertiary enrolments and slightly worse in technical enrolments. Transition economies have the highest intensity of technical skill creation. Lowincome countries have achieved significant growth in general tertiary enrolments, but from very small bases. The distribution of tertiary enrolments in developing regions has a pattern similar to that for MVA and manufactured exports, with a strong lead by East Asia (figure 2.6). South Asias performance is better here because of high enrolments in India, but its shares of global and developing country enrolments fell between 1987 and 1995. In 1995 Sub-Saharan Africa (excluding South Africa) accounted for less than 2 percent of technical enrolments in developing countries though it had the second-fastest growth rate (after East Asia). Per capita, its enrolments were 6 percent of those in East Asia (excluding China) and 9 percent of those in Latin America and the Caribbean. Among developing countries, China and India dominate total and technical tertiary enrolmentstogether accounting for a third of the totalfollowed by Indonesia and the Republic of Korea. Among industrialized countries the United States and the Russian Federation have the most students enrolled in

technical education, accounting for nearly one-quarter of the world total.

Technological effort
Technological effort is a crucial driver of industrial development, even for industrial latecomers. Countries that import technologies must engage in conscious learning to master the technologies and adapt them to local conditions. The more advanced and complex the technology, the greater the learning effort required. Much of this effort cannot be quantified. It occurs in almost all parts of an enterprise, and much of it is informal. Some does take the form of formal R&D, however. R&D becomes more significant as a countrys industrial structure develops and firms use more advanced technologieseven in firms not innovating at technological frontiers, because R&D is needed to understand, adapt, imitate and improve imported technologies. R&D is also vital to keeping track of technological progress elsewhere in the world. These imitative and monitoring functions of R&D are prominent even in industrialized countries. Because comparable information on R&D spending is available across countries, these data are used here as a proxy for technological effort.7 Data on R&D financed by productive enterprisesas defined by the United Nations Educational, Scientific and Cultural Organization (UNESCO)are preferred to data on total national spending on R&D because enterprise

Figure 2.7 Regional distribution of developing world R&D nanced by productive enterprises, 1985 and 1998 1985 1998 Percent 80 70 60 50 40 30 20 10 0 East Asiaa East Asia (without China) South Asia Latin America and the Caribbean Latin America and the Caribbean (without Mexico) Sub-Saharan Africa Sub-Saharan Africa (without South Africa) Middle East and North Africa and Turkey

Source: UNIDO Scoreboard database (see technical annex). a. Data for 1985 are missing because data for China are missing.

Global industrial activity

35

R&D is more directly related to industry (total R&D spending also includes agriculture, defence, health and so on). Moreover, in most developing countries governmentfinanced R&D, often the bulk of national R&D, goes to state or university laboratories, with little impact on industrial innovation. Around the world, enterprise-financed R&D grew 10 percent a year between 1985 and 19951998 (using the most recent available data), reflecting the growing importance of technological effort. Developing country R&D grew faster (18 percent a year) but from a low base, accounting for just 5 percent of the world total in 1998. Per capita R&D in developing countries was only 1.0 percent of that in industrialized countries, up from 0.5 percent in 1985. High- and upper-middle-income developing countries accounted for nearly 90 percent of the groups total. In the poorest countries (excluding China and India) R&D was practically nonexistent. During this period R&D as a share of world MVA rose from around 4 percent to 6 percent. Among developing regions, East Asia has a stronger lead in enterprise-financed R&D than in production, exports or skills (figure 2.7). Including China, the region accounts for nearly four-fifths of the developing world total. The shares of all other developing regions decreased, particularly in South Asia (from 14 percent to 2 percent). In both 1985 and 1998 SubSaharan Africa (excluding South Africa) had no share of the total. The developing worlds enterprise-financed R&D is also highly concentrated (figure 2.8). In 1998 the Republic of Korea accounted for more than half of the total, and the two next biggest sourcesTaiwan Province of China and Brazil together accounted for more than a quarter. Almost all R&D

funded by enterprises occurred in 10 economies, and none occurred in the 30 economies at the bottom of the list. Among industrialized countries the United States, Japan and Germany were at the top, accounting for 82 percent of the world total in 1985 and 75 percent in 1998though their shares changed. Germanys share fell 10 percentage points and that of the United States fell 3 points, while Japan gained about 6 points.

Inward foreign direct investment

Foreign direct investment (FDI) is an important way of transmitting skills, knowledge and technology to developing countries and so is an important driver of industrial performance. Transnational corporations, generally the leading innovators in their industries, are engaging in more technology transferreflecting the rising cost and pace of technical progress and the reluctance of innovators to sell valuable technologies to independent firms.8 Transnational corporations also provide capital, skills, managerial know-how and access to markets. Countries can accelerate their industrial development by plugging into integrated global production systems governed by transnational corporationsand becoming global or regional supply centres, particularly in high-tech activities. Independent firms in developing countries can participate in these systems, but few have the capabilities to meet the extremely high technical standards.9 Most countries that have entered these systems in recent years have done so through FDI. Transnational corporations, though good at transferring operational technologies, are far less so interested in transferring or fostering deeper technological capabilities. It is often not economical for transnational corporations to relocate R&D at the same pace as production. Agglomeration, linkages, cumulative learning and economies of scale make it costly to spread R&D overseasparticularly to developing countries short on high-level skills, advanced infrastructure and supporting research institutions. Thus most developing countries do not attract the research activities of transnational corporations.10 So, although FDI is important for industrial performance, it may not be the best way to deepen technological activity in developing countries. The ideal FDI measure for assessing industrial performance would be inflows into manufacturing (and within that, into domestic and export production). But this kind of disaggregation is generally not possible: for most countries the only available measures are inward FDI flows and stocks. This

Figure 2.8 Leading developing economies in R&D financed by productive enterprises, 1998 Taiwan Province of China 14%

Republic of Korea 53%

Brazil 12%

China 6% Singapore 3% South Africa 3% India 2% Others 7%

Source: UNIDO Scoreboard database (see technical annex).

36

Industrial Development Report 2002/2003

report uses flows, averaged over three years to minimize variations.11 The inability to distinguish FDI for manufacturing and FDI for other sectors should be taken into account, particularly in regionssuch as Latin Americawhere a large portion of FDI goes into utilities and banking. Around the world, FDI inflows grew more than 15 percent a year between 19811985 and 19931997. This growth was faster than that for global MVA (about 7 percent) and manufactured exports (10 percent). Developing countries achieved faster FDI growth (18 percent) than industrialized countries (about 14 percent) and in 19931997 accounted for one-third of world FDI inflows. The Middle East and North Africa has seen a large decrease in its share of world and developing country FDI. Latin America and the Caribbean retained its share of the developing country total (about 30 percent). Sub-Saharan Africa (excluding South Africa) experienced a small decrease in its world share, while its developing country share halved. Per capita flows are lowest in South Asia$2 in 19931997, less than half that in Sub-Saharan Africa (excluding South Africa) and only 3 percent of that in East Asia and Latin America and the Caribbean. Although South Asias shares have increased (it has the fastest growth in flows), it has a long way to go before it catches up with other regions. In India and Pakistan export-oriented FDI is lower than the average for the developing world. The same is true for most of Sub-Saharan Africa, where most FDI goes into resource extraction, which is concentrated in oil-exporting economies.

Latin America and the Caribbean has the highest per capita FDI among developing regions. But much of the regions FDI goes into industrial activities oriented towards the domestic market and into privatized utilities. The main exception is the automobile industry in Argentina, Brazil and Mexico, where transnational corporations have invested large amounts and improved export competitiveness enormously. Apart from Costa Rica and Mexico there has been little FDI in high-tech export-oriented activities in Latin America. Chinas dominates FDI flows to developing countries, followed by Singapore, Brazil and Mexico (figure 2.9). In 19931997 nearly 80 percent of FDI in developing countries went to 10 countries.

Royalty and technical payments abroad

Royalty and technical payments abroad are meant to capture arms-length purchases of know-how, patents, licences and blueprintsimports of embodied technology in nonequity forms.12 Some countries do not publish royalty data, and values have to be imputed by looking at ratios of royalty payments to total service payments by similar countries. Moreover, royalties and technical fees are not always for industrial technologysome of them go for service sector purchases of know-how, brand names and franchises. Royalty payments often include nonarms length transactions, with significant flows between the affiliates and parents of transnational corporations. Many of the largest spenders on licensing (Hong Kong SAR, Ireland, Singapore)

Figure 2.9 Regional distribution of foreign direct investment inflows, 19811984 and 19931997 19811984 19931997 Percent 70 60 50 40 30 20 10 0 East Asia East Asia (without China) South Asia Latin America and the Caribbean Latin America and the Caribbean (without Mexico) Sub-Saharan Africa Sub-Saharan Africa (without South Africa) Middle East and North Africa and Turkey

Source: UNIDO Scoreboard database (see technical annex).

Global industrial activity

37

Figure 2.10 Regional distribution of royalty payments, 1985 and 1998 1985 1998 Percent 70 60 50 40 30 20 10 0 East Asiaa East Asia (without China) South Asia Latin America and the Caribbean Latin America and the Caribbean (without Mexico) Sub-Saharan Africa Sub-Saharan Africa (without South Africa) Middle East and North Africa and Turkey

Source: UNIDO Scoreboard database (see technical annex). a. Data for 1985 are missing because data for China are missing.

have a strong presence of transnational corporations, and a substantial share of the payments are made by these corporations systems. But while the statistical correlation between FDI and royalty payments is positive, it is not very high (0.43)suggesting that arms-length transactions account for a significant portion of the total. Thus this variable is the best available proxy for technology purchases by local firms. Technology licence payments rose 17 percent a year in 19851998, even faster than FDI flows. The world leaders are the United States and Japan, which are also the largest indus-

trial innovators and technology exporters. The growth of their technology exports suggests that innovators are specializing and that technology markets are becoming quite integrated. East Asia pays far more royalty fees than any other developing region (figure 2.10). East Asia (excluding China) also has by far the highest per capita spending on these fees (about $27), in line with its high-tech specialization. That the region also spends the most on R&D and receives the most FDI suggests that these different modes of acquiring and developing technology complement each other. South Asia and Sub-Saharan Africa (excluding South Africa) spend the least per capita on royalties (less than $0.25 each), suggesting a massiveand possibly harmfulgap in accessing world technologies. In 1998, 10 countries accounted for 86 percent of the developing worlds spending on royalty fees (figure 2.11). The bottom 30 countries accounted for almost none.

Figure 2.11 Leading developing economies in royalty fees, 1998 Others 14% Argentina 3% Mexico 3% Thailand 7% Indonesia 7% Brazil 7% Hong Kong SAR 8% Singapore12% Taiwan Province of China 9% Republic of Korea 16%

Malaysia 16%

Modern infrastructure
National data on traditional infrastructurerailways, roads, ports, water suppliesare not readily available. Data on modern infrastructurerelated to information and communication technologiesare easier to collect, so this report uses data on telephone mainlines, mobile telephones, personal computers and Internet hosts for each country. The distribution of modern infrastructure is similar to that for other drivers, with East Asia in the lead (figure 2.12). But Latin

Source: UNIDO Scoreboard database (see technical annex).

38

Industrial Development Report 2002/2003

Figure 2.12 Regional distribution of information and communication technologies, 19982001 Mobile telephones Telephones Percent 70 60 50 40 30 20 10 0 East Asia East Asia (without China) South Asia Latin America and the Caribbean Latin America and the Caribbean (without Mexico) Sub-Saharan Africa Sub-Saharan Africa (without South Africa) Middle East and North Africa and Turkey Internet hosts Personal computers

Source: UNIDO Scoreboard database (see technical annex).

Figure 2.13 Regional distribution of information and communication technologies per 1,000 population, 19982001 Mobile telephones Telephones Percent 150 120 90 60 30 0 Internet hosts Personal computers

East Asia

East Asia (without China)

South Asia

Latin America and the Caribbean

Source: UNIDO Scoreboard database (see technical annex).

Latin America and the Caribbean (without Mexico)

Sub-Saharan Africa

Sub-Saharan Africa (without South Africa)

Middle East and North Africa and Turkey

America and the Caribbean also does well, reflecting its high per capita incomes. South Asia and Sub-Saharan Africa (excluding South Africa) do quite poorly. Adjusted for population size, South Asia deteriorates considerably relative to East Asia (excluding China), while Latin America and the Caribbean does fairly wellparticularly on telephones (figure 2.13). But for advanced information and communication technologies (personal computers and Internet hosts), East Asia (excluding China) retains the lead among developing regions. South Asia performs poorly, even compared with Sub-Saharan Africa (excluding South Africa).

Notes
1. East Asia (excluding China) was catching up rapidly, however, and its 1998 figure is an aberrationreflecting the immediate effects of the previous years financial crisis. 2. Note that the value of manufactured exports from East Asia, particularly East Asia (excluding China), exceeds the regions MVA. This is because the value of manufactured exports is a gross figure (including the value of inputs) while MVA is net of inputs.

Global industrial activity

39

3. MVA data do not distinguish between medium- and high-tech industries, so they are combined. 4. Country classifications by income level are from World Bank (2001b). 5. In some developing countries, however, primary education remains the main source of skill formation. These countries skill have high illiteracy rates, and their industrial sectors are concentrated in simple activities that do not require high-level skills. But including primary enrolment rates in the skill measure does not noticeably change country rankings; all rankings of skill levels are closely correlated. 6. This report recalculates the Harbison-Myers index for 1985 and 1998. The index measures a countrys average share of the relevant age groups enrolled in secondary and tertiary education, with tertiary given a weight of five. In addition, the report uses two other measures to capture the creation of high-level skills: the number of students enrolled at the tertiary level in all subjects and the number enrolled at the tertiary level in technical subjects (defined as mathematics, computing, engineering and pure science). These two measures use population as the deflator (unlike the Harbison-Myers index, which uses the relevant age group as the deflator). The Harbison-Myers index and tertiary technical enrolment measures are significantly correlated, with a correlation coefficient of 0.87. Both measures are also highly correlated with the Barro and Lee (1993, 1996) measure based on countries average number of years of education. In other words, countries that invest in secondary and general tertiary education also produce technically skilled workforces. Harbison-Myers indexes are shown in annex table 3.5 and tertiary technical enrolments in annex table 3.6. The starting year for the enrolment data is 1987 because it provides broader country coverage. Because enrolment indicators tend to change slowly, this does not significantly affect comparisons with 1985 data for other indicators. Harbison and Myers (1964) calculated their index using 1958 data. For a comparison of their rankings for 1958 with data for 1995, see Lall (1999). 7. An alternative measure of effort, patents taken out in the United States, was also calculated. Although industrialized countries often

use this as a measure of innovative output, it is more relevant to frontier innovation than R&D effort. But R&D and United States patenting, appropriately deflated, are closely correlated: countries that spend a lot on R&D also take out a lot of patents overseas. 8. Radosevic (1999). 9. For discussions of technical standards in the electronics industry, see Ernst (1997, 2000), Hobday (1995), Mathews and Cho (2000). 10. See Lall (2001b). Developing countries account for an extremely small share of R&D by transnational corporations. In the mid-1990s less than 1 percent of R&D by transnational corporations based in the United States occurred in developing countries (while 11 percent was in other industrialized countries). Even this small share was highly concentrated, with nearly two-thirds in Brazil, Mexico and Singapore (in declining order of importance). Still, even though the amount spent on R&D in developing countries is low relative to corporate R&D, it can account for a substantial share of national R&D in host economies. 11. Flows are preferred for two reasons. First, stocks are calculated at historical values and can give a misleading picture of the current value of foreign investment assets. Second, flows give a better picture of current FDI activity and so are more relevant to explaining current performance. Still, a comparison of the two datasets for the sample yielded similar rankings, so the choice between flows and stocks does not make much practical difference. 12. Capital goods are also a type of embodied technology import but are not included in externalized technology imports for several reasons. First, the data capture a large element of equipment imports for nonindustrial investment. Second, they include re-export of equipment, particularly by entrept centres like Hong Kong SAR and Singapore, biasing the results in their favour. Third, the data include components of capital goods for export processing, making exportoriented countries in electronics appear to be very large importers of technology.

40

Industrial Development Report 2002/2003

3
P

Benchmarking industrial performance

OLICYMAKERS ACROSS THE WORLD KEEP CLOSE TRACK OF OTHER

countries industrial performance. This trend is growing, as the Web sites of many government ministries, research institutes, consultancies and international organizations show. Industrial performance, productivity, innovativeness, skills, foreign direct investment inflows and the like are constantly compared at varying levels of detail. This concern with comparative industrial performance reflects global competition and the usefulness of comparisons for policy purposes. The systematic use of comparisons, or benchmarking, clearly serves a strong need.1 Benchmarks are needed because it is difficult to assess national industrial performance on the basis of a priori norms. For many facets of performance there are no norms in economic theory. Are manufacturing production, exports and employment growing fast enough, given a countrys resources, industrial structure and level of technology? Are domestic enterprises sufficiently innovative, or workers sufficiently skilled? Is the industrial infrastructure coping adequately with the needs of the new economy? Is the economy participating fully in international knowledge flows? These and many similar questions cannot be addressed using only theoretical parameters. The best guide when addressing such important questions comes from the performance of other (comparable) economies. If they are doing consistently better, something is clearly amiss at home. Even where a priori engineering parameters exist say, for an industrial plantbenchmarking against best practice is still useful. It helps operators to see whether equipment can be "stretched" to perform better or workers reorganized to become more productive. Wherever performance can be improved, benchmarking is a useful tool. The sheer pace of change in the national and international economic and technological environment also makes it far more difficult for governments to assess domestic performance without looking at other economies. The need for benchmarking is all the greater for countries undergoing wrenching internal structural and policy shifts.

Benchmarking can be conducted at many levelsenterprise, industry, institution, government or government department. It can focus on specific matters, such as capital and labour costs, infrastructure, technology, innovation, skills or the environment. The more specific the level, the easier it is to derive quantitative benchmarks; the more general the level, the harder it is to define what is relevant and, often, how to measure it. So, benchmarking industrial performance is easier than benchmarking national competitiveness. Still, even the level of industry is quite general, with complexities and variations that broad benchmarks cannot take into account. National benchmarks should thus be seen as useful preliminary indicators of relative performance. As an aid to policymaking, they have to be supplemented by detailed analysis by country and activity. It is just as important to bring in the qualitative institutional and policy variables that such benchmarks have to leave out. The Scoreboard introduced in this report provides useful information on crucial aspects of industrial development.2 It provides a simple tool that countries can use to assess their position with respect to industrial performance (box 3.1) and its structural features (discussed in chapter 4). The Scoreboard is an analytical tool using published hard data to explain differences in industrial performance and capabilities. It is merely a series of benchmarks (which will be improved and enlarged over time). In the next chapter, however, some simple statistical analysis is conducted using the Scoreboard to check how closely industrial performance relates to a given set of capability factors and whether their structural features have changed over time. The Scoreboard focuses on manufacturing and on a small number of structural variables on which hard data are available.

Benchmarking industrial performance: the competitive industrial performance index

A ranking of economies by the competitive industrial performance (CIP) index reveals a general pattern that is as expected:

41

Box 3.1 The competitive industrial performance index The competitive industrial performance (CIP) index focuses on the national ability to produce manufactures competitively. Since it is difficult to find a single indicator that captures all the dimensions of competitive production, the CIP index is constructed from four basic indicators of industrial performance. 1. Manufacturing value added (MVA) per capita. MVA would automatically capture the competitiveness of industrial activity if all production from all countries were fully and equally exposed to international competitionbut it is not. Trade and other policies limit the exposure of domestic industry to international competition. So do natural barriers to trade, such as high transport costs, poor access to natural resources, differences in taste, legal and institutional variations and information gaps. Production for home markets (particularly in countries with large markets or with strong import substitution policies) faces less intense competition than production for export. 2. Manufactured exports per capita. The export measure indicates how competitive industrial activity is in one set of markets. It also captures another important aspect of industrial performance: the ability of national industry to keep pace with technological change, at least in exported products. Exports can be taken to demonstrate that producers are using competitive (modern) technologies. This is important because the technology measures below do not capture technological upgrading within broad product groups. The export indicator partially offsets this inability. 3. Share of medium- and high-tech activities in MVA. The higher the share of medium- and high-tech activities in MVA, the more technologically complex is the industrial structure of a country, and the more competitive is the countrys industrial performance. Industrial development generally entails moving up from resource-based and low-tech to medium- and high-tech activities (Chenery, Robinson and Syrquin 1986). Technology-intensive structures are better for growth and development. Technologically complex activities offer greater learning potential and lend themselves more to sustained productivity growth (because of the greater potential for applying new scientific knowledge). Many have stronger spillover benefits, especially hub activities that disseminate technology across different activities. Historically this role was played by the capital goods sector; today the subsector of electronic technologies is vital (Mowery and Rosenberg 1989). As seen in the previous chapter, high-tech activities also enjoy better growth prospects. And they often have dynamic international production systems. Even so, structural change is not automatic or easy because of the slow, incremental and path-dependent nature of learning. Many low-tech and resource-based industries can also have bursts of rapid growth; activities within these industries can have high-tech segments. And industries can shift between technological categories over time. Still, the technological complexity measure offers insights into the ability of countries to sustain growth in the new global setting. 4. Share of medium- and high-tech products in manufactured exports. The share of medium- and high-tech products in manufactured exports is considered separate from the share in MVA, because in certain circumstances the two differ significantly. In large import-substituting developing economies, for example, the structure of MVA tends to be more complex than that of exports. The values for each of the four variables are standardized for the sample to range from zero (worst performers) to one (best performers). The composite index is calculated as a simple average of the four standardized basic indicators. No weights are assigned to any of the basic indicators. The effect of each basic indicator on the final rank is also analyzed separately (see annex tables A3.1 and A3.2). The technological breakdown of MVA is far from perfect. The main reason for this is the lack of consistent cross-country data at the level of disaggregation required for fine distinctions by technological intensity. Moreover, as noted in the previous chapter, it is impossible to distinguish between industrial (or export) structures based on genuine technological capabilities and those based on low-tech assembly in high-tech industries. This problem is more acute for export data: countries with large shares of high-tech assembly in total exports appear among advanced industrial performers (the Philippines is a good example). But the developing countries affected by this data problem are relatively few, and their identities are well known.

industrialized countries congregate near the top, transition economies and middle-income developing countries around the middle, low-income developing countries and least developed countries at the bottom (table 3.1). In general, CIP ranks changed little between 1985 and 1998. The correlation coefficient between the index values for the two years is 0.940, supporting the argument that performance is the outcome of slow and incremental processes. Moreover, since all countries are trying to raise their industrial performance, achieving relative improvements is difficult. Leaps in the rankings are nevertheless possible. Between 1985 and 1998, 22 countries changed ranks by 10 or more places. Countries near the top and bottom tend to have relatively stable positions, while those in the middle are more volatile. The main cause of the large upward leaps appears to be participation in integrated global production networks, which

sharply raises the share of complex products in exports (and, over a longer period, in MVA). Among the top 40 ranking economies the largest improvements were in China, Ireland, Thailand, the Philippines, Malaysia, Costa Rica and Hungary, with Mexico, Singapore, the Republic of Korea and Taiwan Province of China close behind (figure 3.1). All of these, except for the Republic of Korea and Taiwan Province of China, have experienced a significant increase in the presence of transnational corporations in export manufacturing. The Republic of Korea and Taiwan Province of China are more strongly linked to integrated global production systems through such arrangements as original equipment manufacturing than through foreign direct investment. Among the bottom 40 countries the largest improvement was in Indonesia, again with a strong presence of transnational corporations in export manufacturing.

42 Industrial Development Report 2002/2003

Table 3.1 Ranking of economies by the competitive industrial performance index, 1985 and 1998

Rank 1998 1985 1 6 2 1 3 15 4 2 5 3 6 5 7 4 8 7 9 8 10 12 11 10 12 11 13 13 14 14 15 19 16 9 17 16 18 22 19 21 20 20 21 17 22 30 23 28 24 .. 25 45 26 26 27 34 28 .. 29 23 30 18 31 24 32 43 33 27 34 25 35 29 36 44 37 61 38 36 39 32 40 33 41 37 42 31 43 42 44 ..

Economy Singapore Switzerland Ireland Japan Germany United States Sweden Finland Belgium United Kingdom France Austria Denmark Netherlands Taiwan Province of China Canada Italy Korea, Republic of Spain Israel Norway Malaysia Mexico Czech Republic Philippines Portugal Hungary Slovenia Australia Hong Kong SAR New Zealand Thailand Brazil Poland Argentina Costa Rica China Turkey South Africa Greece Romania Bahrain Uruguay Russian Federation

Index value 1998 1985 0.883 0.587 0.751 0.808 0.739 0.379 0.696 0.725 0.632 0.635 0.564 0.599 0.562 0.633 0.538 0.494 0.495 0.489 0.473 0.426 0.465 0.450 0.453 0.445 0.443 0.424 0.429 0.398 0.412 0.292 0.407 0.474 0.384 0.379 0.370 0.247 0.319 0.259 0.301 0.290 0.301 0.348 0.278 0.116 0.246 0.125 0.243 .. 0.241 0.044 0.240 0.159 0.239 0.088 0.221 .. 0.211 0.214 0.204 0.320 0.186 0.188 0.172 0.058 0.149 0.140 0.143 0.176 0.140 0.122 0.129 0.053 0.126 0.021 0.108 0.082 0.108 0.096 0.102 0.093 0.095 0.072 0.089 0.099 0.087 0.062 0.077 ..

Rank 1998 1985 45 40 46 35 47 53 48 56 49 65 50 50 51 38 52 57 53 46 54 41 55 49 56 47 57 67 58 48 59 39 60 55 61 58 62 64 63 60 64 66 65 52 66 51 67 69 68 .. 69 71 70 62 71 63 72 .. 73 74 74 54 75 72 76 59 77 68 78 75 79 79 80 70 81 78 82 73 83 77 84 80 85 .. 86 76 87 ..

Economy Tunisia Venezuela Chile Guatemala Indonesia India Zimbabwe El Salvador Morocco Saudi Arabia Colombia Mauritius Egypt Peru Oman Pakistan Ecuador Kenya Jordan Honduras Jamaica Panama Bolivia Albania Sri Lanka Nicaragua Paraguay Mozambique Bangladesh Algeria Cameroon Senegal Zambia Nigeria Nepal Tanzania, United Republic of Malawi Madagascar Central African Republic Uganda Yemen Ghana Ethiopia

Index value 1998 1985 0.068 0.064 0.060 0.085 0.056 0.030 0.056 0.028 0.054 0.012 0.054 0.034 0.052 0.071 0.051 0.027 0.048 0.038 0.047 0.063 0.041 0.035 0.041 0.037 0.038 0.012 0.035 0.037 0.032 0.069 0.031 0.028 0.025 0.025 0.025 0.013 0.024 0.022 0.023 0.012 0.022 0.032 0.022 0.032 0.021 0.009 0.021 .. 0.017 0.008 0.017 0.020 0.015 0.013 0.013 .. 0.011 0.008 0.009 0.029 0.008 0.008 0.008 0.023 0.007 0.010 0.006 0.006 0.006 0.001 0.005 0.009 0.003 0.003 0.003 0.008 0.003 0.003 0.003 0.001 0.001 .. 0.001 0.006 0.000 ..

Source: UNIDO Scoreboard database (see technical annex).

Economies losing rank significantly include Hong Kong Special Administrative Region (SAR) of China, Bahrain, Poland, New Zealand and Canada among the top 40, and Oman, Algeria, Senegal, Zimbabwe, Saudi Arabia, Jamaica and Venezuela among the bottom 40.

Industrialized and transition economies

Switzerland led the industrialized countries in both years (table 3.2). In 1985 it also led the world, but Singapore had overtaken it by 1998. Ireland had the biggest jump in rank in

the group, improving its position from 15th to third. The major industrial powersJapan, Germany and the United Statesnear the top of the scale, saw a deterioration in their positions as a result of the improved positions of Singapore and Ireland. Of the top 20, all but five are mature industrialized countries. The exceptions, apart from Ireland and Israel, are three East Asian Tigersthe Republic of Korea, Singapore and Taiwan Province of China. Among transition economies the leaders are the Czech Republic, Hungary and Slovenia. Hungary, the only one of these three for which 1985 data are available, improved its

Benchmarking industrial performance

43

Figure 3.1 Changes in ranking by the competitive industrial performance index between 1985 and 1998 1998 rank
0

Top 40 economies Singapore Ireland United States Switzerland

10

15

Japan Germany Sweden Finland United Kingdom Belgium France Denmark Austria Taiwan Province of China Netherlands Republic of Korea Spain Israel Italy Norway Canada

20

Malaysia Philippines Hungary Mexico Portugal

25

Australia Hong Kong SAR New Zealand Poland

30

Thailand
35

Brazil Argentina Turkey Romania South Africa Greece Bahrain

China
40

Costa Rica

45 65 60 55 50 45 40 35 30 25 20 15 10 5 0

1985 rank 1998 rank

40

Bottom 40 economies Uruguay Guatemala Chile India Colombia Egypt Mauritius Pakistan Kenya Honduras Bolivia Sri Lanka Ecuador Jordan Nicaragua Paraguay Cameroon Nepal Zambia Nigeria Malawi Senegal Algeria Jamaica Panama Oman Peru Tunisia Venezuela

Indonesia
50

El Salvador

Morocco

Zimbabwe Saudi Arabia

60

70

Bangladesh

80

Uganda
90 85 80

United Republic of Tanzania Madagascar Central African Republic Ghana

75

70

65

60

55

50

45

40

35

30

1985 rank
Source: UNIDO Scoreboard database (see technical annex). Note: Economies on the line had the same rank in both years, while those above the line had a higher rank in 1998 than in 1985, and those below it a lower one.

44 Industrial Development Report 2002/2003

Table 3.2 Ranking of economies by the competitive industrial performance index, by region or country group, 1985 and 1998

Region or country group Industrialized countries

Rank 1998 1985 2 3 4 5 6 7 8 9 10 11 12 13 14 16 17 19 20 21 26 29 31 40 1 15 2 3 5 4 7 8 12 10 11 13 14 9 16 21 20 17 26 23 24 33 .. 34 .. 25 37 .. .. .. 28 27 29 44 42 35 53 56 57 49 48 58 66 52 51 69 62 63

Economy Switzerland Ireland Japan Germany United States Sweden Finland Belgium United Kingdom France Austria Denmark Netherlands Canada Italy Spain Israel Norway Portugal Australia New Zealand Greece Czech Republic Hungary Slovenia Poland Romania Russian Federation Albania Albania Mexico Brazil Argentina Costa Rica Uruguay Venezuela Chile Guatemala El Salvador Colombia Peru Ecuador Honduras Jamaica Panama Bolivia Nicaragua Paraguay

Transition economies 24 27 28 34 41 44 68 68 Latin America and the Caribbean 23 33 35 36 43 46 47 48 52 55 58 61 64 65 66 67 70 71

Region or Rank country group 1998 1985 Economy East Asia and the Pacific 1 6 Singapore 15 19 Taiwan Province of China 18 22 Korea, Republic of 22 30 Malaysia 25 45 Philippines 30 18 Hong Kong SAR 32 43 Thailand 37 61 China 49 65 Indonesia South Asia 50 50 India 60 55 Pakistan 69 71 Sri Lanka 73 74 Bangladesh 79 79 Nepal Sub-Saharan Africa 39 32 South Africa 51 38 Zimbabwe 56 47 Mauritius 62 64 Kenya 72 .. Mozambique 75 72 Cameroon 76 59 Senegal 77 68 Zambia 78 75 Nigeria 80 70 Tanzania, United Republic of 81 78 Malawi 82 73 Madagascar 83 77 Central African Republic 84 80 Uganda 86 76 Ghana 87 .. Ethiopia Middle East and North Africa and Turkey 38 36 Turkey 42 31 Bahrain 45 40 Tunisia 53 46 Morocco 54 41 Saudi Arabia 57 67 Egypt 59 39 Oman 63 60 Jordan 74 54 Algeria 85 .. Yemen

Source: UNIDO Scoreboard database (see technical annex).

Benchmarking industrial performance

45

position by seven places, while those of Poland and Romania deteriorated. The lowest ranked transition economy is Albania, at 68th in the world.

56th),3 both with large declines. Most other African countries congregate at the bottom. Of the 16 Sub-Saharan African countries, only Kenya improved its rank. Of the 20 lowest ranking countries, 12 are in Sub-Saharan Africa.

Developing economies Least developed countries

Among the developing regions East Asia led in both years and also had the greatest improvement in ranks over the period. There is wide dispersion within the region, however. Singapore led East Asia in both years, ranking sixth and first. Its nearest rival was Taiwan Province of China, ranking 13 places lower in 1985 and 14 lower in 1998followed by the Republic of Korea, which joined the top 20 in 1998. Most of the economies that improved their ranks are in East Asia: China (which improved its rank by 24 places), the Philippines (20), Indonesia (16) and Thailand (11). Hong Kong SAR is the only one in the region whose position worsened, by 12 places. In Latin America and the Caribbean, Mexico, Costa Rica, Chile and Guatemala improved their positions while those of other countries deteriorated. Of the 18 countries in this region, 7 improved and 11 worsened over the period. In the Middle East and North Africa only one country improved its ranking, Egypt, by 10 places. Others in the region had the steepest declines in the sample: Algeria and Oman (by 20 places each) and Saudi Arabia (13). In South Asia, India (ranked 50th) leads, followed by Pakistan, Sri Lanka, Bangladesh and Nepal. Only Sri Lanka and Bangladesh improved their rankings, and only slightly. In Sub-Saharan Africa the leader by far is South Africa (at 39th), followed by Zimbabwe and Mauritius (at 51st and Not surprisingly, the 12 least developed countries in the sample are near the bottom of the world rankings, along with Algeria (which suffered a massive deterioration in performance over the period) and three other African countries, Cameroon, Nigeria and Ghana. Among the least developed countries Bangladesh and Nepal had stable rankings over the period (because of a rise in their CIP index values; table 3.3). But for most other countries the picture is less sanguine. Senegal recorded a large fall (one of the largest in the sample), as did Madagascar, the United Republic of Tanzania and Zambia. Since the sample coverage differs between 1985 and 1998, the changes reflect both shifts in relative industrial performance and the entry of new countries. Bangladesh and Nepal maintained their ranks because of an upgrading of the technological structure of their exports, while Senegal lost rank largely because of a downgrading of its export structure.

Basic indicators of industrial performance a useful source of information

Analysis of country rankings for each basic indicator in the CIP index provides useful information. As each new indicator is added to the baseper capita MVAthe position of a country shifts. The ranks at each stage illustrate different aspects

Table 3.3 Ranking of least developed countries by the competitive industrial performance index, 1985 and 1998

Overall rank 59 68 70 73 74 77 78 79 80

1985 Economy Senegal Zambia Tanzania, United Republic of Madagascar Bangladesh Central African Republic Malawi Nepal Uganda

CIP value 0.023 0.010 0.009 0.008 0.008 0.003 0.003 0.001 0.001

Overall rank 72 73 76 77 79 80 81 82 83 84 85 87

1998 Economy CIP value Mozambique 0.013 Bangladesh 0.011 Senegal 0.008 Zambia 0.007 Nepal 0.006 Tanzania, United Republic of 0.005 Malawi 0.003 Madagascar 0.003 Central African Republic 0.003 Uganda 0.003 Yemen 0.001 Ethiopia 0.000

Source: Annex tables A3.1 and A3.2. Note: The 1985 sample is smaller than the 1998 sample because no 1985 data are available for Mozambique, Yemen and Ethiopia.

46 Industrial Development Report 2002/2003

of competitive industrial performance, as examples from the 1998 index show. Take the largest industrialized economy (by total value of production and exports), the United States. It ranks seventh in per capita MVA, below Switzerland and Japanand below even Ireland and Singapore (the use of population to normalize indicators works against large countries but remains a good way to adjust for country size). When per capita exports are added, the United States falls to 13th, reflecting the stronger pull of its domestic market relative to that of other highly industrialized competitors. But its rank improves when the indicators of complexity of MVA and exports are added, showing its relative technological strength. In contrast, New Zealand loses rank significantly: while it ranks 21st in MVA per capita, it ends up 31st, mainly because its manufactured exports are far less technology intensive than its production. Now consider some developing countries. Singapore starts at fourth but rises to first place because of its high exports per capita and large shares of high-tech products in production and exports. The Philippines experiences more volatility in its position: it starts at 60th in MVA per capita and ends up at 25th in the composite index because of the large share of hightech products in its exports (the second largest share in the world, after Japans). Similarly, China starts at 55th and finishes at 37th, again mainly because of the technology intensity of its manufactured exports. Zimbabwe shows a comparable pattern, rising from 69th to 51st. Chile provides a counterexample. Its final rank (47th) is 10 places lower than its starting rank because its manufactured exports are far less technologically sophisticated than its MVA, pulling it down by 11 places. So, rankings by the basic indicators in the CIP index can provide interesting insights into comparative national perform-

ance. The main cause of variation in ranks is the technological structure of exports. When this variable is introduced, the ranks of 16 countries change by 10 or more places, with seven improving and nine worsening. Adding the other two variables causes far smaller shifts in ranks. The countries whose positions improved the most with the addition of the export structure variable are the Philippines, India, Zimbabwe, China and Kenya. Several of these (India, Zimbabwe and Kenya) do not have particularly technology-intensive exportsthey move up the scale because their exports are more complex than those of other countries near them in the rankings. Those whose positions deteriorated the most with the export structure variable are Mauritius, Jamaica, Bahrain and Saudi Arabia, followed closely by Chile, Peru and Algeria. Adding the technological structure of MVA or exports had little effect on the comparative industrial performance of the least developed countries as a group: these countries congregate near the bottom of the ranking even if only MVA per capita is used as the index (annex tables A3.1 and A3.2). Including these variables has some effects on individual countries, of course, but the variables are not biased against least developed countries. For example, adding the export structure to MVA and exports per capita leads to a significant deterioration in the ranks of Senegal and Yemen in 1998, but to a significant improvement for Mozambique, Nepal and the United Republic of Tanzania. The four components of the CIP index are highly correlated with one another. The strongest correlation is between MVA per capita and the technological structure of MVA: the higher the level of industrialization, the more complex is the structure of production (table 3.4). Similarly, the higher the level of exports per capita, the more sophisticated is the MVA structure, and the higher the level of industrialization, the larger are per capita exports. The weakest (though still statistically significant) correlation is between export intensity and export structure.

Table 3.4 Correlation between components of the competitive industrial performance index, 1998

Component MVA per capita Manufactured exports per capita Share of medium- and high-tech activities in MVA Share of medium- and high-tech products in manufactured exports
Source: UNIDO Scoreboard database (see technical annex). ** Significant at the 1 percent level.

MVA per capita 1.000 0.717** 0.968** 0.732**

Manufactured exports per capita 1.000 0.752** 0.574**

Share of mediumShare of mediumand high-tech and high-tech products in activities in MVA manufactured exports

1.000 0.719** 1.000

Benchmarking industrial performance

47

Winners and losers in CIP ranks

Which economies experienced exceptionally large shifts in CIP ranks between 1985 and 1998? Six "winners" gained 10 or more places, and 12 "losers" lost 10 or more (figure 3.2). The main reasons for this volatility are shifts in relative MVA per capita and in the complexity of MVA and exports. Underlying factors may include political turmoil, natural disasters or declining terms of trade. Since many of these are external to industry, they are not germane to the general analytical and policy discussion here. For this report the most important factorone that can lead to great volatility in rankings and is more relevant for analysis and policyis rapid upgrading of technological structures. For all the winners except Egypt (where MVA rose rapidly), the entry into high-tech integrated global production systems has been the most dynamic new element in industry. That is not to say that this strategy is the only way to upgrade technology in developing countries. In fact, it may not even be the best
Figure 3.2 Winners and losers in competitive insustrial performance rankings between 1985 and 1998 China Philippines Indonesia Thailand Ireland Egypt United Republic of Tanzania Peru Ghana Venezuela Hong Kong SAR Zimbabwe Saudi Arabia Jamaica Panama Senegal Oman Algeria 20 15 10 5 0 5 10 15 20 25

and most sustainable way: it introduces new production technologies and raises exports, but it may not develop or deepen local capabilities if the country fails to move beyond final assembly of high-tech products. Latecomers that have managed to build genuine technological capabilities in complex activities have generally done it through a slower, costlier and riskier process of advancing from assembly to real manufacturing, and from there to local design and development. They have done much of this without investment by transnational corporationseven restricting foreign entry to encourage the development of deeper capabilities in local enterprises.

Clustering countries to discern patterns in industrial performance

One useful way to analyze competitive industrial performance is through cluster analysis (for details see the technical annex). The technique allows the number of clusters to be specified in advance.

Industrial performance over time

Cluster analysis is first used here to group countries based on similarities in the behaviour of three variables: the CIP index values in 1985 and 1998 and the change in the index value between those two years. The analysis captures how countries cluster according to the level of industrial performance and the change in that performance. The analysis is conducted separately for industrialized and transition economies and for developing economies. For industrialized and transition economies four clusters were chosen, covering the 25 countries with data for both 1985 and 1998. The first cluster has only one country, Ireland (table 3.5). Ireland stands out in the group because of its high CIP index value in 1998 and the large change in this value since 1985 no other economy has a similar combination. In the second cluster are the industrial giants, Japan, Germany and the United States, and two smaller highly industrialized countries, Sweden and Switzerland. These countries have a high and relatively stable average CIP index value in both years (though the value showed a small decline from 1985 to 1998). In the third cluster are most other Western European countries along with Canada and Israel; these have a lower average CIP index value in both years (but the value rose slightly). In the fourth cluster are the weaker industrialized economies in Europe (Greece and Portugal), all the transition economies for which there are data for both years and Australia and New Zealand. This cluster has even lower CIP index values but a slightly larger increase between 1985 and 1998.

Places lost or gained

Source: UNIDO Scoreboard database (see technical annex).

48 Industrial Development Report 2002/2003

Table 3.5 Cluster analysis of competitive industrial performance for industrialized and selected transition economies, 1985 and 1998

CIP 1998 (mean value) CIP 1985 (mean value)

Cluster 1 0.74 0.38 Ireland

Cluster 2 0.64 0.68 Germany Japan Sweden Switzerland United States

Cluster 3 0.42 0.41 Austria Belgium Canada Denmark Finland France Israel Italy Netherlands Norway Spain United Kingdom

Cluster 4 0.17 0.14 Australia Greece Hungary New Zealand Poland Portugal Romania

Source: See technical annex. Note: The table excludes Albania, the Czech Republic, the Russian Federation and Slovenia, for which no 1985 data are available.

Now let us look at the results of a cluster analysis for developing economies (table 3.6). Five clusters are specified to take account of the larger number of economies in this group (52). Interestingly, the two smaller Tiger economies in East Asia, Hong Kong SAR and Singapore, each form an individual cluster. Each has a performance that is quite different from that of the rest of the developing world and from that of each other. Singapore has the highest CIP index value of all the clusters in both years, and the value rises over time. The Republic of Korea and Taiwan Province of China form another cluster, with moderately high and rising CIP index values. The fourth cluster includes 11 countries, with low to medium CIP index values.4 This group includes several major industrializing economies (Argentina, Brazil, China, Mexico, South Africa and Turkey) as well as the fast-growing new Tigers in

Asia (Malaysia, the Philippines and Thailand) and one in Latin America and the Caribbean (Costa Rica). There is also an outlier, Bahrain. The fifth cluster contains the remaining countries, with very low and stable CIP index values. These countries have weak or stagnant industrial and export values and structures. India falls into this group despite its large industrial sector.

Evolution of industrial and export structures

Cluster analysis can also spotlight differences in the evolution of countries industrial and export structuresthat is, in the shift of their MVA and export structures up the technological scale. First consider four clusters for the industrialized and transition economies (figure 3.3):

Table 3.6 Cluster analysis of competitive industrial performance for developing economies, 1985 and 1998

CIP 1998 (mean value) CIP 1985 (mean value)

Cluster 1 0.88 0.59 Singapore

Cluster 2 0.39 0.27 Korea, Republic of Taiwan Province of China

Cluster 3 0.20 0.32 Hong Kong SAR

Cluster 4 0.16 0.09 Argentina Bahrain Brazil China Costa Rica Malaysia Mexico Philippines South Africa Thailand Turkey

Cluster 5 0.03 0.03 Other developing countries

Source: See technical annex. Note: The table excludes Ethiopia and Yemen, for which no 1985 data are available.

Benchmarking industrial performance

49

 The leading cluster (Germany, Japan, Sweden, Switzerland, the United States), which has the highest MVA per capita and the most technology-intensive production and exports, had a significant rise in MVA per capita and some improvement in the technological structure of both MVA and exports between 1985 and 1998. The second cluster, with Belgium, Ireland and the Netherlands, had a rapid rise in MVA per capita and a rapid upgrading of industrial and export structures. The third cluster, with eight mature industrialized countries, resembles the first cluster but has a smaller average MVA per capita. The last cluster had the lowest average MVA per capita and the least technology-intensive MVA and exports in 1998. But it also had a rapid improvement in the techno-

logical composition of exports (and a somewhat less rapid one for MVA). Now consider the developing economies, with seven clusters (figure 3.4): Singapore and Hong Kong SAR again form clusters of their own, with strong differences in rates of upgrading and growth. Singapore showed rapid improvement in MVA per capita and technological sophistication, while Hong Kong SAR showed practically no change in MVA per capita and export structure, though it had some improvement in MVA structure. The Republic of Korea, Malaysia and Taiwan Province of China form a dynamic cluster, with high rates of growth in MVA per capita and rapid rises in the technological complexity of production and exports.

Figure 3.3 Cluster analysis of technological evolution of industry in industrialized and transition economies, 19851998 Average MVA per capita, 1985 Average MVA per capita, 1998 Share of medium- and high-tech products in manufactured exports (percent) 80

70

60

Austria, Canada, Denmark, Finland, France, Italy, Norway, United Kingdom $4,480 $3,130 Germany, Japan, Sweden, Switzerland, United States $6,372

50

Australia, Greece, Hungary, Israel, New Zealand, Poland, Portugal, Spain $1,950

40 $2,054 30 $1,556 Belgium, Ireland, Netherlands $5,147

20 $981 10

40

45

50

55

60

65

70

Share of medium- and high-tech activities in MVA (percent)

Source: UNIDO Scoreboard database (see technical annex).

50 Industrial Development Report 2002/2003

Figure 3.4 Cluster analysis of technological evolution of industry in developing economies, 19851998 Average MVA per capita, 1985 Average MVA per capita, 1998 Share of medium- and high-tech products in manufactured exports (percent) 90 80 70 Bolivia, Cameroon, 60 Central African Republic, Ecuador, Ghana, Honduras, Jordan, 50 Madagascar, Mauritius, Nicaragua, Oman, Panama, Paraguay, 40 Sri Lanka, Uganda, Uruguay $242 30 20 10 0 10 Other developing economies $236 Brazil, China, India, Saudi Arabia, South Africa $485 Argentina, Costa Rica Mexico, Philippines, Thailand, Turkey $726 Republic of Korea, Taiwan Province of China, Malaysia $2,131 Singapore $6,178

Hong Kong SAR $1,411 $765 $1,322

1,618

$273 $158

$313

$159

10

20

30

40

50

60

70

80

90

100

Share of medium- and high-tech activities in MVA (percent)

Source: UNIDO Scoreboard database (see technical annex).

Another dynamic cluster has two new Asian Tigers, the Philippines and Thailand, along with Argentina, Costa Rica, Mexico and Turkey. The next cluster contains three industrial giants in the developing world, Brazil, China and India, along with Saudi Arabia and South Africa. With medium to low MVA per capita, these countries are upgrading their export structures but less so their industrial structures.5 The last two clusters, with 43 countries, have much lower MVA per capita and, on average, negligible upgrading of MVA and export structures. These are the laggards. The results of cluster analysis clearly illustrate patterns of industrial performance. But the technique glosses over national differences in particular aspects of performance. More detailed analysis is called for in making specific country comparisons. What this analysis does show, however, is how

widely dispersed developing countries areand how they are diverging rather than converging.

Explaining export performance

Enterprises are much more successful at exporting from some countries than from others. A large literature explores the role of trade and industrial policies in export success. Rather than add to this literature, here the intention is to focus on aspects on which the CIP index throws some light. The identities of the leading and lagging exporters in 1998 as assessed by manufactured exports as a share of MVA and medium- and high-tech exports as a share of medium- and high-tech MVAsuggest an interplay of several factors in explaining export performance (table 3.7). The main factors are the size of the domestic economy, competitive capabili-

Benchmarking industrial performance

51

ties, the nature of the trade regime and the intensity of participation in integrated global production systems. There are several large industrialized economies in the group with low export propensities. Since all these countries have open trade and investment regimes, the main explanation must be the size of the domestic market: large economies have an inherent propensity to export less relative to production than do small economies.6 For Japan and the United States, countries with open regimes and strong technological capabilities, market size is clearly the main explanation of the low export propensities (in absolute terms, these are among the worlds leading exporters). For other industrialized countries other factors may also matter. For example, the low export ratio for Australia, a medium-size economy, may reflect competitive weaknesses in industry. Size also matters in the developing world, but trade policies and capabilities probably play a larger role. Thus weak export performance in large countries, such as India, reflects not just the large market size but also the legacy of strong inwardlooking policies, the small presence of transnational corporations and competitive weaknesses in manufacturing. Brazil has a relatively open economy and a large presence of transnational corporations. Nevertheless, it had a weak showing, perhaps because transnational corporations do little high-tech exporting and local enterprises (apart from obvious exceptions like the aircraft producer Embraer) lack strong competitive capabilities.7 As the data on Brazils export structure show, high-tech products play a surprisingly

small role in the developing worlds second largest industrial power. Now consider export performance in medium- and high-tech products. In the top 15 performers foreign direct investment clearly plays an important role. Most of the highly export-oriented countries have a strong presence of transnational corporations. This feature is combined with strong domestic technological capabilities in some countries (Belgium, the Netherlands, Singapore) and with modest domestic technological capabilities in others (Costa Rica, Malaysia, the Philippines). The strong showing by this second group shows that it is possible for newcomers without a technological base to upgrade their industrial structure and performance by leveraging participation in integrated global production systems. Even in relatively large economies, such as Mexico, such participation can offset the pull of the home market. China, despite its size, does not appear among the bottom 15 for medium- and high-tech exports as it does for total manufactured exports. Its medium- and high-tech sector is apparently far more competitive than the rest of its manufacturing. In China, as in Mexico, foreign enterprises in special export zones contribute a significant share of medium- and hightech exports. Chile is among the laggards in technologyintensive exports despite its open (and fairly small) economy, strong presence of transnational corporations and high skill levels. Its weak performance reflects in part its comparative advantage in resource-based activities, but also in part its inability to enter global production systems.

Table 3.7 Leading and lagging exporters, 1998

Manufactured exports as a percentage of total MVA Top 15 exporters Bottom 15 exporters Economy Value Economy Value Singapore 529.5 Russian Federation 54.0 Belgium 338.5 Turkey 51.9 Malaysia 371.4 Morocco 51.0 Hong Kong SAR 245.2 China 47.2 Netherlands 225.0 Australia 46.3 Ireland 222.3 Uruguay 41.9 Hungary 213.0 Japan 41.4 Philippines 197.2 India 40.5 Mauritius 187.0 United States 38.4 Costa Rica 174.4 Colombia 32.3 Czech Republic 159.2 Argentina 26.5 Sweden 158.6 Brazil 25.7 Canada 154.3 Ecuador 22.0 Taiwan Province of China 144.2 Peru 15.5 Denmark 143.4 Egypt 11.2

Medium- and high-tech exports as a percentage of medium- and high-tech MVA Top 15 exporters Bottom 15 exporters Economy Value Economy Value Singapore 566.4 Romania 66.6 Philippines 524.1 Russian Federation 65.9 Costa Rica 478.5 South Africa 65.8 Czech Republic 462.6 Greece 63.4 Hungary 445.7 Japan 56.5 Belgium 385.9 Venezuela 47.7 Malaysia 382.5 New Zealand 43.4 Mexico 380.4 United States 39.3 Hong Kong SAR 296.4 Turkey 38.5 Netherlands 291.7 Argentina 38.1 Poland 232.7 Brazil 35.4 Portugal 221.4 Colombia 31.4 Slovenia 191.3 Chile 28.5 Austria 186.4 Australia 23.7 Thailand 179.8 India 23.5

Source: UNIDO Scoreboard database (see technical annex). Note: The values can be well over 100 because the numerator (exports) is in terms of total value while the denominator (manufacturing value added, or MVA) is only the value added. The table includes only economies with manufactured exports of more than $1 billion in 1998.

52 Industrial Development Report 2002/2003

Linking the structure of exports to their growth

Does the technological structure of a countrys exports affect the growth of that countrys manufactured exports? Econometric analysis suggests that it does. The analysis, covering the 80 countries for which data are available for both 1985 and 1998, regressed the annual growth of manufactured exports on two measures of structure: the technological composition in the base year and the change in structure over time. The first measure is the share of medium- and hightech products in each countrys manufactured exports in 1985. The second is the share of medium- and high-tech exports in 1998 minus the share in 1985. The initial technological structure captures the relative performance of countries in the dynamic segments of trade, while changes in the technological structure capture the effects of technological upgrading. Countries with advanced initial structures are expected to have slower upgrading. The sample was divided into two groups: large exporters, with manufactured exports exceeding $1 billion in 1985 (46 countries), and others (34). This sorting differentiates countries by level of industrial development, based on the premise that the technological structures of large exporters will differ significantly from those of small exporters and have different effects on performance in exports. The means of the variables for the two groups support this premise. The average share of medium- and high-tech products for large exporters in 1985 was eight times that for small exporters. Both groups upgraded the technological structure of their exports over time, but the change for large exporters was smaller.

The regression results suggest that the initial technological structure and subsequent changes affect export growth in the expected direction for large as well as small exporters (annex table A3.3). The equations explain performance better for the large exporters because technological factors are more important for the export performance of relatively advanced economies. But the technological structure of exports also matters for small exportersmore so, in fact, than do improvements in the structure. For large exporters, by contrast, changes over time are more important than the initial technological structure. Even though these exporters are technologically sophisticated in the initial periodand thus well positioned to benefit from the rapid growth of technologyintensive exportsthey are able to accelerate export growth through further technological upgrading.

Linking industrial performance and environmental sustainability

The relationship between industrial performance and environmental sustainability is of growing concern to governments and among the main areas of activity for UNIDO. Although calculating comprehensive national indices of environmental performance is difficult because of data limitations, it is still important to take this aspect of industrial development into account. Yet little cross-country research has been done on the links between industrialization and environmental degradation. There are several possible reasons for this. Appropriate cross-country environmental data are scarce, and isolating the effects of industrial activity on the environment from those of other factors is very difficult. It is also difficult to analyze meaningfully the environmental data that are available. Take industrial emissions of carbon dioxide (CO2). Low emission rates could mean that a country has tackled industrial pollution effectively or that it has low levels of industrial activity. Most environmental indicators do not make such distinctions, so causal connections are difficult to establish. One measure of environmental performance for which data are available for all countries in the sample is CO2 emissions in 1998. An analysis was done to find the correlation between national CO2 emissions, normalized by population and by GDP, and the components of the CIP index for 1998 (table 3.8).8 The two measures of emissions give very different results. CO2 emissions normalized by population are positively correlated (at a 1 percent level of significance) with all components of the CIP index: not surprisingly, the more industrialized a country, the higher are its emissions per capita. CO2 emissions normalized by GDP are negatively (and mostly significantly) correlated with the components of the CIP index. This suggests, again not sur-

Table 3.8 Correlation between industrial performance measures and carbon dioxide emissions, 1998

Variable CIP 1998 MVA per capita Manufactured exports per capita Share of medium- and high-tech activities in MVA Share of medium- and high-tech products in manufactured exports

CO2 emissions per capita 0.655** 0.630** 0.570** 0.495**

CO2 emissions per unit of GDP 0.289** 0.341** 0.07 0.228*

0.530**

0.165

Source: UNIDO Scoreboard database (see technical annex). * Significant at the 5 percent level. ** Significant at the 1 percent level.

Benchmarking industrial performance

53

Table 3.9 Biggest and smallest polluters, 1998

CIP rank 1 42 6 29 21 16 54 44 24 5 9 13 20 8 4

Top 15 by CO2 emissions CIP Per unit Per capita rank of GDP Singapore 44 Russian Federation Bahrain 37 China United States 41 Romania Australia 78 Nigeria Norway 51 Zimbabwe Canada 34 Poland Saudi Arabia 42 Bahrain Russian Federation 50 India Czech Republic 24 Czech Republic Germany 46 Venezuela Belgium 65 Jamaica Denmark 74 Algeria Israel 39 South Africa Finland 54 Saudi Arabia Japan 57 Egypt

CIP rank 72 84 87 81 83 79 80 82 73 86 62 77 69 76 75

Bottom 15 by CO2 emissions CIP Per unit Per capita rank of GDP Mozambique 2 Switzerland Uganda 84 Uganda Ethiopia 7 Sweden Malawi 83 Central African Republic Central African Republic 30 Hong Kong SAR Nepal 11 France Tanzania, United Republic of 4 Japan Madagascar 12 Austria Bangladesh 13 Denmark Ghana 43 Uruguay Kenya 14 Netherlands Zambia 5 Germany Sri Lanka 79 Nepal Senegal 33 Brazil Cameroon 17 Italy

Source: Calculations based on CO2 emissions data from World Bank (2001b).

Figure 3.5 Regression of competitive industrial performance index values on carbon dioxide emissions (log model), 1998 1998 CIP index value 0.9 Average CO2 emissions Singapore

prisingly, that more industrialized economies are more efficient at dealing with emissions relative to their income. A comparison of the top and bottom 15 pollutersranked by CO2 emissions normalized by both population and GDPprovides some interesting detail (table 3.9). Singapore, the United States, Germany, Belgium, Denmark, Finland and Japan are among the largest polluters in per capita terms because of their large industrial basesthese countries rank among the top 20 on the CIP index. Countries with the lowest CO2 emissions per capita, many of them in Africa and South Asia, rank at the bottom on the CIP index. When emissions are normalized by GDP, the identities of the top and bottom 15 polluters change dramatically. The best industrial performers are not among the largest polluters. Most transition economies have very high CO2 emissions relative to their industrial base, with the Russian Federation ranking highest on this measure. The two developing giants, China and India, are among the 15 biggest polluters. And Switzerland is the "cleanest" country in the world. A regression of CIP index values on emissions (normalized by GDP) shows that there is a clear negative relationship between the two (figure 3.5). But the regression explains only 15 percent of the variation in emissionsclearly, other factors are also important in determining emissions (though these cannot be investigated here). Almost all countries with strong performance on the CIP index appear to be relatively "clean", while the largest polluters generally score below the average on the CIP index. As noted, transition economies (the Russian Federation, Romania and Poland) have particularly high emissions relative to the size of their economies.

0.8 Switzerland Ireland 0.7 Japan Germany 0.6 Sweden Finland Belgium United States

0.5

United Kingdom 0.4 Canada Republic of Korea

0.3

Malaysia Czech Republic Hungary Average CIP index value South Africa Poland China Romania Bahrain Russian Federation Zimbabwe 1 0 1 2 3 4 5 6

0.2

0.1

CO2 emissions, industrial (kilogram per international dollar of GDP)

Sources: UNIDO Scoreboard database (see technical annex) and World Bank (2001b). Note: International dollars are U.S. dollars adjusted for purchasing power parity.

54 Industrial Development Report 2002/2003

Interestingly, a comparison with emissions in 1985 (not shown in the figure) shows that countries that have had large jumps in CIP ranks (China, Hungary, Malaysia and Thailand) have also had substantial increases in CO2 emissions relative to GDP. This suggests that rapid industrial growth can raise the propensity to pollute, at least until some level of industrial maturity is attained. This inverted-U-shaped relationship between emissions and industrialization needs to be investigated in detail.

3. Mauritiuss export profile dominated by garments, while the CIP index emphasizes only medium- and high-tech activities. 4. Because the cluster analysis uses CIP scores rather than ranks, some countries appear to be wrongly placed. While Argentina and Brazil lost rank over time, they are clustered in a group with rising performance scores on average. Both these countries in fact raised their CIP scoresArgentina from 0.122 to 0.140 and Brazil from 0.140 to 0.149but fell in the rankings because other countries improved their scores faster. 5. All countries in a cluster need not perform equally well in all aspects. China upgraded its export structure relatively rapidly in comparison with the others in its group, but had similar MVA and exports per capita. 6. There is a presumption that large economies, because they allow domestic economies of scale and scope, will tend to have deeper industrial sectors (with more medium- and high-tech production) and a larger share of technology-intensive exports. The sample data suggest that this presumption is true. The size of the economy (GDP) is correlated positively (and significantly at the 1 percent level) with the share of medium- and high-tech products in MVA (0.39) and exports (0.44) for the sample (coefficients are shown here for 1998, but those for 1985 are similar). The correlations are stronger for developing than for industrialized countries. But size has a stronger correlation with MVA structure (0.60) in developing countries than with export structure (0.49). Size has no statistical correlation with the growth of MVA or exports in industrialized countries, but it has a positive correlation with the growth of exports (but not MVA) in developing countries. 7. Brazil had a weak showing despite the growth of automobile exports by transnational corporations; clearly the amounts are insufficient to offset low export activity in other industries. 8. Manufacturing value added was also used to normalize emissions. The results were essentially the same as those for GDP.

Notes
1. Benchmarking has been widely used by enterprises as a tool for evaluating performance, learning from best practices and understanding how best practices are achieved. In recent years government agencies and other institutions (such as universities) have also discovered its value. Benchmarking has spread beyond Europe and the United States to the developing world, where many countries are conducting competitiveness analyses based on benchmarking of one another and global leaders. 2. The UNIDO Scoreboard complements existing competitiveness indices. The best known are the World Economic Forums current competitiveness and growth competitiveness indices, in the Global Competitiveness Report (http://www.weforum.org), and the International Institute for Management Developments world competitiveness scoreboard, in the World Competitiveness Yearbook (http://www.imd.ch/wcy/wcy.cfm). (For an analysis of these indices see Lall 2001b.) While the UNIDO Scoreboard focuses on manufacturing and a small number of structural variables, other indices use large numbers of variables and rely heavily on qualitative responses. Moreover, the UNIDO Scoreboard is modular, making it possible to easily add new variables.

Benchmarking industrial performance

55

4
I

Benchmarking the drivers of industrial performance

bilities. It is often uneconomical for transnational corporations to set up R&D facilities and build appropriate technological capabilities in host countrieseven in those with the industrial sophistication that makes such efforts feasible. That is why Japan, the Republic of Korea and Taiwan Province of China, with strong technological ambitions, restricted foreign direct investment at critical stages of industrialization, when they sought to develop local innovative capabilities. They forced domestic enterprises to license or copy foreign technologies and invest in the capabilities to absorb and improve on them.2 For many developing countries, however, there is no real conflict between domestic R&D and foreign direct investment. Domestic R&D (formal domestic R&D spending by productive enterprises) is generally low or negligible, so foreign direct investment is often one of the best ways of gaining access to new technologies, information and skills. Nor is there a conflict between R&D and foreign direct investment for industrialized countries: strong local capabilities attract R&D by transnational corporations and benefit from it. It is for newly industrializing economies, in the middle, that there are possible non-complementarities. Even here the conflict is real only for countries with the skills and incentives to build genuine R&D capabilities by restricting foreign direct investment. Many countries that pursued this strategy were unable to build efficient innovative capabilities; instead, they simply suffered from growing technological gaps. Nevertheless, some did build advanced technological and innovative capabilities by restricting foreign direct investment. For these countries, treating R&D and foreign direct investment as complements biases the benchmarking, making their average capabilities appear lower (because of their relatively weak performance in foreign direct investment). Even in these countries, however, the conflict between the two occurs in specific phases of industrial developmentand diminishes thereafter. Taking account of such complexities in building the benchmarks is difficult, since there is no a priori way of distinguishing one

NDUSTRIAL PERFORMANCE IS THE OUTCOME OF MANY SOCIAL, political and economic factors interacting in complex and dynamic ways. These interactions are often specific to each country, reflecting its history, culture, legal system, legal and institutional framework, social capital, political and social conditions and ways of doing business. Industrial performance also reflects macroeconomic policies as well as policies relating to technology and education. These factors need not be only national: the outside world can strongly affect industrial activity and performance. With globalization, the role of external factors and rules is growing rapidly.

It is not possible to benchmark countries on all these factors. The purpose here is more modest: to benchmark countries on their key structural variablesreferred to here as drivers using available data. The drivers chosen for benchmarking are skills, local technological effort (research and development, or R&D), foreign direct investment, licensing payments abroad (royalties) and physical infrastructure.1 As in chapter 2, the objective is not a full econometric explanation of the determinants of industrial performance, but a useful positioning of countries with respect to important structural variables to help policymakers. Benchmarking countries, even by a few structural variables, raises difficulties. There are problems relating to the availability and definitions of variables, discussed in chapter 3. There may also be problems relating to the complementarity of the variables. Benchmarking implicitly assumes that each driver of industrial performance complements the others for the entire sample. For example, it takes for granted that higher skills, R&D and inward foreign direct investment all work towards improving industrial performance. While this might seem plausible, it is easy to think of exceptions. For example, domestic R&D and foreign direct investment may complement each other in some countries but compete in others. Foreign direct investment is effective in transferring and deploying production technology in host countries, but it may be less so in building or transferring deeper innovative capa-

57

58

0.1

0.2

0.3

0.4

10

15

20

25

30

30

Average CIP index value 0.5

Enterprise-financed R&D per capita (dollars) 35

Royalties per capita (dollars)

10

20

Ea

Figure 4.1 Competitive industrial performance and its drivers by region, 19811985, 1985, 19931997 and 1998

Sources: UNIDO Scoreboard database (see technical annex). Note: Data are unweighted averages. For details on telephone mainlines per 1,000 people by country for 1985 and 1998, see table A2.23 on page 175.

Industrial Development Report 2002/2003

1998

1985

1998

1985

1998

1985

Ea st As Ea ia (w st ith As ou ia tC hi na So ) ut La La h a t Ca tin nd in A As ia rib A th m be me e C eri an ric ar ca (w a a ibb ith nd ea ou th n Su tM e bex Sa ico ha S ) ra (w ubn ith Sa Af ou ha r t S ran ica ou A f M th ric an idd Afr a an d le ica d No Ea ) Tu rt s rk h t ey Af ric a st As E ia (w ast ith As ou ia tC hi na S ) La out La an tin h A Ca tin d rib A th Am sia be me e C eri an ric ar ca (w a a ibb ith nd ea ou th n Su tM e bex Sa ico ha Su ) ra (w bn S ith a A fr ou ha t S ran ica ou A f M th ric an idd Afr a an d le ica d No Ea ) Tu rt s rk h t ey Af ric a

In co du un str tri ial e iz T s ed ec ran on sit om ion Ea Eas ies (w st t A ith As si ou ia a tC h L La a atin Sou ina Ca tin nd A th ) rib A th m As be me e C eri ia an ric ar ca (w a a ibb Su itho nd ean b- ut the M S Sa (w ub- har ex ith Sa an ico ou ha A ) t S ran fric o A a M uth fr an idd A ica an d le fric d No Ea a) Tu rt s rk h t ey Af ric a

20

40

60

150

120

Tertiary technical enrolments (percentage of population) 5

Telephone mainlines (per 1,000 people), 1998

Ea st As

Ea
30 60 90

st

19931997

1998

1998

1985

Average annual inflows of foreign direct investment per capita (dollars) 80 19811985

As E ia (w ast ith As i ou a tC h So ina) ut La h a ti L As Ca atin nd n A i t he me a rib A be me Ca ric an ric ri a (w a a bbe ith nd an Su out the bM S e Su aha xic o) (w bra n ith Sa A h ou a f t S ran rica ou A f M th ric an idd Afr a an d le ica d No Ea ) Tu rt s rk h t ey Af ric a E ia (w ast ith As ou ia tC h So ina) ut L h a ati L As Ca atin nd n A ia t m h rib A e e be me Ca ric an ric ri a (w a a bbe ith nd an Su out the bSa Me xi h S (w ub- ara co) n ith Sa ou ha Afr t S ran ica ou A f M th A rica a id f an nd dle rica d No Ea ) Tu rt s rk h t ey Af ric a

Ea st Ea As ia (w st ith As ou ia tC hi na So ) La u t La and tin h A A Ca tin si t rib A he mer a be me Ca ica rib an ric (w a a be ith nd an ou th Su tM e bex Sa ico h S ) (w ub- aran S ith a ou ha Afri t S ran ca ou A t fr M h A ica an idd fri an d le ca) d No Ea Tu rt s rk h t ey Af ric a

set of countries or one development phase from another. The most practical way is to proceed with benchmarking and, as done here, take up these caveats in the analysis. The sample economies were ranked by each driver of industrial performance in both 1985 and 1998 (appendix table A4.1; see figure 4.1 for a snapshot of industrial performance and structural drivers by region). A few highlights of the rankings are worth mentioning. Most of the top 20 economies are industrialized, but there are notable exceptions. The Republic of Korea led the world in skills in both years, because of its high tertiary enrolments and high share of technical students in the population. The Russian Federation ranked sixth in 1998, and Taiwan Province of China eighth. Finland moved up in the skill ranking between 1985 and 1998, displacing the United States of America in second place. In R&D spending per capita Germany fell from first place to fifth, with Switzerland taking the top slot. The leading developing economy was again the Republic of Korea, in 13th place, followed by Singapore (in 14th place, just ahead of the United Kingdom) and Taiwan Province of China (20th). Singapore led the developing world (and the world as a whole) in foreign direct investment per capita in 1998, followed by Hong Kong Special Administrative Region (SAR) of China in fifth place. Other developing economies among the top 20 recipients of foreign direct investment were Malaysia and Chile; one transition economy, Hungary, also ranked among the top 20. Singapore and Hong Kong SAR ranked among the top 5 in royalties per capita, followed, in the developing world, by Malaysia, Taiwan Province of China and the Republic of Korea. Singapore ranked third in physical infrastructure, with Bahrain and Hong Kong SAR also in the top 20.

Box 4.1 Highlights of the Scoreboard analysis The analysis of industrial performance and drivers points to the following main messages:
The correlation between the CIP index and the drivers of industrial

performance is positive and significant.

There is a surprising degree of consistency in the relationship

between the CIP index and the drivers over time.

Most drivers are also related to each of the others, so that it makes

sense to have more of everything.

There are many ways of combining the drivers, however, and suc-

cessful countries have used varying strategies.

The rankings of economies by the drivers are stable over time, as

would be expected for such structural variables.

Despite the general stability in rankings, some countries changed

their relative position significantly.

The impact of the drivers on industrial performance also changes

over time, with foreign direct investment in particular gaining in significance because of the rise of integrated production systems.
While foreign direct investment remains a small share of global

investment, it plays a vital role in the industrial performance of a growing number of countries. At the same time, it remains highly concentrated, particularly in its deployment of sourcing for hightech products and components.
Domestic technological effort, as measured by R&D financed by

productive enterprises, is the most consistent and significant of the drivers. But this R&D variable should not be assessed in isolation: the ability to undertake technological effort clearly depends on the availability of skilled manpower and access to foreign technologies. The analysis shows that these are crucial factors in industrial performance, though the importance of technology licensing appears to be declining.
Physical infrastructure is strongly associated with industrial growth

and technology upgrading, but probably as a permissive rather than a causal factor.
The drivers are unevenly distributed in the developing world, and

the distribution is growing more uneven. East Asia dominates in almost every variable, while Sub-Saharan Africa is consistently the weakest.

Stability and complementarity of drivers

The ranking of economies by each driver of industrial performance shows considerable stability over time (just as the ranking by the competitive industrial performance, or CIP, index does). Thus the ranking of economies by R&D spending per capita for 1998 is highly correlated with that for 1985, and so on.3 This is not surprising: it is naturally difficult for economies to significantly shift their relative position with respect to structural drivers in the short to medium term.4 Even so, some countries changed their relative position significantly between 1985 and 1998, such as Uruguay in the skills index, Ecuador in R&D per capita and Tunisia in foreign direct investment per capita (box 4.1). Each driver is positively correlated with the others, suggesting that there are underlying structural relationships between them and that they generally reinforce one another. Thus greater innovative activity is related to higher technical skills, greater foreign direct investment inflows to higher royalty payments and so on. But the strength and significance of the relationships between the drivers vary (table 4.1). Moreover, such correlations do not reflect causal links, though some may well reflect an indirect linksay, through higher incomes. Thus richer countries tend to have more of each driver on a per capita basis. This is why the infrastructure variable has such high correlations with most others. That being said, it is

Benchmarking the drivers of industrial performance

59

Table 4.1 Correlation between drivers of industrial performance, 1998

Driver Skills Research and development Royalties Foreign direct investment Infrastructure

Skills 1.000 0.537** 0.249* 0.380** 0.828**

R&D 1.000 0.197 0.396** 0.687**

Royalties

Foreign direct investment

Infrastructure

1.000 0.430** 0.295**

1.000 0.611**

1.000

Source: UNIDO Scoreboard database (see technical annex). * Significant at the 5 percent level. ** Significant at the 1 percent level.

also likely that many drivers do feed into one another, with skills and R&D an obvious example. Whatever the underlying mechanisms, the correlations suggest that overall in the sample of economies the drivers complement rather than offset one another. This supports a general presumption that industrial development requires all structural drivers to growbut not necessarily in tandem at all stages of development. It is still possible, of course, that countries need different combinations of drivers at different levels of industrialization. Less industrialized economies may need more infrastructure and basic skills, for example, while more industrialized ones need more R&D and advanced skills. It is also possible to combine the drivers in different ways in line with different development strategies. Recall, for example, the tradeoff between deepening technology through domestic R&D and importing readymade technology through foreign direct investment. Countries have responded in different ways to this tradeoff. Some, like the Republic of Korea, restricted inward foreign direct investment and promoted domestic R&D. Others, like Ireland and Singapore, have targeted high-tech foreign direct investment and used policies to increase innovative activity by transnational corporations. Still othersthe majorityhave had no explicit technology strategies for R&D or foreign direct investment, leaving technology upgrading to market forces. Aggregate analysis of this type cannot capture differences across particular industries and countries in the patterns of competitiveness and globalization. Each industrial activity in a country may perform at a different level, and each certainly faces different technological and competitive conditions. The organization of global value chains differs significantly, with different structures and agents dominating and coordinating activity. Each industrial activity needs different drivers and institutions. Policy has to be based on these specifics; the broad benchmarks in the Scoreboard provide only the starting point.5

Trends in industrial capacity building

The drivers of industrial performance tend to grow together as countries develop and their industrial sectors mature. But countries can combine the drivers in different ways: recent economic history reveals that there is no single road to industrial success. The differences in strategy reflect many factors that cannot be explored heremarket size, geographic location, natural resources, external pressures, political economy and history, initial base of skills and capabilities and so on. Even so, it is instructive to illustrate the main differences in strategies for building industrial drivers. Cluster analysis can again be used to do this (see the technical annex on the technique).

Skills, R&D and infrastructure

The cluster analysis looks first at the three domestic capability driversskills, R&D and infrastructure. For this exercise R&D spending is normalized by GNP to capture national differences in the allocation of resources to technological effort. The skill and infrastructure drivers are measured by composite indices so that each economys position reflects its relative rank rather than an absolute value for the driver. Consider these highlights of the results for developing economies (figure 4.2): Each of the four mature Asian Tigers forms a cluster of its own, showing significant differences from one another and from the rest of the developing economies. While all four raised their R&D effort between 1985 and 1998, Hong Kong SAR lags behind the other threeits strategy is clearly to use technologies developed elsewhere rather than to invent new technologies locally. Hong Kong SAR also lags behind the other three in skills, but has relatively strong physical infrastructure. The Republic of Korea leads strongly in both skills and R&D, though its relative skill

60

Industrial Development Report 2002/2003

Figure 4.2 Cluster analysis of skills, infrastructure and R&D in developing economies, 1985 and 1998 Average R&D spending as a share of GNP, 1985 Average R&D spending as a share of GNP, 1998 Infrasturcture index 0.9 Singapore 0.69%

0.8

0.7

0.6 Hong Kong SAR 0.01% Algeria, Bolivia, Brazil, Egypt, El Salvador, Jamaica, Malaysia, Mauritius, Morocco, Oman, Saudi Arabia, South Africa, Thailand, Tunisia, Turkey 0.07% 26 developing economies 0.01% 0 Taiwan Province of China 0.99%

Republic of Korea 2.10%

0.5

0.4

0.20%

0.3

0.60% 0.48%

0.2

0.1 0.04% 0.03% Argentina, Chile, Colombia, Costa Rica, Ecuador, Jordan, Mexico, Panama, Peru, Philippines, Uruguay, Venezuela 0.05% 0.2 0 0.2 0.4 Skills index
Source: UNIDO Scoreboard database (see technical annex). Note: Where bubbles have no value, R&D spending was negligible. The infrastructure index is an average of the standardized scores for traditional infrastructure (commercial energy use) and modern infrastructure (telephone mainlines). The skills index is an average of the scores for the Harbison-Myers index (see chapter 2) and tertiary technical enrolments.

0.1

0.6

0.8

1.0

intensity declined slightly over time. Singapore leads in physical infrastructure. Taiwan Province of China improved its position for all three drivers, but lags behind the Republic of Korea in each one. The next two developing country clusters, well below the Asian Tigers in each of these drivers, are about equal in infrastructure and R&D effort but one has significantly higher skill levels than the other. But this clustercomprising Argentina, Chile, Mexico, the Philippines and othersalso had a relative decline in skills, though its infrastructure index improved over the period. The bottom cluster, with 26 countries (including all the least developed countries) is weak in all three drivers, with minuscule R&D spending, low skills and poor infrastructure.

The findings for industrialized and transition economies show a similar variation in performance (figure 4.3): All the clusters had rising R&D spending except for that comprising the three transition economies (Hungary, Poland and Romania) and Portugal. The decline in R&D spending in transition economies, possibly a temporary response to liberalization, may reverse itself in time. The two clusters with the largest R&D efforts comprise the major industrialized countriesJapan, Germany and the United Statesalong with Finland, Sweden and Switzerland. The group with the United States and the two Scandinavian countries had the largest average R&D effort of all the industrialized countries in 1998. Both clusters show small relative declines in the skills and infrastructure indices.

Benchmarking the drivers of industrial performance

61

Figure 4.3 Cluster analysis of skills, infrastructure and R&D in industrialized and transition economies, 1985 and 1998 Average R&D spending as a share of GNP, 1985 Average R&D spending as a share of GNP, 1998 Infrasturcture index 1.0 Austria, Belgium, Canada, Denmark, France, Israel, Netherlands, Norway, United Kingdom 0.01% 1.52% 0.66% 0.6 Germany, Japan, Switzerland 1.78% Finland, Sweden, United States 2.03%

0.9

0.78%

0.8

0.7

0.5

0.4 Hungary, Poland, Portugal, Romania 0.01% 0.30%

0.22%

0.3

Australia, Greece, Ireland, Italy, New Zealand, Spain 0.51%

0.2

0.1

0.2

0.3

0.4 Skills index

0.5

0.6

0.7

Source: UNIDO Scoreboard database (see technical annex). Note: The infrastructure index is an average of the standardized scores for traditional infrastructure (commercial energy use) and modern infrastructure (telephone mainlines). The skills index is an average of the scores for the Harbison-Myers index (see chapter 2) and tertiary technical enrolments.

The third group (including the Benelux countries, Canada, France, Israel and the United Kingdom) has a lower average R&D effort but improved in all three drivers. A fourth cluster with low R&D (including Australia, Ireland and Italy) has even lower average levels of R&D but registered a significant rise in relative skill levels.

Technological effort and inventiveness

Now consider a composite index of technological effort and inventiveness based on two indicators, R&D financed by productive enterprises and patents taken out internationally (in this case in the United States).6 The index is the average of the two variables, which have been standardized, with equal weight assigned to both. The value of the composite index ranges from zero to one.

As expected, there is a strong and significant correlation between the input measure, R&D effort, and the output measure, patents (reflected in a coefficient of 0.85). But the two measures yield somewhat different rankings of economies (see annex table A4.1 and table 4.2). For example, Hong Kong SAR ranks low in R&D spending (40th) but high in patents (16th), as does Taiwan Province of China (20th and 4th). By contrast, Brazil ranks 27th in R&D spending and 42nd in patents, while China ranks 44th and 56th. The variations may be due to several factors, such as foreign companies affiliates patenting technology based on R&D elsewhere, differences in the quality or orientation of R&D and differences in the propensity to take out international patents. Without more detailed country analysis, deciphering the underlying forces is difficult.7 The 59 economies with positive values for the index of technological effort and inventiveness can be divided into three groups by performance: high, moderate and low (figure 4.4):

62

Industrial Development Report 2002/2003

Table 4.2 Patents taken out internationally, 1998

Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

Economy United States Japan Switzerland Taiwan Province of China Sweden Israel Germany Finland Canada Denmark Netherlands Belgium Korea, Republic of France United Kingdom Hong Kong SAR Austria Norway Australia Singapore New Zealand Italy Ireland Slovenia Spain Hungary South Africa Malaysia Greece Bahrain Venezuela Russian Federation Argentina Chile Uruguay Portugal Mexico Czech Republic Saudi Arabia Ecuador Costa Rica Brazil Jordan Poland Jamaica Philippines Thailand Guatemala Colombia Honduras Bolivia Tunisia Sri Lanka India Morocco China Turkey Indonesia Peru

Patents per 1,000 people 3.297 2.412 1.884 1.622 1.421 1.275 1.134 1.118 1.090 1.005 0.817 0.699 0.657 0.650 0.601 0.540 0.511 0.490 0.402 0.386 0.356 0.305 0.200 0.076 0.072 0.045 0.030 0.017 0.016 0.016 0.013 0.012 0.011 0.011 0.009 0.009 0.009 0.008 0.006 0.006 0.006 0.005 0.004 0.004 0.004 0.003 0.002 0.002 0.002 0.002 0.001 0.001 0.001 0.001 0.001 0.001 0.000 0.000 0.000

Japan leads the world and the high performers, with Switzerland and the United States close behind. This group includes most of the industrialized countries (exceptions include Greece, Portugal and Spain), but also four Asian Tigers (Taiwan Province of China, the Republic of Korea, Singapore and Hong Kong SAR). The moderate performers include most transition economies and the largest Latin American economies (Brazil, Argentina, Chile and Mexico) along with Costa Rica, Venezuela and Uruguay. Malaysia is the only Asian country here. South Africa, Turkey and Bahrain are also in this group. The low performers include large countries with complex industrial sectors and a high absolute value of R&D activity (China, India) and export-oriented economies with relatively modest R&D activity and a high reliance on transnational corporations (Indonesia, Thailand). They also include countries with small industrial sectors, low exports and little R&D activity (Panama, Jamaica, Bolivia, Kenya).8

Strategies for enhancing industrial performance

Turn now to the different patterns of reliance on domestic R&D effort (financed by productive enterprises) and foreign direct investment. Using R&D spending as a share of GNP and foreign direct investment as a share of gross domestic investment allows a more direct assessment of national allocations of resources between these two modes of acquiring new technology. While both modes have resulted in greater use of technology by the industrial sector, they may have led to different outcomes in performance. Cluster analysis, with the sample economies sorted into seven groups in 1985 and 1998, reveals interesting patterns in the economies relative reliance on R&D or foreign direct investment for each year and over time. The patterns for 1985 show the following (figure 4.5): The most successful countriesthose with the highest CIP indicesformed four clusters. Of these, the cluster containing Germany, Japan and Switzerland had the highest CIP score. This group had the greatest reliance on R&D and the least on foreign direct investment. Singapore, which formed the second most successful cluster on its own, had the opposite characteristics: very low R&D spending and extremely heavy reliance on foreign direct investment.

Source: U.S. Patent Office. Note: In this case internationally refers to the United States.

Benchmarking the drivers of industrial performance

63

Figure 4.4 Economies by technological effort and inventiveness index, 1998

High performers Japan Switzerland United States Sweden Germany Finland Denmark Taiwan Province of China Netherlands France Israel Belgium Canada Norway Republic of Korea Austria United Kingdom Singapore Australia Ireland Italy New Zealand Hong Kong SAR 0

Moderate performers Slovenia Spain Czech Republic Hungary South Africa Greece Portugal Brazil Argentina Malaysia Russian Federation Poland Chile Costa Rica Venezuela Turkey Bahrain Mexico Uruguay Romania 0.2 0.4 0.6 0.8 1.0 0 0.02 0.04 0.06

Low performers Saudi Arabia Ecuador Panama Jordan China Jamaica Philippines Indonesia Thailand Colombia India Guatemala Honduras Sri Lanka Bolivia Mauritius Morocco Tunisia Egypt Peru Algeria Nicaragua Kenya 0 0.0005 0.001

Source: UNIDO calculations based on data from U.S. Patent Office (see technical annex).

Two other successful clusters were fairly close to each otherone comprising Canada, Finland, the Republic of Korea and the United States, and the other consisting of most other industrialized countries plus Taiwan Province of China. Both clusters had fairly high R&D spending, higher than the average for the successful industrial performers. But the first, with a higher CIP score, had a distinctly lower orientation towards foreign direct investment, while the second struck a balance between R&D effort and foreign direct investment. The rest of the developing world and several industrialized countries fell into three clusters. Each had low R&D spending but different degrees of reliance on foreign direct investment. The best performing of these clusters included Hong Kong SAR, Greece, New Zealand, Portugal and Spain, along with the Philippines, Argentina, Mexico and some other Latin American and Caribbean countries. These economies had a moderately high reliance on foreign direct investment, but also R&D spending higher than the average for the lowperforming clusters. The next group, combining the lowest R&D spending with high reliance on foreign direct investment, included 14 developing countries, ranging from Brazil and Malaysia to least developed countries like the Central African Republic and Ghana.

The remaining 25 developing countries (and Albania) combined very low R&D spending with similarly low foreign direct investment. By 1998 there was a general shift towards greater reliance on foreign direct investment in all groups, a clear indication of the growing role of transnational corporations in the world economy. In the leading economies there was also a greater propensity to invest in R&D, but this did not hold for many developing countries lower on the industrialization ladder. For 1998 there were five successful clusters, with a composition that was quite different (figure 4.6). The cluster with the highest average CIP score included Belgium, Ireland and Singapore. In contrast with the group that had the highest CIP index in 1985, this cluster depended on foreign direct investment rather than R&D. It did, however, have moderate R&D activity. The second cluster, very close in average CIP score to the first, contained Germany, Japan and Switzerland (this was the leading cluster in 1985). This group continued to rely on domestic R&D, and their R&D spending rose as a share of GNP. But foreign direct investment also played a larger role in domestic investment than before. The third cluster included only Sweden, with high R&D spending as well as high reliance on foreign direct

64

Industrial Development Report 2002/2003

Figure 4.5 Cluster analysis of industrial performance, R&D and foreign direct investment, 1985 Average CIP index value R&D spending as a percentage of GNP 2.0

1.5

Germany, Japan, Switzerland 0.72 1.0 Taiwan Province of China, other industrialized countries Argentina, Australia, Chile, 0.41 Ecuador, Greece, Hong Kong SAR, Jordan, Mexico, New Zealand, Panama, Peru, Philippines, Portugal, Spain, Uruguay, Venezuela 0.11

0.5

Canada, Finland, Republic of Korea, United States 0.45

25 developing economies 0.03

Singapore 0.58

Bolivia, Brazil, Cameroon, Central African Republic, Colombia, Costa Rica, Egypt, Ghana, Guatemala, Malaysia, Nigeria, Oman, Tunisia, Zambia 0.04 0.5 5 0 5 10 15 20 25

Foreign direct investment as a percentage of gross domestic investment

Source: UNIDO Scoreboard database (see technical annex).

investment. But the high level of foreign direct investment was something of an aberration: Sweden has traditionally had a very limited foreign presence, though recent waves of foreign mergers and acquisitions have raised the level of foreign direct investment. The underlying trend for Sweden is greater reliance on domestic R&D. The fourth cluster with a high average CIP score included Finland, the Republic of Korea and the United States. This cluster was similar to the second: it had high reliance on domestic R&D and relatively low reliance on foreign direct investment (though slightly higher than the second clusters). The fifth cluster contained most other industrialized countries (except New Zealand, which appeared in the sixth cluster) with Taiwan Province of China and three transition economies. This group showed a balance between R&D and foreign direct investment.

The sixth cluster had a relatively high reliance on foreign direct investment and a very weak domestic R&D effort. It included New Zealand along with Albania, Chile, China, Hungary, Malaysia, Poland and eight other developing countries. The cluster with the weakest performance contained the other 45 developing countries. These countries had little domestic R&D and moderate foreign direct investment (but foreign direct investment played a larger role in this group than it did in the bottom group in 1985). The cluster analyses lead to four main conclusions. First, while the leading industrialized economies rely heavily on domestic R&D efforts, their reliance on foreign direct investment has increased. In most of these countries inward foreign direct investment serves two purposes: it brings in new technology, but it also helps transnational corporations to tap local R&D. Technology flows strongly in both directions between these countries as they increasingly specialize in innovation. The

Benchmarking the drivers of industrial performance

65

Figure 4.6 Cluster analysis of competitive industrial performance, R&D and foreign direct investment, 1998 Average CIP index value R&D spending as a percentage of GNP 3.5

Sweden 0.56 3.0

2.5

Germany, Japan, Switzerland 0.69

2.0

1.5 Finland, Republic of Korea, United States 0.41 1.0

Belgium, Ireland, Singapore 0.70

0.5

Czech Republic, Russian Federation, Slovenia, Taiwan Province of China other industrialized countries 0.32

0.0 45 developing economies 0.05 0.5

Albania, Bolivia, Chile, China, Costa Rica, Ecuador, Hungary, Malaysia, New Zealand, Nicaragua, Nigeria, Panama, Peru, Poland, Venezuela 0.09 20 25 30 35

10

15

Foreign direct investment as a percentage of gross domestic investment

Source: UNIDO Scoreboard database (see technical annex).

Republic of Korea had the strongest technological position among the developing economies in both years, with a low reliance on foreign direct investment. Second, the strong showing by Ireland and Singapore suggests that industrial latecomers can achieve an impressive performance by relying heavily on foreign direct investment. This reliance does not preclude growth of their domestic R&D effort, though their R&D levels still lag behind those in countries with more autonomous strategies. Third, most developing countries continue to languish at the bottom of the technological ladder, with no perceptible rise in domestic R&D. Some have managed to attract fair amounts of foreign direct investment (as a share of domestic investment), but only a few have managed to break into integrated global production systems. Fourth, where successful, both the strategy based on R&D and that based on foreign direct investment involve acquiring for-

eign technology, but in different ways. The strategy based on domestic R&D is more autonomous and involves large investments in skills. For industrial latecomers it is also a riskier strategy, because it tends to involve extensive use of industrial policy. The strategy centred on foreign direct investment can take countries a long way without a need for strong local R&D. But countries that succeed with this strategy tend to raise their investments in R&D over time, with transnational corporations shifting some innovative functions to these countries. Relatively few countries have managed to combine heavy dependence on foreign direct investment with strong growth in innovative capabilities (domestic R&D), and those that have done so relied extensively on industrial policy (as in Ireland and Singapore).

R&D, foreign direct investment and hightech exports

The relationships between R&D, foreign direct investment and high-tech exports show more clearly the differences between

66

Industrial Development Report 2002/2003

the two modes of acquiring technology. Given the importance of high-tech exports in industrial performance, analysing the roles of domestic R&D and inflows of technology through foreign direct investment separately is instructive. The competitiveness of high-tech exporters (particularly in electronics) can be traced to domestic innovation or to participation in integrated global production systems. Comparing the intensity of R&D and foreign direct investment in a country with its high-tech export performance gives an indication of the relative importance of these drivers of industrial performance (table 4.3). An analysis of R&D spending per unit of high-tech exports and per unit of inward foreign direct investment was conducted for all major exporters of high-tech productsthose with high-tech exports of more than $5 billionin 1998.9 These 26 economies include nine developing economiesall those in East Asia (except Indonesia) plus Mexico.

As expected, economies with high R&D spending per unit of high-tech exports and per unit of inward foreign direct investment have a strong technological base. Those ranking highest on these measures, not surprisingly, are the major industrial powers; they also generally lead in high-tech exports (in value terms). At the bottom of the scale are developing countries, specializing in assembly and testing. This method of distinguishing competitive strategies clearly has some merit. The analysis leads to some interesting findings: Japan, which has followed an autonomous R&D-based strategy, led the world in R&D spending per unit of hightech exports in 1998. In 1985, however, Germany held this place, followed by the United States. Clearly, global production systems have spread faster to other industrialized countries than to Japan. The degree to which Japan

Table 4.3 Reliance of major high-tech exporters on domestic R&D and foreign direct investment, 1985 and 1998

R&D per dollar of high-tech exports Ranka (dollars) 1998 1985 Economy 1998 1985 1 3 Japan 0.937 0.635 2 2 United States 0.622 0.686 3 1 Germany 0.368 0.816 4 6 Switzerland 0.331 0.282 5 9 Sweden 0.283 0.231 6 8 France 0.266 0.245 7 18 Korea, Republic of 0.264 0.119 8 5 Austria 0.233 0.284 9 11 Denmark 0.225 0.228 10 10 Spain 0.213 0.229 11 15 Italy 0.210 0.141 12 4 Finland 0.200 0.342 13 7 Canada 0.177 0.278 14 19 Belgium 0.159 0.105 15 13 United Kingdom 0.134 0.167 16 12 Israel 0.113 0.211 17 14 Netherlands 0.098 0.164 18 17 Taiwan Province of China 0.068 0.131 19 26 China 0.033 0.000 20 21 Ireland 0.022 0.019 21 23 Singapore 0.010 0.008 22 16 Mexico 0.004 0.134 23 25 Malaysia 0.004 0.001 24 24 Hong Kong SAR 0.002 0.003 25 20 Thailand 0.001 0.043 26 22 Philippines 0.000 0.014

R&D per dollar of inward foreign direct investment (dollars) 1998 1985 100.40 62.42 1.75 1.68 5.01 13.09 1.35 2.33 0.71 3.33 0.76 1.67 5.90 3.50 0.65 2.19 0.57 5.50 0.28 0.20 1.45 0.97 1.45 2.95 0.52 0.98 0.26 0.20 0.49 0.72 0.67 2.76 0.34 0.78 1.50 1.37 0.03 0.00 0.38 0.31 0.07 0.02 0.02 0.28 0.03 0.00 0.00 0.00 0.01 0.03 0.01 0.07

High-tech exports (billions of dollars) 1998 1985 114.9 36.6 196.9 53.3 92.7 24.3 18.3 4.7 20.4 4.1 65.1 14.3 36.0 3.7 7.4 1.6 7.6 1.8 10.2 1.5 24.5 7.5 10.5 0.8 23.8 6.2 17.4 3.5 76.3 17.9 6.6 1.1 40.8 6.9 38.6 4.7 33.5 0.3 25.2 2.7 62.3 4.7 31.3 1.9 34.3 2.3 6.0 2.4 15.6 0.2 19.0 0.3

High-tech products as a share of manufactured exports (percentage) 1998 1985 29.6 20.8 31.0 25.8 17.1 13.2 23.2 17.0 24.7 13.4 21.6 14.6 27.2 12.2 12.2 9.2 16.0 10.9 9.3 6.0 10.1 9.5 24.4 5.7 11.1 7.1 9.7 6.4 28.2 17.6 28.3 17.0 24.3 10.2 35.0 15.4 18.2 1.2 39.3 25.8 56.7 20.4 26.6 8.6 46.9 14.8 24.5 14.2 28.3 2.4 64.3 5.8

Source: UNIDO Scoreboard database (see technical annex). Note: Includes only economies with high-tech exports of more than $5 billion in 1998. a. Based on R&D spending per unit of high-tech exports.

Benchmarking the drivers of industrial performance

67

relies on R&D rather than foreign direct investment is strikingly illustrated by figures for R&D spending per unit of foreign direct investment: in 1998 this figure for Japan ($100) was 20 times that for Germany ($5). The United States has maintained a stable profile in both ratios. By contrast, Germany had a sharp decline in R&D spending per unit of both high-tech exports and foreign direct investment, indicating its growing participation in global production systems. The United Kingdom, the fourth largest high-tech exporter in 1998, had a surprisingly low R&D ratio, indicating its growing importance as a base for the operations of transnational corporations in electronics. Ireland ranked lowest among industrialized countries in R&D spending per unit of high-tech exports, bearing out the dominant role of transnational corporations in building its competitiveness. The Republic of Korea had the second highest R&D spending per unit of foreign direct investment, after Japan, in 1998.10 Taiwan Province of China also had relatively high R&D spending per unit of foreign direct investment, followed by China (though with a very low value of $0.03 in 1998). Other developing countries depend heavily on transnational corporations for their high-tech exports, though Singapore has a relatively strong R&D base compared with others in this group.11 Four developing countries increased the share of hightech products in their manufactured exports by 26 or more percentage pointsSingapore, Malaysia, Thailand and the Philippines. Each is producing and selling within integrated global production systems. A scatter diagram illustrates the relationship between R&D spending per unit of foreign direct investment in 1985 and 1998 (figure 4.7). Economies on the lineincluding Hong Kong SAR, Japan and Switzerlandhad no change in rank by R&D spending per unit of foreign direct investment. Those above the line increased their relative reliance on R&D for the export of high-tech manufactures, while those below it increased their reliance on foreign direct investment. Italy, Taiwan Province of China and the United States show the biggest rise in ranks by dependence on domestic R&D. But some countries ranking high in reliance on foreign direct investment and very low in R&D in 1985China, Malaysia and Singaporealso had big increases. In Malaysia and Singapore the rise in R&D ranks is due mainly to technological deepen-

Figure 4.7 Ranking of economies by R&D spending per unit of foreign direct investment, 1985 and 1998 1998 rank 0 Republic of Korea 5 United States Taiwan Province of China Italy Switzerland France Germany Finland Sweden Israel 15 United Kingdom Ireland Spain Belgium Singapore China Malaysia 25 Thailand Austria Canada Netherlanda Denmark Japan

10

20

Mexico

Philippines Hong Kong SAR

30 30

25

20

15 1985 rank

10

Source: UNIDO Scoreboard database (see technical annex).

ing by transnational corporationsand in China it is also due to R&D by local enterprises. In Singapore foreign enterprises perform around 57 percent of the R&D in manufacturingin Malaysia, 50 percent.12 Ireland, the European economy with the highest rank by reliance on foreign direct investment, also had an increase in local R&D, again led by transnational corporations. Foreign enterprises accounted for 68 percent of Irelands manufacturing R&D in 1996.13 Countries that rose in the ranking by reliance on foreign direct investment over the period include Denmark, Sweden and Israel at the high end and Mexico, the Philippines and Thailand at the low end. Again, however, the high level of foreign direct investment in Sweden in 1998 was something of an aberration. These differences in strategy can be further explored with the help of cluster analysis for R&D, foreign direct investment and high-tech exports in 1985 and 1998 (see figures 4.8 and 4.9, which include all economies in the sample). The changes in the clusters between the two years throw new light on the strategies. In 1985 the largest bubble was for Singapore, for which hightech products accounted for 20 percent of manufactured exports. Singapore stood apart from other clusters, with very

68

Industrial Development Report 2002/2003

Figure 4.8 Cluster analysis of R&D, foreign direct investment and high-tech exports, 1985 High-tech products as a share of manufactured exports R&D spending as a percentage of GNP 1.8

1.6

Germany, Japan, Switzerland 17%

1.4 Austria, Denmark, France, Israel, Italy, Netherlands, Sweden, Taiwan Province of China 12% Hong Kong SAR, Ireland, United Kingdom, United States 1% Argentina, Australia, Belgium, Chile, Ecuador, Greece, Jordan, Mexico, New Zealand, Panama, Peru, Philippines, Spain, Uruguay, Venezuela 2% Bolivia, Brazil, Cameroon, Central African Republic, Colombia, Costa Rica, Egypt, Ghana, Guatemala, Malaysia, Nigeria, Oman, Tunisia, Zambia 1% 4 6 8 10 12 14 16 18 20

1.2

1.0

0.8

0.6

0.4

Singapore 20%

0.2

Canada, Finland, Republic of Korea, Norway 7%

0 28 developing economies 1% 0.2 2 0 2

Foreign direct investment as a percentage of gross domestic investment

Source: UNIDO Scoreboard database (see technical annex).

high foreign direct investment and relatively low R&D. Economies in another cluster, including Ireland, the United Kingdom and the United States, had fairly large shares (averaging 21 percent) of high-tech products in their manufactured exports, with a balanced mix of foreign direct investment and R&D. Germany, Japan and Switzerland (with high-tech products averaging 17 percent of their manufactured exports) clustered together as countries highly dependent on R&D. Most developing countries had small shares of high-tech exports and low R&D, with differing degrees of reliance on foreign direct investment. By 1998 there had been a general move towards greater reliance on foreign direct investment and larger shares of high-tech products in manufactured exports. The cluster with the largest share (52 percent) included Ireland, Malaysia, the Philippines and Singaporeall highly dependent on foreign direct investment. The economies most dependent on R&D (including the Republic of Korea and Taiwan Province of China) were also more reliant on foreign direct investment in

1998 than in 1985. Most developing countries, however, continued to have very small shares of high-tech products in their manufactured exports.

Do the drivers explain performance?

That there is a positive relationship between industrial performance and its drivers is apparent from even a cursory look at the data (see figure 4.1). Regions that do well in one tend to do well in the other. This conclusion is borne out by the results of statistical analyses of the relationship between industrial performance and its drivers (box 4.2). Technology in the generic sensedomestic R&D as well as access to foreign technology through foreign direct investment and licensinghas a powerful influence on industrial performance.

Benchmarking the drivers of industrial performance

69

Figure 4.9 Cluster analysis of R&D, foreign direct investment and high-tech exports, 1998 High-tech products as a share of manufactured exports R&D spending as a percentage of GNP 3.5

3.0 Finland, Germany, Japan, Republic of Korea, Switzerland, Taiwan Province of China, United States 26% China, Costa Rica, Hong Kong SAR, Hungary, Mexico, Poland, Thailand 21% Ireland, Malaysia, Philippines, Singapore 52%

2.5

Sweden 24%

2.0

1.5

1.0

0.5

Other industrialized countries, Czech Republic, Russian Federation, Slovenia 13%

0 41 developing economies 1% 0.5 5 0 5 10 15 20

Albania, Bolivia, Chile, Ecuador, New Zealand, Nicaragua, Nigeria, Panama, Peru, Venezuela 2%

25

30

Foreign direct investment as a percentage of gross domestic investment

Source: UNIDO Scoreboard database (see technical annex).

Among the drivers of industrial performance, R&D is statistically the most important, in both 1985 and 1998 and over time. This finding highlights the need for domestic technological effort even at low levels of industrial development. While the causation may run in both directions (the more industrialized countries become, the more they invest in R&D), theory does suggest that the causation from R&D to industrial performance is likely to be predominant. The capability building literature shows that technological effort (formal and informal) is as critical a driver of competitive industrial performance in developing countries as it is in industrialized countries.14 Licensing foreign technology is also statistically significant, but its role appears to be diminishing. The role of foreign direct investment, by contrast, has grown in significance. This corresponds with the evidence on the increasing role of integrated production systems in the world economy, on the rising importance of technol-

ogy transfer by transnational corporations and on their export activity as a dynamic element in the industrial competitiveness of developing countries. The significance of skills is also increasing, again entirely in line with the conventional wisdom on the importance of human capital and technology for competitive industrial performance. It is reassuring, however, to see that the statistical findings confirm this for such a broad sample.15 Infrastructure remains important in both 1985 and 1998. In sum, the results show that the set of structural drivers is strongly associated with industrial performance and that the association is broadly in the expected direction for the entire sample. Clearly, a full exploration of cross-country differences in industrial performance and its drivers would require a much more ambitious effort, with many more qualitative variables, data on many more years to capture lags and econometric tests for feedback and simultaneity.

70

Industrial Development Report 2002/2003

Box 4.2 The relationship between industrial performance and its drivers: results of statistical analyses Multiple regression analysis was used to explore the relationship between industrial performance and its drivers. The dependent variable was the CIP index in 1985 or 1998, and the independent variables were per capita R&D, foreign direct investment, royalty payments and the indices for skills and infrastructure in the appropriate years. To control for differences arising from levels of development not captured by other variables, a dummy variable was added, taking the value 0 for industrialized and transition economies and 1 for developing, countries. Regressions were conducted separately for the two years. Performance in 1998 was also regressed on drivers in 1985 to capture the impact of the initial stock of drivers on subsequent performance. (See the table for the three sets of results.) Results for 1985. The equation explains 93 percent of the variation in the CIP index. R&D per capita shows up as the most important influence, followed by royalties and infrastructure. The skill variable is significant at the 10 percent level. Foreign direct investment is not significant and has a negative sign. The dummy variable for developing countries has a significant and negative effect. This result suggests that with the structural drivers taken into account, being a developing country has an independent negative effect (capturing a range of other potential factors) on industrial performance. Results for 1998. All independent variables except the development dummy variable are now positive and significant, explaining 88 percent of the variation in the CIP index. The dummy variable for developing countries is no longer significant, suggesting that the level of development does not affect performance. In other words, the only significant effects arise from the drivers. R&D is again the most important driver, followed by royalties. Foreign direct investment is now significant and positive, suggesting that the contribution of transnational corporations to industrial performance has grown over the period. The skills index is also significant and positive, and its coefficient is higher than in 1985, suggesting that high-level skills are becoming increasingly important to industrial competitiveness. Results for 198598. The results are broadly similar to those for 1985, with interesting variations. Skills are far more important and significantthe base in 1985 seems to have a strong positive influence on performance in 1998. R&D remains significant and important, suggesting continuity and cumulativeness. Foreign direct investment is insignificant; clearly, its positive impact grows over the period. Infrastructure loses significance, suggesting that current patterns of infrastructure investment are more closely related to industrial performance. The dummy variable has a significant negative effect; being a developing country in 1985 held back industrial performance in 1998.

Regression results for competitive industrial performance and its drivers, 1985 and 1998

Independent variable Skills Research and development Foreign direct investment Royalties Infrastructure Development dummy variable

75 economiesa Standard coefficient t-statistic 0.090* 1.832 0.443*** 9.300 0.112 1.575 0.384*** 5.228 0.204** 2.240 0.203*** 3.188 Adjusted R2 = 0.928

85 economiesb Standard coefficient t-statistic 0.130* 1.822 0.466*** 8.846 0.183*** 3.379 0.253*** 5.986 0.196** 2.018 0.024 0.401 Adjusted R2 = 0.881

75 economiesc Standard coefficient t-statistic 0.261*** 2.911 0.493*** 5.270 0.074 0.651 0.342** 2.902 0.125 0.851 0.299** 2.922 Adjusted R2 = 0.809

Source: UNIDO Scoreboard database (see technical annex). * Significant at the 10 percent level. ** Significant at the 5 percent level. *** Significant at the 1 percent level. Note: All statistical tests for functionality, heteroskedasticity and collinearity are satisfied. The potential problem raised by the high correlation between the drivers does not affect the result. a. The dependent variable is the CIP index for 1985; the independent variables refer to 1985. b. The dependent variable is the CIP index for 1998; the independent variables refer to 1998. c. The dependent variable is the CIP index for 1998; the independent variables refer to 1985.

Balance between drivers and performance

The positive relationship between performance and its drivers holds broadly for the entire sample. Behind that overall relationship, however, lie interesting differences among countries. Some countries perform in line with their stock of drivers; some do better than warranted by the stock; and some do worse. To explain these differences requires a composite index of drivers for each country, derived by averaging the normalized

scores for the five drivers. A scatter diagram illustrates how the sample economies are spread according to this composite index and the CIP index in 1998 (figure 4.10): Most economies have a balance between industrial performance and its drivers. Industrialized countries are largely balanced at the high end, with performance and drivers above the average for the sample. Most developing economies are balanced at the low end, with a large cluster at the bottom left.

Benchmarking the drivers of industrial performance

71

Figure 4.10 Competitive industrial performance index and average drivers of industrial performance in selected economies, 1998 CIP index 1.0 0.9 0.8 0.7 0.6 0.5 Germany United Kingdom Average value for composite index of drivers Singapore Switzerland Ireland Japan United States

Sweden Belgium Finland France Austria Denmark Netherlands 0.4 Italy Canada Malaysia Hungary Republic of Korea Spain Norway 0.3 Israel Mexico Portugal Hong Kong SAR Philippines Australia 0.2 Thailand New Zealand Slovenia Brazil Greece Average CIP index value 0.1 Bahrain Russian Federation 0 Czech Republic Taiwan Province of China Panama Chile 0.1 0.05 0.05 0.15 0.25 0.35 0.45 0.55 0.65

conditions (an obvious example is the effect on Mexico of the North American Free Trade Agreement). The divergence may be due to ephemeral factors, such as wars, civil unrest or natural disasters. And it could be related to measurement problems. The measures may not properly capture the underlying structural variables. For example, the R&D measure might not properly capture differences among countries in technological effort, or the figures on foreign direct investment might misrepresent participation by transnational corporations in manufacturing, particularly in export-oriented activities. Or the measures may not capture strategic differences in the use of structural drivers. For example, economies with similar levels of foreign direct investment might target different types of investorsas Singapore targets high-tech transnational corporations but Hong Kong SAR does notwith corresponding effects on the structure and growth of production and exports. A matrix that sorts developing economies into four groups balanced high, balanced low, overperformers and underperformerssheds further light on the relationship between performance and its drivers (table 4.4). Balanced economies are those in which industrial performance is consistent with their composite index of driverswith both at the high or low end. Overperformers have an industrial performance higher than expected given their drivers (based on the average for the entire sample)and underperformers an industrial performance lower than expected. In 1998 three-quarters of the economies in the sample were balanced low. The four Asian Tigers were in the top left-hand quadrant, with high performance and high average drivers (like most mature industrialized countries). Of these, the Republic of Korea was balanced high, while Singapore and Taiwan Province of China were overperformersthat is, their CIP indices were higher than the level commensurate with their stock of drivers. Hong Kong SAR, by contrast, was an underperformer, though it had been an overperformer in 1985. Argentina had been in the balanced high group and Brazil among the overperformers in 1985, but both were in the balanced low group by 1998. Chile joined Hong Kong SAR as an underperformer in 1998, along with Bahrain and Panama. The overperformers in 1998, apart from Taiwan Province of China, have undergone rapid export growth and technological upgrading in recent years by plugging into global production networks as major supply bases. The strong foreign presence in high-tech export activity has enabled many of them to overcome gaps in domestic industrial capabilities. While this strategy is creditable and may offer lessons to other countries, overperformance may also indicate vulnerability. Take the Philippines, with a high-tech export structure that involves little local value added and is essentially driven by low wages. The country is, moreover, highly dependent on semi-

Composite index of drivers

Source: UNIDO Scoreboard database (see technical annex). Note: The range around the central diagonal, given by the standard deviation of the sample mean, is the normal range within which an economys industrial performance is in balance with its drivers. Economies above the top diagonal line perform better than warranted by their composite index of drivers (stock of drivers), those below the lower diagonal line worse. The average values for the indices are derived for the entire sample of economies.

Four developing economies have above-average performance and driversthe mature Asian Tigers, which earlier analysis also showed to be outliers in the developing world. Some developing countries have above-average CIP indices but below-average drivers (Malaysia, Mexico, the Philippines and Thailand). Some countries have above-average drivers but belowaverage CIP indices (Bahrain, Chile, Greece and the Russian Federation). Some economies lie outside the normal range, performing better or worse than warranted by their drivers. This divergence between performance and drivers could stem from a range of factors. It could be caused by factors not quantified here, such as macroeconomic or political factors, industrial policies, institutional differences or external market access

72

Industrial Development Report 2002/2003

Table 4.4 Developing economies by industrial performance and average capabilities, 1985 and 1998

Year 1998

1985

Overperformers (industrial performance higher than drivers) Singapore Taiwan Province of China Malaysia Mexico Philippines Thailand Brazil Hong Kong SAR Zimbabwe

Balanced high Underperformers (industrial (industrial performance performance and drivers lower than in balance) drivers) Korea, Bahrain Republic of Chile Hong Kong SAR Panama

Argentina Korea, Republic of Taiwan Province of China Singapore

Ecuador Panama Jordan

industrial performer, with large and rapidly growing exports backed by a strong base of skills, infrastructure and foreign direct investment. But in the past decade or so its engine of growth has shifted to services, and much of its industrial activity has shifted to other, lower-wage countries. As a result, its performance in the CIP index does not match its drivers (which continue to improve): its export structure remains relatively low tech. The share of manufacturing in GDP has declined dramatically, from nearly 30 percent in the 1960s to 5.7 percent by 1999; moreover, manufacturing output stagnated or declined in the 1990s.16 So, it is not surprising that Hong Kong SAR underperformed in 1998 relative to its base of drivers. But its service industry has the unique advantage of access to the giant Chinese market, allowing respectable economic growth despite industrial decline. Chile is similar in some ways, but it has not undergone such marked deindustrialization. It also has a relatively strong base of drivers, particularly skills and foreign direct investment, in which it ranks highest in Latin America. Unlike Hong Kong SAR, it is a resource-rich economy, with copper accounting for most of its traditional exports. In recent years, with government assistance, the economy has developed other resourcebased activities for export (mainly wine, farmed fish and pulp and paper). Chile lags in manufacturing value added (MVA) per capita and in the technological structure of MVA and exportsthe reason that it underperforms in the Scoreboard. Even so, Chiles manufacturing has grown at reasonable ratesat least by Latin American (if not East Asian) standards; it grew by 4.6 percent a year in the 1990s (well below its GDP growth of 6.7 percent).17 The share of manufacturing in GDP has fallen from 21 percent in the mid-1980s to 15 percent today.18

Balanced low All other developing economies (45 in 1998 and 44 in 1985)
Source: UNIDO Scoreboard database (see technical annex). Note: The analysis is based on the difference between the CIP index and the composite index of the five drivers. Balanced economies have values within the range defined by the standardized mean plus or minus its standard deviation. Overperformers have values above that range, and underperformers values below that range. Balanced high and low economies have values above and below the standardized mean of the CIP index and the composite index of drivers.

conductors for much of its export growth. That dependence makes it vulnerable to an erosion of competitiveness: rising wages, changing technologies, failing fortunes for the semiconductor industry and the like can easily lead to a rapid falloff in performance. Thus strong performance based on weak drivers immediately raises questions about sustainability and signals a need to expand the base of drivers. Some economies that have been successful in raising GDP such as Hong Kong SAR and Chileare nevertheless in the group of underperformers. Why? One reason is that the Scoreboard deals with manufacturing rather than GDP performance. The second is that the categorization is based on the ratio of industrial performance to drivers. An economy may rank high in industrial performance yet still be an underperformer if its CIP index falls below the band for countries with similar stocks of drivers. This might occur, for example, if its drivers are directed into non-manufacturing activities that yield relatively high incomes without showing up in the CIP index. Such factors explain the outcomes for Hong Kong SAR and Chile. Hong Kong SAR has had high income and respectable rates of growth over a long period. Some years ago it was also a strong

Using the Scoreboard for formulating industrial strategy

There are three main steps involved in using the Scoreboard and going beyond it to formulate industrial strategy (table 4.5). The first is to identify the main comparators that provide the benchmarks for a country. These can be of four types: Neighbours that share similar advantages and disadvantages of location or resources. For example, India might choose Pakistan as one benchmark because of similarities in the costs of transport to their main markets, their common specialization in cotton-based manufactured exports and their similar wage costs. Immediate competitors in industrial activities relevant to the country. Some of these may be neighboursothers

Benchmarking the drivers of industrial performance

73

Table 4.5 Using the Scoreboardand going beyond it

Step 1. Identify comparators

2. Benchmark performance

3. Benchmark drivers

Some important issues Which comparators can provide useful information? 1.1 Identify neighbours 1.2 Identify immediate competitors For which activities are the comparators useful? 1.3 Identify potential competitors What is a manageable number of comparators? 1.4 Identify role models How has the country performed over time in global 2.1 Compare overall industrial performance 2.2 Compare basic indicators of industrial performance or regional rankings? 2.3 Trace competitive strengths and weaknesses Is the industrial structure suited to growth and the with respect to different sets of comparators best use of local resources and capabilities? Which comparators have been more successful than the country or vice versa? How far from or close to selected benchmarks is the country? In which aspect of performance does the country lead or lag? Does the performance of comparators suggest cause for concern about any aspect of performance? Is there a need for more detailed technical benchmarking of particular industries, clusters or technologies? What are the relative strengths and weaknesses in 3.1 Compare individual elements of drivers 3.2 Trace competitive strengths and weaknesses the capabilities of the selected country? with respect to different sets of comparators Do the general indicators capture the underlying 3.3 Assess which drivers are most important drivers at work? If not, how can they be refined? for improved performance Which drivers constitute the most critical constraints 3.4 Add new data and analysis as necessary to industrial growth and competitiveness? Is there enough information to evaluate nonquantifiable variables such as linkages, institutions and governance? If not, how can more information be obtained?

may be located across the world. In automobiles Brazil might compete directly with Mexico in some products and with Europe or Asia in others, while in shoes its competitor might be India or China. Potential competitorscountries likely to emerge as challengers in the near future. Many economies in East Asia regard the entry of China into technology-intensive activities as a major threat. Role modelscountries more advanced in industry and technology and thus able to provide benchmarks to which to aspire. Many developing countries look to the East Asian Tigers or the new Tigers (the second wave of export-oriented countries, such as Malaysia and Thailand) as countries that have successfully overcome latecomer disadvantages. Others look to mature industrialized countries for long-term benchmarks. Once comparators have been identified, the next step is to compare the countrys industrial performance (the CIP index) with its benchmarks. Since data on each basic component of the index are given separately, each element can be benchmarked and evaluated separatelybreaking down the com-

ponents of performance is useful to identify where strengths and weaknesses lie. This general benchmarking can be supplemented by more detailed benchmarking at the level of industry, technology or cluster. The third step is to benchmark the drivers. This can be done using the data in the Scoreboard, for the current position as well as for changes over time. Bear in mind that some measures are aggregate and may need to be extended and refined to draw comparisons with other countries. Even as they stand, however, the measures allow analysis of drivers and performance that can show broad areas of strength and weakness. But if the analysis is to lead to policies, the Scoreboard must be supplemented by deeper analysis of the policy and regulatory regime, institutions, linkages and factors that could not be taken into account in the quantitative comparisons. Many of these can also be benchmarked against selected comparators, though it is difficult to do this for the large sample used in the Scoreboard. Most country competitiveness analyses do just this, but such analyses have to be based on painstaking collection of detailed information and careful qualitative analysis.

74

Industrial Development Report 2002/2003

Appendix 4.A. Statistical table

Table A4.1 Ranking of economies by the drivers of industrial performance, 1985 and 1998 R&D spending per capita by productive enterprises 1998 1985 Switzerland Germany Japan Sweden United States Germany Finland Denmark France

Skills index Rank 1998 1985 1 Korea, Korea, Republic of Republic of 2 Finland United States 3 Australia Canada 4 Canada Finland 5 6 7 8 United States Argentina Russian Federation Spain Taiwan Province of China New Zealand United Kingdom Norway Ireland Austria Sweden New Zealand Spain Sweden

Infrastructure index 1998 1985 United United States States Switzerland Belgium New Zealand Singapore Hong Kong Canada Canada SAR Japan Sweden Switzerland Netherlands Netherlands Singapore Sweden United New Zealand Australia Hong Kong Switzerland Sweden Finland States SAR Sweden Hong Kong Belgium Switzerland Belgium Norway Norway SAR Denmark Netherlands Oman Malaysia New Bahrain Denmark Zealand Norway Norway Netherlands Belgium Sweden Finland Germany Netherlands Denmark Hong Kong Sweden Canada Switzerland Switzerland SAR Canada France Switzerland Ireland Australia United Kingdom France Austria United States United Kingdom Malaysia Canada Ireland Greece United Kingdom Austria Finland Norway Japan New Zealand Canada Taiwan Province of China Germany Australia Korea, Republic of Spain Australia Finland Ireland Japan Norway France Denmark Australia

Foreign direct investment per capita 1998 1985 Singapore Singapore

Royalties per capita 1998 1985 Ireland Singapore

9 10 11 12 13 14

Germany Denmark Belgium France Norway Taiwan Province of China Uruguay

Norway Belgium

Netherlands Netherlands Australia France Belgium Germany France Belgium New Zealand United Kingdom Japan Austria

Netherlands Austria Austria Finland Korea, Republic of Singapore Israel United Kingdom Belgium Taiwan Province of China New Zealand Italy Singapore

15 16

Germany

Netherlands Israel

United Kingdom Ireland

Canada United States Finland Hungary Malaysia

Spain Norway

Germany Austria

United Kingdom Japan

17 18 19

France Denmark Greece

Ireland Australia Ecuador

Australia Canada Israel

France Sweden Tunisia

United Kingdom Argentina Taiwan Province of China Korea, Republic of Spain Israel Italy Panama United States Ecuador Egypt

New Zealand Hong Kong SAR Austria

Singapore Czech Republic Italy

20

Belgium

Austria

21 22 23 24 25

Portugal Japan Israel Italy Chile

Panama Greece Japan

Taiwan Province of China Italy Slovenia Spain

Hungary

Chile

Jordan

Greece

Greece

Ireland Australia Korea, Republic of Spain

Israel Panama Spain Argentina

Austria Costa Rica Israel Chile Italy

France United States Israel Portugal Hungary

Korea, Republic of Israel Ireland Italy Taiwan Province of China Czech Republic Spain

Israel Hong Kong SAR Saudi Arabia Ireland Taiwan Province of China Spain Poland Romania Korea, Republic of

Czech New Republic Zealand Netherlands Czech Republic Switzerland United Kingdom Philippines Venezuela Portugal Brazil

South Africa Portugal

26 27 28 29

Argentina Switzerland Slovenia Panama

Poland Greece

Czech Republic Costa Rica Mexico Greece

Portugal Colombia Finland Cameroon

Italy Slovenia Thailand Argentina

Greece Mexico South Africa Romania

South Africa Slovenia Costa Rica Portugal

Benchmarking the drivers of industrial performance

75

Table A4.1 Ranking of economies by the drivers of industrial performance, 1985 and 1998 (continued) R&D spending per capita by Skills index productive enterprises Rank 1998 1985 1998 1985 30 Philippines Singapore Hungary Jordan 31 Singapore Italy Argentina Chile 32 33 34 35 Peru Bahrain Poland Czech Republic Hong Kong SAR Romania Venezuela Colombia Jordan Costa Rica Uruguay Mexico Bolivia Algeria Turkey Ecuador Hungary Peru Chile Jordan Hong Kong SAR Mexico Poland Russian Federation. Malaysia Costa Rica Turkey Portugal Brazil Ecuador

Foreign direct investment per capita 1998 1985 Slovenia Argentina Peru Mexico Venezuela Poland Germany Taiwan Province of China Italy Egypt Brazil Denmark Panama

Royalties per capita 1998 1985 Jamaica Portugal Czech Denmark Republic Denmark Uruguay Brazil Egypt Panama Malaysia Indonesia Chile

Infrastructure index 1998 1985 Hungary Hungary Mauritius South Africa Russian Venezuela Federation. Saudi Arabia Portugal Poland Turkey Argentina Oman

36

Chile

India

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

Costa Rica Romania Poland Colombia

Turkey Romania Venezuela Hong Kong SAR Egypt Mexico Bolivia Panama Portugal Uruguay El Salvador China Hungary Indonesia South Africa India Turkey Mauritius Morocco Thailand Egypt

Tunisia Malaysia Thailand Philippines Indonesia Kenya Madagascar Malawi Jamaica Argentina Costa Rica Venezuela

Colombia Jamaica Brazil Bolivia Ecuador Uruguay Tunisia Paraguay Thailand Oman South Africa Korea, Republic of China Mauritius Romania Philippines Indonesia Albania Morocco

Taiwan Province of China Germany Guatemala S Arabia Venezuela Ecuador Thailand Nigeria Paraguay Bolivia Honduras Mauritius Uruguay Korea, Republic of Zambia South Africa Japan Sri Lanka Morocco Central African Republic El Salvador Yemen Turkey Indonesia Senegal Peru Philippines

Morocco

Mexico

Malaysia

Mexico

Costa Rica Ecuador Greece Mexico Poland Indonesia South Africa Chile Peru Philippines Turkey Uruguay Kenya Colombia El Salvador Romania Honduras Madagascar Bolivia

Thailand Poland Jamaica Greece Zimbabwe Honduras Morocco Bolivia Philippines El Salvador Brazil Colombia Peru Turkey Tunisia Cameroon Kenya Pakistan Romania

Uruguay Argentina Chile Romania South Africa Jamaica Venezuela Oman Colombia Costa Rica Panama Mexico Brazil Thailand Jordan Tunisia China El Salvador Ecuador

Jordan Malaysia Uruguay Panama Turkey Costa Rica Brazil Colombia Algeria Chile Jamaica Ecuador Tunisia Mauritius Egypt Peru Bolivia Zimbabwe China

South Africa Thailand Thailand El Salvador Tunisia Egypt Morocco Indonesia

Hong Kong SAR Saudi Arabia Colombia Panama Algeria Jordan Uruguay Brazil Guatemala China Honduras Algeria Mauritius Nicaragua Saudi Arabia Egypt Tunisia Peru Colombia

56 57 58 59 60 61 62

Saudi Arabia Malaysia Nicaragua Sri Lanka Brazil Paraguay Malaysia Honduras Jamaica China India Jamaica Indonesia Guatemala

Morocco Philippines Honduras Nicaragua Sri Lanka Yemen Tunisia

Guatemala Nicaragua Algeria Jordan Saudi Arabia Russian Federation Peru Saudi Arabia Morocco Nigeria Honduras Egypt Nicaragua Turkey

Zimbabwe China Tunisia Senegal India Pakistan Tanzania, United Republic of Paraguay Cameroon Bangladesh Russian Federation Albania Algeria

Algeria Senegal India Nigeria Ghana Madagascar Paraguay

Algeria Peru Paraguay Bolivia Egypt Morocco Guatemala

El Salvador Nicaragua Thailand Paraguay Philippines Morocco Guatemala

63 64 65 66 67 68

Sri Lanka Paraguay Guatemala Albania India Zimbabwe

Oman China Zimbabwe Nepal

Malawi Sri Lanka Madagascar Yemen Kenya Albania Jamaica Bahrain Bangladesh Bolivia

Honduras Sri Lanka Guatemala Ghana Yemen Japan

Kenya Pakistan China Ghana

Bangladesh China Russian Federation Albania Bahrain Central African Republic Ethiopia Guatemala

Honduras Philippines Indonesia Zimbabwe Nicaragua Sri Lanka

Zambia Yemen Honduras Indonesia Pakistan India

Mauritius Ecuador Madagascar Albania

69 70

Mauritius Oman

Bangladesh Ghana

Bahrain Bangladesh

Cameroon Central African Republic

Zambia Senegal

Poland Tanzania, United Republic of Madagascar Bahrain Malawi Central African Republic

Albania India

Nigeria Kenya

76

Industrial Development Report 2002/2003

Table A4.1 Ranking of economies by the drivers of industrial performance, 1985 and 1998 (continued) R&D spending per capita by Foreign direct Skills index productive enterprises investment per capita Royalties per capita Infrastructure index Rank 1998 1985 1998 1985 1998 1985 1998 1985 1998 1985 71 Nepal Pakistan Bolivia El Salvador Uganda India Ethiopia Jordan Pakistan Senegal 72 Nigeria Nigeria Cameroon Ethiopia Pakistan Mozambique Ghana Malawi Nigeria Cameroon 73 Bangladesh Cameroon Central Ghana Zimbabwe Zimbabwe Guatemala Mauritius Zambia Sri Lanka African Republic 74 Cameroon Senegal El Salvador Mozambique Tanzania, Ethiopia Jordan Mozambique Kenya Ghana United Republic of 75 Pakistan Kenya Ethiopia Nepal Mozambique Nepal Malawi Nepal Senegal Malawi 76 Madagascar Zambia Ghana Nigeria El Salvador Bangladesh Mauritius Nicaragua Mozambique Mozambique 77 Senegal Ethiopia Mozambique Oman India Uganda Mozambique Oman Ghana Madagascar 78 Yemen Uganda Nepal Pakistan Bahrain Algeria Nepal Saudi Arabia Tanzania, Tanzania, United United Republic of Republic of 79 Ghana Central Nigeria Paraguay Cameroon Jamaica Nicaragua Sri Lanka Cameroon Bangladesh African Republic 80 Zambia Malawi Oman Senegal Madagascar Bahrain Nigeria Uganda Nepal Uganda 81 Kenya Tanzania, Pakistan Tanzania, Nepal Hungary Oman Venezuela Yemen. Central United United African Republic of Republic of Republic 82 Uganda Mozambique Paraguay Uganda Kenya Czech Saudi Arabia Zambia Ethiopia Ethiopia Republic 83 Central Albania Senegal Zambia Central Slovenia Sri Lanka Hungary Malawi Nepal African African Republic Republic 84 Ethiopia Bahrain Tanzania, Zimbabwe Algeria Romania Uganda Slovenia Bangladesh Albania United Republic of 85 Mozambique Russian Uganda Russian Bangladesh Albania Venezuela Czech Madagascar Bahrain Federation Federation Republic 86 Malawi Slovenia Zambia Slovenia Malawi Nicaragua Yemen Tanzania, Uganda Russian United Federation Republic of 87 Tanzania, Yemen Zimbabwe Czech Ethiopia Russian Zambia Yemen Central Slovenia United Republic Federation African Republic of Republic
Source: UNIDO Scoreboard database (see technical annex). Note: Economies in italics have negligible values, and those in bold italics have missing data. These economies are not ranked in the analysis; their location in the table is incidental.

Benchmarking the drivers of industrial performance

77

Appendix 4.B. Country cases from the UNIDO Scoreboard

This appendix illustrates how the Scoreboard can help in analyzing industrial performance and its drivers. It uses selected country cases from different regions and occasionally supplements the Scoreboard data with other information to enrich the analysis. As noted in the chapter, however, using the Scoreboard in country analysis would require much more detailed information than that here.

Figure A4.1 Technological structure of manufacturing production and exports in selected industrialized countries (percent) Share of medium- and high-tech products in manufactured exports 1985 1998 90 Japan 80 Japan

Industrialized countries
As would be expected, industrialized countries as a group lead the sample in both industrial performance and its drivers. Their rankings have remained relatively stable over time again as would be expected for countries with mature industrial sectors that cannot change structural parameters much over the medium term. Nevertheless, there are some interesting shifts, so some structural change is clearly possible. The most striking case is that of Ireland, which improved its ranking by the competitive industrial performance (CIP) index by 12 places between 1985 and 1998. At the other end of the spectrum is New Zealand, which lost 7 places.

70 United States 60 United States 50 Ireland

40 Ireland 30

20 New Zealand Australia 10 New Zealand Australia 0 30 40 50 60 70 80 Share of medium- and high-tech activities in manufacturing value added
Source: UNIDO Scoreboard database (see technical annex).

IRELAND
One of the most dynamic industrialized economies, Ireland improved its CIP rank from 15th in 1985 to third in 1998, just behind Singapore and Switzerland. Not only has it significantly increased industrial production and manufactured exports per capita, it has also achieved high levels of technological sophistication in both (figure A4.1). Yet despite the fairly large share of medium- and high-tech products in its manufactured exports, its rank by this measure fell from 12th to 17th over the period. The main reason is the rise in the ranks by such developing economies as Singapore, the Philippines, Mexico, Malaysia and Taiwan Province of China, with their even more technology-intensive export structures. Irelands drivers match its dynamic performance. Figure A4.2 shows shifts in Irelands ranks for each indicator of industrial performance and each of the five drivers (a move towards the centre signifies an improvement in rank). Irelands manufacturing value added (MVA) and manufactured exports per capita rose rapidly in 19851998, propelling the country to close to the top of the ranking for these measures. Total MVA grew by 13.6 percent a year and manufactured exports by 15.8 percent (two or more percentage points higher than the average for industrialized countries). The share of medium- and high-tech products in MVA rose by 12 percentage points, and their share in manufactured exports by 11 percentage points.

Irelands success was driven mainly by foreign direct investment combined with a massive upgrading of human capital. Starting from a weak inherited technological and industrial base, Ireland used incentives and targeting to attract transnational corporations into high-tech activities.19 Access to the European Unions market was, of course, an important magnet for transnational corporations. Ireland drew heavily on EU assistance to develop its physical infrastructure. Inward foreign direct investment grew rapidly in 19851998 (at 19 percent a year on a per capita basis). Payments of royalties and licensing fees to foreign firms also rose sharply, reaching $6 billion by 1998, around 9 percent of global royalty payments (and the highest level in per capita terms). Irelands drive into sophisticated manufacturing and software activities required many skilled workers, and the country became a global leader in enrolments. It now ranks higher than Germany, the United

78

Industrial Development Report 2002/2003

Figure A4.2 Changing ranks in industrial performance indicators and drivers for Ireland 1985 1998 MVA 30 Royalties 20 10 FDI 0 Exports

Figure A4.3 Changing ranks in industrial performance indicators and drivers for New Zealand 1985 1998 MVA 50 Royalties 40 30 20 Exports

MHT in MVA

FDI

10 0

MHT in MVA

R&D

MHT in exports Skills Infrastructure

R&D

MHT in exports Skills Infrastructure

Source: UNIDO Scoreboard database (see technical annex).

Source: UNIDO Scoreboard database (see technical annex).

Kingdom or the United States in tertiary students enrolled in technical subjects as a share of the population. The Industrial Development Authority of Ireland, an organization with highly qualified staff and a strong private sector orientation, managed the promotion and targeting of foreign direct investment. But the authority (later, Enterprise Ireland) was more than an investment promotion agency. It also managed industrial policy and so was able to ensure that the needs of foreign investors were met. This strategy making and coordination function allowed it to mastermind the technological upgrading of the industrial sector (much as the Singapore Economic Development Board has done in that country). The authority later mounted strategies to strengthen links between local suppliers and affiliates of transnational corporations and to induce affiliates to deepen local technological activity and undertake research and development (R&D).20 As a consequence, enterprise-financed R&D per capita rose from $14 in 1985 to $153 in 1998; total enterprise-financed R&D grew by 20 percent a year. Ireland now ranks ahead of Australia, Canada, Israel and Taiwan Province of China in enterprise-funded R&D. Although affiliates of transnational corporations perform most of the R&D, local companies are also increasing R&D efforts.

exports per capita have been lower than the average for industrialized economies, pulling it down in the rankings by these measures by three and six places. More important, the share of medium- and high-tech products in its manufactured exports (14.5 percent) is now the lowest in the industrialized world. And the share of complex activities in its MVA exceeds only that of Greece and Portugal among industrialized countries. A number of newly industrializing and transition economies now rank higher than New Zealand by various measures of industrial performance. Still, New Zealand is strong in some drivers of industrial performance. It has one of the largest skill bases in the world (though its ranking by the skills index has declined over time). Its foreign direct investment inflows are among the largest in the world on a per capita basis, and as a share of GDP (4.8 percent) they exceed those to Belgium (3.9 percent), Sweden (3.6 percent) and the Netherlands (2.6 percent). Other drivers are not as strong. New Zealand lags behind other industrialized countries in infrastructure and R&D. Although enterprise-financed R&D per capita doubled in nominal terms between 1985 and 1998, the level in 1998 was an eighth of the average for industrialized countries and only a third of that in Australia, another resource-rich industrialized country. What is more, R&D declined as a share of GDP in New Zealand. Foreign direct investment goes less into manufacturing than into services, and little is aimed at high-tech activity.

NEW ZEALAND
New Zealand presents something of contrast to Ireland (figure A4.3). Its growth rates for MVA and manufactured

Benchmarking the drivers of industrial performance

79

Developing countries LATIN AMERICA AND THE CARIBBEAN

Industrial performance has been variable in Latin America and the Caribbean. Of the 18 countries from that region in the sample, 7 (Bolivia, Chile, Costa Rica, El Salvador, Guatemala, Honduras and Mexico) improved their CIP ranks between 1985 and 1998, while 11 saw their ranks deteriorate. The laggards include large, industrialized economies like Argentina and Brazil. Compared with the rest of the developing world, the region is relatively well placed in most drivers of industrial performance but lags in domestic technological effort. Mexico and Jamaica illustrate the range of experience in the region: the first performs relatively well in the Scoreboard, while the second does not. Their ranks by performance and

drivers show, at least in part, the reasons for the differences in outcomes for the two. Mexico. Mexico led the region in 1998 in industrial performance, having overtaken Brazil in the period since 1985. It recorded an impressive increase in manufactured exports, with annual growth of 21.4 percent, and vigorously upgraded the technological structure of these exports (figure A4.4). In fact, Mexico now has one of the most technology-intensive export structures in the world, ranking ahead of the United States and behind only Japan, the Philippines and Singapore. But Mexicos production record is less impressive. While its annual growth in MVA in 19851998 (6.2 percent) exceeded the average for Latin America and the Caribbean (5.9 percent), it fell short of the average for East Asia (9.3 percent). More important, the share of medium- and high-tech products in MVA has declined over time. This differentiates Mexico from Brazil, the other Latin American industrial giant. While Brazil has a less technology-intensive export structure, its MVA structure is far more complex (ranking Brazil 13th globally) and has become more so over time. Mexicos domestic industrial sector seems to be on a different trajectory than its export sector. The main driver of Mexicos industrial performance is activity by transnational corporations in the maquiladoras on the U.S. border. Average annual per capita inflows of foreign direct investment grew from $16.5 in 19811985 to $102.4 in 19931997. In 1998 Mexicos foreign direct investment inflows (as a share of GDP) exceeded those to Hong Kong
Figure A4.5 Changing ranks in industrial performance indicators and drivers for Mexico 1985 1998 MVA 60 Royalties Mexico Brazil Argentina FDI Argentina 40 20 0 Exports

Figure A4.4 Technological structure of manufacturing production and exports in selected countries in Latin America and the Caribbean (percent) Share of medium- and high-tech products in manufactured exports 1985 1998

70
Mexico

60

50

40
Costa Rica Brazil

30

20

MHT in MVA

10

Costa Rica Jamaica Jamaica

10

20

30

40

50

60

70

R&D

MHT in exports Skills Infrastructure

Share of medium- and high-tech activities in manufacturing value added

Source: UNIDO Scoreboard database (see technical annex).

Source: UNIDO Scoreboard database (see technical annex).

80

Industrial Development Report 2002/2003

Special Administrative Region (SAR) of China, Thailand, the Philippines and Indonesia. But this high dependence on internalized technology transfer has done nothing to enhance domestic innovation; on the contrary, enterprise-financed R&D spending per capita in 1998 was half that in 1985, and Mexicos rank dropped by 13 places (figure A4.5). Enterprisefinanced R&D fell as a share of GNP by nearly 8 percent a year over the period. Just as worrying is that Mexicos skill base and physical infrastructure are deteriorating relative to those of competitors. Moreover, the maquiladoras have only weak links to local industry for inputs and technology.21 The domestic content of export production in Mexico remains low, though it is rising slowly.22 The real drivers of its industrial success have been low labour costs, locational advantages and the privileges gained through the North American Free Trade Agreement (NAFTA). Despite its large and well-established industrial sector, Mexico makes little use of the modern drivers of industrial competitiveness. Unless the domestic content of industrial activity increases, with a better match between export and MVA structures, the sustainability of the countrys industrial growth remains open to question. Jamaica. Jamaica lies almost at the other end of the technological spectrum. It suffered the largest deterioration in CIP ranks in the region between 1985 and 1998, falling from 52nd to 65th. Although its MVA and manufactured exports have grown at respectable rates since 1985 (6.7 percent and 6.9 percent), it has not kept pace with competitors. Moreover, the technological structure of Jamaicas MVA and exports has regressed. The country now has one of the smallest shares of medium- and high-tech manufactured exports in the sample, leading it to lose 22 places in the rankings by this measure. Jamaicas experience can be contrasted with that of neighbouring Costa Rica, which has managed a rapid transition from resource-based and labour-intensive exports to technology-intensive industrial activity. Costa Rica has achieved this success through targeted foreign direct investment policies and skill creationthe leading example of Irish- or Singaporestyle industrial policy in Latin America and the Caribbean. It has attracted a $500 million semiconductor plant from Intel,23 which has transformed its production and export structure and is leading to extensive spillover benefits.24 Jamaica performs poorly in some industrial drivers, ranking fairly low on the skills index (where its rank has stagnated) and with negligible enterprise-financed R&D (figure A4.6). But its foreign direct investment rank has improved sharply (by 41 places), with per capita inflows rising from negative figures in 1985 to almost $60 in 1998. This record is misleading, however, since the negative inflows in 1985 were clearly an aberration. More important is that the investment seems to have

Figure A4.6 Changing ranks in industrial performance indicators and drivers for Jamaica 1985 1998 MVA 80 Royalties 60 40 FDI 20 0 MHT in MVA Exports

R&D

MHT in exports Skills Infrastructure

Source: UNIDO Scoreboard database (see technical annex).

done little to improve Jamaicas industrial structure or competitiveness; most probably went into services and utilities. Jamaica has used its locational advantages and low wages to build up some (labour-intensive) exports, but its export industry has remained concentrated in low-tech activities (the countrys export structure is among the least technology intensive in the sample). With no technological upgrading over time, Jamaica has a weak base for future industrial growth. Its exports will come under increasing threat from lower-wage countries, particularly those in other regions, as trade is further liberalized and its special access to the U.S. market is reduced. Without a move up the technological ladder (probably best achieved by emulating Costa Rica), Jamaica may find itself increasingly marginalized in industry.

EAST ASIA
East Asia leads the developing world in industrial performance as well as in its drivers. While there is much diversity among East Asian economies, the leaders in the region are among the global leaders in several indicators (figure A4.7). Since the experience of the Republic of Korea, Singapore and Taiwan Province of China is fairly well known,25 the focus here is on two other economies, China and the Philippines. China is of obvious interest: it is the leading industrial power in the developing world and poses a strong competitive challenge to other economies in the region and elsewhereacross the entire technological spectrum. The Philippines is not a major industrial power, but its experience offers interesting lessons,

Benchmarking the drivers of industrial performance

81

Figure A4.7 Technological structure of manufacturing production and exports in selected economies in East Asia (percent) Share of medium- and high-tech products in manufactured exports 1985 1998

Figure A4.8 Changing ranks in industrial performance indicators and drivers for China 1985 1998 MVA 80 Royalties 60 40 FDI 20 0 MHT in MVA Exports

80
Philippines

Singapore

70
Taiwan Province of China Republic of Korea

60
R&D Republic of Korea Singapore MHT in exports Skills China Infrastructure

50

40

Taiwan Province of China

30

Source: UNIDO Scoreboard database (see technical annex).

20
Philippines

10
China

complex products. Exports are balanced almost evenly between local and foreign firms: in 1998 local firms accounted for 52 percent, and foreign firms for 48 percent, of total exports.
50 60 70 80 90

10

20

30

40

Share of medium- and high-tech activities in manufacturing value added

Source: UNIDO Scoreboard database (see technical annex).

both positive and negative. Much like Mexico, it shows that integration into high-tech global production systems can boost national performance. China. China rose in the CIP ranks by 24 places between 1985 and 1998the largest jump in the sample. Its rank in three of four performance indicators improved significantly, with only its rank in the technological structure of MVA deteriorating (figure A4.8). China combines strong growth in per capita MVA and manufactured exportsaveraging 9.8 percent and 29.1 percent a year in 19851998with rapid upgrading of its export structure. China exploits competitive advantages in both labour-intensive and technology-based (assembly) activities. Its special economic zones, with a strong presence of exporters from Hong Kong SAR and Taiwan Province of China, drive the low-tech activity. And transnational corporations and some dynamic domestic firms drive exports of technologically

China sharply increased its imports of foreign technology in 19851998, as reflected in its improved ranks in foreign direct investment and royalty payments. In 1998 only the United States received more foreign direct investment (in absolute terms). China also has slightly improved its ranks in skills and R&D. It invests almost as much in enterprise-financed R&D (in absolute terms) as the Russian Federation and nearly three times as much as India. China accounts for 17 percent of tertiary enrolments in the developing world and lags behind only the United States and the Russian Federation in the number of tertiary students enrolled in technical subjects. Even so, Chinas skill base appears weak relative to its size. Moreover, its infrastructure rank did not improve in 19851998. While China is rapidly strengthening both these drivers of industrial performance, it still has some distance to go before it matches the regions leading industrializing economies. Philippines. The Philippines moved up 20 places in the CIP ranking, from 45th in 1985 to 25th in 1998higher than industrialized countries such as Australia, New Zealand and Portugal and transition economies such as Hungary and Slovenia. Rapid growth in manufactured exports (more than 20 percent a year in 19851998) and improvements in its

82

Industrial Development Report 2002/2003

Figure A4.9 Changing ranks in industrial performance indicators and drivers for the Philippines 1985 1998 MVA 80 Royalties 60 40 FDI 20 0 MHT in MVA Exports

There is a lack of connectivity between the Philippines high-tech export activity and its domestic drivers, reflected in the fact that its rank fell by 17 places in R&D, 4 places in the infrastructure index and 2 places in the skills index. The only driver of industrial performance in which its rank improved is foreign direct investment, underlining the fragility of the countrys industrial success.

SOUTH ASIA
The industrial performance of South Asia is strikingly different from that of East Asia. South Asia ranks low in the CIP index and has weak drivers of industrial performance, comparable to those of Sub-Saharan Africa except for skills. This may seem surprising in view of the regions history of heavy industrialization and the recent surge in software exports by India, its main industrial economy. But even India has a relatively weak base of industrial capabilities relative to the size of its economy. Moreover, small islands of technological success, as in software, do not reflect the dynamism and competitiveness of the industrial sector as a whole. Export structures in the region are relatively weak, though India has a complex MVA structure (figure A4.10). Other economies

R&D

MHT in exports Skills Infrastructure

Source: UNIDO Scoreboard database (see technical annex).

export structure account for its success, rather than growing MVA or better capabilities (figure A4.9). Like Mexico, it shows that countries with relatively weak drivers can upgrade rapidly by plugging into high-tech global production systems. And like Mexico, it is not assured of continued success, because its domestic base for growth remains weak. In fact, lacking the locational advantages of Mexico and with a much narrower export specialization, the Philippines is more vulnerable to technical change and emerging competition.26 Medium- and high-tech products increased from 10.5 percent of the Philippines manufactured exports in 1985 to 74.4 percent in 1998, and the countrys recent export growth has been among the strongest in the region (it fared better than most other large exporters during the East Asian financial crisis). The Philippines has overtaken Malaysia as the main base in the region for assembly and testing of semiconductors for transnational corporations, and this product accounts for more than 70 percent of manufactured exports. But manufacturing production has grown slowly (by 4.8 percent a year), leading to a loss of three places in the MVA per capita ranks. The structure of MVA is biased towards low-tech and resource-based activities, creating a sharp disparity in technological complexity between exports and domestic production (see figure A4.7). The local content of semiconductor assembly remains low, and traditional labour-intensive exports (such as apparel) are weakening. Other technology-intensive export activities have yet to take root.

Figure A4.10

Technological structure of manufacturing production and exports in selected countries in South Asia (percent)

Share of medium- and high-tech products in manufactured exports 1985 1998

20
India

16

12
Pakistan India Pakistan Sri Lanka

4
Sri Lanka

Bangladesh Bangladesh

10

20

30

40

50

60

70

Share of medium- and high-tech activities in manufacturing value added

Source: UNIDO Scoreboard database (see technical annex).

Benchmarking the drivers of industrial performance

83

have less advanced industrial sectors, though their export patterns look surprisingly similar to that of India. India. India led South Asia in the CIP ranking in 1998, though its rank had changed little since 1985. During this 13-year span Indias industrial structure remained static, and the country was overtaken in the CIP ranking by such economies as China and Indonesia, mainly because of their greater orientation towards technologically complex export industries. India had annual growth of MVA (5.1 percent) lower than the regions average (5.5 percent), and its MVA per capita in 1998 ($65.2) was similar to that of Cameroon ($64.6) and Nigeria ($62.2). Its manufactured exports reached only $26.4 per capita in 1998, less than in Senegal ($34.5), Cameroon ($34) or Kenya ($28.3). The annual growth of its manufactured exports (11.6 percent), while healthy, was less than the regions average (12.4 percent), which was in turn lower than the average for all developing economies (13.3 percent) and for East Asia (14.5 percent). What really distinguishes India from other countries in South Asiaand from Africais the depth and complexity of its industrial structure (see figure A4.10). Medium- and hightech products accounted for almost 60 percent of Indias MVA in 1998, compared with less than 10 percent for its neighbours. In fact, India ranked 12th globally in the technological complexity of MVA, ahead of Taiwan Province of China, Brazil or Chinaa legacy of its prolonged import substitution strategy emphasizing heavy industry. But this strategy has also left Indias diverse manufacturing sector with large technological and competitive gaps relative to world frontiers (figure A4.11). Indias manufactured exports have a strikingly different structure than its manufacturing production, with low-tech products (textiles and clothing) and resource-based activities (diamond cutting) accounting for 62 percent. The gap between the two structures reflects the competitive gap in large sections of Indian industry. Indias most dynamic export activitysoftwareis not captured by the figures, which exclude services. This competitive gap in manufacturing is the result not only of past trade policies but also of weaknesses in industrial drivers.27 The countrys skill base is weak and deteriorating relative to those of competitors. The absolute numbers are, of course, largeIndia accounted for 16.3 percent of tertiary enrolments in the developing world in 1998, and it ranked behind only the United States, the Russian Federation and China in tertiary enrolments in technical areas. But the number of tertiary students enrolled in technical subjects is small relative to Indias population, and in contrast to the numbers in most other regions, it has declined since 1985.

Figure A4.11 Changing ranks in industrial performance indicators and drivers for India 1985 1998 MVA 80 Royalties 60 40 FDI 20 0 MHT in MVA Exports

R&D

MHT in exports Skills Infrastructure

Source: UNIDO Scoreboard database (see technical annex).

Still, the absolute number of technically qualified people is large, allowing India to foster nodes of skill-intensive activity. That explains the presence of software exports and of some competitive firms in complex activities. The real skill constraint will arise when India tries to upgrade technologies in a large range of activities to compete in liberalized markets. Other industrial drivers in India are also weak. Lags in infrastructure are widespread, a well-known problem confirmed by the Scoreboard data. More striking is that India is falling further behind its competitors in this driver of industrial performance. Moreover, enterprise-financed R&D was the same on a per capita basis in 1998 as in 1985 and declining as a share of GNP. Inward foreign direct investment has risen significantly in recent years, but the growth has not been enough to reverse a decline in Indias rank by this indicator. And in sharp contrast to foreign direct investment in China, little in India goes into export-oriented manufacturing. That puts India at risk of losing out on links with the most dynamic part of industrial export activity, integrated production systems in technology-intensive activities. Bangladesh. Bangladesh ranked 73rd in the CIP index in 1998. Despite respectable industrial growthMVA per capita rose by 6.6 percent a year between 1985 and 1998, from $33.5 to $59.6, and manufactured exports per capita by 14.7 percent, from $8.1 to $37.3it had gained only one place in the performance ranks. Part of the explanation lies in its overwhelming specialization in low-tech products, primarily clothing (which accounts for 90 percent of manufactured exports).

84

Industrial Development Report 2002/2003

This specialization grew rather than diminished over the period, as other exports failed to grow. This stagnation in technological development has worrying implications, not only because of what it denotes about industrial upgrading but also because clothing exports have grown on the back of quota protection under the Multi-Fibre Arrangement (MFA), which ends in 2004. Like other countries that have built manufactured exports on the basis of the MFA, Bangladesh faces a serious adjustment problem once competition with more efficient developing countries opens up fully. While some clothing exporters will clearly survive, they may not drive export growth in the way they have in the past and there are few signs of other dynamic manufacturing exporters replacing them. Bangladesh also does poorly in other drivers of industrial performance (figure A4.12). Enterprise-financed R&D is negligible. Foreign direct investment inflows have risen, but reached only $0.3 per capita in 1998 (ranking Bangladesh third from the bottom in the sample). Some foreign direct investment goes into export-oriented activity, but since clothing is a labour-intensive activity, the values involved are quite small. Infrastructure is weak, with Bangladesh ranking fourth from the bottom by this indicator. It had only three telephone mainlines per 1,000 people in 1998, even fewer than Malawi, Mozambique, Nigeria and the United Republic of Tanzania. Bangladesh lost four places in the combined infrastructure index, ranking 84th in 1998. Its skill base is weak and is deteriorating relative to that of competitors. Like Mauritius, Morocco, Sri Lanka and other countries (particularly in Central America) that have reached
Figure A4.12 Changing ranks in industrial performance indicators and drivers for Bangladesh 1985 1998 MVA 100 Royalties 80 60 40 FDI 20 0 MHT in MVA Exports

Figure A4.13

Technological structure of manufacturing production and exports in Turkey and selected countries in the Middle East and North Africa (percent)

Share of medium- and high-tech products in manufactured exports 1985 1998

25
Turkey

20
Tunisia Turkey Tunisia

15
Morocco

10
Morocco

Egypt Saudi Arabia

5
Saudi Arabia Egypt

10

20

30

40

50

60

Share of medium- and high-tech activities in manufacturing value added

Source: UNIDO Scoreboard database (see technical annex).

the first rung of industrial development on the basis of clothing, Bangladesh will need to greatly improve its industrial capabilities if it is to sustain industrial growth.

MIDDLE EAST AND NORTH AFRICA AND TURKEY

In industrial capabilities the Middle East and North Africa typically ranks at about the middle among developing regions, generally just after Latin America and the Caribbean. It has a relatively good base of skills and infrastructure, but lags in technological activity. Turkey specializes heavily in low-tech exports, with relatively slow upgrading over time (figure A4.13). Some countries, like Egypt, have a long history of import-substituting industrialization led by the state. Saudi Arabias export structure continues to be dominated by resource-based products. An increase in the share of mediumand high-tech products in manufactured exports over the years is indicative of the move towards export diversification. Turkey. Turkey placed 38th in the CIP ranking in 1998, leading the Middle East and North Africa despite having lost two places since 1985. It ranked just behind China and a little ahead of Greece, Romania and the Russian Federation. Despite reasonable industrial growth in 19851998, Turkey

R&D

MHT in exports Skills Infrastructure

Source: UNIDO Scoreboard database (see technical annex).

Benchmarking the drivers of industrial performance

85

has not kept up with many newly industrializing countries the Philippines, Thailand, Costa Rica and China have all jumped ahead in the CIP ranking. Turkeys total MVA grew by 10.3 percent a year between 1985 and 1998, and its total manufactured exports by 11.2 percent. Technologically complex industries raised their share in manufacturing production from 32 to 38 percent and in manufactured exports from 18 to 23 percent. That gave Turkey a more technology-intensive export structure than that of countries in the Middle East and North Africa, with the exception of Saudi Arabia. Saudi Arabias technology-intensive export structure is due largely to its petrochemical facilities. Even so, Turkey is not technologically advanced by the standards of East Asia or Latin America and the Caribbean. Most of its exports come from the clothing sector, where it has capitalized on low wages relative to those in Europe and on access to the EU market. It has high wages by Asian standards, however, and will soon face severe competition from lowercost countries, particularly China (whose exports of clothing and textiles rose from $2 billion in 1985 to almost $53 billion in 1998). Turkey will need to move into much higher quality segments, which require advanced design and marketing skills, in order to sustain a competitive edge here. In industrial drivers Turkey has both strengths and weaknesses (figure A4.14). It falls very low in the foreign direct investment ranking, despite its strong EU connection. Although per capita inflows of foreign direct investment rose

from $1.7 in 1985 to $12 in 1998, the average for developing countries is twice as high, and the average for the world six times as high. Tertiary enrolments in technical subjects are relatively low, with Turkey ranking 46th in the skills index in 1998 (up one place since 1985), lagging behind Bahrain and Jordan in the region. It fares better relative to the rest of the region in enterprise-financed R&D. But it lags behind much of Europe and several developing economies, such as the Republic of Korea, Taiwan Province of China, Brazil, South Africa, Malaysia, Chile and Costa Rica. Infrastructure, especially for information and communication technology, has improved greatly, with Turkeys rank by the infrastructure index rising by nine places (to 35th) between 1985 and 1998. Egypt. Egypt improved its CIP rank by 10 places, reaching 57th in 1998, the only country in the Middle East and North Africa to climb in this ranking. Its MVA grew by 9.5 percent a year between 1985 and 1998, more than the average for the developing world and for the region (7.7 percent). Manufactured exports grew even faster (13 percent), though the level remains low$36.5 per capita in 1998, less than those of Tunisia, Turkey, Pakistan, Bangladesh or Zimbabwe. Its domestic market orientation, a legacy of socialist and import substitution policies, is changing. But as in India, the policy of import substitution left a sharp divergence between MVA and export structures. Medium- and high-tech products account for a much larger share of MVA (39 percent) than of exports (8.8 percent). While Egypt is abandoning its past industrial strategies, it still has a long way to go in industrial drivers (figure A4.15). Per capita inflows of foreign direct investment remain very low ($13.3 in 1998) and have fallen over time (from $15.5 in 1985). As a result, Egypts rank by this indicator dropped by 29 places. Enterprise-financed R&D per capita, while rising, reached only $0.2 in 19984.3 percent of the average for developing countries and 14.3 percent of that for the Middle East and North Africa. Egypt is also losing ground to other developing countries in skills and infrastructure. The number of tertiary students enrolled in technical subjects declined from 75,000 in 1985 (0.15 percent of the population) to 70,000 in 1995 (0.12 percent), pushing Egypt nine places lower in the ranks. In the infrastructure index Egypt ranked 63rd in 1998, just ahead of the Central African Republic and Morocco, but it had lost nine places since 1985. This record suggests that Egypt can take advantage of its long industrial history, low wages and favourable location near Europe only if it is able to attract more foreign direct investment and improve its base of skills, technology and infrastructure.

Figure A4.14 Changing ranks in industrial performance indicators and drivers for Turkey 1985 1998 MVA 80 Royalties 60 40 FDI 20 0 MHT in MVA Exports

R&D

MHT in exports Skills Infrastructure

Source: UNIDO Scoreboard database (see technical annex).

86

Industrial Development Report 2002/2003

Figure A4.15 Changing ranks in industrial performance indicators and drivers for Egypt 1985 1998 MVA 80 Royalties 60 40 FDI 20 0 MHT in MVA Exports

Figure A4.16

Technological structure of manufacturing production and exports in selected countries in Sub-Saharan Africa (percent)

Share of medium- and high-tech products in manufactured exports 1985 1998

30
South Africa

25

20
R&D MHT in exports Zimbabwe Zimbabwe South Africa

15
Skills Infrastructure

10
Source: UNIDO Scoreboard database (see technical annex).

Kenya

SUB-SAHARAN AFRICA
Sub-Saharan Africa lags behind the other developing regions in all indicators of industrial performance and capabilities. Of the 15 African countries included in the Scoreboard, 14 fell in the CIP ranking between 1985 and 1998 (the exception was Kenya). Even the regional leaders dropped substantially, South Africa by 7 places, Zimbabwe by 13 and Mauritius by 9. The regional laggardsUganda, Ghana and Ethiopia have the weakest industrial performance in the world. Moreover, Kenyas improvement in the ranking had little to do with its industrial performance; instead, it was due primarily to the greater deterioration in the ranks of other countries in the region. The regions drivers are also weak. Take tertiary enrolment in technical subjects. In 1985 Morocco had almost the same number of tertiary students enrolled in technical subjects as all of Sub-Saharan Africa excluding South Africa. In 1998 Turkey had more students in technical subjects than SubSaharan Africa, this time including South Africa. In the composite skills index, 15 of 16 countries in the region ranked among the bottom 20. Sub-Saharan Africa also ranks last among regions in the infrastructure index. It accounts for less than 1 percent of the worlds telephone mainlines, mobile phones and computers. According to the Economist (9 September 2000), only 3.1 million of the worlds 360 million Internet users are in Africa, with most of these in South Africa and northern Africa.

United Republic of Tanzania United Republic of Tanzania

Kenya Nigeria Nigeria

10

20

30

40

50

Share of medium- and high-tech activities in manufacturing value added

Source: UNIDO Scoreboard database (see technical annex).

Foreign direct investment has largely passed Africa by. Although total and per capita inflows have grown since the mid-1980s, the regions share of flows to the developing world fell from 3.7 percent in 1985 to 3 percent in 1998. All this is not say that there has been no progress. The region has seen improvement in most indicators, but its ranks by these indicators are deteriorating. In a competitive world that is what matters more. Economies must improve faster to stay in the same place. Zimbabwe and the United Republic of Tanzania illustrate different aspects of the disappointing industrial performance of Sub-Saharan Africa. Zimbabwe, the most industrialized country in the region (after South Africa), has failed to build on inherited advantages. The United Republic of Tanzania, one of the least industrialized, has failed to build new advantages. Both countries experienced a wholesale worsening in industrial performance and its drivers between 1985 and 1998. Take industrial production. MVA per capita fell from $31 to $15.8 in the United Republic of Tanzania and from $123 to $77 in Zimbabwe. This record of decline is not

Benchmarking the drivers of industrial performance

87

uncommon in the region: MVA per capita also decreased in Cameroon, the Central African Republic, Madagascar, Nigeria and Zambia. Zimbabwe. Zimbabwe built a relatively deep and diverse industrial structure during the period of economic isolation under the Unilateral Declaration of Independence, with the capability to manufacture a range of simple capital and intermediate goods in addition to basic consumer products.28 But it has failed to upgrade this technological base since gaining independence in 1980. In the 1990s there was some liberalization in Zimbabwe and manufactured exports grew by 7.1 percent a year, with mainly resource-based (mineral) products going to industrialized countries and simple machinery and intermediates to neighbouring countries. But many firms were unable to upgrade enough to cope with import competition at home and in neighbouring export markets. Growth slackened, MVA declined, and exports failed to diversify and take off. Recent political uncertainties have exacerbated the situation. Between 1985 and 1998 the share of medium- and high-tech products in Zimbabwes MVA and manufactured exports declined (figure A4.16). As a result, the country lost 16 places in the ranking for export structure and 21 places in the ranking for MVA structure. Zimbabwe ranks relatively well in the region in human capital, coming second (after South Africa) in the skills index in 1998. But that ranked it only 68th in the world, three places lower than in 1985 (figure A4.17). As in most Sub-Saharan African countries, enterprise-financed R&D

is negligible in Zimbabwe, and it has fallen in the R&D ranking because of the entry of countries above it with some R&D. Foreign direct investment inflows rose from zero per capita in 1985, but remain at very low levelswith little going into manufacturing and almost none into export-oriented manufacturing. Royalty payments have declined marginally despite exposure to competition. And infrastructure has declined in relative terms. Thus Zimbabwe has failed to capitalize on its initial industrial advantages. Part of the reason for this is political, but structural factors also explain why its enterprises have failed to cope with international competition.29 The liberalization of the economy has not been accompanied by measures to improve drivers, and there is a serious risk that the existing base of learning and capabilities will be eroded. United Republic of Tanzania. In the United Republic of Tanzania MVA declined by 2 percent a year between 1985 and 1998. The countrys manufactured exports grew by 3.8 percent a year, but from a tiny base. Medium- and high-tech products accounted for only 1.5 percent of manufactured exports in 1998 (down from 2.3 percent in 1985) and for 25 percent of MVA. The country fell in the rankings in each of these indicators (figure A4.18). The base of capabilities in the United Republic of Tanzania is very weak. The country ranked lowest in the world in the skills index in 1998, having fallen in the ranking since 1985. Its R&D effort is negligible. Inflows of foreign direct investment amounted to
Figure A4.18 Changing ranks in industrial performance indicators and drivers for the United Republic of Tanzania 1985 1998 MVA 100

Figure A4.17 Changing ranks in industrial performance indicators and drivers for Zimbabwe 1985 1998 MVA 80 Royalties 60 40 FDI 20 0 MHT in MVA Exports

Royalties

80 60 40

Exports

FDI

20 0

MHT in MVA

R&D

MHT in exports Skills Infrastructure

R&D

MHT in exports Skills Infrastructure

Source: UNIDO Scoreboard database (see technical annex).

Source: UNIDO Scoreboard database (see technical annex).

88

Industrial Development Report 2002/2003

only $3.3 per capita in 1998; though the investments flows had increased since 1985 ($0.4), they have been directed largely to resource-based activities. Royalties are minuscule, and infrastructure among the poorest. But while Bangladesh, with even weaker infrastructure, has managed to build manufactured exports on the basis of clothing, the United Republic of Tanzania has failed to do this despite even lower wages. Studies show that capabilities at the enterprise level are weak even by Sub-Saharan African standards. Simple entry-level activities such as clothing and footwear are dying in the face of import competition from Asia.30 The countrys institutional structure for supporting industry is fragmented and ineffective.31 The United Republic of Tanzania seems to epitomize the problems facing the least developed countries in industrializing in an open environment with weak industrial structures and capabilities. There is no quick fix for these structural problems: the drivers of industrial growth have to be improved in all dimensions. This is difficult, in part because least developed countries also often lack the capabilities needed to design and implement the necessary strategies. Merely opening to global market forces is not enough. That is exactly what the United Republic of Tanzania has done, with meagre reward. The challenge is to find strategies within the reach of least developed countriesand the financial and, even more important, human resources that they need to mount those strategies.

coincide with a change in the orientation of the investment from domestic to export production or from manufacturing to services. 5. With all its simplifications the Scoreboard manages, with parsimonious use of data, to identify and provide information on important structural features of industrial performance. The trends revealed are reassuring, since it would be difficult to take seriously an exercise that yielded intuitively implausible results. Even so, some detailed findings are unexpected. Much can be done to improve the Scoreboard, of course. New variables can be added as meaningful and comparable cross-country data emerge. The measures can be refinedsay, with finer disaggregation of technological categories in production and exports. And they can be extended over longer periods, which would give a clearer and more robust picture of trends and lags. Future editions of this report will carry improved versions of the Scoreboard. 6. Patents taken out internationally are often used as a measure of inventive output of high quality (for example, by Cantwell and Janne 1998 and Porter and Stern 2000). But the use of patentsdomestic or internationalas an indicator of inventiveness is subject to caveats. Many patents are not exploited and so do not constitute technology in terms of practical application. Even the number of patents used in production may not indicate their economic significance. But because taking out patents in the United States is expensive, the practice tends to be confined to large innovators likely to exploit the patents in their activities or use them as a legal instrument for trading technology with other companies. This makes international patents a better indicator of inventiveness than local patents. 7. The Scoreboard uses R&D as the variable for technological effort because it is more relevant to developing countries than patenting in the United States. 8. This diversity is not surprising, as benchmarking (with indicators normalized by population) leads to fairly low scores for all these countries. But the underlying structural differences have to be borne in mind when using the Scoreboard for further analysis. 9. The two ratiosR&D spending per unit of high-tech exports and R&D spending per unit of inward foreign direct investmentare strongly correlated, with a coefficient of 0.745 in 1998. 10. But note the enormous difference in R&D per capita between the Republic of Korea ($5.90) and Japan ($100.40). 11. See, for example, Best (2001), Hobday (1995), Lall (1996) and Mathews and Cho (2000). 12. Wong (1999b); Rasiah (2000). 13. OECD (1999a).

Notes
1. Macroeconomic variables extensively analysed by other institutionsinflation rates, real effective exchange rates, interest rates and so onwere left out of the benchmarking analysis. And several more structural variables were quantified but discarded because of poor data or lack of theoretical justification. The list of drivers may, of course, be enlarged in subsequent editions of the report. 2. See Amsden (2001), Rodrigo (2001) and Lall (1996). This conflict may resolve itself once R&D capabilities are established. The experiences of Ireland and Singapore also show that with careful targeting, incentives and skill building, it is possible to induce transnational corporations to invest in local R&D (see chapter 2). 3. The correlation coefficients between the two years are 0.950 for Harbison-Myers index, 0.876 for tertiary technical enrolments, 0.896 for R&D, 0.895 for foreign direct investment, 0.399 for royalties and 0.958 for infrastructure. Each is significant at the 1 percent level. 4. The ranking of economies by structural drivers cannot capture changes in the nature or quality of a driver, however. A stable rank by inward foreign direct investment for a country, for example, may

Benchmarking the drivers of industrial performance

89

14. Although the R&D measure is perhaps a crude one with respect to informal technological effort, there is still likely to be a real correlation between R&D and the intensity and quality of informal effort. 15. The unmeasured influences captured by the developing country dummy variable grow less important over time. Thus while being a developing country had a negative effect on performance in 1985, this effect vanishes by 1998, when structural drivers explain much of the variation in performance. 16. This is based on data from the Hong Kong Census and Statistics Department (http://www.info.gov.hk/censtatd/eng/hkstat/hkinf/ production/production_index.html). The index of production for all manufacturing industry stood at 103.1 in 1999 and 100.9 in 2000, with 1986 = 100. 17. The data are from Banco de Chile (http://sie.aplicaciones.cl/ basededatoseconomicos/900base.asp?usuIdioma=I) and are in constant Chilean pesos. 18. These examples illustrate how the balance between industrial performance and its drivers may be interpreted and used. But the simplifications and approximations inherent in many of the measures must always be borne in mind: the Scoreboard is meant as a preliminary guide, not as a honed tool for precise measurement. Moreover, its practical use requires the analyst to adapt it to specific needs and to seek more information.

19. Wells and Wint (1990); Barry (1999); OHearn (1998). 20. UNCTAD (2000). 21. Cimoli (2000). 22. UNCTAD (2000). 23. Spar (1998). 24. UNDP (2001, box 4.2). 25. See for example Lall (1996). 26. Lall (2001b). 27. Lall (2001b). 28. Lall and others (1997). 29. Lall (1999b). 30. Lall (1999b). 31. Lall (2001b).

90

Industrial Development Report 2002/2003

Part 2 Laying the foundations for industrial competitiveness

5
C

Innovation and learning to drive industrial development

to help enterprises undertake the kinds of effort they cannot manage on their own. The learning process for the enterprise remains difficult and uncertain, but its progress and dynamism depend on the system in which it is embedded.1 Countries undergo costly, uncertain, prolonged and unpredictable learning, even when the technologies are well known abroad. Entry into different technologies involves different innovation and learning processes, and the difficulties rise with the depth of capabilities acquired. Just opening to international trade, investment and knowledge flows does not ensure that countries learn efficiently, though many policy analysts often assume that it does. And when a new technology is launched, using it in production does not by itself ensure that enough learning occurs. In other words countries need to engage in purposive and directed effort to create new capabilities and to capture the externalities generated by collective learning. They have to provide ample access to new information and knowledge, but they also have to promote domestic investment in learning. National industrial innovation and learning are path-dependent and cumulative. The initial base of capabilities and learning determines how well countries cope with new technologies. Patterns of specialization are difficult to change quickly. Growing national technological maturity requires the industrial sector to move from easy to complex technologies and, within given technologies, from know-how to knowwhy. At each stage, there may be costs, risks, delays and externalities, all likely to rise at higher levels of technology and capability development. In the less developed countries incentives are often skewed, factors lacking and support organizations missing or irrelevant. This makes triply daunting the challenges facing the smaller enterprises that dominate these economies. This chapter looks at the process of innovation and learning at the enterprise and cluster level. It shows what it takes for enterprises and clusters of enterprises to become competitive

OUNTRIES CAN SUSTAIN INDUSTRIAL GROWTH TODAY ONLY IF THE

key playersindividual enterprisesare able to develop competitive capabilities. Building capabilities requires conscious technological and other effort. And this effort is not very different whether an enterprise is creating new technologies or learning the efficient use of technologies brought from other countries. Whether such effort is undertaken, and how well it is managed, varies from enterprise to enterprise, according to its management, strategy and resources. But it also depends vitally on the economic environment in which the enterprise functions. The environment reflects complex interactions among the incentives, factor markets and institutions facing an enterprise. Incentives are the signals emanating from the market: competitive pressures at home and abroad, growth prospects and the like. Factor markets include all the inputs that enterprises need, from information and capital to skilled and unskilled labour, components and infrastructure. Institutions are both the rules of the game (legal and cultural) and the intermediary agencies (standards, quality, training and so on) that supplement or embody factor markets. Each country has its own mixture of incentives, factors and institutions, reflecting its history, policies and business practices. These form an innovation and learning system. A strong system tends to produce a larger number and diversity of competitive industrial enterprises, a weak system relatively few. The primary incentives for technological effort arise from competition, which reflects the structure of the market and government policies on trade, foreign direct investment, ownership and domestic competition. A more competitive setting generally results in greater technological effort, with one important caveatthere may be legitimate reasons to protect infant industries to help them overcome the initial costs of learning (see below). Flexible and responsive factor markets are vital to building and deploying industrial capabilities. Clear and supportive rules of the game induce enterprises to make long-term investments in innovation and learning. And efficient intermediary institutions are essential

93

in todays fast-moving global economy. It focuses on efforts to link up with global players to acquire new technology and to leverage those acquisitions in ways that get as much as possible from the new relationships.

industrialized countries because of their far greater market and institutional deficiencies. In many developing countries, building competitive capabilities even in simple, low-tech activities is demanding. And technical change is a process of continually raising the thresholds. The success of enterprises (and clusters of enterprises) in upgrading their capabilities depends on many factors, not all of them under the enterprises control. Apart from its own efforts, which count for a lot, it needs access to information (accurate and current) and to knowledge and expertise. It also needs infrastructurein the form of, say, basic and dependable power supplies. If it is an exporter, it needs access to export facilities, such as a reliable airport, and honest and efficient customs services. It also needs intangible infrastructure, such as efficient corporate administration agencies free of graft and corruption. Lack of these features all too often bring down small, local enterprises, despite their best endeavours. The interactions inherent in innovation and learning constitute a coherent (if chaotic) system of incentives, factors and institutions to which firms and clusters respond in their technological effort. Reflecting the quality, density and interaction of the various elements, such systems can differ greatly in their ability to stimulate or retard innovation and learning. And given market and institutional failures, their efficacy also reflects the impact of government policy on the various elements. Governments can improve some elements of the systems. More important, they can coordinate the many elements to form a coherent strategy for industrial and technological development. Where this has been done efficiently, as in some East Asian economies, the results have been spectacular.4 If technology exists elsewhere and it can be transferred, surely all that developing countries need to do is to allow it to flow in? According to conventional wisdom, this requires countries only to open their economies to inflows of investment and knowledge, set the framework conditions right and allow competitive markets to allocate foreign and domestic resources to their appropriate uses. Competitiveness must follow if these things are done. In this view technology is considered something of distinct and separate concern only to economies at the frontiers of basic research and innovation: processes of technological change and innovation are not thought central to development efforts. This greatly oversimplifies the industrial process. It assumes that technologies can be fully embodied in machines, blueprints, instructions and so on, and moved like physical products to new locations and deployed efficiently without further effort. This can be very misleading, for technology has many tacit elements that require a new user to build skills, knowledge and institutional routines (capabilities). Mastery of these

Becoming competitive takes a lot of effort

Becoming competitive does not mean seeking temporary advantages that can raise market share for a short period through currency depreciation and wage cuts that tend to forgo longer run possibilitiesknown as the low road to competitiveness. Sustaining growth in the long run entails strengthening competitive advantages and building new capabilities on the high road to competitiveness. In todays technology-driven world, whether a country takes the high road depends mainly on the extent and efficiency of its enterprises enhanced adaptive capabilities to deploy new production and management technologies, upgrade them over time and ultimately create new technologies. This whole set of activities and capability building is here termed innovation and learning. Enterprises do not build industrial capabilities on their own. They react to signals from markets. They draw on information, skills and inputs from othersnot just suppliers and customers but also competitors. They raise funds in capital markets to finance innovation and learning activities. They seek skilled workers from labour markets, at least where the local economy has the new skills they need. They similarly seek specialized technical assistance from technology institutions like standards and quality agencies, research and development (R&D) institutes and universities and technical extension bodies. Innovating and learning are interactive processes that work best in dense networks of efficient enterprises, institutions and markets. Enterprises thus innovate and learn as parts of a collective larger, interconnected group.2 They have to manage and adapt their internal systems of control, interaction, information and effort to absorb or create new understanding. Each of them has its own culture and way of doing things, some better suited to innovation and learning than others. Adapting enterprise routines to competitive pressures is slow and evolutionarynot instantaneous, as is assumed in economic models of enterprise behaviour.3 Since capability building occurs in specific market and institutional contexts, its pace, depth and spread depend on the efficiency of markets and institutions. Not surprisingly, the process faces more difficulties in developing countries than in

94

Industrial Development Report 2002/2003

tacit elements is needed everywhere, but it is particularly difficult in developing countries, where enterprises lack the initial base of technical skills and knowledge on which to graft new technology. Enterprises have to learn to learn.5 And there often is fairly little in the local industrial environment surrounding them that can help them in their learning process. The capability-building process is uncertain and incremental, thus far removed from the optimizing process of textbook theory with known and certain outcomes. How efficiently technologies are used depends on the efficacy of this evolutionary effort. There is no predictable learning curve for all enterprises to travel down. Enterprises do not have a clear knowledge of the available set of technologies or of how to operate any new technology efficiently. Finding the right technology at the right price involves cost and risk, particularly for enterprises in a developing country. Using the new technology involves further effort to master its features: it entails search, experimentation and new information and learning. Making the technology work efficiently under new conditions involves further effort to adapt it to local demand, scales of production, worker skills and raw materials. Technological effort does not end with mastery. All technologies can be improved by minor adjustments, calling for further effort and new capabilities. If the international frontier in the technology moves aheadas it nearly always doesthe firm has to adapt and master the new versions to stay competitive. At some stage, a dynamic enterprise may separate its monitoring, learning process, improvement and related functions into a separate formal R&D department, with its own budget and management. This department need not devote itself to creating breakthrough innovations. Much R&D, even in developed countries, is devoted to monitoring, copying and adapting the innovations of others. And in developing countries the main function of R&D is to leverage, learn, adapt and improve imported technologies. While using any new technology effectively involves building new capabilities, the uncertainty and difficulty vary from case to case. And they rise with the complexity of the technology and the novelty of the effort. The effort depends on three things: the initial capabilities of the enterprise, the support it can draw upon from the enterprises environment and the novelty of the technology relative to its existing stock of knowledge. The same technology may be almost costless to absorb for an enterprise in an industrialized country but very costly and difficult to learn for an enterprise in an industrializing one. Technological efforts can occur almost anywhere in the enterprisein R&D, on the shop floor, in process or product

engineering departments. The ideas and impetus may come from these units or from marketing, procurement or quality management. They may also come from outside the enterprise. Indeed, much technical information comes from equipment and material suppliers, contractors and buyers (particularly from discriminating foreign buyers). And some can come from extension services, technology institutions, universities, trade fairs or even competitors. Ultimately, however, capability building is institutional: it has to reach through the whole enterprise and reach all membersotherwise it cannot affect performance. A very similar process of capability building has to take place in such non-technical functions as marketing, procurement, training and financial management. The technical and other processes have to interact and influence each other, since building capability often involves changing institutional processes or routines and launching new ways of managing information and people. The organizational aspects of innovation may be as important as the technical.6 The whole innovation and learning process is a conscious and purposive activityand often uncertain, costly and difficult activityby the enterprise interacting with the environment. That is why this report treats innovation and learning as different aspects of the same process of building new capabilities (table 5.1).

Start small and see what you can become

Enterprises in developing countries generally start the innovation and learning process by importing new technology; they then invest in building their capabilities to master the tacit elements (figure 5.1). How much they invest depends on the incentives thrown up by markets, mainly by the competition faced in foreign and domestic markets. Enterprises draw upon internal and external resources to build their capabilities. The process starts with capabilities needed to master the technology for production purposes and may deepen over time into improving the technology and creating new technology. Linking, leveraging and learning captures what firmsand countries as wellhave to do to foster their technological development.7 Linkingconnecting with outsiders to acquire needed technologies and skills. Leveraginggoing beyond arms-length transactions to squeeze as much as possible from the new relationships with those outsiders.

Innovation and learning to drive industrial development

95

Table 5.1 Technological and organizational capabilities within firms

Production Nature of capacity building strategy and effort Basic Simple, routine: based mainly on internal effort and experience Investment Preinvestment Project execution Process engineering Product engineering Industrial engineering and HRD

Pre-feasibility and feasibility studies, site selection, scheduling, arranging finance

Routine engineering of civil works, ancillary services, erection and commissioning

Debugging plant; routine process coordination; quality management; routine maintenance; process quality certification

Assimilation of basic product design; product quality management and certification; minor adaptations to meet market needs

Workflow scheduling; time/motion studies; innovative management and optimization; skill upgrading and training

Intermediate Adaptive, duplicative: based on search, experimentation and inter-firm and other cooperation Search for sources of technology, equipment. Contract negotiation Equipment procurement, detailed engineering, staff recruitment and training Capacity stretching; adapt/improve technology; use new techniques (JIT, TQM, etc.); routinized process engineering; preventive maintenance Product quality/ design improvement; licensing new technology; reverse engineering; continuous monitoring of global technologies Continuous and systematic productivity analysis and benchmarking; skill audit and formalized training; supply chain/logistics management; advanced inventory control

Advanced Innovative, risky: based on purposive effort, R&D and advanced forms of collaboration Own project outline and design capability. World class project management capabilities Basic process engineering, equipment design and start up. Turnkey capability Continuous process improvement; process innovation; basic research; use of new process design methods. Organizational capacity for generating, codifying, socializing knowledge Mastery of product design methods; new product innovation; basic research. Strategic alliances. Organizational capacity for innovation and risk taking World-class industrial engineering and supply chain capabilities, training systems, inventory management

Source: Based on Lall (1992). Note: HRD is human resources development. This is only an illustrative list of capabilities within a manufacturing firm. It does not include several types of capability, such as financial management, labour relations, logistics and so on.

Learningmaking the many efforts to master process and product technologies, consciously building the foundation for improving current technologiesand creating new ones. Enterprises have to start somewhere. Most latecomer enterprises in developing countries start with few resources and few connections. They need to acquire a minimal complement of skills, resources and capabilities just to be players. How this is done depends on a variety of circumstances. An enterprise may be a traditional supplier to an enterprise in a traditional industrial sector like textiles that is suddenly hit by

policy reforms that bring in foreign competitionor gets taken over by a more enterprising entrepreneur. It may be a staid family operation that is taken under new management by a son or daughter who has received a business education abroad. It may be a government-owned enterprise that is privatized, beginning life anew in a competitive world. Whatever the process, enterprises have to start with its initial complement of resources, technologies, skills and capabilities. And it is what it does with these things that counts. The most important thing that an enterprise can do is to accelerate its acquisition of capabilities by looking overseas to obtain

96

Industrial Development Report 2002/2003

Technology Domestic Foreign

Marketing Domestic Foreign

Local procurement systems and procedures, drawing in available knowledge from institutions

Foreign sourcing; information from suppliers; industry networking; accessing public information

Market research; distribution and servicing systems; some advertising

Export market analysis; links with buyers and other export channels. Design/ packaging capability

provide answers to the questions: Where are we, and where do we want to go? This step also includes the identification of partners with whom capability enhancement is feasible. The leveraging phase requires a strategic choice and specifies the means of knowledge acquisition. It answers questions of the type: How do we get what we want? The final step involves the actual process of capability enhancement through learning. Different forms of learning are feasiblethere is learning by doing, learning by interacting, learning by monitoring, learning by formal training. The choice will depend on the type of linkage and leverage involved. The learning process is difficult and complex; it lies at the heart of the arduous process of industrial innovation and development. Capturing linkage opportunities is thus just half the story for latecomer enterprises in developing countries. The other aspect is the way that enterprises use the links established to leverage skills, knowledge and technology from the enterprises contracting with them. This is what drives the process of technological innovation within a networked system. So, if resources are lacking, leveraging them from external sources is the obvious way to proceed. The concept of resource leverage comes from strategic management of firms.8 The concept has been used as a means of explaining how the best competitors in the world stay abreast of new developments, by ensuring that through alliances and various forms of joint ventures, they identify and secure access to the resources needed to keep diversifying their product portfolio. But the same idea underpins the strategy of enterprises in developing countries.9

Technology transfer to and from local suppliers/buyers; coordination in design and manufacture; links with technology and other institutions. Capacity to take collective action

Vertical technology transfer; systematic coordination of international knowledge sources; links with technology institutions overseas

Dedicated marketing department. Systematic monitoring, feedback analysis. Branding and differentiation

Systematic market building and analysis of foreign markets. Alliances and networks abroad. Brand introduction. OEM arrangements

Continuous links with R&D institutions and universities. Licensing own technology to others. Deep innovative links with other firms. Specialization in context of networks and clusters

Cooperative R&D; strategic alliances; advanced leveraging strategies for new technologies. Foreign acquisitions, direct investment

Advanced brand creation; coordination with retailers/ buyers; advanced distribution systems

Brand deepening. ODM and OBM arrangements. Own marketing and design channels and affiliates abroad

Using technology better

For enterprises using existing technologies at levels below best practice, innovation and learning can lead to better use of technologyto higher capacity use, lower reject rates, reduced inputs or improved product quality. Much innovation in developing countries is of this type. And it does not require much in new investmentcritical here are the learned capabilities. Bedi Investments in Kenya exemplifies an enterprise in a new phase of its technological learning trajectory. A textiles and garments producer, Bedi has been building its basic competences, using international standardsin this case International Organization for Standardization (ISO) 9000 quality manufacturing standardsas its benchmark. Such standards represent the minimum attributes an enterprise needs to enter world markets and claim a place in global networks. Bedi was established in 1972 by a Kenyan entrepreneur as a small family-run garment firm producing for the local market.10 Over the years, it integrated backward into making fab-

information, purchase machinery, acquire bits of technology, bring in consultants and so on. An important part of this can be linking up with other enterprises or institutions, locally or overseas, through formal or informal ties (box 5.1). Strategically it makes a lot of difference what choice is madebut the choices are also heavily constrained by the enterprises competence and the options available to it. The linking, leveraging and learning strategy of industrial innovation has to start with an in-depth analysis of key factors of competitiveness, and the various options for linking a developing firm to sources of knowledge. This linking stage has to

Innovation and learning to drive industrial development

97

Figure 5.1 Enterprise innovation and learning

Technology import Internalized (FDI) Externalized (licensing equipment, copying)

Market conditions Nature of local demand Trade regime Domestic competition Other

Embodied in Technical assistance Patents Hardware Knowhow Blueprints

Enterprise innovation and learning process

Enterprise technological capability development

Innovating Improving Learning Acquiring (simple to complex functions)

Internal resources Human resources Technology effort Management skills Organization skills

External resources: Inputs from other firms Technology Skills Finance Infrastructure

rics and yarnsand emerged as one of the most modern integrated textile-garment plants in the country. Now managed by the founders three sons, all graduates in engineering or business from schools in the United Kingdom, it has a good base of technical staff by local standards (2 percent of employees are engineers and technicians). Bedis adoption of ISO 9000 in the early 1990s was stimulated by a foreign buyer that provided Bedi with information about the ISO programme and helped with implementation. Initially, the buyer arranged for an audit by a qualified consultant from abroad. It then helped Bedi implement the post-audit changes, including purchases of new equipment, calibration tests, training of workers and quality personnel and a detailed monitoring system. The buyer also helped Bedi with verification and certification by an independent accredited agency. In 1994 Bedi had a 26-strong quality control department (7 percent of employees), and its internal reject rate was under 1 percent. The ISO 9000 system doubled Bedis labour productivity growth to 6 percent a year (between 19841989 and 19891994).

Bedi now has good capacity to search and negotiate terms for imported technology. It has one of the best production capabilities in the Kenyan garment industry (a strong emphasis on quality control, well-maintained equipment, negligible equipment breakdown rates and frequent changes in plant layout). And it has good technological linkages with foreign buyers and equipment suppliers. But it still lacks independent design capabilities, relying heavily on foreign buyers for product designs, common for enterprises in the early stages of export development. Another good example comes from Tema Steel, set up by the Ghanaian government in the late 1970s to smelt steel from scrap and to make billets and rods for the construction industry.11 A British firm set up the plant on a turnkey basis with almost no participation by local personnel. The turnkey supplier recruited and trained staff, but the local base of steelmaking capabilities was practically nil. One of the two blast furnaces could not be made to work, the layout was highly inefficient and the training was insufficient to ensure the smooth operation of the technology.

98

Industrial Development Report 2002/2003

Box 5.1 Linking up with othersto start the processes of leveraging and learning The interlinked character of economic systems provides the opportunities for enterprises to make connections with potential sources of technology and skills. Enterprises operate in clusters, or in value chains of one kind or another, some implicit, others explicit, some tightly organized, others not. Some links involve providing services directly to a leading enterprise, either upwards in the value chain as a supplier, or downwards, as a distributor (based frequently on FDI arrangements). Some involve outsourcing and original equipment manufacturing (OEM) contracting. Some involve technology licensing. Each of these represents a strategic choice for the lead enterprises, but it is precisely these choices that create opportunities for developing country latecomers, which they are quick to seize and to turn into opportunities for leveraging and learning. FDI links: forward and backward linkage The simplest linkages involve being engaged in contractual supply of goods or services. Transnationals moving into a new market usually need local firms as suppliers of maintenance services or as suppliers of simple materials and components. These can be upgraded to encompass more demanding tasks, involving more added value, as the incumbent builds longer term relations. These are captured in many countries as vendor partnership programssuch as the Intel vendor partnership program with supplier firms in Malaysia. These are all described as backward linkages. Correspondingly, forward linkages can be established when a local enterprise takes on the distribution of a product range for an existing incumbent, often on an exclusive basis in the local enterprises domestic market. Many of todays giants from the developing world, such as the information technology firm Acer from Taiwan Province of China, began life as such simple distributors of advanced enterprises products, taking advantage of these forward linkages to leverage knowledge and market access from their partner. Outsourcing In the 1960s and 1970s advanced firms in industrialized countries
Source: UNIDO.

exported some of their manufacturing of mature products to low-cost production platforms in Asia and Latin America, as many garment companies did with the cutting and sewing parts of their value chains. Latecomer countries benefited from these decisions in raising their social capital, and soon individual latecomer enterprises were bidding for parts of the manufacturing cycle, as for testing and packaging activities in the semiconductor industry. OEM contracting can be considered a form of outsourcing where the activities contracted to a third party are critical, high value-adding parts of the process entailed in contracting to manufacture the entire product. Electronics and information technology companies like IBM, Apple, or Texas Instruments outsourced the production of entire products like personal computers to latecomer firms in East Asia, securing strategic advantages in low-cost production, but also offering the latecomers valuable learning and leveraging opportunities. Many personal computer firms in Taiwan Province of ChinaAcer, Mitac and Tatungestablished themselves with the help of such OEM contracting. Second sourcing is a related concept, where a supplier of critical hightech products like memory chips or logic chips seeks an outside second source to back up its own supplies (in case of problems in meeting a customers order) or to take over the more mature products as the innovator moves on to the newer products. Again, this strategy on the part of incumbents creates numerous opportunities for an agile latecomer to grasp a business opportunity and use it as a leveraging and learning experience. Technology licensing Samsung from the Republic of Korea got into semiconductors by securing a product design from Micron, a small U.S. firm. Later, Samsung secured access to microprocessor technology by licensing it from DEC (which became part of Compaq, which is now part of HP). DECs interest was served by widening the field of applications for its Alpha chip.

Later, an Italian consultancy was brought in, again on a turnkey basis, to set up an expensive casting machine and a foundry. It also changed the rolling mill to enlarge its product range, making its operation even more complex. Again, given the lack of local capabilities, neither the casting machine nor the foundry could be brought into operation by Tema Steel, and operational efficiency declined even further. Some 17 years after starting, the plant was running at only 10 percent of design capacity, with much of its main equipment unused, resulting in high costs and enormous losses. Much of the plant in use was badly in need of repair and upgrading. Very few qualified steel technicians or engineers were on staff, and the staff received practically no training. There was no systematic attempt to learn the technology better or to bring in new knowledge from outside. There was no supplier or consultancy industry locally for steel making, so there was no networking with other enterprises. As a result, few competitive capabilities were built over the 17 years of operation.

In 1991 the plant was sold to an Indian steel company, which brought in 17 employees with experience in steel (only two were graduate engineers) to take charge of the technical functions of the plant. With very little new investment in equipment, the new employees started to refurbish the machinery and improve maintenance and every area of operations. Within a year capacity use was up threefold, with the rolling mill working two shifts and the furnaces three shifts for six days a week. The blast furnace that had never been used was commissioned by changing a few controls. The first furnace was upgraded to run continuously. The continuous casting machine was also brought into operation by inserting some missing items that the previous technicians had not even been able to identify. Various motors that had died from neglect were refurbished and put into use. The foundry was commissioned, and new refractory products developed. Quality management was improved to match UK standards (and so meet import competition). Training programmes were launched in-house for local

Innovation and learning to drive industrial development

99

staff in all technical functions, because most of the Indian temporary staff were due to return home in another year or two. Tema shows what capabilities do for the operation of imported technology. It also shows how learning can be transfered from one developing country (where considerable learning had taken place) to another (where almost no learning of that kind had occurred).

implement continuous improvements in production technology. By the 1990s the firm had a 500-strong team of engineers and technicians doing applied R&D. Having established sales offices in industrialized countries in the 1970s, Tatung later set up overseas production facilities to improve its competitive position in those markets, reinforcing its brand image and acquiring advanced technologies and skills. By the mid-1990s it had eight manufacturing operations abroad making televisions, washing machines, refrigerators and other household items. Thus by establishing plants overseas, Tatung now exports technology.

Adapting and improving processes and products

Other manifestations of innovation and learning are improvements and adaptations to products and processes, drawing on in-house technical efforts, outside sources of knowledge and interaction with leading international enterprises. Minor improvements and adaptations are part of gaining capabilities in efficient production. But major improvements and adaptations require higher levels of enterprise skill and competenceand generally a more advanced industrial system and infrastructure. Tatung, the largest electronics manufacturer in Taiwan Province of China, was one of the countrys leading industrial conglomerates in the 1990s.12 In 30 years it had advanced from making simple consumer electronics (black and white televisions) to computers, colour displays and television monitors. By the mid-1990s its electronics sales exceeded $1 billion and it had eight overseas subsidiaries (including those in the United States, Japan, Germany and Ireland). Tatung initially assimilated manufacturing know-how by acquiring mature process technologies for household appliances and consumer electronics from U.S. and Japanese companies through technical assistance deals and joint ventures. It later learned many of its technological skills through original equipment manufacture or OEM, whereby Tatung undertook contract manufacturing of electronics products for foreign transnationals, for sale under their own brand names. Transnationals often assisted Tatung with selecting equipment, training engineers and providing advice on technology and management for OEM. By the late 1980s Tatung was exporting about half its colour televisions, personal computers and hard disk drives under OEM. Most of this embodied little original R&D, but the company invested considerable effort in assimilating technology and closing process and product technology gaps with its competitors. It developed the ability to absorb and adapt advanced foreign technology and modify, re-engineer and redesign consumer goods for different types of customers in regional markets. Its in-house engineering capabilities were used to scale down production processes, adjust capital to labour ratios and

Moving up the value chain

An enterprise in the initial phases of industrial development in the typical developing country starts with simple assembly and packaging operations. If very successful in deepening its skills and capabilities, it can move into more complex manufacture say, of critical components, which can be at a much higher technological level, involving far more sophisticated process management and larger scales. Semiconductor enterprises in Malaysia made such moves up the value chain, as affiliates of leading transnational corporations upgraded their facilities to run state-of-the-art process technologies at massive capacities.13 Further moves up the value chain are possibleright up to invention. The enterprise can do process or equipment design in complex technologies, take on product design and engineering and finally do basic or blue skies R&D. Most leading local enterprises in the Republic of Korea and Taiwan Province of China have reached this stage, as have many affiliates of transnational corporations in high-tech industries in Singapore. A similar progression is also possible in services and does not need a highly developed industrial network or infrastructure. But there is still a discrete learning process that has to be undergone and financed. And having a base of skills outside the enterprise is critical. ArtinSoft is one of the worlds leading providers of database and programming language migration software. Its main product (Freedom) is a programming tool that uses artificial intelligence technologies to enable enterprises to change their software systems into newer computing languages (Java, the Internet language; and Visual Basic, Microsofts prime language). In eight years the firm has grown from three penniless new graduate engineers and their professor to a 200-employee enterprise, which has attracted the attention of global players such as Oracle and Intel. Founded in 1993 and located, with two computers, in a rented farmhouse in the hills, the company was a spin-off of the

100

Industrial Development Report 2002/2003

Institute of Technology in Cartago, Costa Rica. A teacher with a Ph.D. in artificial intelligence encouraged three of his best masters students in computer science to pursue his vision the development of fast, secure and cost-effective migration software. To support their research, the foursome developed a bread and butter product, an Enterprise Resource Planning and Scheduling software product to support larger Costa Rican businesses. Close proximity to the clients, low cost and accessibility (in Spanish) made it a commercial success in Costa Rica and the region. With this product ArtinSoft acquired the experience and capital to develop Freedom. In 1998 Oracle asked one of its main clients, a large European software company, to approach ArtinSoft to manage an important migration project. The project was completed successfully and at very low cost. This gave it the opportunity to interact with world players and build a global reputation. It later attracted the attention of Intel Communications fund, which invests in companies working on porting applications. In February 2001 the Fund invested in ArtinSoft to upgrade and develop new migration systems to bring the applications closer to Intels technologies. ArtinSoft benefited from being in Costa Rica, which has one of the most qualified information technology workforces in Central and Latin America. This is so in part because the governments Informatica Educativa scheme, which provides each school with at least one computer and helps youngsters to develop skills in the new information technologies.

to organization. The assembly of clothing for export is easy in this sense. With a semi-skilled and disciplined workforce and a few knowledgeable technical and managerial staff, it is possible to set up a plant and reach competitive standards in a few weeks or months. By contrast, difficult technologies need a long time and considerable accumulation of skills, knowledge, experience and organizational changes to master. The ability to efficiently run a large steel plant, a modern automobile factory (not a simple assembly plant) or a complex machine-tool manufacturing facility can take many years of cumulative capability building. Enterprises cannot become competitive without going through this processno matter what their level of development. Interactions among the three main areas of technological capabilityproduction, investment and innovationare also essential in moving into more complex activities. Operating productive facilities makes it easier to expand those facilities or to establish new ones. Such production and investment capabilities make it easier to develop new technologies that are cheaper and better suited to local conditions. And that innovation capability makes it easier to improve operations and undertake new investments. Activities differ not only in their technological needs but also in the economic rewards they yield: in general, technologically complex activities yield greater development benefits to successful adopters, at both the firm and economy levels. Complex industrial activities offer greater scope for learningand lead to learning more advanced and diverse skills. Technology-intensive activities generally offer better prospects for the continued application of new knowledge to production. So, activities that rely heavily on R&D (and have greater technological opportunity) enjoy a larger stream of innovations and so faster rises in productivity. Note that the borders between these capabilities are blurredand that their difficulty is in the eye of the beholder. Once possessed, the capabilities look easier. And as firms learn, what seemed difficult, or looked impossible, becomes easier. Unless they run into a technological brick wall.

Moving into more complex activities

While all technologies need effort to leverage and learn, clearly there are differences in the intensity, cost, risk and duration of effort involved. All industries, even the simplest, have relatively difficult functions (clothing design) and all have relatively simple ones (final assembly and packaging). However, it still makes sense to describe an activity as easy on average if it has a preponderance of relatively simple functions. This is the justification for the divisions of high, medium and low technology. The greater the role of difficult technologies, the higher the level of technology in this sense. Final assembly and packaging are relatively easy functions in most industrial activities. Making critical components in large volume is more difficult. Designing new products or processes is even more difficult, while innovating on the frontiers of knowledge is the most difficult of all. At each stage, the building of capabilities becomes more difficult and costly, with growing demands on the supporting industrial environment and institutions. Easy technologies can be learned fairly quickly, with few new skills, small amounts of new information and minor changes

Combinations and progressions

Brazils Usinas Siderrgicos de Minas Gerais SA (Usiminas) shows how successful technological development depends on a long-term effort to build systematically on foreign technological inputs and on accumulated experience. It also highlights the uncertainty and unpredictability of the technological trajectory. Usiminas started with foreign technical

Innovation and learning to drive industrial development

101

assistance but quickly absorbed it, adopting the imported technology and then building a capacity to adapt to the local operating conditions.Usiminas shows that an enterprise (or a country) does not need to do everything itself to expand and modernize its industrial production. It is possible to start with only the barest production capability and, with that as foundation, to build increasingly the base of other technological capabilities.14 But this takes timedecades rather than years. There are no dramatic short-cuts to acquiring technological competence. Usiminas traces its origins to the 1950s, when the Brazilian government and firms in the industrial province of Minas Gerais wished to build a steel firm capable of adding value to the regions iron ore production and of feeding steel to downstream industrial users. After much deliberation and intelligence-gathering (such as sending teams to visit foreign steel plants to gain first-hand experience of different processes) a decision was taken to start with coke-based blast furnace technology using a basic oxygen process rather than electric arc technology. Usiminas then entered a consortium with 30 Japanese steel makers and steel equipment suppliers. The Japanese arranged generous export credits for the required steel-making equipment and agreed to pass full responsibility to the Brazilians after they had developed sufficient capability to run the plant. The firms in the Japanese consortium undertook the basic engineering and project management of the construction of the mill, working closely with local managers and providing training for Brazilian engineers in Japanese steel mills. The basic capabilities were acquired in this fashion. Start-up operations proceeded for three years under Japanese oversight. With the Brazilian takeover a new administrative structure was put in place, and there followed six years of financial stringency during which the Brazilian engineers managed to double capacity without further investment. This was accomplished through a variety of testing and experimental approaches that allowed the Brazilians to tweak the technology inherited from the Japanese. New special steels, such as thick ship-plate, were introduced through licensing and technical assistance contracts. These efforts were undertaken in conjunction with intensive benchmarking of the plants technical performance and procedures against foreign plants, in an effort to close the gap between the Usiminas operation and the world technological frontier. Next came a period of expansion, as the government invested in greater capacity and as the market for the plants steel expanded. The company had to develop a set of capabilities associated with investment, planning and project implementation, quite different from the capabilities involved in plant operation. The com-

panys specialist engineering department was expanded to achieve these goals, working this time with a variety of equipment suppliers as well as the Japanese, and having foreign experts review the expansion plans as a precautionary measure. The company was by now providing technical advice to its downstream users, so much that it spun off a new business, Usiminas Mecnica, as a subsidiary to produce capital goods for the steel industry and foster the use of steel in Brazils machinery and construction sectors. By then the company was in a position to export technical services to neighboring countries. Technological capability within Usiminas reached the point that early in the 1990s the company could be successfully passed across to the private sector, where it has continued to perform well. By the late 1990s it was one of the top steel groups in Latin America.

Nine features of innovation and learning

The need for innovation and learning exists in all cases, even when the seller of the technology provides advice and assistance. But the extent and costs of learning vary by technology, enterprise and country.15 Innovation and earning call for conscious, purposive and incremental effortsto collect new information, try things out, create new skills and operational routines and forge new external relationships.16 This process is strikingly different from textbook depictions of how technology is transferred and used in developing countries. Summarized here are the nine most important features of technology capability development.17 1. Conscious and purposive. Learning is a real and significant process. Vital to industrial development, innovation and learning are primarily conscious and purposive rather than automatic and passive. 2. Risky and costly. Enterprises do not have full information on technical alternatives. Instead, they function with imperfect, variable and rather hazy knowledge of technologies they are using. As a result, there is no uniform, predictable learning curve. Each enterprise has a different innovation and learning experience, depending on its initial situation and subsequent efforts. Each faces an element of risk, uncertainty and additional cost in innovation and learning. 3. Not obvious. Enterprises may not know how to build up the necessary capabilities. In a developing country, knowledge of traditional, stable and simple technologies may not be a good base for knowing how to master modern technologies. So, enterprises may not be able to predict

102

Industrial Development Report 2002/2003

if, when, how and at what cost they would learn enough to become fully competitive, even if the technology is well known and mature elsewhere. This adds to the uncertainty and risk of the learning process. 4. Path-dependent. Firms cope with these conditions by developing organizational and managerial routines, which they adapt over time as they collect new information, learn from experience and imitate other firms. So, technological trajectories tend to be path-dependent and cumulative. Once embarked on, they are difficult to change suddenly (for countries and for enterprises), and patterns of specialization tend to persist over long periods. 5. Highly specific. The innovation and learning process is technology specific, since technologies differ in their learning requirements. For instance, some technologies are more embodied in equipment while others have more tacit elements. Process technologies (like chemicals or paper) are more embodied than engineering technologies (machinery, automobiles or electronics), and demand different (often less) effort. Capabilities built in one manufacturing activity may not be easily transferable to another, and policies to promote innovation and learning in one may not be very useful in another. Similarly, different technologies can involve different breadth of skills and knowledge, with some needing a relatively narrow range of specialization and others a very wide range. 6. Many complex interlinkages. Technological innovation and learning in a firm do not take place in isolation: the process is rife with externalities and interlinkages. The most important direct interactions are with suppliers of inputs or capital goods, competitors, customers, consultants and technology suppliers. Technological linkages also occur with firms in unrelated industries, technology institutes, extension services and universities, industry associations and training institutions. Many such linkages take place informally and are not mediated by markets. Not all are deliberate or cooperative: some learning involves imitating and stealing knowledge. Where information and skills flow around a set of related activities, clusters of enterprises and industries come together. Tapping these cluster effects can be very effective in accelerating technological competence. Different technologies have different degrees of interaction with outside sources of knowledge (enterprises, consultants, equipment suppliers or technology institutions). These differences in turn lead to different learning costs, risks and duration. A set of policies conducive to the development of one set of capabilities may therefore not be suited to another.

7. Many levels of effort. Capability building involves effort at all levels: procurement, production, process or product engineering, quality management, maintenance, inventory control, outbound logistics, marketing and other outside links. What appear to be routine and easy technical functions, like quality management or maintenance, can be very difficult to master in a developing country. Most learning in developing countries arises in such mundane technical activities. But formal R&D becomes important in complex technologies, where even efficient absorption requires search and experimentation. 8. Many depths of development. Technological development can take place to different depths. The attainment of a minimum level of operational capability (know-how) is essential to all industrial activity. This may not lead automatically to deeper capabilities, the ability to understand the principles of the technology (know-why). The deeper the levels of technological capabilities, the higher the cost, risk and duration involved. It is possible for an enterprise to use imported technologies without developing the ability to decode the processes to significantly adapt, improve or reproduce themor to create new products or processes. But this is not optimal for long-term capability development. Without technological deepening the enterprise or country remains dependent on external sources for major expansion or improvement to its technologiesa costly and possibly inefficient outcome. The development of know-why is an important part of overall innovation and learning. It allows a firm to select the new technologies that it needs, lower the costs of buying them, adapt and improve on them more effectively, add more value by using its own knowledge in production and develop autonomous innovative capabilities. The lack of these deeper capabilities may also restrict an enterprises ability to move up the technology scale even in using higher levels of know-how in its given activity, diversifying into other activities or coping with unexpected demands of technological change. Note that even good follower strategies, in which enterprises efficiently imitate and adapt technologies developed by others (common for efficient enterprises in developing countries), require good know-why capabilities. 9. Foreign plus domestic. Technological interactions occur within and across countries. Imported technology provides the most important initial input into technological innovation and learning in developing countries. And since technologies change constantly, access to foreign sources of innovation remains vital to continuing technological progress. But technology imports do not substitute for the development of indigenous capabilitythe effi-

Innovation and learning to drive industrial development

103

cacy with which imported technologies are used depends on local efforts. Domestic technological effort and technology imports are largely complementary. But not all modes of importing technology are equally conducive to indigenous learning. Much depends on how the technology is packaged with complementary factors: whether it is available from other sources, how fast it is changing, how developed local capabilities are and the policies adopted to stimulate transfer and deepening. Transfers internal to a firm, as from a transnational corporation parent to its affiliate, are efficient means of providing the latest know-how, but they tend to be slow in building know-why in the affiliate. To sum up, considerable technological effort is involved in industrial development. This effort can be called innovation, since it differs only in intensity and emphasis (not in kind) from the effort to create new products and processes. Such innovation arises at any point in the value chainfrom design and procurement to production, R&D and marketing. Chapter 6 shows how firms and countries can build a foundation for ongoing innovation and learning by competing in global value chains. Chapter 7 describes what governments can dolegitimately and effectivelyto help firms grasp new opportunities and solve technological problems. Chapter 8 makes the case that building competitive industrial capabilities needs extensive policy supportspelling out the framework imperatives, the elements of industrial strategy and the principles for government conduct of that strategy.

4. There is a large literature on the role of governments in industrial development in East Asia. See World Bank (1993), Amsden (1989), Wade (1990), Stiglitz (1996), Rodrik (1996), Westphal (forthcoming), Lall (1996) and Rodrigo (2001). 5. Stiglitz (1987). 6. Nonaka (1994); Teece (2000). 7. The linking, leveraging and learning framework comes from introducing a strategic perspective (from the literature of strategic management) to the prospects for a special kind of firm, namely the latecomer firm (see Mathews and Cho 2000 and Mathews 2001, background paper). Each of the terms has a long historylinkage from the literature on networks and global value chains, leverage from management strategy (Prahalad and Hamel 1990) and learning with sources too numerous to mention. 8. Prahalad and Hamel (1990); Hamel and Prahalad (1994); Mathews and Cho (2000). 9. Mathews (2002). 10. Wignaraja and Ikiara (1999). 11. Lall and others (1994). 12. Hobday (1995). 13. Rasiah (2000). 14. Dahlman, Ross-Larson and Westphal (1987).

Notes
For further details on sources, information and the literature on subjects covered here, see the background papers. 1. Lall (1992).

15. Ernst, Ganiatsos and Mytelka (1995). 16. The theoretical antecedents are evolutionary theories developed by Nelson and Winter (1982). For extensions and related approaches, see Dosi, Teece and Chytry (1998) and Metcalfe (1995). 17. Based on Lall (2000).

2. Enos (1992). 3. Nelson and Winter (1982).

104

Industrial Development Report 2002/2003

6
G

Innovation and learning in global value chains

LOBAL VALUE CHAINS SPANNING FUNCTIONS, PROCESSES AND

countries provide a means for accelerating the development of enterprises and countries, providing openings that developing country enterprises can exploit to upgrade their capabilities. For such enterprises, or local clusters of enterprises, the task is to insert themselves into the wider networks. This takes discipline, to attain the higher world standards. It also takes an initial base of technological capability, built through purposive innovation and learning. But the effort should be worth it, for it offers access to markets and the knowledge of players in the world economy. The advantage of global value chains is that enterprises can seek involvement at their level of technological competence. In Mexico garment producers were vertically integrated in supplier networks that did not offer much scope for skills enhancement and innovation. With the North American Free Trade Agreement (NAFTA), however, buyer groups from the United States started to create alternative global value chains that offered enterprises much greater scope for expanding their functional responsibilities (from narrow job completion to design and manufacture), termed full package production. This replicates the experience decades earlier when electronics and garment contract firms in East Asia pulled themselves up the capability ladder to higher and higher levels in global value chains. Competing in a global value chain can build a foundation for the industrial innovation and learning described in chapter 5. There are many paths for this: Process innovation, improving the efficiency of transforming inputs into outputs. Internal processes become significantly better than those of rivals, both within links in the chain (more inventory turnovers, less scrap) and between links (more frequent, smaller and on-time deliveries). Product innovation, leading to better quality, lower priced and more differentiated products, as well as shorter times to market for new products.

Functional innovation, assuming responsibility for new activities in the global value chain. That can involve extending involvement from contract manufacturing to design and marketing or incorporating logistics within the contracted work. Interchain innovation, moving to new and more profitable chains. Enterprises in Taiwan Province of China moved from the manufacture of transistor radios to calculators, to televisions, to computer monitors, to laptops and now to Wireless Application Protocol telephones. Some enterprises even latch onto several global value chains, providing further opportunities for linking to local enterprises connected with them (box 6.1). Such firms lift themselves and those connected with them in supply chainsto new levels of performance and quality, driving forward the momentum of collective industrial development. Such industrial learning is a long and strenuous process, with no short-cuts or magic solutions. The global value chains offer convenient structures to fashion this process, but they really offer only a starting point for the enterprises technological effort.

Links in the chains

The metaphor of global value chains captures the links among enterprises spread across a variety of locations around the world (figure 6.1). These enterprises perform a sequence of related dependent activities to bring a product or service from conception through the different phases of production to delivery to final consumers and to final disposal after use. The metaphor of global value chains is now being joined by the metaphor of value networks of specialist enterprises, suggesting rays fanning out from nodes rather than links in a chain. The global value chains are not just a teeming mass of complementary enterprises. They are an organized set of interconnected

105

Box 6.1 Jumping into the leadin global value chains The Ammar and Sarah knitwear group, founded in 1982, has been Pakistans leading knitwear group since 1988 because of its international connections in global buyer- and producer-led global value chainsand its commitment to state-of-the-art technology and management. It now has a knitted products line encompassing mens and boys wear and womens and girls wear in cotton, knitted tops and bottoms, with and without Lycra, in a broad range of fabric finishes. Ammar and Sarahs key to securing contracts from global buyers and producers is using the most advanced computer-aided technology, whether for washing, dyeing, cutting, knitting or stitching. That gives it great flexibility and lead time of 45 to 75 days in responding to new orders (not the full season lag in traditional firms). The companys major customers are all global buyers (Target, Arrow, Nautica, Haggar, Eddie Bauer, Vantage, Timberland, Alexander Julian Colors, Land Rover, Tommy Hilfiger, Nike and Damani Dada) or global producers (Levi Strauss and Sarah Lee).

Figure 6.1 Simple value chain Disposal and recycling

Marketing and sales

Manufacturing

Design and product development

Source: UNIDO.

Why the attraction of Ammar and Sarah? It can exploit its low labour and infrastructure costs (for as long as they last). It can purchase the most advanced technical equipment to provide flexible and highquality manufacturing services. It can recruit graduates trained in advanced technical institutes abroad. And its entrepreneurial founders can put their Harvard business training to good effect. These are all ways for a new firm to accelerate its catchup. Unlike the incumbents, it is not burdened with technologies and practices inherited from earlier eras.

arms-length transactions or intra-enterprise trade. The lead enterprises in global value chains play a major role in organizing trade. They range from the transnational producers that source inputs from suppliers around the globe to the retail chains that do not make goods but organize production at locations around the world. The impetus for the formation of these global value chains lies with enterprises in the advanced countries, either as buyers or producers. But the decision to latch onto these global value chains, or networks, lies with the latecomer enterprise in the developing country (figure 6.2). Links are not just between enterprises. Depending on the specific needs of a firm, it might be appropriate to link with universities or nonmarket institutions. The links can also be vertical (backward to suppliers or forward to clients) or horizontal (in consortia). Crucial factors for latching onto the global supply chain are not only the hard facts of price, quality and punctuality but also the willingness to learn and to absorb advice from the lead enterprises. Global value chains can thus unleash enterprisesbut they can also constrain them (box 6.2). A strategic perspective sees such linking and innovation possibilities as opportunities for enterprises, not as barriers to further development. Particularly in manufacturing, the insertion of local activities in wider networks is a great opportunity for developing countries to upgrade their capabilities. The types of global value chains depend on speed of change, learning needs, scale economies, transaction and coordination costs, value-to-weight ratios and logistics. Easy technologies can give rise to buyer-driven chains, while difficult technologies with close coordination needs, proprietary technologies and the like can promote supplier-driven chains coordinated by transnational corporations. Some analysts distinguish global value chains that are buyer-driven from those

The prospects for Ammar and Sarahprepared to make the international global value chain connections, to invest in state-of-the-art equipment and to compete on quality and speed of service rather than least costare promising.
Source: Ammar and Sarah marketing brochure (2000).

networks of enterprises linked with each other through multiple interactions and linkagesa worldwide web of inter-enterprise connections. The focus of interest is not just on the enterprises. It is also on the shifting links and contractual relations among them. Enterprises expand their product lines, and expand internationally by forging new links with enterprises already active in the global economy, dominated by criss-crossing global value chains encompassing research and development, production, logistics, marketing and exchange, where all the links are between enterprises rather than between countries. The global value chain provides two insights about innovation and learning. First, creating value is not confined to production. Products are brought to market through a combination of activities. So, innovation can involve improving capabilities in production, developing new capabilities outside production (design and marketing skills), diversifying customers and market destinations, developing the capacity to introduce new products or to imitate leading innovators quickly and successfully. Second, more international trade is taking place between formally independent enterprises in networks than through

106

Industrial Development Report 2002/2003

Figure 6.2 Linking local producers and global buyers

Box 6.2 Pluses and minuses of being in a global value chain In the late 1960s the Sinos Valley shoe cluster, in the South of Brazil, was made up mainly of small firms producing for the domestic market. With the arrival of buyers from the United States, and stimulated by local initiatives and Brazilian government export incentives, the characteristics of the cluster began to change. The buyers looked for large volumes of standardized products and encouraged a rapid increase factory size. They also helped their suppliers raise process standards and product quality. They also eased the considerable risks of entering export markets. They studied the market, developed models, worked out the product specifications, helped choose technology and organize production, inspected quality on site and set up transport and payment arrangements. The firms in Sinos Valley concentrated on production and the organization of their own local supply chains, while the buyers were responsible for product definition (and hence, market knowledge) and logistics. This greatly reduced the investment and risks in entering export markets, but it also confined firms in the Valley to a narrow range of functions. Becoming very competent in these functions, they benefited from rapid growth in export sales in the 1970s and 1980s. But they also depended on the buyers, evident when Chinese producers undercut Brazilian products in the United States market in the early 1990s. This is a danger inherent in global value chains. Global buyers actively scout for new sources of supply, and substitution by new sources is always a threat to existing suppliers. Indeed, some of Brazils main buyers in the United States helped to build Chinese export capability. As a result, the Brazilian producers were faced with sharply declining prices for their products in North America. But by reorganizing their factories and local supply chains, they raised quality, reduced batch size and increased speed. Indeed, the buyers helped them switch to a new way of producing.

Global customer

Large-scale, multi-outlet retailer

Small-scale retailer

Buyer (including specilized buyers and transnational corporations)

National boundary

Buyer and export agent

Local customers

Large-scale or multi-plant manufacturer

Small-scale manufacturer Local cluster

Source: Kaplinsky and Readman (2000).

Small-scale manufacturer

The advances in production were not matched, however, by advances in marketingeven though firms tried. The Brazilian producers worked out a collective strategy of raising Brazils image in the world footwear markets, strengthening design capabilities, and exhibiting in significant numbers at the worlds main trade fairs. But the proposed strategy was not put into practice, mainly because a small number of very influential export manufacturers did not support it. They feared that advancing into design and marketing would upset the relationship with their main foreign buyer, which accounted for more than 80 percent of their output and close to 40 percent of the clusters output.
Source: Schmitz (1995, 1999b).

that are producer-driven (table 6.1). Global value chains can also be regional or national, providing a local latecomer enterprise with opportunities to be pulled into a wider network of activities through contracting its services to enterprises beyond its immediate environment.

Staying nimble in the turmoil of global value chains

Entering global value chains does not provide an automatic move up the capability ladder. It is often a fast track to acquir-

ing production capabilities, but moving further up the chain can lead to conflicts with existing customers.1 Some enterprises even have had their capabilities downgraded as a result of their integration in global value chains. So, it makes sense for latecomers to use all the resources they can acquire from the advanced world, in return for providing such services as low-cost manufacturing. But the tradeoff can be exploited to the advantage of the latecomer only if there is a strategic choice to use the links to gain knowledgeto learn. Innovation within global value chains moves along two dimensions of leverage strategies: market expansion and technological capabilities. Own brand manufacturing, usually the most

Innovation and learning in global value chains

107

Table 6.1 Characteristics of producer-driven and buyer-driven global value chains

Figure 6.3 Leverage paths within two dimensions Market expansion

Characteristics Driver of global chains Core competencies

Producerdriven chains Industrial capital Research and development (R&D), production Consumer goods, intermediate goods, capital goods Automobiles, computers, aircraft Transnational corporations Investment-based

Buyerdriven chains Commercial capital Design, marketing Global logistics contracting

A Own brand manufacturing B

Sectors

Typical industries Ownership

Main network links

Source: Gereffi (1999b).

Non-durable consumer goods Apparel, footwear, toys Local enterprises, predominantly in developing countries Trade-based

Original equipment manufacturing Starting point

Own design and manufacturing

profitable segment of a global value chain, requires both market and technological competencies (figure 6.3).2 Path A represents a trajectory along which many of the activities entailed in original equipment manufacturing, all of them initially accomplished domestically along with key activities, are relocated to production facilities in third countries, giving rise to triangle manufacturing. Capability enhancement is centered on mastering the complex of logistical functions required when sourcing and combining inputs from a number of different producers and locations. Path B by contrast focuses on capability enhancement through expanding functional responsibilities, from original equipment manufacturing to including some responsibility for design, leading the enterprise to then market its own designs under its own brand. Enterprises pursue market niches by developing unique production capabilities, often of a technological form. But the process of developing such capabilities creates new market opportunities in the form of a redesigned product to meet customer needs better. The interactive process is endless. The insertion of a local enterprise in a global value chain instigated by a buyer or a producerputs great pressure on the enterprise to meet demanding quality, reliability and logistics standards. But the buyer or producer also wants to be able to make rapid product adjustments (in response to shifting patterns of consumer demand in their stores, for example), and so there is also great pressure to change product lines quickly and reliably. The endpoint is an enterprise that has attained full lean production capabilities in flexibility and agility. Then there is the all-important step of moving from one functional specialization to another. The move from production to design might seem a small step in itselfbut it is a huge step

Technological capabilities
Source: Mathews and Cho (2000).

for a latecomer enterprise looking to build its capabilities. It is the first step towards self-sufficiency, where the enterprise might no longer be entirely dependent on the global value chain for its survival. This step is sometimes taken by the individual enterprise itselfas with East Asian electronics firms. They moved through phases of original equipment manufacturing, where the buyer enterprise gives all specifications to contracting firms, to own design and manufacture, where the buyer enterprise simply gives broad specifications and allows the contractor to fill in the details, to own brand manufacturing, where the enterprise is fully fledged and produces its own line of branded products. Last is the all-important break from one global value chain to another. Of course inserting an enterprise or local cluster into a global value chain is an important stepbut the smart enterprise or cluster does not have to see its horizons limited. Always seeking ways of spreading its involvement across two or more global value chains, it looks to expand its options and capabilities. This leverages skills, enhances capabilities and reduces the risk of being tied to a single global value chain. In Taiwan Province of China, television producers in the electronics industry used global value chains instigated by buyers in the United States like J.C. Penney and Kmart to leverage the skills in mass production of television sets. They then transferred these skills to produce computer monitors for such computer producers as Hewlett Packard, IBM and Apple, which were building quite different global value chains. This cross-insertion builds a vari-

108

Industrial Development Report 2002/2003

ety of capabilities and provides a platform of independence for the developing enterprise. The chapter now turns to the global value chains for garments and for wood furniture. The first shows the dynamics within a global value chain, dynamics that demand considerable nimbleness from the enterprises and local clusters working in them. The second shows what a local cluster has to do to move into a global value chain.

Figure 6.4 Apparel value chain

Raw materials

Machinery

Natural and synthetic fibres

Dye stuff/ bleacher

Trust and triangles in garments

The apparel industry is labour-intensive, with labour accounting for 60 percent of production costs.3 Asia has become the dominant region of production. This trend started in the 1950s and 1960s as the industry shifted from Europe and the United States to Japan. The second shift was from Japan to Hong Kong Special Administrative Region (SAR) of China, Taiwan Province of China and the Republic of Korea, which dominated in the 1970s and 1980s. In the 1990s production shifted to China and other Asian countries and to some Latin American countries. The big export drivers in the apparel business have been quotas and preferential tariffs. Quotas on apparel and textiles items will continue to be regulated by the Multi-fibre Arrangement (MFA) until it expires in 2005. Used by the United States, Canada and various European nations since the early 1970s to impose quantitative limits on imports, the clear intent was to protect industrialized country enterprises from a flood of low-cost imports that threatened to disrupt major domestic industries. The long-run result was just the opposite. Protection in industrialized countries heightened the competitive capabilities of developing country manufacturers, who learned to make sophisticated products that were more profitable than simple ones. In recent years the European Union and NAFTA have granted preferential tariffs within regional markets, shifting global sourcing dynamics in these regional markets. The clothing global value chain ranges from raw materials processing and production of textiles and manufacturing garments, to marketing and retail (figure 6.4). Aside from the upstream activities there are four stages of moving up the chain:4 1. Assembly of imported products (typically in export processing zones near major ports). 2. Original equipment manufacturing. Production for transnational corporations (design specification comes from foreign company, which is responsible for marketDesign

Textile manufacturer

Logistics, quality advice

Apparel buyers

Marketing/ branding

Domestic wholesale Domestic retail Consumers

Foreign wholesale Foreign retail

Source: Appelbaum and Gereffi (1994).

ing and branding). Supplier lacks control over distribution. A variant is global logistics contracting. 3. Own design manufacturing. Design of products sold under the brand of foreign firms. 4. Own brand manufacturing. Sale of own branded products. Entry barriers are low for most garment factories, but they get progressively higher in the move upstream to textiles and fibres.

Three global buyers

The apparel chain has three categories of buyers: retailers, branded marketers and branded manufacturers. The retailers account for 50 percent of imports, branded marketers and branded manufacturers 20 percent each, and various others for the rest.

RETAILER
Such international retailers as Wal-Mart and Sears Roebuck, once the apparel manufacturers main customers, are now

Innovation and learning in global value chains

109

their competitors. In the 1980s many retailers began to compete directly with the national brand names of apparel producers and marketers by expanding their sourcing of private label merchandise. These products are sold more cheaply than the national brands, yet they are also more profitable to the retailers, who can eliminate the middlemen in the chain. Private label goods were about 25 percent of the apparel market in the United States in 1993. While retailing and marketing are becoming more concentrated, manufacturing is splintering. Todays superior information flows give retailers far better day-to-day market knowledge about consumer purchasing decisions, allowing them to demand more from their suppliers in better inventory management, faster responses and more frequent deliveries. As each type of buyer in the apparel commodity chain has become more active in offshore sourcing, the competition between retailers, marketers and manufacturers has intensified, blurring the traditional boundaries between these enterprises and realigning interests within the chain.

they are unable to compete with the low cost of foreign-made goods, they are defecting to the ranks of importers. The decision of many larger manufacturers in industrialized countries is no longer whether to engage in foreign production but how to organize and manage it. They supply intermediate inputs (cut fabric, thread, buttons and other trim) to extensive networks of offshore suppliers, typically in neighbouring countries with reciprocal trade agreements that allow goods assembled offshore to be reimported with a tariff charged only on the value added by foreign labour. This kind of international subcontracting system exists in every region of the world. It is called the 807/9802 program or production sharing in the United States (USITC 1997), where the sourcing networks of U.S. manufacturers are predominantly in Mexico, Central America and the Caribbean because of low wages and proximity to the market. The trend for the branded manufacturers is to de-emphasize production in favour of marketing by capitalizing on brand names and retail outlets. Sara Lee Corporation, one of the largest apparel producers in the United States, recently announced its move out of making the brand-name goods it sells.

BRANDED MARKETER
Well-known manufacturers without factoriessuch as athletic footwear companies (Nike, Adidas, Puma) and fashionoriented apparel companies (The Gap, Liz Claiborne)carry out no production. Instead, they just design and market their goods. As pioneers in global sourcing they provided knowledge that later allowed overseas suppliers to upgrade their own positions in the apparel chain. To deal with new competition, branded marketers are discontinuing some support functions (such as pattern grading and sample making) and reassigning them to contractors. They are instructing contractors where to obtain needed components, reducing their own purchase and redistribution activities. They are shrinking their supply chains, using fewer but more capable manufacturers. They are adopting more stringent vendor certification systems to improve performance. And they are shifting their sourcing configuration from Asia to the Western Hemisphere. Marketers (and retailers even more) now recognize that overseas contractors can manage all aspects of production, which offers linking and leveraging opportunities for contractors to move into designing and branding.

Latching onto the global value chain

The first step for garment manufacturers in developing countries is to become linked to branded manufacturers. The easiest way has been to engage in contract manufacturing under U.S. tariff schedule provision 807/9802. But those activitiesoften performed in export processing zoneshave low value added. Enterprises in the United States engaged in production sharing have an incentive to minimize locally purchased inputs since only components made in the United States are exempt from import duties when the finished product is shipped back to the United States (box 6.3). There is a similar system in Europe, known as outward processing trade, with the principal suppliers located in North Africa and Eastern Europe. The same holds in Asia generally, where manufacturers from relatively high-wage economies like Hong Kong SAR have outward processing arrangements with China and other low-wage nations. The next stage after export processing is to link with global retailers or branded marketers in original equipment manufacturing or full-package production. Compared with the mere assembly of imported inputs, full-package production fundamentally changes the relationship between buyer and supplier in a direction that gives far more autonomy and learning potential for industrial innovation to the supplying enterprise. Full-package production is needed because the retailers and marketers that order the garments have limited knowledge of their manufacturing details. Hong Kong SAR, Taiwan Province of China, Republic of Korea and China used the full-package route to create an enduring edge in export-oriented development.

BRANDED MANUFACTURER
Apparel manufacturers, such as Levi Strauss, have been caught in a squeeze because foreign producers can often provide the same quantity, quality and service as domestic producers, but at lower prices. In the United States and Europe, the attitude among many smaller and mid-sized apparel manufacturers is If you cant beat them, join them. Feeling that

110

Industrial Development Report 2002/2003

Box 6.3 Races to the bottom The Dominican Republic has an especially large dependence on export processing zone assembly using the U.S. 807/9802 trade regime. The share of export processing zones in official manufacturing employment increased from 23 percent in 1981 to 56 percent in 1989, when they generated more than 20 percent of foreign exchange earnings. Investors in the United States account for more than half (54 percent) of the companies operating in the zones, followed by firms from the Dominican Republic (22 percent), Republic of Korea (11 percent) and Taiwan Province of China (3 percent). The rivalry among export processing zones in neighbouring countries to offer transnational companies the lowest wages fosters a perverse competitive devaluation, where currency depreciations are seen to increase international competitiveness. Export growth in the Dominican Republics export processing zones skyrocketed after a very sharp depreciation of its currency against the dollar in 1985. Similarly, Mexicos export expansion was facilitated by recurrent devaluations of the peso, most notably in 199495. Hourly compensation rates for apparel workers in the early 1990s were $1.08 in Mexico, $0.88 in Costa Rica, $0.64 in the Dominican Republic and $0.48 in Honduras, compared with $8.13 in the United States. It may make sense for one country to devalue its currency to attract users of unskilled labour to their production sites. But the advantages quickly evaporate when other nations simultaneously engage in wage-depressing devaluations, which lower local standards of living while doing nothing to improve productivity.
Source: Kaplinsky (1993); ILO (1995).

Box 6.4 Linking to the leaders The key factor in Mexicos ongoing transition from assembly to original equipment manufacture (or full-package) production has been NAFTA, which began to remove the U.S. restrictions that had virtually locked Mexico into assembly. The maquiladora system effectively conditioned Mexicos access to the U.S. market on the use of its inputs. More of the apparel supply chaincutting, washing and producing textilesis relocating to Mexico as U.S. restrictions on each of these stages is eliminated. But NAFTA does not guarantee Mexicos success. While the massive peso devaluations of 199495 made Mexico very attractive as a production site for U.S. apparel manufacturers with international subcontracting operations, Mexico has traditionally lacked the infrastructure and supporting industries to do full-package production of garments. Textile and apparel companies in the United States have been expanding their investments in Mexico at a rapid and accelerating pace. So Mexico is now better positioned to provide the quantity and quality of inputs needed for original equipment manufacture of standard apparel items, such as jeans, knit shirts, trousers and underwear. But Mexico is still lagging in the fashion-oriented, womens wear categories. The solution to completing the transition to full-package supply, and developing new production and marketing niches, is to forge links to the kinds of lead enterprises that can supply technology and tutelage. Mexico needs to develop new and better networks to compete with East Asian suppliers for the U.S. full-package market. Enterprises in the United States have already shown a strong interest in transferring missing pieces of the North American apparel supply chain to Mexico. A real problem to be confronted, though, is who controls critical nodes of the chain and how to manage the dependency relationships this implies. Thus far, enterprises in the United States are in clear control of the design and marketing segments of the apparel chain, while Mexican companies are in a good position to maintain and coordinate the production networks in apparel. But textile manufacturers in the United States, and to a lesser degree Mexico, are making strong bids to integrate a broad package of apparel services that would increase their leverage over smaller garment contractors. Mexico is likely to retain a mix of assembly plants linked to U.S. branded manufacturers and a new set of full-package producers linked to private-label retailers and marketers. As more of the critical apparel inputs become available in Mexico, inputs from the United States will decline and traditional Mexican assembly plants will be replaced by full-package manufacturers or by clusters of related enterprises that compete through localized networks, such as the jeans producers in Torren.
Source: UNIDO.

But NAFTA, along with a relative decline in the importance of East Asian apparel exports to the United States, has now created favourable conditions for extending full-package production to the North American setting (box 6.4). Prominent apparel suppliers to Europe, such as Turkey and several East European economies, also appear to be adopting the full-package model. Manufacturers from those countries need to acquire the skills and resources to move into the more diversified activities associated with full-package production. The arrangement offers further innovation opportunities towards own brand manufacture. It enhances the ability of local entrepreneurs to learn the preferences of foreign buyers, including international standards for the price, quality and delivery of export merchandise. It also generates substantial backward linkages in the domestic economy because original equipment manufacture contractors are expected to develop reliable sources of supply for many inputs, including those to be imported. The supplier learns much about the downstream and upstream segments of the apparel commodity chain from the buyer. This tacit knowledge can later become a powerful competitive weapon. One of the most important mechanisms facilitating the shift to higher value-added activities for mature export industries like apparel in East Asia is the process of triangle manufacturing (global logistics contracting). The essence of triangle manufac-

turing, initiated by the East Asians in the 1970s and 1980s, is that global buyers place their orders with the manufacturers they have sourced from in the past; those manufacturers then shift some or all of the requested production to affiliated offshore factories in low-wage countries (China, Guatemala, Indonesia). These offshore factories can be wholly owned subsidiaries, joint-venture partners or simply independent overseas contractors. The triangle is completed when the finished goods are shipped directly to the overseas buyer under the U.S. import quotas issued to the exporting nation.

Innovation and learning in global value chains

111

Triangle manufacturing thus changes the status of original equipment manufacture from established suppliers for retailers and designers in the United States to middlemen in buyerdriven commodity chains that can include as many as 50 to 60 exporting countries (box 6.5).

Box 6.5 From trust to triangles to own brand manufacturing The East Asians did not employ the production-sharing provisions established by the 807/9802 U.S. trade regime in apparel because their great distance from the United States made textile imports from the United States impractical. In addition, textile mills in the United States did not have the production capability or mentality to supply the diverse array of fabrics favoured by the designers of womens wear and fashion-oriented apparel, which became the specialty of the East Asian exporters. Both factors created an original equipment manufacture niche, adroitly exploited, for East Asian apparel companies. Highly successful textile and apparel exporters from Hong Kong SAR, Taiwan Province of China and Republic of Korea (preceded by Japan, followed by China) progressed through a sequence of export roles from assembly to original equipment manufacture to own brand manufacture. They developed and refined their original equipment manufacture capabilities in the 1960s and 1970s by establishing close ties with retailers and marketers in the United States, and then learning by watching to use these foreign partners as role models to build East Asias export capabilities. The performance trust built up through many successful business transactions with these U.S. buyers enabled East Asian suppliers to internationalize their original equipment manufacture expertise through triangle manufacturing. The East Asian manufacturers became intermediaries between the buyers in the United States and hundreds of apparel factories in Asia and other developing regions to take advantage of lower labour costs and favourable quotas around the world. The creation of these global sourcing networks helped the East Asians sustain their international competitiveness when domestic economic conditions and quota constraints threatened the original, bilateral original equipment manufacture relationships. The East Asians have been moving beyond original equipment manufacture in many ways. They have shifted to higher value upstream products in the apparel commodity chain (exports of textiles and fibres, rather than apparel). They have been moving downstream to own brand manufacture in apparel. And they have been aggressively investing in efforts to switch to other global product chains. The Republic of Korea is the most advanced of the East Asians in own brand manufacture, with its brands of automobiles (Hyundai), electronic products (Samsung) and household appliances (Samsung and Goldstar), among other items, being sold in North America, Europe and Japan. Companies in Taiwan Province of China have pursued own brand manufacture in computers, bicycles, sporting equipment and shoes, but not in apparel. Clothing companies in Hong Kong SAR have been the most successful in shifting from original equipment manufacture to own brand manufacture. The womens clothing chain, Episode, controlled by Hong Kong SARs Fang Brothers Group, one of the foremost original equipment manufacture suppliers for Liz Claiborne in the 1970s and 1980s, has stores in 26 countries, only a third of which are in Asia. Giordano, Hong Kong SARs most famous clothing brand, has added to its initial base of garment factories 200 stores in Hong Kong SAR and China, and another 300 retail outlets scattered across Southeast Asia and Republic of Korea. Hang Ten, a less-expensive line, has 200 stores in Taiwan Province of China, making it the largest foreignclothing franchise on the island.
Source: UNIDO; Granitsas (1998); Gereffi (1997, 2000).

Opportunity and initiative in wood furniture

In 1998 the furniture industry, with global trade of close to $45 billion, was the largest traditional, low-tech sector, exceeding both apparel ($41 billion) and footwear ($34 billion).5 Although furniture is a resource- and labour-intensive product, many of the major furniture exporting countries are industrially advanced (table 6.2). Italy is far and away the leader, with net exports of $7.8 billion in 1998. Developing countries in the top 10 are China, Mexico, Malaysia, Romania and Indonesia. The wooden furniture global value chain starts with the provision of seed inputs, chemicals, equipment and water for the forestry sector (figure 6.5). Cut logs pass to the sawmill industry, which obtains its primary capital inputs from the machinery industry. Sawn timber moves to the furniture manufacturers who, in turn, obtain inputs from the machinery, adhesives and paint and other industries and also draw on design and branding skills from the service sector. Depending on the market served, the furniture then passes through various intermediary stages until it reaches the final customer, who after use consigns the furniture for recycling. The chain is very heterogeneous due to the many market segments (office, kitchen, bedroom, dining room and living room) and within these segments the many market niches (high volume, price sensitive, design intensive, brand-intensive, and so on). Three major buying agents facilitate the entry of wood furniture producers into final markets: Large multinational retailers, with both retail outlets and suppliers in many countries. (For example, IKEA sources from 2,000 suppliers in 52 countries and has more than 300 outlets in three continents) Small retailers, purchasing directly from a limited number of suppliers in a limited number of countries. Specialized medium-size buyers, sourcing from many countries and on-selling to retail outlets, predominantly in a single country or region. It is not atypical for these buyers to have more than 1,500 suppliers, in many coun-

112

Industrial Development Report 2002/2003

Table 6.2 Global furniture tradetop 10 net exporting countries, 1994 and 1998

Figure 6.5 Links in the wood furniture value chain

Italy China Canada Denmark Mexico Malaysia Spain Sweden Romania Indonesia
Source: http://www.intracen.org.

Net export value (millions of dollars) 1994 1998 6,105 7,831 1,381 2,725 32 1,804 1,412 1,323 259 1,190 698 1,052 251 741 254 494 375 382 754 339

Seeds

Machinery

Water

Chemicals

Forestry

Machinery

Sawmills Design Furniture manufacturer Machinery Buyers Domestic wholesale Domestic retail Consumers Foreign wholesale Foreign retail Paint, adhesives Logistics, quality advice

tries; even the smaller specialized buyers will typically source from more than 100 suppliers. In general, buyers serve different market segments. Often these segments are distinctively different, but the growing capabilities of world class manufactures means that there is a diminishing trade-off between critical success factors. For example, the large retailers are increasingly able to offer low prices and high quality, and low prices and variety. Suppliers confront a much more demanding set of critical success factors when they sell to global retailers than when they sell to small retailers and specialist buyers. Not only are almost all the critical success factors considered important, but they are also all ranked as being of higher order importance. The innovation challenge confronting part of the wood furniture global value chain in South Africa is symptomatic of a more general challenge facing other furniture exporting countries. South Africas wood furniture global value chain has been on a suboptimal trajectory since its pine furniture has faced increasing price competition in overseas markets. The unit prices of its exports, measured in dollars, fell by 250 percent between 1992 and 1999. Moreover, South African products have been considered cheap, but of low quality and poor delivery reliability. As a consequence IKEA, the major global buyer, decided to move out of South Africa (to Eastern Europe and East Asia). This has placed the South African wooden furniture firms in a dilemma. An effective response was found, after much searching, in the context of the global trend towards environmental responsibility. South Africa is the home of a commercially grown semi-hardwood named saligna. Furniture based on saligna offered the potential to become a low-cost and environmentally acceptable alternative to increasingly scarce and highly priced traditional hardwoods such as teak and mahogany.

Recyling

The opportunity
One of the key dynamic market forces in the global timber products industry is the move (primarily by the industrial countries) towards environmental responsibility. For most developing countries, this threatens their exports because their timber product industries have traditionally drawn on indigenous hardwood forests. South Africa, however, happens to be uniquely placed to take advantage of this opportunity. The most outstanding feature of saligna (a species of Eucalyptus hardwood) is that in South Africa it is a commercially grown semi-hardwood distinguishing it from other hardwood species grown in indigenous forests in the developing world. Although saligna is not a traditional hardwood, it has the ability to take colouring well and can therefore be treated to look like virtually any wood, including all the species of threatened hardwoods. Traditionally saligna was grown for use in the local mining industry, but the changeover to concrete mining supports has

Innovation and learning in global value chains

113

led to a sharp decline in domestic demand. In the context of growing environmental concerns in final markets, therefore, the existence of the previously low-priority saligna hardwood plantations, with unused capacity, offers unexpected potential for exporting of furniture to Europe and North America. It is also an opportunity that offers the potential to move furniture producers into new market niches, with higher unit prices.

paper and pulp meant that unless the final furniture products could be positioned within a relatively higher product niche than South Africas pine furniture exports, the manufacturers would not be able to survive paying the market price for the timber input. An additional product innovation challenge was that the specific properties of saligna (when compared with pine), and especially of young saligna, meant that the designs used for pine furniture could not always be translated into the new type of wood. Product redesigndesign for manufacture was therefore a necessity, which required many furniture manufacturers to venture into new territory, and this could not be done in isolation from the sawmills. Finally, one of the virtues of saligna was its ability to absorb finishes, and this required the manufacturers to work closely with lacquer and paint suppliers, particularly because environmental pressures in Europe are forcing a move to water-based finishes (one of the main areas of competitive advantage of Italian producers).

The innovation challenge

Grasping this opportunity requires inter-chain innovations, a reorientation from the previous trajectory of the wooden furniture global value chain, which has traditionally focused on the export of pine furniture into increasingly price-competitive markets. This reorientation entails substantial inter-chain innovation through simultaneous and carefully coordinated process, product and funtional innovations.

PROCESS INNOVATION FUNCTIONAL INNOVATION

The primary challenge was to increase the supply of clear saligna hardwood, at an affordable price. This challenge exists both because of competing uses (in pulp and paper), for which clarity is unimportant, and because the sawmills serving furniture manufacturers were geared for cutting softwoods (pine) rather than hardwoods (saligna). The mills had also operated in a sellers market for many years, and consequently were unresponsive to the needs of the manufacturers, delivering at unpredictable intervals, with varying quality and in inconvenient take-it-or-leave-it product specifications. An additional processing problem was that manufacturers needed to learn how to work with saligna, and to be effective, this required close collaboration with the sawmills (for example, in regard to knowledge about timber density). Perhaps most important, the key determinant of timber costs was the gestation period of the trees. Traditionally, saligna had been cut at an age of 23 years, but it was thought possible to reduce this considerably, to around 12 years; given high interests costs (a real interest rate of more than 10 percent), the financial benefits to this innovation would be considerable. But to be effective it required close collaboration between growers, the sawmills and the manufactures. Thus, process innovation could only be achieved in the saligna furniture global value chain through a combination of enterprise-specific innovations and inter-enterprise collaboration to enhance communication within the chain and to address important chain-specific problems. If new designs were to be introduced, who would take responsibility for this high value-added activity? Would the saligna industry fall back on the pattern in the pine industry, where global buyers provided design templates for manufacturers, or where manufacturers continued to produce standard items such as garden benches? Alternatively, would there be a surge in domestic design capabilities, and if so, would these be lodged in South African buyers, furniture manufactures or in specialized design houses? Just as saligna furniture represented a transition within the wood furniture chain from softwood to a hardwood, were there also opportunities to move from saligna furniture to other saligna-based products such as garage doors (a big export item), industrial products and toys?

The initiative
To stimulate innovation, a first saligna network workshop was organized in late 1998 by a university-based research project. It was well attended by government departments, manufacturers, timber traders, industry specialists (both academic and consultants) and timber growers and sawmills. It successfully brought together stakeholders from all levels of the saligna global value chain with a view to promoting cooperative problem resolution. The involvement of a number of competing enterprises at each level of the global value chain created a situation where a failure to cooperate held the risk of missing out on benefits enjoyed by competitors. The workshop gave birth to the Saligna Global Value Chain Group (SVC Group), a cooperative national network of stakeholders spread throughout the global value chain.

PRODUCT INNOVATION
In itself, process innovation would not produce sufficient benefits. The problem was that alternative uses for saligna in

114

Industrial Development Report 2002/2003

Technical working groups, led by the sawmills, worked on a variety of issues for improving knowledge flows. A questionnaire was sent to all timber product customers to try to establish optimal sizes and to get consensus on a range of dimensions that manufacturers felt most comfortable with. The mills then experimented with new grading systems to see if this could increase the total availability of clear wood. They also began to collect more accurate data on total demand in order to determine overall existing and potential supply and usage of saligna in South Africa. The full flowering of a Salinga furniture global value chain remains a work in progress, with much yet to be achieved to realize the promise that may inhere in exports of salinga furniture to advanced country markets. So far, the activities of the SVC Group have yielded the greatest efficiency gains in the areas of: Generating information in all three of the innovation trajectoriesprocess, product and function. Markedly improving inter-enterprise process and supply chain efficiency between the mills and manufacturers. Important product development occurring both within and between linkages through the young tree and wood density experiments. Internal enterprise process innovation primarily of a technical nature. Some gains in changing the mix of activities within enterprises and up the global value chain through emphasizing design, finishing and marketing. Innovation in the production processes of the firms in the global value chain was not an explicit focus of the activities of the SVC group. But work on the numerous supply issues between the sawmills and the manufacturers in the global value chain did in fact have an innovation impact on the internal production processes of the manufacturers, through challenging the technical parameters of what they could produce.

Working local to go global

Enterprises are thus part of a local industrial fabric. Despite globalization and new communication technologies, geographical proximity and local sources of competitiveness are still important. The local advantages of synergy have been well documented in recent case studies on industrial clusters. They show the passive and active gains that clustering can provide to enterprises. The passive gains arise from agglomeration economies; the active, from inter-enterprise cooperation. The success and failure of clusters depend on achieving dynamic synergy within the cluster and on being nimble in the interactions with the outside world. For developing countries the capturing of cluster benefits is difficult and elusive. If it is hard to start individual enterprises, it is harder still to start clustersor to get enterprises to cooperate locally, as customers and suppliers of each other, rather than as cut-throat competitors. In the long run developing country clusters will have to be inserted into a wider cluster into a global value chainif they to survive in the face of continuing global competition.

Notes
For further details on sources, information and the literature on subjects covered here, see the background papers. 1. Schmitz (1999b). 2. Similar innovation models are offered by scholars in the formerly developing countries, such as Korea (Kim 1998, 1999). 3. This section draws on Mathews (2001, background paper). 4. Gereffi (1999b). 5. This section draws on the Web site of the UNCTAD-WTO International Trade Centre: http://www.intracen.org and Kaplinski (2001, background paper).

Innovation and learning in global value chains

115

7
S

Supporting innovation and learning by firms

UPPORT INSTITUTIONS AFFECT HOWAND WHETHERFIRMS MEET

the information, skill, finance and other needs that are difficult to satisfy in open markets. Infrastructure determines the cost of operation and interacting with the outside world. A nurturing environment is required to foster vibrant industrial development. And ensuring access to vital services that support innovation and learning is a critical part of establishing that environment. Many of these services are supplied through the market in advanced countries, but even these countries find it necessary to augment what is supplied through the market with subsidized services. Various considerations provide ample justification for the provision of subsidized services to support the process of innovation and learningeven more for developing countries. Most important: what is being provided is, to a greater or lesser degree, depending on the service, a public goodin short, knowledge (or information) in one form or another. It is widely appreciated that knowledge is indeed a public good, one to be made available at a price no greater than the marginal cost of its dissemination. But the costs of searching for and translating even freely available information into terms useful to local firms are not trivial. And there are great economies in centralizing these activities in organizations with special capabilities to carry them out. Efficiency requires that these costs, separate from the vastly lesser variable costs of dissemination, be borne but once. Otherwise each potential beneficiary of the same information would have to replicate the search and translation costs that would far better be shared, as fixed costs, in some way among all the potential beneficiaries. There are good reasons for not imposing on beneficiaries the full, or even partial, sharing of the fixed costs of establishing and maintaining technology support organizations. One is found in the pronounced economies of scale inherent in their operations. These justify providing the services well in advance of the point where the market would be large enough to sustain the delivery of their services by private entities. Other reasons are found in the externalities generated through the use of these services to achieve higher productivity with existing

resources. In many instances the benefits from higher productivity cannot be fully captured by the firms that receive the services. Some (often most) of the benefits spill over to other economic agents in the form of externalities.1 These reasons have particular force in developing countries, where markets for industrial services are only beginning to be developed and where externalities related to technology transfer and effective absorption are particularly pronounced.2 Market failures not directly related to the provision of industrial services typically afford additional justification in less developed countries. The most obvious example involves financial sector failings in lending for technological efforts. Financial institutions in most developing countries are illequipped to appraise properly projects that entail technological efforts of kinds not previously undertaken locally. Even where they can conduct proper appraisals, they typically require collateral in forms that greatly raise the cost of borrowing or preclude it. So the services that ought to be used will go unused unless provided at lower cost to the recipient. Finally, constraints on public policy often mean that the provision of industrial services is the only practicable means of subsidizing technological efforts that should, according to first-best principles, be subsidized directly. Indeed, such constraints lie at the heart of the rationale for infant-industry promotion through protectionist means. Given that such means are no longer tolerated under international trade conventions, the promotion of innovation and learning through subsidized industrial services provision has to be thought as having much greater importancemuch greater. The World Trade Organization (WTO) affords little scope for policies that have been used successfully in the past to accelerate industrial and technological development. In an important sense, all that is left in the way of major policies to developing countries is to provide industrial services in the forms discussed in this chapter. That makes the direction and management of organizations that provide these services all the more important.

117

Many types of institutions are essential in supporting the innovation and learning by firms. Training and specialized education are very important, as are financial services. The focus here, however, is on the institutions directly supporting the innovation and learning efforts of firms.

often not the absence of such institutionsit is that existing institutions cannot be justified on the basis of the benefits from the services they provide (where services are indeed provided). If there is no demonstrated need for the delivery of some service, demonstrated within the parameters of the national strategy, the service should not be provided. Second, as a general rule, subsidized provision of industrial services has greater justification the more widely shared the specific services rendered. The closer some service comes to serving only one or a few firms, the more difficult it becomes to justify subsidized delivery on the grounds just enumerated. Third, the services should not be supplied solely by government. As quickly as is feasible, they should be supplied in public-private partnerships or by private firms and associationswith subsidies, if justified, or without, if the market can supply the services. Indeed, the reliance on

Helping firms grasp opportunities and solve problems

What principles, then, should guide provision of subsidized services for innovation and learning? Three are paramount. First, support institutions should be established and managed, and subsidized services should be provided in strict accord with the framework of the national strategy for industrial development (chapter 8). This is in line with the observation by many commentators that the problem is

Box 7.1 Institutional support to technological efforts of firms Basic industrial services
Promote inward investment Provide export services Provide management services

Productivity Centres
Improve quality Improve productivity, efficiency Provide training

Collect marketing information Collect data on exports and imports Provide managerial consulting

Technological Extension Agencies

Extend available technology to businesses lacking technical

Provide financial services (accounting, tax assistance, investment

capabilities
Help firms use cleaner production technologies

advice) Technology Information Centres

Provide information on available technology Provide information technology to firms, including networks, soft-

ware, Internet capabilities, intranet, and databases

Perform troubleshooting, assistance, and repair to firms

Identify problems and use access to technology sources to solve

problems
Serve as external consultants and assist firms with trouble-shooting Provide training in informational technology applications Promote cooperation of small and medium-size enterprises with

Metrology, Standards, Testing, and Quality Control Centres

Define domestic standards Assist firms in meeting International Organization for Standard-

larger research and cluster initiatives (South Africa MAC program) Research and Development Laboratories
Design new processes and products. Train

ization (ISO) compliance standards Train firms in ISO standards and regulatory requirements Test products to ensure compliance with standards Provide technical assistance to firms
Help firms with calibration of instruments

businesses through demonstration, participation and extension

Implement new technologies

Maintain calibrated standards and calibration equipment Calibrate firms machinery

Import and learn foreign technology Adapt foreign technologies to local needs Integrate these technologies into economy in collaboration with firms

Source: UNIDO.

118

Industrial Development Report 2002/2003

private associations to deliver subsidized services has greatly improved the efficiency, relevance and quality of public servicesand strengthened the cooperation between firms and the support organizations. Institutions that support the technological efforts of firms reflect the wide variety of industrial activities and needs related to them (box 7.1). Some organizations offer general services, meeting needs that are not specific to particular industries. Information centres that are effectively gateways to the vast pool of knowledge that is available at no cost from the source are an example. So too are organizations devoted to the identification and fostering of latent entrepreneurial talent. Other organizations offer specific services of general use across the industrial sector. Productivity centres, for example, have often focused on disseminating modern methods of quality control or assurance. Still other organizations are closely identified with the industries they serve, for they require specialized human and physical assets to deliver services specific to the needs of these industries. Among these organizations there is further differentiation by degree of involvement in solving technical problems specific to individual firms. There is also differentiation according to the breadth or depth of the organizations capacities. Typically, the more highly specialized is the organization, the deeper are its capabilities and the closer are its competencies to the global technological frontier. Research and development (R&D) institutes actively involved in the initial transfers of foreign technology to domestic industries have the most sophisticated capabilities. Many developing countries have set up such support institutions copied from developed countries. Unfortunately, a large number of such institutions do not function well (box 7.2). They tend to be of poor quality, with inadequate equipment and poorly motivated and remunerated staff. Their services are often out of touch with the needs of the industrial sector and are offered passively. Their objectiveslike creating new or appropriate technologiescan be unrealistic. Frequently, their incentive and management systems are not geared to providing services to firms.

Box 7.2 Reforming poorly performing organizations It is possible to reform long-irrelevant public sector technology institutions, and many developed and newly industrializing countries are doing just that. In the late 1980s the World Bank launched an Industrial Technology Development Project in India. One important component was promoting industry-sponsored research at a number of public research institutes as well as at the Indian Institutes of Technology, other universities and private research foundations. This component, the Sponsored R&D (SPREAD) Fund, was aimed at promoting research awareness especially among small and medium-size companies and changing the culture of research laboratories and higher education establishments towards an emphasis on serving the needs of industry as articulated by the firms themselves. The fund is administered by a newly established technology cell in the Industrial Credit and Investment Corporation of India (ICICI), a private sector development bank started by the World Bank. This technology cell helped firms identify the appropriate research institute, develop their business plans, coordinate with the institute and generally hold the hands of new entrepreneurs (as a venture capitalist would). The funds were given as conditional loans rather than grants, and enterprises had to provide matching funds from their own resources. By the end of 1997 around 100 firms had contracted 95 projects under this programme, with an average project size of $400,000 and an average loan component of $170,000. So far, there have been no failures, though three or four projects were likely to be cancelled. Most of the companies using the programme had never contracted research to a public research institute before; most were small and medium size. Some 50 technology institutes were involved, including 5 institutes of technology or science, 12 universities, 5 private research foundations and 28 government laboratories. Overall, the project has been highly successful in technological terms; the subsidy element has been minimal and most firms claim that they will continue their links with the research institutions in the future. A number of elements account for the success of the project. A private sectororiented matchmaking intermediary (ICICI, a well-established private financial institution with intimate knowledge of industry) administered the funds and overcame information and trust barriers between researchers and business. A technically oriented unit in this intermediary assessed the viability of applications and remained involved as the projects developed (more like venture capitalists than bankers). The finance was given in the form of loans rather than grants, with a substantial matching contribution by entrepreneurs. There was a significant effort to help technology institutions understand the needs of industry and change their operating culture.
Source: World Bank.

Investment promotion agencies

Support for technological development does not come only from organizations that directly assist technological efforts by firms. It comes as well from organizations that support industrial development more generally. Investment promotion agencies play a particularly important role. Investment is, after all, the vehicle through which technology embedded in phys-

ical capital is acquired. But attracting foreign direct investment (FDI) is a complex matter, and this is so not simply because it may serve as an important conduit of disembodied technology into local firms. Properly conducted, the promotion of foreign investment can serve as a vital tool for enriching the indigenous base of technological capabilities. Attracting the right kind of foreign direct investment for the development of a thriving industrial sector requires a good deal more than simply establishing the proper general policy environment, one that is business friendly. It also requires dedicated effort in a wide range of activities ranging from the

Supporting innovation and learning by firms

119

identification of suitable inward-investment prospects (or foreign partners) to the active servicing of the strategic needs of foreign-invested firms once established in the country, including development of skills, recruitment services and identification and upgrading of local suppliers. This is in part because of the vigorous competition among countries to attract foreign direct investment, but it is equally because investment promotion can be an instrument of overall industrial development by exploiting the potential complementarities between local and foreign undertakings. Prospective investors can seize opportunities, yielding mutual advantage to them and the host country only if they are aware of them. But knowledge of foreign investment opportunities is necessarily highly imperfect, subject to severe misperceptions and lack of essential information. In short, the market for foreign direct investment does not function effectively unless the countries and regions that seek its benefits devote sufficient resources to publicizing business opportunities in terms meaningful to prospective investors who have a sophisticated understanding of their own needs. Investment opportunities that will dynamically change and enrich the local industrial base must be distinguished from those that simply take advantage of the existing base. The first warrant special priority in promotional efforts. Such opportunities do not simply exist. They are created by first identifying the elements requiredspecialized skills, specific infrastructural requirements, particular university resources, and so onand then by seeing to the coordinated realization of each element. Coordination is crucial. It requires that the promotion agency have the authority to ensure meaningful cooperation among all entities whose activities must be coordinated to achieve a successful outcome. The agency must be subject to proper oversight at the highest levels of government, achieved in part through audits of the agencys own disciplined and continual monitoring and evaluation of its activities. The advanced countriesIreland and Singapore, also the United States (at the state level), for examplepractise a highly sophisticated form of investment promotion designed to achieve strategic industrial development objectives. Extensive study of successful agencies in both advanced and developing countries shows that an effective program of investment promotion entails many activities (box 7.3).3 Success is the result of following a coherently integrated approach that responds effectively to the industrial sectors evolutionary development. The Malaysian Investment Development Agency (MIDA) is an example of a successful investment promotion agency.4 Established in 1967 it serves as the lead agency for orchestrat-

ing the countrys industrial development and has responsibilities for meeting the needs of local and foreign investors in manufacturing. It assesses and advises on industrial and trade policies affecting the sector, formulates detailed plans for the sectors continued development, oversees industrial site development and handles applications for investment incentives. To attract foreign direct investment, MIDA maintains an extensive network of overseas and local regional offices and conducts investment missions to countries from which it sources carefully targeted inward investment. MIDA consults widely and regularly with stakeholders on the design and implementation of its various activities. Malaysias success in attracting foreign direct investment is manifest. More than 3,000 projects involving firms from 40 countries have been implemented since the mid-1960s. The manufacturing sector has grown rapidly and now accounts for 30 percent of GNP and 85 percent of total exports. MIDA has contributed to this record since its inception. Early it seized the opportunity to attract electronics production, which became the basis of Malaysias flourishing high-tech export industries. With its regionally focused sister institution, the Penang Development Corporation, MIDA fostered the evolution of a thriving high-tech cluster in Penang. Over time MIDA has also worked aggressively to create industrial clusters sparked by inward investment in other regions of Malaysia. It has stimulated the development of local firms capable of supplying a variety of high-quality inputs to its foreign-invested firms, establishing a base of industrial competencies that enables foreign-invested firms to operate at ever higher levels of productivity as they move to higher value-added activities.

Industrial parks and export processing zones

The establishment of industrial parks, especially export processing zones (EPZs), is a crucial initiative. Industrial parks are developed and subdivided into plots according to a comprehensive plan with provision for roads, transport and public utilities for the use of a group of industrialists. Parks may go beyond physical infrastructure to satisfy the corporate and technological needs of tenants, providing a variety of common facilities and support services, such as consulting, financial services, training, technical guidance, information services, joint research facilities and business support services (hotel and conference rooms). Parks may also provide a pleasant environment to attract skilled employees and foreign investors. This entails providing facilities such as housing, medical services, shopping and educational establishments. While economic or political reasons

120

Industrial Development Report 2002/2003

Box 7.3 Activities involved in successful investment promotion 1. Laying the proper foundations
Establishing the economic development policy context

Targeted contacts with individual investors

Identifying the most attractive potential investors Establishing and maintaining relationships with these investors

Implementing macroeconomic, trade and industrial policies Providing for the establishment of supportive infrastructure Articulating regional development priorities Determining the nature and scope of special incentives to attract foreign direct investment Determining the objectives for industrial development and inward investment

3. Meeting the needs of interested investors

Ensuring proper reception and facilitating detailed assessments by

Agreeing on the rationale for attracting inward investmentits

developmental role Identifying key sectors or industrial clusters where foreign direct investment is to be sought Discriminating among new ventures, expansion of existing operations Setting priorities among formssubsidiaries, joint ventures, mergers and acquisitions
Ensuring an adequate structure of investment promotion

prospective investors Assigning a single staff member to provide continuity and coordination among local entities Developing a thorough understanding of the potential investors needs Providing information on how the investors specific needs can be met Facilitating site visits and meetings with relevant local parties Specifying the incentives to be made available, including supporting activities Enabling all necessary arrangements and clearances to be made on a one-stop basis
Facilitating implementation of inward investment

Deciding on a single national agency or regionally focused agencies Ensuring the autonomy required for independent operation responsive to business interests Determining mechanisms of coordination with responsible government departments

At a minimum, ensuring that the investors reasonable expectations are fully met Trouble-shooting as necessary to remove obstacles

Providing post-investment service to ensure that potential benefits

Assessing competitive positioning

Knowing potential locations pluses and minuses in attracting priority investments Developing tailored selling points for specific possibilities within key sectors

are realized Supporting follow-on investments in expansions and upgrading Encouraging development and growth of local suppliers Embedding the foreign-invested firm within cooperative networks of supporting institutions Encouraging firm efforts to attract other foreign investments 4. Moving to the next stage of industrial development

2. Mounting a well targeted promotional campaign among selected foreign investors

Setting the stage for contacts with individual foreign prospects

Building a strategy based on past promotional activity Implementing the strategy through strategic activities

Establishing an image that is both accurate and appropriately inviting Maintaining and managing a permanent presence overseas where warranted Participating in conferences and events at which foreign investors congregate Undertaking specifically designed missions to make contact with prospective investors

Continuing to upgrade infrastructure and establish industrial sites Tending to the broadening and deepening of related local value chains Supporting related cluster developments to increase valueadded levels Ensuring continued innovation and learning within existing firms and networks Working with firms on skills development to ensure continued industrial progression

Source: Based on and adapted from Loewendahl (2001).

may mean that nationwide provision of such infrastructure is impossible, the creation of these conditions is feasible within the confines of a park. By providing adequate physical infrastructure and a legal and institutional framework, industrial parks reduce costs and risks of all kinds. They pool resources to make and market goods and meet large orders. And they breed off-shoot companies and provide fertile ground for cross-fertilization of ideas. EPZs are useful for countries working to establish an exportoriented manufacturing sector while lacking the technical or

administrative capacity to develop a countrywide system to allow exporters duty-free access to imported equipment and materials. Science and technology parks are intended for technologically advanced industries and emphasize the highlevel support services such activities need: marketing, technical consultancy through networking with local R&D institutions, advisory services on finance and venture capital and search for joint venture partners. The types of facilities, services and amenities that a park provides depend on the types of industries targeted and the failures the parks are intended to overcome. These vary with a

Supporting innovation and learning by firms

121

countrys level of development. Taiwan Province of Chinas early EPZs, for example, provided basic infrastructure and, especially important, freedom from red tape. These incentives were targeted at light industries, such as textiles and apparel, plastic products and electrical appliances, that could use the countrys plentiful and cheap labour. As countrywide infrastructure improved and administrative procedures with it, parks providing these features became redundant and unnecessary. Taiwan Province of China established its first EPZ in the southern port city of Kaohsiung in 1965 as part of an outward-looking export-oriented industrialization strategy.5 Two other zones were established in Nantze and Taichung in 1969 when applications from investors to set up in the Kaohsiung EPZ flooded in in excess of the space available. The purpose of these EPZs was to increase exports by attracting foreign investment. This was to be accomplished by combining in one place the advantages of a free trade zone, an industrial estate, and all the relevant administrative offices of the government. Firms in the EPZ were offered complete exemption from customs duties, commodity and sales taxes, as well as other incentives. In addition to duty- and tax-free imported inputs, the EPZ provided good infrastructure facilities and simplified procedures for trade and remittances. This simplification and one-stop access to incentives were particularly valuable. Whereas firms outside of the EPZ had to obtain duty- and tax-free imported inputs by means of a rebate system, location within the EPZ allowed firms to avoid all the formalities connected with obtaining these rebates. EPZs thus provided the important benefit of the cutting red tape so that investors could start their projects quickly and could run them with minimum bureaucratic fuss. The main constraint facing firms within the EPZ was that they were required to export all of their production, keeping it out of the domestic market. EPZs were very important in placing Taiwan Province of China squarely on the path of export-led industrialization. Arrivals of foreign direct investment roughly doubled within the first year after the Kaohsiung zones establishment. Total arrivals averaged $12 million a year from 1961 to 1965 and $44 million a year over the five years after 1966.6 But as infrastructure facilities improved rapidly and regulatory procedures were rationalized the importance of EPZs diminished over time, accounting for only 79 percent of the countrys cumulative exports since the 1960s. Bonded factories, which are like mini-EPZs located outside the formal zones, have been responsible over time for a much larger share. Since the 1980s, little new investment has occurred in the EPZs, reflecting their redundancy as infrastructure and duty-free procedures have improved outside of them. The administrative costs of the tax rebate system have been substantially

reduced, particularly as the result of the establishment bonded factories and warehouses, lowering the value of circumventing this red tape. Science and technology parks are at the other end of the spectrum from most EPZs in level of services. An example is Chinas Suzhou Technology Park.7 Suzhou Park is made up of three institutions: Suzhou New and High-Technology Service Centre, Suzhou International Business Incubator and China Suzhou Pioneering Park for Overseas Chinese Scholars. The first incubator was set up in 1994, and the China Suzhou Pioneering Park for Overseas Chinese Scholars was created in 1998. The park now houses 300 enterprises. Ninety percent of these firms were set up by overseas Chinese and 10 percent by R&D institutes and universities. Twenty percent are high technology enterprises. In 2000 the park employed 3,000 people, 100 with Ph.D.s. Suzhou Parks success is linked to the services it provides. The government provided seed money and attracted venture capital from abroad; additionally, banks and financial organizations provide flexible loans to firms. The park includes an incubation site of 38,000 square meters. It provides Internet connections every 10 square meters, conference rooms, a multi-media room, a technical trading room, information centres, product testing centres and public laboratories. The park includes an accounting office, law firm, business planning space and other services. Import-export services, such as customs declaration and a bonded warehouse, are provided free. Human resources support in the form of recruitment events and a database help firms identify people with the right skills. Additional services are provided in the form of management and business training by university professors and successful entrepreneurs, assistance in introducing new products and membership in the Shanghai Technology Stock Exchange.

Information services
Providing information services requires information specialists who are also technically savvy. These services are the least dependent on prior targeting and the like. Serving as an intelligent gateway to globally available, searchable knowledge basesintelligent by virtue of their trained staffinformation services offer a truly generic service, of equal potential use to all comers (box 7.4). As such, they are the closest among service organizations to providing a public good having universal value. But many information centres also routinely produce material to disseminate the results of the continuing search. An example is Sri Lankas Industrial Technology and Market Information Network (ITMIN), a commercially managed and

122

Industrial Development Report 2002/2003

Box 7.4 Available on the Internet Information services online can now provide instant, free or for-fee access to technical information and International Organization for Standardization (ISO) compliance requirements. As an experiment, a student with little knowledge of technology (neither an engineer nor a scientist) spent three hours looking for specific technical information on the Internet and in that brief time was able to find a great deal of technical information. The hunt began at the National Institute for Standards and Technology (NIST) in the United States. While the bulk of highly technical, specific material is available only through purchase, the volume of information on the site and relevant links gave a sense of what information is available and what information a manufacturer would need to stay compliant and competitive. NIST publishes a Ceramics WebBook that offers free databases on most materials. For example, a manufacturer that uses alumina could find measurements on porosity, density and flexural strength for varied sizes of grains and different pressures, information that would help the manufacturer meet compliance requirements and learn how the product would hold up under various stresses. The following additional links were found: American National Standards Institute (ANSI). ANSI maintains a list of international and regional standards institutions with links to their sites, as well as a list of domestic standards. International Electrotechnical Commission (IEC). IEC provides information on electrotechnical materials and their categorization. It offers an online database of the new, standardized electrotechnical vocabulary so manufacturers can understand terms used in ISO standards and begin using a common vocabulary. British Standards Online. This site offers bibliographic information about standards and sells BSI publications. For a fee a business or individual could subscribe to the Web site and have access to substantial information online. Singapore Productivity and Standards Board (PSB). The PSBs Web site houses a full eShop where pamphlets, information on standards, and documentation can all be purchased. Sparks and plugs were searched and eShop found pertinent documents available for S$20 and S$54. The site also links to many other Singapore programs that assist businesses, including the new A*STAR, Agency for Science, Technology, and Research (previously National Science and Technology Board). While the bulk of technical information is available only for a fee, these sites all contained relevant information for each country on compliance, regulation, necessary paperwork and the like. For a manufacturer needing information on a particular product or process, online information might be the fastest, most efficient and most reliable source. Similarly, government agencies and policy analysts looking for useful knowledge of how things are done by other governments could find a wealth of information from the same kind of search.
Sources: http://www.ceramics.nist.gov/webbook/webbook.htm; http://www.ansi.org/ public/library/internet/intl_reg.html; http://domino.iec.ch/iev/iev.nsf/Welcome?OpenForm; http://www.bsi-global.com/index.html; http://www.psb.gov.sg/index.html.

it provides information services to a broad range of businesses that previously lacked access to global information networks. The information groups staff receive extensive training in processing and marketing industrial and technical manufacturing information. ITMIN provides current information on technology transfer, business intelligence, electronic publishing, market surveys and investment opportunities. It researches technologies relevant to Sri Lankan firms and publishes a monthly list of adoptable technologies for Sri Lankan firms. These information services provide Sri Lankan companies with tools for decisionmaking on available technology, production, investment and export markets. ITMIN offers Internet hosting, training and customized services ranging from Web page development to networked intranets). ITMIN offers general classes in computer applications and designs customized private training programs for firms. There are special programs in information technology for managers, accountants and secretaries.

Standards and metrology

The globalization of value chainswith a multitude of firms acting as interconnected suppliers, intermediaries and marketershas been sustained by the parallel drive towards the standardization of practices and procedures. Firm interactions along the value chain require conformity with agreed standard business practices in contracting, accounting, project management, environment management and the communication of product design and engineering information. Standards would be meaningless in the absence of the ability to make precise measurements of the various attributes chemical, electrical, physical, and so onof the produced outcomes at each stage along the value chain, using common modes of measurement across international boundaries, with assurance that measured magnitudes are precisely correct within agreed error tolerances. Metrology is thus the essential foundation upon which standards rest. This foundation is maintained through a carefully linked hierarchy of metrological agenciessome autonomous and responsible only for metrology, others embedded with organizations having linked responsibilitiesat the international, regional, national, and intranational levels. At the apex is the Bureau International des Poids et Mesures (BIPM) in Paris, France, which has been assigned by international convention the task of assuring an evolving basis for a single, coherent system of

oriented public company funded by public and private shareholders and originally developed by UNIDO under a United Nations Development Programmefinanced project.8 Operating in three areasinformation, Internet service and training

Supporting innovation and learning by firms

123

measurements thought out the world, one traceably linked to the International Standard of Units which is continually augmented as technologies are developed over time. The BIPM is responsible for certifying and calibrating the units of measure that are maintained by national agencies, which are in turn ultimately responsible for certifying and calibrating the units of measure and instruments of measurement that are employed in producing firms, research laboratories, universities, and the like. In turn, just as the BIPM must ensure the establishment of new standard units to keep abreast of evolving technologies, so too must each national agency ensure that its metrological assets and competencies develop sufficiently in advance of the evolving needs of the industrial base that it serves. At the same time it must provide training and consultancy services sufficient to develop the requisite measurement capabilities of the various entities functioning within its client base. The International Organization for Standardization (ISO), a worldwide federation of national standards bodies, introduced the best-known standards being used today, the ISO 9000 series.9 While the vast majority of ISO standards are specific to a particular product, material or process, ISO 9000 refers to a generic management system standard, specifically one centred on quality control. Quality management is what the organization does to enhance customer satisfaction by continually striving to meet customer and applicable regulatory requirements in the most cost-effective manner possible. ISO 9000 certification is quickly becoming an imperative for potential exporters, signaling quality and reliability to foreign buyers, retailers and transnational corporations seeking local partners and sub-contractors. A high-ranking priority for a developing country is therefore to have a local institution entitled to grant ISO 9000 certification to domestic firms. A standards institution can disseminate best practice in an industry by encouraging and helping firms to understand and apply new standards. Such institutions not only manage the process of certification but also provide consultancy services for firms preparing to meet those standards (box 7.5). India is an example of a country that has a formal certification program for ISO 9000 compliance.10 The Bureau of Indian Standards (BIS), authorized by the government to certify compliance with ISO standards, has the following process for licensing a firm for ISO 9000 standards. Firms are instructed to: Establish a documented quality system and ensure its effectiveness. Submit an application on prescribed forms along with a completed questionnaire and necessary fees.

Box 7.5 Programmes to help domestic firms achieve standards Singapores member body in ISO is the Singapore Productivity and Standards Board (PSB - http://www.psb.gov.sg/). PSB has several programs which are designed to ensure that domestic firms are producing quality products. In order to help firms achieve ISO standards, PSB Certification was created in 2001as a separate corporatized entity born out of PSB. Its purpose is to assess customers management systems based on national, international, or industry standards, such as ISO 9000. PSB Certifications web site (http://www.psbcert.com/home. htm) provides details for firms as to the application process for becoming certified, even allowing online completion of the initial application form. PSB Certification provides training events such as workshops focusing on providing participants with the knowledge and skills to implement ISO 9000. As PSB Certification is a private sector company, one of its goals is to achieve $12 million revenue in three years time. As a public entity PSB itself has several programs to assure quality. For example, the People Developer Standard is a three-year certification scheme that attempts to assure that firms maintain high standards in human resource development. Launched in 1997, the People Developer Standard provides organizations with a systematic process to review their human resources practices, develop staff and improve training effectiveness. This initiative provides a good example of the ways that a country can bring firms up to a desired standard. A variety of assistance programs are offered to this end. Training programs designed to impart the basic knowledge on how to set up and implement the People Developer systems are provided by PSB, private companies, and educational institutions; PSB has a Skills Development Fund (SDF) which supports 90 percent of course fees. It also has an Enterprise Development Fund (EDF) which assists small and medium-size enterprises to hire external consultants to help set up appropriate systems; the EDP supports up to 70 percent of the consultancy fee. The web site provides a list of consulting firms. Additionally, two-hour, one-on-one discussions with National Assessors are provided free-of-charge in order to clarify the standards requirements, ensure the correct implementation of the systems, and ascertain readiness to apply for an audit. Half-day, free-of-charge workshops covering application and audit procedures, common pitfalls and critical success factors, and the writing of assessments reports are also held. Similarly, half-day, free-of-charge sharing sessions are held quarterly in order that firms desiring to adopt the People Developer Standard can learn from the experiences of newly certified organizations. Detailed instructions on how to start the process of certification are provided online, as well as information on how to apply (including downloadable application forms) and how to maintain the standard once certified.
Source: http://www.psb.gov.sg.

Submit the quality control manual and related documents, when requested. Arrange for an audit by a BIS assessment team. Take corrective actions on non-conformities observed by assessment team and get them verified. Obtain the license, which will enable the company to compete effectively in national and international markets. After registration of the application BIS examines the firms quality control manual and quality plan to verify conformity to the

124

Industrial Development Report 2002/2003

relevant standard. After that, BIS arranges a preliminary visit to learn the size, nature of operation and the firms readiness for the initial audit. An audit team from BIS then visits the firm for the initial audit of compliance with the procedures and activities enumerated in the documents provided and with the relevant ISO standards. If the audit report is satisfactory, the firm is granted a license by BIS for a period of three years. Grant of a license is followed by surveillance audits every six months by BIS auditors to verify implementation and maintenance of the quality system established by the firm. The phase of export promotion that began in Taiwan Province of China in the 1960s brought with it a concern for quality control of manufactured goods.11 The reason for this can be understood by looking at the example of bicycles. Although Taiwan Province of China exported virtually no recreational bicycles in the 1970s, as bicycles become popular in the United States, many wholesalers placed orders in Taiwan Province of China solely on the basis of price. Thus Taiwan Province of China began producing bicycles for export. Many of these early bicycles had quality problems and defects, and Taiwan Province of Chinas chance to enter the world market was at risk. The government responded by commissioning the Metal Industries Development Centre to design testing equipment for bicycle quality control. Combined with technical assistance, this action was instrumental in improving the quality of bicycles. By the mid-1970s, Taiwan Province of China was exporting several million bicycles, and by the early 1980s, it accounted for about half of the world supply. This concern for quality was present as far back as 1953, when the Taiwan Government first began inspecting the quality of goods that its firms were exporting. Canned foods are another early example of why this was important: the poor quality of some brands was adversely affecting the international reputation of others. Inspecting each good that left the country, or even a sampling, clearly posed a practical difficulty; the magnitude of exports was too great. The system was redesigned in the mid-1970s to focus on a firms quality control procedures rather than on individual goods. For example, instead of examining each bicycle shipment for defective bikes, the government now examined the way that bicycleproducing firms ensure that only high quality bicycles are exported. Factories were classified into three categories based on inspection of their quality control procedures. Those scoring below the minimum were not given a license to export. Those scoring the highest were allowed to export without inspection of their merchandise; only their quality control system was reinspected yearly. In addition to having their systems inspected twice a year, those scoring in the middle had a one-in-thirty chance of having individual shipments inspected. The quality control systems of those in the third

category were inspected three to four times a year and individual shipments faced a one-in-fifteen chance of being inspected. The inspections were carried out by the Controls Bureau of Standards, part of the Ministry of Economic Affairs, in cooperation with the Bureau of Commodity Inspection and Quarantine (these bureaus are now integrated into the Bureau of Standards, Metrology and Inspection which has an English language Web site12).

Productivity centres
Productivity centres are broadly focused, geared more to industrial development than to technological development alone. They work with firms to promote efficiency and productivity in manufacturing, often changing their focus as the problems needing attention change over the course of development. Productivity centers operate on a national and regional level. At the national level they are generally funded initially by government and promote nationwide awareness of the need for productivity enhancement. Most of these campaigns focus on the positive relationship between employment and productivity growth to combat the fear that increased productivity displaces labourers. The Japan Productivity Centre, founded in 1955 with funds from the United States as an organization of labour, management and academia, merged in 1994 with the SocioEconomic Congress of Japan to form the Japan Productivity Centre for Socio-Economic Development (JPC-SED).13 The original guiding principles emphasized that: Productivity gains increase employment. Labor and management must work together. Gains from productivity should be shared by labour, management and the public. In the 1960s and 1970s the primary focus was on improving the relationship between labour and management through consulting services and seminars. JPC borrowed many of its productivity ideas from Europe and the United States, bringing in foreign experts for seminars and reading foreign publications. After aid from the United States ended in 1962, businesses covered most of the expense for the JPCSED, with some government help. Programs increased to include graduate courses, information technology programs, robotics and automation training, database assistance, increased managerial training and mental health research on employee reaction to various workplace environments.

Supporting innovation and learning by firms

125

The JPCSED created many spin-off organizations, many focusing on technology-specific improvements, such as the Japan Industrial Engineering Association. Japan soon became a productivity model, creating and hosting a series of international conferences, study tour programmes and workshops for Europeans and Americans. Today the JPC-SED has expanded its productivity models to public agencies, environmental concerns and the social welfare system. Productivity centres can thus provide vital information and services to private firms. Institutionally, they appear to be quite adept at spawning a network of related support organizations. Together, they are changing as the industrial sector evolves and its needs change. The Asian Tigers invested heavily in setting up and developing such institutions. Even laissez faire Hong Kong Special Administrative Region (SAR) of China provides subsidized technology support to its exporting firms, most of them small or medium size (box 7.6). The national productivity centre provides a general agenda to regional productivity centres. These regional centres provide productivity assistance to firms and may have more local, satellite locations. Productivity centres directly consult and advise firms on management strategies, efficient floor layouts, labour-management relations and workplace environment and environmental concerns, among others. Programs often include a training component, which ranges from single-day seminars to longer courses in business administration, to financed trips and site visits, to appropriate foreign examples of successful factories and plants. In more developed countries productivity centres shift from providing direct advice to helping firms network and find private market consultants and programs for solutions.

Box 7.6 Technology support from the Hong Kong Productivity Council The Hong Kong Productivity Council (HKPC) was started in 1967 to help the myriad small firms that constitute the bulk of the industrial sector. Its focus has been helping firms upgrade from declining labour-intensive manufacturing to more advanced, high value-added activities. It provides information on international standards and quality and provides training, consultancy and demonstration services on productivity and quality to small firms at subsidized rates, serving over 4,000 firms each year. Its on-line information retrieval system has access to over 600 international databases on a comprehensive range of disciplines. Its technical library subscribes to more than 700 journals and has more than 16,000 reference books. The HKPC acts as a major technology import, diffusion and development agent for all the main industrial sectors. It identifies relevant new technologies in the international market, builds up its own expertise in those technologies and introduces them to local firms. Successful examples of this approach include surface-mount technology and three-dimensional laser stereo-lithography. HKPC has also developed a number of computer-assisted design, manufacturing and engineering systems for the plastics and moulds industry, of which over 300 have been installed. HKPC provides a range of management and technology courses, reaching some 15,000 participants a year. It also organizes in-house training programmes tailored to individual needs. To disseminate information technology, HKPC has formed strategic alliances with major computer vendors and provides specially designed software for local industry, consultancy and project management in computerization. HKPC provides consultancy services in ISO 9000 systems and has helped several firms in Hong Kong obtain certification. It assists local firms in automation by designing and developing special-purpose equipment and advanced machines to improve process efficiency. Because small firms have difficulty getting information on and adopting new technologies the HKPC has always had to subsidize the cost of its services. Despite the growth in the share of revenue-earning work, the government still contributes about half its budget. Market failures affecting access to technical information occur even in a highly sophisticated export-oriented economy with highly developed financial services like Hong Kong SAR.
Source: Lall (1996).

Extension services for small and medium-size enterprises

Technology extension programs, sponsored and coordinated through governments, create networks that help small and medium-size enterprises stay viable in a competitive economy through the use of technology to increase productivity. Technology extension is concerned with creating small but profitable improvements by extending already established technology to smaller firms. While the design of technology extension organizations differs, all of them have relationships with small and medium-size enterprises and with sources of technology. Technology extension programs either provide resources that enable firms to identify needs and find appropriate technological solutions or engage in the actual identification and provision of solutions by means of targeted technical assistance.14

An example of an extension program that performs both functions in relation to a critical, narrowly focused technological objective is the national cleaner production centre, which assists firms in achieving best-practice standards in the prevention of pollution (box 7.7). These centres are like productivity centres in that they are intended to serve the full range of a countrys industries for a particular objective, but they are otherwise like, and closest to, extension agencies. In the United States many extension services have been provided locally by university systems. But with the inception of the Manufacturing Extension Partnership (MEP) in 1988 and its coverage of all 50 states by 1997, field personnel from 400 MEP offices connect small and medium-size firms to a large network of both public and private industrial service providers, including university extension services. MEP per-

126

Industrial Development Report 2002/2003

Box 7.7 National cleaner production centres The joint UNIDO/United Nations Environment Programme (UNEP) sponsored national cleaner production centres aim to enable developing countries to promote, gain access to and use cleaner production technologies, including technology transfer from abroad. The NCPCs build capacity within a country for the adoption of cleaner production to take place, promoted and chosen by professionals in that country and adjusted to local conditions. These centres can be considered part productivity centre (environmental control is not unlike quality control in have cross-cutting, generic elements) and part extension service (focused on the specific needs of specific enterprises). As theyve so been operating so far, they appear closer to the productivity centre side. In January 1995 UNIDO began NCPC programmes in Brazil, China, the Czech Republic, India, Mexico, Slovakia, the United Republic of Tanzania and Zimbabwe. As of September 2001, 23 centres have been funded. The total budget allocated since 1991 amounts to approximately $21 million, with 85 percent of the budget coming from donor countries, 10 percent from multilateral organizations and the remaining portion from self-financing by countries receiving the centres. UNIDO acts as the executing agency with the UNEPs Industry and Environment Programme Activity Centre providing methodology and information. It takes approximately three to five years to fully establish each centre. An NCPC within a particular country should eventually be able to offer the following six services:
Short seminars to raise awareness of cleaner production and dis Advice on financing and funding sources for cleaner production. Policy advice.

All the centres are engaged in short-term training, dissemination of information on cleaner production and in-plant assessments. Six of the original NCPCs have conducted in-plant assessments at 316 enterprises per year, with most falling in the 79 range. Some of these assessments were actually several separate in-plant assessments, undertaken by different teams for different production lines. Some NCPCs, such as Indias, assist only small and medium-size enterprises, some work with all sizes of enterprises, while others, like those in Brazil and China, deal primarily with large enterprises. Enterprise staff and local, NCPC-contracted consultants jointly identify problems within an enterprise and choose appropriate technological changes. In 1997 and 1998 half the 439 identified cleaner production options were improvements in housekeeping, changes that are easily identifiable and have a high benefit/cost ratio. Overall, only 64 percent of the identified cleaner production options were implemented. Implementation rates vary, with housekeeping measures having the highest implementation rate (76 percent) and changes in process technology having the lowest rate (37 percent). The need for investment and the time required for major technological change help to explain this discrepancy. This tendency to choose low-investment options becomes reinforced when in-plant assessment staff know of financing constraints and thus recommend low or noninvestment options. Only about 3 percent of the cleaner production options included a transfer of technology, which were all purchases of capital goods rather than purchases of licenses: new equipment, hardware modifications and replacement equipment. However, the cleaner production methodology itself can be viewed as technology transfer as can the information and know-how transferred from foreign experts, consultants, and counterpart (twinned) institutions. The NCPCs did, however, seem to improve local transfer of technology within countries. In Slovakia, enterprises acquired and adapted technology from a local university, and in Brazil, China and Mexico, companies purchased environmentally sound technologies from local producers.

seminate general information.

Training in cleaner production methodology. Dissemination of technology information needed for in-plant

assessments.
Technical

assistance, including in-plant cleaner production assessments.

Source: UNIDO.

sonnel identify a firms needs and problems and assist it in finding appropriate solutions. While MEP was intended to bring cutting-edge technology to small firms, in practice it focuses on bringing more realistic help on existing technologies and management. Funding for this partnership is provided by state, federal and private funds; firms receiving assistance also pay a portion of the cost. Their fees at most seem to cover 40 percent of MEP operating costs; thus federal and state funding seems necessary to continue providing this consulting service. Many manufacturers who have benefited from MEP programs report increased profits; surveys also suggest that extension services increase employment and generate business growth. Comparative studies have shown that receiving extension services from MEP offices increases the rate of growth and adoption of technology over that of firms not receiving assistance. In Japan 170 Kohsetsushi centres provide technological support for businesses that have fewer than 300 employees.

Unlike extension services in the United States, they provide only technological servicesmanagement and financial services are left to other agencies. Charging only nominal fees to their clients, the centres were created and sponsored by the central government, but maintain relationships with local and prefectural governments. They conduct research, have open laboratories for training, test and examine products for compliance, provide advice and guidance and promote technology diffusion and information dissemination. Because of the longterm relationships between large manufacturers and the smaller manufacturers of their inputs, the centres meet the demands of both-sized firms by focusing on the testing and analysis of materials and productspromoting quality, performance and precision while ensuring standards among the input suppliers. Managers of small firms appreciate and depend on the personalized services that the centres provide and prefer dealing with them than with universities. Traditionally, these small firms simply produced intermediate inputs with new product design

Supporting innovation and learning by firms

127

and engineering coming down from the larger firms. These same small firms, guided by the Kohsetsushi centres, are now designing new products and spurring technological growth among themselvesinnovating themselves rather than simply following larger firms leads. Without the government-funded, highly localized centres, most of these small firms would not have access to technological advice or capability, to their detriment and that of the larger companies that use their products. Many developing countries have recognized the need for extension services to help small and medium-size enterprises produce better products with better technology.15 Chiles Technical Cooperation Service (SERCOTEC) has provided technical support since the 1950s. One successful project provided technical assistance to honey manufacturers in Litueche, replacing traditional hives with modern ones and adding more hives, increasing the overall yield from 600 kilograms a year to 18,000. In Pomaire, Chile, SERCOTEC helped local pottery craftsmen implement the technology of gas ovens that would mimic the traditional wood-burning ovens yet efficiently produce lead-free pottery. This change increased productivity and led to exports to markets demanding lead-free products. South Africas National Productivity Institute and Industrial Research Organization have formed a partnership to create the National Manufacturing Advisory Centre (NAMAC). NAMAC is to eventually consist of nine manufacturing advisory centres (MACs)regional centres that target and advise small manufacturing firms (under 200 employees) in various industries. MACs strive to increase competitiveness and efficiency by upgrading firms technological capabilities and providing other business support services including financial services, plant layout redesigns and export and marketing information. Individual MACs provide their own services on a for-fee basis and direct firms to appropriate service providers or sources of export intelligence. NAMAC is additionally funded by partnering member organizations and the Danish foreign ministry. Currently, two of the proposed nine centres are fully functional and operating. The Government of Taiwan Province of China also provides an extensive range of support to its myriad small and mediumsize enterprises, allowing them to compete in extremely skilland technology-intensive industries without being able to invest large amounts in in-house R&D. Taiwan Province of Chinas technology infrastructure for supporting its many small and medium-size enterprises is perhaps one of the best anywhere. There are around 700,000 small and medium-size enterprises in Taiwan Province of China, accounting for 70 percent of employment, 55 percent of gross national product (GNP) and 62 percent of total manufactured exports. In 1981 the government set up the Medium and Small

Business Administration to support small and medium-size enterprise development and coordinate the several agencies that provided financial, management, accounting, technological and marketing assistance to small and medium-size enterprises. The government covered up to 5070 percent of consultation fees for management and technical consultancy services for small and medium-size enterprises. The CentreSatellite Factory Promotion Programme of the Ministry of Economic Affairs integrated smaller factories around a principal one, supported by vendor assistance and productivity raising efforts. By 1989 there were 60 networks with 1,186 satellite factories in operation, mainly in the electronics industry. The Taiwan Province of China Handicraft Promotion Centre supports handicraft industries, particularly those with export potential. Its main clients have been small entrepreneurs, most with under 20 employees. In addition, the Programme for the Promotion of Technology Transfer maintains close contact with foreign firms with leading-edge technologies to facilitate the transfer of those technologies to Taiwan Province of China.

Serving clusters of small and medium-size enterprises

Small and medium-size enterprises can often work with support agencies to propel their innovation and learning. The fast spread and absorption of new ideas have been documented for todays developing country clusters.16 Examples range from the autoparts cluster in Kumasi (Ghana), to the clothing cluster of the Western Cape (South Africa), the footwear cluster in Leon (Mexico) and the surgical instrument cluster in Sialkot (Pakistan). Individual small and medium-size enterprises rarely have the resources or connections to tap the global wealth of product and process ideas. So, where can the new ideas come from? Small and medium-size enterprises have three main ways to gain access to knowhow. The first is by pooling resources and acting together. Joint stands at key international trade fairs are an example. Participating in these fairs is not just about sellingit is also about learning through direct contact with potential customers. The second possibility is to rely on a local technology institute, funded by government or foreign donors, for the import of new technology. A successful example is CETMAM, a technology institute in Curitiba (Brazil), which helped the furniture clusters in the state of Paran accelerate their product and process innovation. The third possibility is learning through foreign buyers. This route deserves particular attention because some of the most successful developing

128

Industrial Development Report 2002/2003

country clusters operate in buyer-driven chains (chapter 6). In many cases, the global buyers have enabled local producers to innovate in the sphere of production, particularly to raise quality and speed. Such improvements in products and processes are clearly in the interests of the buyers. Unfortunately, the majority of small and medium-size enterprises clusters in developing countries are underperformers, unable to link, leverage and learn more advanced product and process technologies and to integrate with value chainson a regional, national or global basis. Moreover, they may be locked into a vicious cycle of cut-throat competition, falling profit margins, deplorable working conditions and mounting environmental degradation. Cluster development institutions can drive firms to take certain kinds of collective actions, such as collaborating to acquire certain new competencies, while remaining fierce competitors in other product markets (box 7.8). There are many examples of such institutions in East Asia, where the linkage and leverage strategies have been best developed. In Taiwan Province of China, small firms have been encouraged to ally themselves in R&D consortia, where technological guidance is provided by a public sector laboratory. The key organizations within these R&D-promoting institutions are private firms and the laboratories. But important guiding and catalytic roles are also played by government ministries, particularly the Ministry of Economic Affairs (for funding and audit), and by trade associations, which legitimize the R&D consortia in the eyes of their member firms and recruit new firms for membership of the consortia (box 7.9). Seldom is one form of service sufficientcomplementary services required to deal with intertwined or linked problems and opportunities in distinct areas of firm behavior. The more targeted the package of services on specific types of firms, the more likely their usefulness to firms. The greater the participation of firms in defining and designing the services offered, the more likely are better results. Services may be provided publicly (government) or collectively (industry association). Institutional support for clusters can cost-effectively increase quality and production while supporting a collaborative, innovative environment. Governmental and private sector institutions and organizations can provide necessary training and technical advice, promote cooperation between firms and help link local industries with foreign export markets. Just as most clusters arose spontaneously, so have the private institutions that support themin organizations of firms, for example. In both the Leon and Guadalajara footwear clusters in Mexico, the local Camara del Calzado shoe making entre-

Box 7.8 Cluster development in Jaipur, India Colorful hand-block printing enjoys a long tradition in Jaipur, the capital of Rajasthan, where approximately 550 small and very small firms engage in both hand-block and screen printing and provide employment to almost 10,000 workers. But the ability of the local artisans to penetrate profitable national and world markets was severely constrained. As a result, the artisans were locked in a vicious cycle of cutthroat competition, falling labor standards and mounting degradation of the environment. An action plan (developed by cluster stakeholders with UNIDO support) for the cluster of Jaipur envisaged enhancing the design, production and marketing capacity of local firmsand developing a product image (including a common brand) in line with current market demand. Manufacturers already in international markets joined in an export consortium (COTEXConsortium of Textile Exporters) for the realization of common initiatives. Artisans previously relying on traders for marketing, organized the Calico Printers Co-operative Society (CALICO) to expand their market both domestically and abroad.
An ad hoc training program was launched to strengthen the mar-

keting competence of CALICO members.

A demonstration fair was arranged in Jaipur to allow the artisans

to make full use of their newly acquired skills.

Several training courses were organized (on marketing, design,

merchandising) in collaboration with key Indian technical institutes.

Participation in national (New Delhi) and subsequently interna-

tional fairs (in Florence and in Osaka) was arranged so that artisans could assess their capacity to handle direct meetings with potential buyers on a much larger scale.
A common brand was created to identify traditional Jaipur prod-

ucts that satisfy strict standards of product quality and production techniques. One outcome of these activities is that the traditional bitterness and conflict have disappeared. Local producers have learned to trust each other and to cooperate with their local partners (small and mediumsize enterprise support institutions, providers of business development). Structural problems are now tackled through collective ventures planned and implemented by the cluster actors themselves. Such ventures include initiatives to enhance technological competence (introduce new production technology, improve inventory management, reduce drudgery), curb pollution (waste processing, cleaner production technology) and increase access to credit (mutual credit guarantee schemes). Moreover, several producers associations and self-help groups have been established in the cluster, along the lines of COTEX and CALICO. As the collaboration among cluster actors has increased, these representative bodies have started to collaborate with greater frequency, providing a forum for a sustainable clusterwide governance framework.
Source: http://www.unido.org/33112.htmls.

preneurs association provides numerous support services to drive the success of each cluster. Leon hosts 51 percent of Mexicos shoe-production, most of it mens and childrens footwear. Guadalajara has 22 percent market share, concentrating on womens shoes. Services of the Leon and

Supporting innovation and learning by firms

129

Box 7.9 Leveraging advanced technologies from abroad Unlike Japan and Republic of Korea, both dominated by giant firms, the most significant players in Taiwan Province of China have been small and medium-size enterprises whose entrepreneurial flexibility and adaptability have been the key to their success. Underpinning this success is the effort of public sector research and development institutes, such as Taiwans Industrial Technology Research Institute (ITRI). Since its founding in 1973 ITRI has acted as a prime vehicle for leveraging advanced technologies from abroad and rapidly diffusing or disseminating them to Taiwans enterprises. This cooperation between public and private sectors, to overcome the scale disadvantages of Taiwans small firms, is characteristic of the countrys technological innovation strategies and of new high-tech sectors. Taiwans current dominance of mobile personal computers (PCs), for example, rests at least in part on a public-private consortium that rushed a product to world markets in 1991. Taiwans strong performance in communications products such as data switches, which now dominate in PC networks, similarly rests on a consortium that worked with ITRI to produce a switch to match the Ethernet standard, in 1992/93. When IBM introduced a new PC based on its PowerPC microprocessor, in June 1995, Taiwan firms exhibited a range of computing products based on the same processor just one day later. Again this achievement rested on a carefully nurtured R&D consortium involving both IBM and Motorola, joint developers of the PowerPC, as external parties. These successes were followed by many more R&D alliances in digital communications and multimedia areas. Taiwan Province of China is emerging as a potentially strong player in the automotive industry, particularly in the expanding China market, driven by its development of a 1.2 litre 4-valve engine; again, this is the product of a public-private collaborative research endeavour involving three companies, which have now jointly created a new Taiwan Engine Company to produce the product. The R&D consortium is an inter-organizational form that Taiwan has adapted as a vehicle for catch-up industry creation and technological innovation. Some of these consortia have been more successful than othersbut all seem to have learned organizational lessons from the early cases where government contributed all the funds, and research tasks were formulated in generic and overly ambitious terms for the companies to take advantage of them. The more recent R&D alliances have been more focused, more tightly organized and managed, and have involved participant firms much more directly in co-developing a core technology or new technological standard which the firms can incorporate, through adoption and adaptation, in their own products. The basic model of the alliances a process in which R&D costs can be shared, and risks reduced, by bringing many small firms together to work with ITRI, the main vehicle for leveraging. The goal is the rapid adoption of new technological standards, products or processes developed elsewhere, and their rapid diffusion to as many firms as possible.
Sources: Mathews and Cho (2000); Mathews (forthcoming).

of measurement systems and by forming agrupamentos industriales. These loose groups are composed of many firms that agree to visit each others factories. These visits allow knowledge exchange, promote discussion and trust and increase innovation and collaboration. Member firms are also required to have external experts audit their plants, which leads to greater efficiency as problems are diagnosed and resolved. The camaras are important in helping their members respond to the cheap imports that have flooded Mexicos no longer protected domestic shoe market, enabling members to increase the quality of inputs and become more efficient. The camaras have also provided an effective means of looking to higher end shoe markets in the United States. Because the success of shoe manufacturers often depends on the current shoe fashion and what sells, the camaras pay special attention to fashion trends and help firms adjust their manufacturing to meet the markets needs. A Ministry of Industries program additionally encourages joint marketing, joint brand names and empresas integradoras that form groups to buy inputs and sell output collectively. Local credit unions have developed within the clusters to facilitate better loan access and encourage this collective input purchasing. Research has shown that the local initiatives taken by these various institutions have significantly influenced the sector and resulted in numerous positive externalities. Indias Tiruppur cluster of cotton knitwear manufacturers uses local institutions for assistance with marketing, exportation and design innovation. The governmental Apparel Export Promotion Council administers a quota system that limits what producers can export under bilateral trade agreements, promotes exports and also helps local companies understand bilateral trade agreements. This organization creates market survey teams, actively finds new markets, organizes trade delegations and collects data on knitwear trade. The Tiruppur Exporters Association also explores new markets and collects marketing data. With 248 regular members and 134 associate members the association has also put in place a selffinanced industrial complex for export knitwear producers with production facilities for 157 firms. It is now working to improve local infrastructure by financing sewage treatment and more telephone lines. To complement marketing and export promotion by these institutions, the autonomous South Indian Textiles Research Association (SITRA) researches and tests cotton fibers to create finer-count cotton and develops new spinning and weaving techniques. It plans to build a training centre and research laboratory that would test cloth and dyes for compliance with standards and would research incorporating design technology into Tiruppur knitwear manufacturing. Government

Guadalajara camaras include trade fairs; business support in financial, legal, and managerial advice and training; and such technical assistance as bringing in foreign experts. Membership fees from firms and profits from the trade fairs fund the camaras. The associations also promote cooperation and closer integration among firms by working with them on the standardization

130

Industrial Development Report 2002/2003

grants, revenue from services and assets and membership fees fund SITRA, with member mills paying both admission and recurring fees per spindle, rotor and shuttleloom. Despite labour being cheaper elsewhere in India, this knitwear cluster continues to grow and successfully export, aided by local institutional support. Thus the public and private sectors together have created local institutions to provide support that collectively benefits the cluster. This emergence of private, membership-based institutions signals that firms in the cluster see the potential gains from collaboration and find it cost-effective to join institutions. Similarly, government bodies see cluster support as an investment in the sectors global viability and the countrys economic health.

It later established a computer-aided design and manufacturing centre to provide training and software to firms in this industry. The Precision Instrument Development Centre fabricated instruments and promoted the instrument manufacturing industry, and later moved into advanced areas like vacuum and electro-optics technology. At the apex stands the Industrial Technology Research Institute (ITRI), established in 1973 under the direction of the Ministry of Economic Affairs, which has specialized in the transfer of sophisticated, frontier-or-close to frontier, technology to introduce entirely new industries into the fabric of the economys industrial sector. The technology is sourced (searched for and assessed) by ITRI, then mastered by its staff, in the course of carrying out what adaptations appear warranted in the local context, then extended to firms. Sometimes the transfer involves formal modes of foreign supply (licensing), other times it does not. ITRIs role has progressed from being the singular implementer of the first steps (singular in the sense of without much cooperation by firms, some of which were accomplishing similar transfers on their own) to being the coordinator of collaborative consortia of key firms involved in cooperatively undertaking transfers. An example of a success story of upgrading the technology of an existing industry is the bicycle industry. In 1984 the Materials Research Laboratories division of ITRI, working in close cooperation with a local producer, Ih Ching Company, developed a carbon fiber-epoxy composite rapier wheel for shuttle-less weaving machines. This technology was adapted to another use in 1987, again in a successful collaboration. Together, Materials Research Laboratory and a local bicycle company, Giant Machine Co., developed an advanced bicycle frame made of a carbon-epoxy composite. This technology was systematically transferred to local firms. This, together with its upgrading, led to a revival of Taiwan Province of Chinas bicycle exports in the 1990s. Taiwan Province of chinas bicycle industry is now one of the worlds most advanced and successful.18

Research and development laboratories

Most of the industrial R&D that is relevant to the needs of developing countries is what might be termed transferrelated R&D. Unlike inventive R&D, which takes place on the global technological frontier and seeks truly novel products and processes, transfer-related R&D aims at the assimilation, adaptation and improvement of technologies transferred from elsewhere. Simple in-plant experimentation to learn the optimal configuration of process and product parameters is but one form of low-level transfer-related R&D. At the highest level, such R&D takes place in specialized, sophisticated labs that are very much like their counterparts in industrialized countries in terms of equipment and staffing. Because of their limited resources, small and medium-size enterprises are generally unable to engage in other than relatively low-level transfer-related R&D. Government-sponsored R&D laboratories thus have a critical role to play in developing and transferring technology to small and mediumsize enterprises. Recognizing this, the government of Taiwan Province of China established a number of such laboratories that have very effectively served the needs of the many small and medium-size enterprises that form the bulk of the industrial sector.17 The activities of these laboratories, most of which are specialized to serve a particular industry, have over time encompassed many levels of transfer-related R&D. Several technology research institutes provide R&D support to the private sector. The China Textile Research Centre, set up in 1959 to inspect exports, expanded to include training, quality systems, technology development and directly acquiring foreign technology. The Metal Industries Development Centre was set up in 1963 to work on practical development, testing and quality control work in metal-working industries.

Sequencing priorities
In thinking about sequencing, having an ITRI comes later. By its nature, it is highly sophisticated and requires highly qualified human capital. Sectorally focused R&D institutes meant to complement extension services prove useful much earlier, in sequential terms, than an ITRI clone. The highest priority at the outset should probably be given to general service organizations having relatively limited requirements for highly skilled technical personnel. This would enable

Supporting innovation and learning by firms

131

serving the greatest number of potential clients without need to predetermine which industrial sub-sectors should be given priority. That would in part be learned through the operations of the initial organizations, as could the details of administering service organizations and linking them effectively to individual firms and their needs and problems. Certainly, high priority should go to the reform of existing organizations that serve, or could be reformed to serve, industries in which the country should readily be able to realize a competitive advantage. As a general rule, organizationswhether newly formed or being reformedshould not seek staffing at a level of technical expertise too far in advance of that in the firms to be served, certainly not at the outset. The point is to gain the advantages of incremental learning, starting with modest means and expectations, learning as experience is accumulated on what works and what does not. The point is equally to ensure the capability for effectively serving firms in small, manageable ways before investing large sums to secure technical expertise without knowing that it can be effectively deployed in ways that will increase firms levels of productivity.

4. MIDA and Malaysian government documents on the Web; UNIDO; Lall (1996). 5. This discussion of EPZs in Taiwan Province of China draws on Dahlman and Sanaikone (1990), Galenson (1979) and Wade (1990). 6. Schive (1990). 7. Dahlman and Aubert (2001). 8. For more information, see http://www.itmin.net/, http://www. itmin.net/information/services.html.Unido, or http://www.unido.org/ doc/100438.htmls. 9. http://www.iso.ch. 10. http://www.bis.org.in/cert2.htm. 11. This section draws on Dahlman and Sananikone (1990). 12. http://www.bsmi.gov.tw/english/e_n_hpg.htm. 13. http://www.jpc-sed.or.jp/eng/. 14. This section draws on Shapira (1992, 1998) and Kolodny and others (2001). 15. This discussion of SERCOTEC and NAMAC draws on Shapira (1992, 1998) and Kolodny and others (2001). 16. This section draws on World Development, special issue on Industrial Clusters in Developing Countries; Institute of Development Studies Research, GlobalizationIndustrial Clusters in Developing Countries; http://www.unido.org/doc/331101.htmls. 17. Hou and Gee (1993). 18. Mathews and Cho (2000).

Notes
For further details on sources, information and the literature on subjects covered here, see the background papers. 1. Benefits that accrue to service recipients increase the general tax base from which government revenues are derived, providing an indirect means of overtly recovering fixed costs, one that figures prominently in some discussions of service provision in the developed countries. 2. See Pack and Westphal (1986, pp. 102126). 3. Loewendahl (2001).

132

Industrial Development Report 2002/2003

8
I

The way forward

NDUSTRY CAN BE A POWERFUL ENGINE FOR GROWTH AND STRUCTURAL

transformation in developing countries. It remains vital to innovation and to the creation of new skills, organizations and attitudes. It lies at the core of technology transfer, learning and diffusion. It is thus essential for ensuring sustained productivity growth. And its importance is increasing. To be emphasized is the enormous productive potential of new technologies and organizational methods (such as plugging into global value chains). There is enough productive knowledge around to transform standards of living in many poor countries, if they could build the capabilities (and raise the investment resources) to exploit them. The continuing disparity in competitive capabilities raises urgent problems, and it is vital to reverse this. The widespread liberalization of trade, investment and information flows is making it possible for industrial activity to encompass the developing world and to transfer resources to their enterprises. In short: the potential of harnessing industrialization for sustained development has never been greater. And the costs for countries that fail to realize this potential have never been larger. So far, only a small number of developing countries are realizing the full benefits of industrialization. The data clearly show that industrial performance is diverging within the developing worlda few successful economies are pulling away from the rest. And there are few signs of reversal. Nor does this appear to be simply a delayed reaction to globalization and liberalization. If it were, it would have corrected itself by now. The divergence of the groups of developing countries reflects the development of strong drivers of industrialization in only a handful of them. And it is highly likely that countries will diverge even more. Vexing and undesirable, this needs to be reversed. The developing countries can build competitive industrial capabilities in the current setting. This is not in doubt. Also clear is

that building these capabilities, faced by pervasive market and institutional failures, needs extensive policy support. But policy interventions in developing countries do not have a happy history: inefficiency and waste have marked the post-war experience of planning, import-substitution and state-led industrialization. Even so, the countries that employed industrial policy in export-oriented environmentswith complementary policies to build skills, technological capabilities and supporting institutions and to leverage foreign resourcesshow that such strategies can radically transform the industrial landscape in less than a generation. A natural starting point in formulating national strategies and policies is for countries to benchmark their industrial performance along the lines detailed in chapter 3. They can also benchmark the drivers of that performance by looking at the key structural variablesat local technological effort, at foreign direct investment, at licensing royalties paid abroad, at physical infrastructure (chapter 4). That way, they can position themselves to see what technological capabilities to develop, what global value chains to latch onto and what services to support for innovation and learning. These efforts cannot be left to detached policymakers alone. Needed are broad coalitions of public, private, civil and academic players, committed to agreeing on a vision that can give direction to their industrial strategy. As this chapter stresses, however, a countrys industrial policies have to be couched in the broader developmental perspective of creating wealth and enhancing welfare. The idea is not just to promote industry. It is to promote efficiency throughout the economyto sustain productivity growth and to ensure that the benefits are distributed equitably. That requires paying great attention to the framework conditions of political, social and macroeconomic stabilitynot just for industry but for all of society. It also requires putting in place the institutional foundations, again not just for industry but for all of society.

133

Framework conditionscannot be ignored

Successful, sustained industrial development can take place only if the economic and political conditions are right. On this the consensus is widespread. The most fundamental conditions are clearly political, social and macroeconomic stability. Without them, investors of all kinds, local and foreign, will shy away, and the signals investors respond to may be distorted. High inflation rates will induce investors to eschew long-term projects in favour of short-term onesand to shift from productive activities to those where payoffs are quicker and larger (property or stocks). Overvalued exchange rates will discriminate against exportable activities. An unstable political or social climate can induce short-term investments or drive them overseas. And so on. Along with stability is an equally important need for clarity and predictability in the policy environment. Otherwise, capacity building and capability accumulation will suffer. These factors are even more important for internationally mobile resources, such as foreign direct investment. Some international investors may be willing to accept high levels of risk. But they will demand high premiums, tending to focus on quickly extracted resources rather than on building longterm competence. Good policymaking goes beyond clarity and predictabilityit increasingly entails transparency and participation. Transparency in the policy process inspires confidence. Participation ensures trust and the flow of information. Indeed, networking between governments and the other main actors is perhaps as important for effective policy design and implementation as it is for innovation. Why? Because policies, while predictable in broad terms, also have to be flexible and adaptable in their details and application. In a world of constant change, it is not always possible to correctly anticipate policy needs or to predict the consequences of particular measures. For processes as complex as industrial restructuring, upgrading and innovation, it is imperative to build in policy learning and flexibility.1 Coherent strategy has, in other words, to be a learning process involving major stakeholdersexchanging information and sharing in implementation. Being competitive entails greater openness to markets, to imported technology, to information flowsindeed, to new ideas. Export competitiveness, in particular, requires close and frictionless contact with foreign sources and customers. It also requires low business transaction costs (dealing with rules, regulations and the bureaucracy) and good governance. To attract export-oriented foreign direct investment, especially in hightech industries, requires a very efficient business environment. The growing significance of efficient supply chain management

means that locations must offer more than good transport infrastructure. Also needed are rapid import and export procedures. In general there is a move away from discretionary procedures in dealing with the private sector and towards simple, universal rules that are easy to understand and comply with. The Asian Tigers built up these framework conditions. They had a leadership committed to competitive industrial development, complemented by a broad education base and a fairly equitable income distribution. The government bureaucracy was skilled and highly respectedmore important, it was relatively insulated from day-to-day politics and able to respond pragmatically to change. These attributes were not inherent to Asian society. Quite the contrary, these policy capabilities were built up in a long process of experimenting, making mistakes, changing and learningvery similar to the process of building industrial capabilities.2 This policy learning may not be replicable in its entirety. But as studies note, it does offer lessons to other countries.3 Improving the bureaucracy, its base of skills and information, its coherence and linkages with the private sectorall these are things that governments can do elsewhere. The pace will depend on circumstances, but the process is cleargradual and cumulative, advancing a step at a time. Of the many framework imperatives required for dynamic industrial development, each is necessary but collectively they cannot be shown to be sufficient (box 8.1). Indeed, the list in box 8.1 is not a list of policiesit is a list of framework imperatives to pay attention to. Each country has to use that list as a starting point for designing the policies best suited to its conditions and aspirations. The East Asian countries paid careful attention to each, but no two countries attained each in precisely the same way. Indeed, the considerable diversity among the East Asians attainment of the imperatives reinforces the previous message: each country must design its own strategy with the expectation that it will, in the details, contain unique elements. The diversity among all countries will be as apparent in policy design and implementation as in institutions and organizations. The framework imperatives are of fundamental importance for achieving and maintaining internationally competitive production, at the outset and as industrial development unfolds. Their vital significance can be seen wherever there is an apparent competitive advantage that is not being realized. Clothing production in Senegal (box 8.2) provides a fruitful illustration of this point, fruitful because it highlights two critical corollaries. One is that the imperatives are not easily or trivially put into practice. Reforms to implement them have more often been problematically incomplete. The other is that the framework imperatives must be seen as vital instruments of a coherently framed strategy, aggressively pursued.

134

Industrial Development Report 2002/2003

Box 8.1 Framework imperatives for effective industrialization 1. Policies assuring macroeconomic stabilitywithin rather narrow limits, and both in reality and in expectationare important to encourage rapid factor accumulation and allocation in accord with comparative advantage (dynamic and static) as well as to make possible quick and effective responses to disruptive shocks. They are reflected in: Relatively low inflation rates and positive real interest rates. Fiscal balance (between government revenues and expenditures). Real (purchasing power parity) exchange rates maintained at levels not greatly overvalued, if at all, relative to free-trade exchange rates. 2. Policies have to ensure resource allocation in accord with dynamic (or potential) comparative advantage, where ensuring is to be understood in the sense of maximizing the likelihood that allocative decisions are made on the basis of rapidly achieving and then maintaining internationally competitive production, whether sales are in open competition on world markets or on the domestic market in unprotected, unsubsidized competition with imports. 3. Rapid accumulation of physical and human capitalthat is, the rapid growth of factor inputsrequires: Forward-looking provision of infrastructure, sufficient to avoid problematic bottlenecks. Expeditious attainment, first, of universal primary education then, secondary education. Attention to technical training and to technical education (engineering and scientific) at the tertiary level. 4. Successful agricultural development is important for equitable developmentand to ensure that appropriate balance is maintained across sectors as each develops. 5. Institutions are required to enable effective commerce among economic agents: Contractual arrangements, explicit or implicit, having adequate sanction, formal or informal. Incentives, whether rooted in individualism and private property or in social solidarity pacts of one kind or another, free from being undermined by capricious authority. Mechanisms fostering adaptive institutional and organizational change in the context of underlying social stability. 6. Competent bureaucracies are needed to orchestrate the development process effectively.
Source: Westphal (forthcoming).

Box 8.2 Comparative advantageto be realized Senegal should, from all appearances, have a strong competitive advantage in exporting clothing. At least that is the opinion of knowledgeable experts who have examined its prospects closely. Among its advantages are a location close to European markets and, very important, a vibrant informal sector of thousands of highly skilled, hard working tailors who produce for the high end of the local market. Senegal has also undertaken reforms to bring its policies into closer conformity with the fundamental imperatives. It has put in place what appears to be a typical package of investment and export incentives designed to attract foreign investment to its export processing freetrade zone. But Senegals export performance in clothing is below what would be expected on the basis of the reforms already undertaken. Why? A careful field study, including interviews, found that formal sector clothing producers were frustrated in their attempts to export. Among the reasons:
Difficulties obtaining on-time deliveries from local fabric produc-

ers, and deficiencies in their quality control.

Inadequate access to finance required to upgrade plant and equip-

ment and to increase employment; insufficient technological capability to export in large volume.
Absence of managerial and marketing knowledge to sell in export

markets, with no practicable notion of how to attract foreign buyers or partner with foreign firms to enter clothing value chains.
Poor quality of infrastructure (leading, for example, to frequent

electricity blackouts) and government services (in contrast to the favored treatment of large fabric producers).
Locational disadvantages due to being far removed from sources

of imported fabrics.
Training institutes incapable of providing useful training, and other

supporting institutions ineffective in providing useful support. The Government of Senegal has initiated consultations with the private sector to address these deficiencies. With UNIDO assistance, it plans to formulate a strategy to upgrade the competitiveness and capabilities of the textile sector, including establishing a textile center to promote exports.
Source: UNIDO and Golub and Mbaye (2000).

The objectives of policy reform must be stated in precise, operational terms, terms sufficient to permit meaningful monitoring of their achievement, this to enable the revision of policy or its implementation when necessary to accomplish the objectives sought. Strategies aggressively pursued are no less relevant to accomplishing sufficient policy reform than they are to achieving continued technological development, this even though the purpose in one case is to unleash market forces toward productive ends and in the other to supplement those forces so as to overcome market failures. Regardless of their necessity, the framework imperatives are not sufficient to enable sustained innovation and learning

leading to continued industrial development. And they certainly are not sufficient if the imperative of resource allocation in accord with dynamic comparative advantage is conceived in narrow terms, adequate only to ensure the realization of static comparative advantage based on existing resources and competencies. This can be seen wherever the achievement of a significant competitive advantage in one area does not unleash a chain-reaction of innovation and learning that leads to a deepening of that advantage and a broadening of competitive advantage in unrelated areas (box 8.3). The framework imperatives reflect two successive generations of international consensus about the conditions required for

The way forward

135

Box 8.3 Broadening competitive advantage is far from automatic The Mauritius clothing industry embarked on a successful path of exporting in the 1970s, and clothing exports continue to be the mainstay of its industrial sector. Careful reading of the studies to ascertain the basis of Mauritian success leaves no doubt that its exports are in large part the result of adherence to the framework imperatives. Not only do exporters in Mauritius benefit from a virtual free-trade regime governing their activitythey also enjoy an institutional setting that is free of the features that militate against profitable private entrepreneurial activity. But adherence to the fundamentals in Mauritius entails export promotion in the presence of high levels of import protection. Moreover, serendipitous factors, some the product of its history, have played an undeniably important role. Its favorable institutional conditions derive from its particular social and political history, and it enjoys important trade preferences under the Multi-Fibre Arrangement. In turn, its colonial heritage includes a small community of well-connected ethnic Chinese who were instrumental in attracting Hong Kong Special Administrative Region (SAR) of China firms to initiate clothing exports from Mauritius. Hong Kong SAR-owned firms remain an important presence in the sector. It is not enough simply to have succeeded at the outset; an evolving strategy is required if the momentum once gained is not to be lost. Here Mauritius has failed. Having created a successful albeit narrowly focused export enclave, the government has no promising strategy to exploit its potential linkages into the rest of the industrial economy or to foster deeper competitive advantages in clothing exports or to broaden the scope of the countrys industrial competitive advantages beyond clothing. In short, Mauritius seems to lack a strategy for ensuring that innovation and learning will lead to continued industrial success.
Sources: Romer (1993); Rodrik (1990); Subramanian and Roy (2001); Lall and Wignaraja (1998).

policies and institutions, but dynamic forces leading to rapidly increasing productivity have not taken hold in its industrial sector. What is missing in such cases is what was present in East Asia: a recognition of the necessity to serve the innovation and learning needs of firms. A third generation of reforms, one that emphasizes the critical importance of innovation and learning, is at hand.4

What is needed
Industrial catch-up has been accelerating. What the Republic of Korea and Taiwan Province of China achieved in three decades took Japan much longer; Japan industrialized much faster than early predecessors, and today China seems set to overtake the records set by the Republic of Korea and Taiwan Province of China. Yet many latecomers are failing to catch up at all in the same technological, trade, investment and information environment. Their industrial capabilities are inadequate to the challenge of competitive growth. The explanation for the different development of national capabilities lies first in the presence or absence of the framework imperativesthen in the attention to innovation, learning and industrial development in a coherent country strategy and policy framework. Policies must thus be changed, reoriented to focus squarely on domestic innovation and learning, on the building of industrial capabilities by linking to global markets and leveraging foreign resources. The objective of industrial strategies and policies is to develop and sustain competitiveness and productivity growththe only viable way to promote industrialization today. This simple but vital objective has many ramifications. At the outset, it usually entails the restructuring and upgrading of industrial activities. This in turn involves developing new capabilities, productive facilities and links with global value chains. To sustain long-term growth, leading to higher wages, also entails moving up the quality and technological ladder, within existing activities and across them, from simple to complex. Industrial maturation inevitably involves such structural upgrading in manufacturing, with the promise of significant benefits. Over the past half century far-reaching institutional and technological changes have fostered the extensive vertical separation of production into separable, sequenced activitiesfrom raw materials extraction through intermediate stages of production to sale of finished products. These changes have been associated with the appearance of

accelerated economic development and thus about what should be the objectives of reform. The first generation centered on the so-called Washington Consensus regarding the necessity for macroeconomic stability andusing the popular aphorismgetting the prices right. The results of first generation reformsof policies affecting resource accumulation and allocationin the countries that seriously undertook them were distinctly mixed. Thus were born the second generation of reform imperatives focused on achieving a constellation of enabling economic, political and social institutions. Policy reform sometimes requires a degree of radical institutional reform, while fundamental changes in the overall institutional setting often take place only over comparatively long periods in what is at best a loosely coordinated fashion. But there is ample reason to believe that policy and institutional reforms taken together are insufficient to trigger the activities of innovation and learning required to achieve rapid productivity growth in the industrial sector. Chile, for example, has realized substantial development gains from its reforms of

136

Industrial Development Report 2002/2003

global value chains spanning many of the worlds most important industries, and they have opened multiple entry points for less developed countries to engage in export processing. These changes have also been associated with greatly enhanced flows of technology among countries, enabling developing countries to take advantage of many of those entry points. Thus a number of developing countries have in recent years succeeded in becoming major players within dynamic global value chains. A few have done so by building skills and technological capabilities within indigenous firms, to achieve entry as participants in globalized production. But most developing countries have done it by undertaking labour-intensive functions for transnational corporations in more formally integrated production systems. Transnational corporations have always established production facilities in other countries, but the traditional mode has been to replicate entire facilities overseas. The forces of globalization, including the emergence of new technologies facilitating information communication and organizational innovation, have radically transformed the ways in which transnational corporations operate. Transnational corporations now separate production processes (and such functions as accounting, marketing, servicing and even research and development) into small slices and locate them across the globe to take advantage of fine differences in labour cost, delivery, skill, innovation capabilities, suppliers and so on. They can manage far-flung sites as a coherent whole to further the competitive position of the corporation. For newcomers participating in such systems opens enormous opportunities which can be more readily seized than can entry through more autonomous means. Newcomers can take on functions for which they are suited rather than the entire manufacturing or service processall the while enjoying access to massive new markets. They can also enter dynamic activities with great opportunities for technological learning and spillovers. Choosing points of entry to promote the development of new activities requires great care, however. The economic and policy context is very different today than when the Asian Tigers mounted their industrial policies. Innovation has accelerated, and economic space has diminished. The rules of the game are also very different. These changes constrain countries from committing some of the more egregious policy mistakes of the pastbut they also preclude the use of tools that have proved very effective in early stages of industrialization (in the mature industrialized countries, not just the newly industrializing ones). For example, promoting industries is now more circumscribed by trading rulesbut the criteria for determining what to promote remain unchanged (box 8.4).

Strategy starts with a clear vision

Governments across the world now have to mount strategies to enhance the competitiveness and support the productivity growth of their firms. And if economies with the most advanced markets and institutions feel the need to undertake competitiveness strategies, the need on the part of poor countries, with much weaker markets and institutions, must be correspondingly greater. Public efforts require direction. And without vision, there can be no focused direction. Nor can the direction be fixedit demands constant monitoring and revision, as every success story details. Vision is not only about the broad dimensions of strategy, it is also very importantly about the technologies and industries to be promoted. The basis of any industrial strategy is a national vision of industrial development. Vision is needed to coordinate and direct policies because it is possible to adopt a range of different development paths. In the textbook case of market failures, policies aim to restore a unique competitive equilibrium. In a world of imperfect markets, externalities, cumulativeness and pathdependence, there is no unique optimum but a range of possible multiple equilibria, some producing low growth or stagnation and others dynamism and high growth.5 Industrial success is the result of countries ability to move across these equilibriato keep jumping to the next curve. Each move needs coherent policies across a range of markets to exploit technological spillovers, scale and scope economies and dynamic learning. The government has to decide on the broad national objectiveseconomic and non-economicthat cannot be thrown up by markets. The strategy may be explicit or implicit, but it provides the parameters within which all other allocation decisions are made. For instance, one government may decide explicitly to promote industrial deepening, greater indigenization of technological activity or the creation of large conglomerates to internalize various markets.6 Another may opt for leaving all choices to free markets (not intervening is as much a strategic choice as intervening). These choices reflect many things: resources, location, external pressures, political circumstances and ideologies as well as rational calculations of how to develop national industrial and technological resources. The four mature Asian Tigers adopted completely different strategies in these respects (box 8.5). The next level of strategy has to do with the design and implementation of specific policies and programmes.7 Policies may be market friendly in that they seek to improve markets and institutions without favouring some over others. For instance, the government may subsidize R&D by enterprises for a period to create a technology culture in local industry or strengthen

The way forward

137

Box 8.4 What to promote? There is no sense denying that the formulation and implementation of industrial development strategy are an imperfect art. Efforts to quantify the costs and benefits of industrial promotion activities can at best yield rough approximations to what is in principle wanted. There are various reasons why this is true, but none is more pertinent here than the fact that many of the benefits sought by industrial promotion are imperfectly foreseen, while others are not foreseeable in any detail. Technological efforts leading to innovation and learning take place in an extensive profusion of cascading changes. Individual changes begetsometimes in compulsive sequences, sometimes as serendipitous outgrowthsfollow-on efforts and changes. And that process continues over the evolutionary course of a vibrant industrial sector. But this reality is not reason enough to eschew cost-benefit determinations. Careful attention to the explicit enumeration of costs (typically relatively well-perceived) and foreseeable benefits (many poorly perceived) and to the quantification of those amenable to some degree of quantification are the only practical way of imposing discipline on actions taken in pursuit of industrialization. And some means of discipline is required to ensure some success in the pursuit of strategy. Blind faith in outcomes to justify costs incurred is no guide to effective action. The accepted test for whether a particular activity should be promoted is the so-called Mill-Bastable test. The box figure illustrates its application in the traditional context of import substitution, but here with the additional expectation of eventual exports. The pronouncedly downward trending curve ABC represents the trajectory of the unit cost of domestic production, while the slightly downward tending line DBE shows the trajectory of the world price of the product in question; both trajectories are with respect to cumulative domestic output, measured on the horizontal axis. The unit cost of domestic production is initially above the world price owing to the absence of mature capabilities in the local industry; it falls with cumulative production as technological effortsassumed here to occur, as occur they must to achieve maturityleading to adaptive innovations and technological learning bear increasing fruit. The world price falls because of technological changes continuously occurring in other producing countries. The test is passed in its most stringent form, but for the neglect of time discounting, only if the area representing initial excess costs, ABD, is exceeded by the area of eventual gains due to the competitive advantage reflected in unit cost being less than world price, BCE extended into the future until the point where the competitive advantage is lost, typically to a more lately developing country.a But the test just stated is in fact too stringent, because it neglects the externalities that may spillover as technological efforts undertaken to achieve competitiveness in the activity contribute in multiple, cascading ways to technological developments in relation to other activities. Required is some estimate of the discounted value of these spillovers, which must be added to the direct value of discounted competitive gains in order to arrive at the proper magnitude for comparison with the value of excess costs. The Mill-Bastable test reflects two essential principles that must, if only qualitatively, guide the prior assessment of promotional undertakings if they are to be sensibly pursued. The first is that the expected competitive gains must exceed the initial excess costs, both taken in total magnitude. Thus it is not sufficient simply to become minimally competitive in the sense of unit cost equal to world price (including transport and transaction costs); a true competitive advantage must be foreseen. The second principle is that conditions in other economies must not be thought static and unchanging; the target to be achieved is a moving target, one moving continually against the countrys advantage. Thus it is imperative to take account of what is likely to happen elsewhere as it may affect production costs in other countries. It may be obvious that these principles as here illustrated have immediate application only to the direct promotion, as through import protection, of activities to produce existing tradable goods. But the scope for direct promotion is now greatly restricted compared with what was possible in the past, so that reliance must now be placed on indirect means of promotion, through such means as the provision of industrial services. This makes the task of benefit-cost analysis all the more difficult. But it does not in any way invalidate the principles just stated or render them any less important. The foregoing statement applies equally when outcomes do not simply replicate what is available from foreign sources. The cost, benefit and duration of infancy Unit values A

Domestic output

Source: Bell and others (1984, pp. 102106). a. Time discounting of excess costs and competitive gains is not reflected in the figure but must be applied in computing the net benefit of the activitys promotion.

incentives for employee training. Or policies may be selective, aiming to promote particular activities or clusters to tap dynamic learning possibilities, capture exceptional spillover benefits or attract the most promising global value chain activities. Both approaches are theoretically justifiable in the presence of market failures, and they are entirely complementary. The choice of appropriate instruments depends on the nature

of the failures and the capabilities of the government to undertake policies effectively. However, the more selective are the policies chosen, the greater are the competence, information, objectivity and flexibility required of the bureaucracy. The choice of appropriate measures involves creating the mechanisms to implement policies. Implementation may

138

Industrial Development Report 2002/2003

Box 8.5 Four Tigersfour broad visions The four Asian Tigers were the first countries in the developing world to launch an export-oriented manufacturing strategy. While Hong Kong SAR was always free trade, the other threeSingapore, the Republic of Korea and Taiwan Province of Chinaturned from import substitution to export orientation in the early 1960s. They led the first wave of labourintensive industrial exports: garments, textiles, toys, footwear and the like. Over the 1970s and 1980s they upgraded their export structures in different ways, depending on their differing visions of what they wanted their development paths to consist of. This depended in turn on their social and political structures, resource endowments, size and history. In Hong Kong SAR the vision of the colonial government was marketdriven resource allocation, with no particular ambition to develop local manufacturing. Although Hong Kong SAR was once the leader in the developing world in manufactured exports, this vision led to quality improvement in labour-intensive exports but to relatively little structural deepening. As a result, with rising wages, most manufacturing shifted to lower wage countries, and industrial and export growth stagnated or turned negative. The export structure remained at low technology levels, the lowest among the Tigers. In Singapore, by contrast, the government had a strong vision of technological upgrading and deepening. This led it to intervene extensively in investment patterns, skill development and infrastructure building while retaining a free trade setting. The result was considerable deepening, allowing Singapore to combine high and rising wages (nearly 20 percent higher than in Hong Kong SAR) with output and export growth. Singapore moved rapidly from low-tech activities to petrochemicals and then producer electronics and equipment, simultaneously raising its technological levels from simple assembly to high-end manufacturing, design and development. Transnational corporations, providing state-of-the-art technologies and access to their global networks, dominated the process. While Singapore developed a very high-tech export structure, however, its research base stayed small and the main sources of innovation remained overseas. The Republic of Korea and Taiwan Province of China also had very strong visions of industrial development, this time with larger ambitions for national enterprises. Transnational corporations were allowed a much smaller role, though foreign technology was tapped extensively in other forms. Their governments used infant industry
Sources: Amsden (1989); Wade (1990); Lall (1996); Westphal (forthcoming).

protection (offsetting its harmful effects by strong export incentives), credit allocation and subsidies, foreign direct investment restriction and skills and technology support. And they did this in ways to induce local firms to enter difficult activities, raise local content and take on advanced technological functions. The Republic of Koreas interventions were more pervasive and detailed. They involved fostering the chaebol, the conglomerates that spearheaded its heavy industry and high technology drive, learned the most advanced technologies and became major transnational corporations in their own right. Taiwan Province of China intervened less directly in the industrial structure, though it used public enterprises to enter several heavy industries. It supported its small and medium-size enterprise dominated structure with an array of technology, training, finance and export marketing policies and institutions. Because of their far-reaching efforts, the Republic of Korea and Taiwan Province of China have the greatest technological depth in the developing world, and their exports embody the most intense learning. A vision gone sour Some of the Republic of Koreas seeming policy successes later turned sour. The Asian financial crisis dramatically exposed the substantial risk inherent in the pursuit of an ambitious vision by aggressively interventionist means. The risk is that policies, while proving to be highly effective in the medium term, can have seriously detrimental consequences for long-term institutional development. The Korean governments practice of directing the allocation of credit severely retarded the development of modern financial institutionsand stifled the establishment of a adequate regulatory system. In turn, its promotion of the chaebol ultimately created a number of differently dysfunctional private entities whose exploitation of the policy regime resulted in unwise investmentsleading in some cases to excess capacity, in others to bankruptcy in whole or in part. In a very real sense, then, the high costs imposed on the people of the Republic of Korea by the crisis-induced recession and the subsequent vigorous pursuit of systemic policy reform are the price paid for the governments earlier inability to rectify growing institutional deficiencies. Taiwan Province of China did not foster chaebol-like enterprises, but otherwise followed a similar though distinct set of policies. It has paid a far lesser price by managing the long-term consequences of directed credit more effectively.

need new institutions (in the public or private sector) to support, interact with and link market agents. In the public sector, for example, the government has to provide the technological public goods needed by industry, such as basic research, extension services, standards and metrology. In the private sector institutions may include business associations, consortia or large private conglomerates (like the chaebol in Korea) that can overcome deficient markets for capital, skills, information and entrepreneurship. Keep in mind that it is the process that is criticalnot the instruments. The actual policies used must be specific to each strategy and context. That makes the policy process more an art than a science. Since mistakes are inevitable, the government has to be flexible and responsive to changing circumstancespolicy has to build in learning and adjustment.8

Some final points on the strategy process. First, policy needs vary with the level of development. As markets and institutions become more efficient and complex, the need for direct interventions falls and their potential costs rise. Second, industrial policy must be systemic. No strategy can succeed unless it dovetails physical investment in capacity with technology development, skill building, cluster strengthening and so on. Third, policies must correspond to the phase of learning and so must change accordingly: policies in the infant phases of capability building must differ from those in the mature phase, when R&D and frontier innovation become vital. Governments require disciplined means of strategy formulation, implementation and monitoring, with monitoring being imperative for determining whether and in what respects an ongoing strategy warrants revision. Two very important tools

The way forward

139

that impose discipline within a process that embeds global knowledge in the articulation and pursuit of development objectives are finding increasing acceptance and use among developing countries: benchmarking and foresight exercises. Benchmarking has long been used by successful industrial firms, first in the industrialized countries, as means of achieving best-practice levels of productivity, which can only be determined on the basis of comparative information across the universe of similar firms. It relies on the identification of measurable factors that are critically related to overall productivity, factors that are subject to the firms control either directly or indirectly. Knowledge of the best values attained in relation to these factors across other firms provides targets that, if achieved, will result in the best-practice level of overall productivity within the individual firm. Targets set in this fashion act as powerful devices for focusing technological and related efforts toward the achievement of objectives that can be realized by searching for and using global knowledge pertaining to the factor and its relationship to overall productivity. Benchmarking has immediate application where relationships between contributing factors and desired results are well known and easily quantified in simple terms; for example, in industries where engineering relationships dominate in determining productivity. It is less readily applied by organizations that provide services, such as those discussed in chapter 7, where the relationships between service norms and outcomes are complex owing to many factors beyond the control of the organization. Even so, benchmarking exercises are prevalent among the exemplars of global best-practice in serving the needs of industry. Benchmarking is yet more difficult in the policy arena, and this for a variety of reasons, central among them the necessity to recognize that policies must often be tailored in conformity with national values and the institutional setting. Nonetheless, there has recently been substantial progress in bringing the discipline of benchmarking to the service of policy analysis and formulation. Just as information centers, extension agencies and the like play a vital role enabling benchmarking practices by industrial firms, so too can international agencies exercise a profound influence by providing the information and supporting technical assistance required for agency and policy benchmarking by developing country governments.9 In the process of selecting industries for promotion, technology foresight exercises, done hand in hand with the private sector, are particularly useful as a means to comprehend emerging global trends, enabling both firms and the government to formulate detailed strategies in areas that seem sensible.10 Having originated decades ago in Japan and France, they are now in widespread use in the industrialized countries,

where their application focuses on forecasting the course of global technological change in relation to the countrys industrial strengths and weaknesses, to guide public science and technology policies and expenditures (Martin 1996). Even comparatively non-interventionist governments, like that of the United States, recognize that foresight exercises are vitally important owing to the fact that firms cannot remain competitive without relying extensively on complementary private and public sources of knowledge whose continuing development must in some fashion be coordinated within a common vision of how the future might unfold. Similar exercises are now also being undertaken by many developing countries, with UNIDOs assistance. The focus of these exercises differs in that the objective relates to steps being taken to catch up with the global technological frontier, not to steps necessary to remain on, or at the forefront of, the changing frontier. Even so, developing countries require foresight in relation to existing industries, not simply for keeping up but also for catching up to a shifting frontier, and in relation to industrial activities for which potential competitive advantage is within grasp. They additionally require foresight not only about technological trends, but also about pending changes in the international ordering of economic activity where, for example, new modes of accomplishing the division of labor among countries may seriously affect the way in which opportunities may be seized. But foresight exercises, wherever conducted, are not simply about external factors, or emerging global trends. They are also fundamentally about internal factors, or assessing a countrys industrial strengths and weaknesses in sufficient detail to ascertain where change efforts are required, and this even if only to come closer to existing best practice. Indeed, in developing countries much of the work most useful for foresight exercises goes into developing a vision of a future in which existing resources are used with greater productivity as a consequence of a diverse variety of technological and other efforts enabled by technology transfers from sources both internal and external to the country. In days gone by, some countriesthe Republic of Korea in the 1960s and 1970s, for examplepracticed a form of visionary benchmarking, comparing their industrial structure with that over the past of some exemplary, more advanced country (for the Republic of Korea, Japan), to determine the activities next in line for development. Such benchmarking may still provide some, albeit limited, useful guidance. But vastly more important, indeed fundamentally essential, are targets and actions determined by the collaborative engagement of industrialists, technologists, academics, government agency officials and other importantly involved parties in coordinated deliberations based on intimate knowledge of the reality that is and

140

Industrial Development Report 2002/2003

that could be. Indeed, such deliberations lay at the heart of the Republic of Koreas planning activity in the mid-1960s and were, in a very real sense, instrumental in the Republic of Koreas embarking on the path of rapid industrialization.11 Then, as now (see box 8.6), the value of the exercise inheres not so much in the formulated vision but in the common understanding by both public and private actors engaged in its achievement of the steps that each must take if it is to be realized.

Box 8.6 Foresight in Hungary Hungarys Technology Foresight Program adapted the conventional advanced-country type of exercise to its strategic needs in becoming fully integrated into the European socio-economic system. Seven sectorally focused panels made up of industrialists, academics, government officials, and other stakeholders were formed to assess strengths and weaknesses together with opportunities and threats in their sectors, with the aim of formulating visions and recommending policies and programs necessary for their fulfillment. Complementing their effort was a large scale survey using the Delphi method (a disciplined means of aggregating individual forecasts of emerging trends stated in detailed terms) that was focused on probable external changes. The program was carefully orchestrated to achieve the greatest possible awareness and participation, toward the end of developing a strong consensus centered around its outcome. Thus the core exercise, which required two years to complete, was preceded by an initial stage devoted to promoting understanding of its importance, and followed by a terminal stage during which results were disseminated and widely discussed. To enable meaningful discussion and to further the objective of achieving consensus around a plan of vigorous action to achieve effective integration and robust economic development, three alternative visions of the future were presented, with only one of them having been fleshed out in considerable micro detail. Hungarys experience in its foresight exercise demonstrates that the methodology can be of great value to developing countries and transition economies alike. Individuals having important responsibilities within the private and public sectors gained in their understanding that innovation is an inherently collaborative enterprise that entails mutual effort and learning among cooperating entities. The exercise additionally led to an enhanced awareness of the importance of communication within and across organizational boundaries, while at the same time it strengthened existing, and established new, network relationships, both formal and informal, among the parties engaged in innovation. Moreover, it reinforced the comprehension that innovation and learning do not only relate to technical matters, but also to economic, organizational, and social factors as well. Hungarys experience is also significant in showing that foresight exercises both can and must, if they are to yield fruitful results, be tailored to the capabilities and institutional settings of the countries in which they take place. And to the central matter at issue, which in Hungarys case was enhanced integration within the larger European system; in many developing countries it would importantly involve greater integration into the world economy to enable greater levels of productivity. Foresight exercises by no means need be as complex and long-lasting as was Hungarys. Simpler, shorter, and more narrowly focused exercises can add real value in formulating and implementing intelligent strategies for industrial development.
Source: UNIDO.

Five principles for government conduct of national strategy

1. Set priorities for policy intervention in line with the vision. Resourcesfinancial and humanfor policy intervention are necessarily limited, even in rich countries. They are far more constrained in poorer countries. That makes it essential to mobilize resources and set priorities in line with the national development vision. The priorities will depend on the circumstances of each country, of course, and these are likely to vary with the level of development. At higher levels the most pressing needs tend to revolve around innovation and specialized skills and infrastructure. At lower levels they revolve around building entrylevel competencies, strengthening resource-based and labour-intensive industrial activities, upgrading smaller enterprises or providing basic infrastructure. 2. Leverage national resources with foreign onesin global value chains. The most effective means of connectivity to technological resources will often be through value chains, and, as noted in chapter 6, it is vital to understand the nature of each chain and the lead players in it. But sustaining linkages requires building complementary capabilities locally. Linking and leveraging strategies will again vary with the level of industrial development and with national strategic priorities. Countries with strong industrial and technological bases can follow strategies pioneered by the Asian Tigers. Those with weaker bases have to use more modest strategies, helping firms and clusters to connect with global players and climb up the value chain from modest levels, relying more heavily on direct foreign investment. 3. Coordinate the vision, the framework conditions and the drivers. Competitiveness-enhancing strategy involves close coordination of the vision, the framework conditions and the drivers. Most governments do not formally exercise such a coordinating function, since it cuts across traditional lines of responsibility in domestic ministries

and agencies. Coordination tends to be dispersed and ad hoc, based on implicit rather than explicit objectives and strategies. This may be effective where drivers are fairly well developed, when decisionmakers agree on priorities and actions and when the line ministries exchange information to support each other. But it is not effective in other situations. In particular, where a country has to mount major policy changes and embark on significant

The way forward

141

structural change, there is a need for a formal mechanism to formulate strategy and coordinate the development of industrial drivers. The coordinating function does not necessarily need a new agency; it may be exercised by a group of existing ministries and institutions. But the function can be carried out only if its execution is placed close to the apex of policymaking. Effective coordination, wherever it is located, needs regular secretariat support for the collection and analysis of datalocally and in international benchmarks and immediate competitors. The data need to cover production, trade and productivity performance, the major drivers and the lead institutions. Everything has to be based on a sound understanding of technological and market trends. 4. Build skills, knowledge and bureaucratic competence. The conduct of strategy can be very demanding in skills, information and bureaucratic competence. But many measures can reduce the pressures on national governments. The private sector can contribute much to the design and implementation of policy, relieving government of many difficult data collection and analysis functions. Indeed, the private sector is much better placed to gauge productivity, technological and market trends at the individual activity level than is the government. What the government needs to do is provide a higher strategic and structural perspectiveand to distill diverse private sector views into a coherent vision of development for the medium term. 5. Enlist the key actors in the international community in strategy formulation. The international community can help in strategy formulation. Apart from material assistance, it can provide valuable information onand analysis ofbenchmarks, institutions and policies in other countries. Many competitiveness analyses by industrialized countries are publicly available. And technology and training institutions are often willing to provide aid or sell their services. Aid agencies furnish technical assistance, often drawing upon the services of industrial experts. Consultants provide analyses of competitiveness as a whole and of its various components; their services tend to be expensive, but they possess a wealth of experience and data. In addition, there is a need for analytical support for governments at a higher level, particularly in evaluating different strategic approaches and the lessons of experience in other countries.

removal of all such policies as local content rules and performance requirements, as envisioned in negotiations for trade-related investment measures (TRIMS). Indeed, the optimal level of openness and the ideal pace of trade liberalization remain a matter for debate.12 Accepting that many countries have intervened excessively in trade (and to the detriment of their industrialization), it does not follow, given the market and institutional failures facing the acquisition of technological capabilities, that completely free trade is a desirable objective, certainly not in the near future, for developing countries. That some interventions were wrongly designed or implemented does not imply that all interventions are inefficient or distorting. Theory suggests that where deficient markets give distorted signals to economic actors, intervention is needed to restore efficiency. Careful trade interventions, set in the context of strong export orientation and balanced by stringent performance requirements can work well. Going further back in history, trade and other interventions were used extensively to promote industrial catch-up in the presently developed countries. Qualifications of the same nature apply to TRIMS-related policies. Similar considerations apply to the widespread application of stricter intellectual property rights in the developing world (under agreements for trade-related intellectual property rights, TRIPS). There is a growing feeling that the universal application of TRIPS offers little to countries at low levels of industrial and technological development, while imposing additional short-term costs on them as importers of technology.13 There may be long-term rewards to them in accepting TRIPS, but the gains may well be negative in present value terms (after discounting future gains at a reasonable interest rate). Careful analysis is needed of whether existing rules are flexible enough to allow the losers to prolong their grace periods or whether the rules need to be changed. Building capabilities is a costly, demanding and continuous processand no amount of good policy can get around the problem of severe resource constraints in most developing countries. As the gap widens between the more successful and less successful countries, the benefits of modern technology and globalization appear further out of reach to many. This raises social and political stresses, threatening the pace of economic reform and integration and affecting the stability of the international economic system. Current development aid practice attaches less weight to the industrial sector than it did in the past. Perhaps donors assume that market forces (liberalization and globalization) will suffice for industrial development. This is wrong. The upgrading and regeneration of manufacturing need support from aid donors. True, the current donor emphasis on edu-

International dimensions
The desired, appropriate level of openness may not entail completely free markets for trade and investment or the

142

Industrial Development Report 2002/2003

cation, infrastructure and micro or small enterprises does feed into industrial development. But it does so at one remove. It does not directly address the needs of industrial restructuring by the formal enterprise sectoror the specific skill and technological needs of modern industry. And for reasons just noted, it cannot be taken for granted that national governments in developing countries will on their own be able to meet these needs. Without a substantial increase in assistance, many viable activities may go underand many more promising activities may never be launched. The time is ripe for a new international agenda on industrial developmentand for a new vision of how developed countries and international agencies can best assist industrial development. Countries have to be helped in their efforts to build competitiveness, attract resources, use more productive technologies and enter larger, more dynamic markets. Otherwise the enormous potential of economic integration and globalization may be lost to a large part of the developing world. Most of the effort has to come from within countries, providing the right environment for capability building and investing in the necessary factors and institutions. But such local efforts should be helped from outside. Opening markets completely in industrialized countries will help greatly, but much more is needed to narrow the widening gap between countries and to build industrial capabilities in developing countries. Indeed, this is the mission of UNIDOall of our activities deal directly with building and enhancing industrial capabilities. We will continue working to narrow that gap and to ensure support for that work with financial and other resources.

4. See Magarios (2001). Magarios discussion also identifies other key aspectsrelating to environmental sustainability and the equitable distribution of productivity gainsplus additional elements missing from the current consensus that for similarly now-obvious reasons require inclusion in the third-generation consensus. 5. On the possibility of multiple equilibria and the need to shift across them see Redding (1999), Rodrik (1996), Stokey (1991) and Stiglitz (1996). 6. Most strategic choices at this level have to do with identifying groups of activities that have the greatest potential for dynamic growth or create most beneficial externalities for other activities (Lall and Teubal 1998). To the extent that such activities involve greater risk and learning costs, coordination problems and capital market failures, free markets cannot lead to their development. Only deliberate promotion would lead private agents to enter such activities in a coherent fashion and on a scale necessary to make the local value chain efficient. Examples would be the strategic targeting of heavy or hightech industries in Japan and the Republic of Korea, or the targeting of information technology in many industrialized countries. Once activities and clusters are identified, governments have to set priorities between competing uses, taking into account complex feedbacks and linkages. 7. Industrial policy needs are not discovered by computing a substitute for a perfectly competitive equilibrium from an innumerable number of shadow prices. They are built up in a more mundane manner by looking for ways to build on the existing base of technical and other capabilities and to push them to exploit future opportunities offered by markets, technologies, externalities and international value chains. Industrial policy in the Asian Tigers did not rely on collecting vast amounts of information to calculate the optimal set of activities. It did, however, involve, considering in detail available technological and market information and the experience of more industrialized countries. But beyond this, the process was akin to creating rather than picking winners: the governments acted rather like venture capitalists. And, despite mistakes, on average they achieved good results in their strategies (Stiglitz 1996). 8. Teubal (1996, 1997).

Notes
For further details on sources, information and the literature on subjects covered here, see the background papers. 1. Lall and Teubal (1998). 2. See Cheng and others (1999) and Evans (1999) for analyses of this process. The Tigers set up specific institutions to manage industrialization strategy (such as the Economic Development Board in Singapore, the Economic Development Bureau in Taiwan Province of China and the Economic Planning Board in the Republic of Korea). They reformed traditional bureaucratic structures, focusing on the few critical ministries responsible for industrial policy. They had frequent and active interaction with private companies and associations. 3. World Bank (1993); Evans (1999).

9. For extensive discussion along these lines, see Sercovich and others (1999). 10. Martin (1996). 11. Adelman and Westphal (1979). 12. Lall (2001b); Rodrik (2001). 13. See chapter 1 and McCulloch and others (2001), UNCTAD (1996), UNDP (2001), World Bank (2001a).

The way forward

143

Technical annex

UNIDO Scoreboard database

Exports
Data source: UN Commodity Trade Statistics (Comtrade) database. The technological classification of exports is based on the Standard International Trade Classification (SITC) revision 2 (table A.1). Data adjustments: Export data refer to 1985 and 1998 except for Panama and South Africa (1986 rather than 1985) and Ethiopia, Malawi, Senegal, Yemen and Zambia (1997 rather than 1998). Export data for Singapore for 1998 were scaled down by 40 percent to reflect re-exports. Data were adjusted to account for re-exports for Bolivia (high-tech products in 1998), the Central African Republic (medium-tech process engineering products in 1998), El

Salvador (high-tech electronics in 1985), Guatemala (high-tech products in 1985), Jordan (medium- and hightech products in 1985) and Mozambique, Nigeria and Uganda (all three for medium-tech products in 1985). Data for 1985 were unavailable for Albania, the Czech Republic, Romania, the Russian Federation, Slovenia and Yemen.

Manufacturing value added TOTAL MANUFACTURING VALUE ADDED

Data source: UNIDO National Accounts database. Data adjustments: Data for total manufacturing value added (MVA) refer to 1985 and 1998 and are based on national accounts statistics from the United Nations Statistics Division, supplemented by national statistics. Missing values were now cast using the best econometric model.

Table A.1

Technological classification of exports according to SITC revision 2

SECTORAL MANUFACTURING VALUE ADDED

Data source: UNIDO Industrial Statistics database.

Type of exports Resource based

Low tech

Medium tech

High tech

SITC sections, divisions or groups 01 (excl. 011), 023, 024, 035, 037, 046, 047, 048, 056, 058, 06, 073, 098, 1 (excl.121), 233, 247, 248, 25, 264, 265, 269, 323, 334, 335, 4, 51, 512 (excl. 512 and 513), 52 (excl. 524), 53 (excl. 533), 551, 592, 62, 63, 641, 66 (excl. 665 and 666), 68 61, 642, 65 (excl. 653), 665, 666, 67 (excl. 671, 672 and 678), 69, 82, 83, 84, 85, 89 (excl. 892 and 896) 266, 267, 512, 513, 533, 55 (excl. 551), 56, 57, 58, 59 (excl. 592), 653, 671, 672, 678, 711, 713, 714, 72, 73, 74, 762, 763, 772, 773, 775, 78, 79 (excl. 792), 81, 872, 873, 88 (excl. 881), 95 524, 54, 712, 716, 718, 75, 761, 764, 77 (excl. 772, 773 and 775), 792, 871, 874, 881

Data adjustments: Because only some of the sample economies report industrial statistics according to the International Standard Industrial Classification of All Economic Activities, Third Revision (ISIC revision 3), data reported according to ISIC revision 3 were converted to ISIC revision 2. To fill in missing values, the ISIC revision 2 series was supplemented with the ISIC revision 3 series. The data were now cast to 1998 using the best econometric model. The data were then aggregated using the technological classification of ISIC revision 2 (table A.2). Because reporting of data at the group (four-digit) level of ISIC is inadequate to allow separation of medium- and high-tech products, the category "high-tech manufacturing" was not

145

Table A.2

Technological classification of manufacturing value added according to ISIC revision 2

(http://www.ricyt.edu.ar); and central banks and other national statistical sources. Data adjustments: Data refer to 1985 and 19971998. Where data were unavailable for those years, values for the closest year available were used. Values for OECD countries for 19971998 were calculated based on data from OECD, Science, Technology and Industry Scoreboard 1999. Values for Latin American countries were calculated based on data from the Iberoamerican Network of Science and Technology. Data for 1985 were unavailable for Albania, Bahrain, the Czech Republic, the Russian Federation and Slovenia. Many countries, particularly in Sub-Saharan Africa, do not report data on R&D financed by productive enterprises. Because of the weak industrial structures of these countries, R&D per capita was assumed to be negligible.

Type of manufacturing Resource based Low tech Medium and high tech High tech

ISIC divisions, major groups or groups 31, 331, 341, 353, 354, 355, 362, 369 32, 332, 361, 381, 390 342, 351, 352, 356, 37, 38 (excl. 381) 3522, 3852, 3832, 3845, 3849, 385

used; instead, medium- and high-tech products were combined in one category. The sectoral shares of value added were then calculated in relation to the total for all manufacturing sectors. Data on MVA by technological classification refer to 1985 and 1998 except for the Central African Republic (data for resource-based MVA refer to 1993 rather than 1998), the Czech Republic and Nigeria (data for lowtech MVA refer to 1995 rather than 1998), Jamaica (data for resource-based MVA refer to 1996 rather than 1998), Jordan (data for resource-based MVA refer to 1997 rather than 1998), Madagascar (data for mediumand high-tech and low-tech MVA refer to 1993 rather than 1998), Mauritius (data for medium- and high-tech and resource-based MVA refer to 1997 rather than 1998), Mexico (data for medium- and high-tech MVA refer to 1994 rather than 1998), Pakistan (data for lowtech MVA refer to 1996 rather than 1998), Saudi Arabia (data refer to 1989 and 1997) and Zimbabwe (data for medium- and high-tech MVA refer to 1995 rather than 1998). Data for 1985 were unavailable for Albania, the Czech Republic, Ethiopia, Mozambique, Romania, the Russian Federation, Slovenia and Yemen. Note: Because of differences in compilation methods and statistical definitions, the figures for sectoral value added from the Industrial Statistics database do not sum to the manufacturing value added reported in the national accounts data.

Foreign direct investment inflows

Data sources: World Bank, World Development Indicators 2000; UNCTAD, World Investment Report 1995 and World Investment Report 1999; and national statistical sources. Data adjustments: Data refer to average annual inflows of foreign direct investment during 19811985 and 19931997. Data for 1998 for Bahrain, Belgium, Saudi Arabia and South Africa are from UNCTAD, World Investment Report 1999. Data from that source may refer to periods that do not correspond exactly with 19811985 and 19931997. Data for Taiwan Province of China are from Taiwan Province of China, Council for Economic Planning and Development, Taiwan Statistical Data Book 1998. Data for 1985 were unavailable for Albania, Bahrain, the Czech Republic, Hungary, Nicaragua, Romania, the Russian Federation and Slovenia.

Research and development financed by productive enterprises Technology licensing payments

Data sources: Calculated on the basis of data from UNESCO, Statistical Yearbook 1994 and Statistical Yearbook 1998; OECD, Science, Technology and Industry Scoreboard 1999; Iberoamerican Network of Science and Technology Indicators Data sources: World Bank, World Development Indicators 2000; central banks; and International Monetary Fund, Balance of Payments Statistics Yearbook 1999.

146

Industrial Development Report 2002/2003

Data adjustments: Data refer to 1985 and 1998. Where data were unavailable for those years, values for the closest year available were used. Data for 1985 were unavailable for Japan; data for the closest year available (1984) were used instead. Countries for which data for 1998 were unavailable and data for the closest year available were used instead are Albania (1994), Algeria (1991), Bahrain (1995), Cameroon (1995), the Central African Republic (1992), Greece (1997), Guatemala (1993), Jordan (1994), Malawi (1994), Mozambique (1992), Pakistan (1997), Senegal (1997), Sri Lanka (1995), Uganda (1997) and Zimbabwe (1994). Balance of payments data from the International Monetary Funds Balance of Payments Statistics Yearbook 1999 and national central bank reports were used to calculate licensing payments for Denmark, Hong Kong SAR, Switzerland, Taiwan Province of China and Turkey. For countries that do not report technology licensing payments in their balance of payments (Indonesia, Malaysia and Singapore), a proxy value was calculated based on the ratio of licensing payments to payments for other services for similar economies. For Malaysia and Singapore royalty payments were assumed to be 25 percent of other services (a ratio similar to that for Taiwan Province of China); for Indonesia they were assumed to be 11 percent (the same ratio as that for Thailand). For countries reporting data for 1998 but not 1985, the ratio of royalty payments to other services in 1998 was applied to 1985. Data on payments for other services are from the International Monetary Funds Balance of Payments Statistics Yearbook 1999. Data for 1985 were unavailable for Albania, the Czech Republic, Hungary, Romania, the Russian Federation, Slovenia and Yemen.

Data for the Harbison-Myers index in 1985 were unavailable for Albania, Bahrain, the Russian Federation and Slovenia. Data for tertiary technical enrolments were unavailable for Albania, Bahrain, the Czech Republic, the Russian Federation, Slovenia and Yemen.

Infrastructure
Data sources: Calculated based on data from World Bank, World Development Indicators 2001; OECD, Science, Technology and Industry Statistics (http://www.oecd.org/ statistics); Telecordia Technologies (http://www.netsizer.com); and African Internet Connectivity (http://www.sn.apc.org). Data adjustments: Data refer to 1985 and 1998. Where data were unavailable for 1998, values for the closest year available were used. Countries for which data for telephone mainlines in 1998 were unavailable and data for the closest year available were used instead are Cameroon (1997), Guatemala (1997), Jamaica (1997), Kenya (1997), Yemen (1997) and Zimbabwe (1997). Countries for which data for mobile phones in 1998 were unavailable and data for the closest year available were used instead are Cameroon (1997), Ghana (1997), Jamaica (1996) and Kenya (1997). Countries for which data for computers in 1998 were unavailable and data for the closest year available were used instead are Algeria (1997), Cameroon (1995), Ghana (1997), Jordan (1997), Kenya (1997), Madagascar (1997), Morocco (1997), Mozambique (1997), Nigeria (1997), Senegal (1997), Sri Lanka (1997), the United Republic of Tanzania (1997), Uganda (1997), Yemen (1997) and Zimbabwe (1997). Data for commercial energy use in 1985 were unavailable for Albania, Bahrain, the Russian Federation, Slovenia and Yemen. Data on Internet hosts refer to 2001 and are from Telecordia Technologies. Data on information and communication technology for Africa not available in the World Banks World Development Indicators 2001 are from African Internet Connectivity.

Skills
Data sources: UNESCO, Statistical Yearbook 1994 and Statistical Yearbook 1998; World Bank, World Development Indicators 2000; and national statistical sources. Data adjustments: Data refer to 1985 and 19971998 (latest year available). Where data were unavailable for those years, values for the closest year available were used.

Technical annex

147

 Data for Taiwan Province of China are from Taiwan Province of China, Council for Economic Planning and Development, Taiwan Statistical Data Book 1998.

where Ij,i represents the ith value of the four individual indices, wn the weights given to the indices and a parameter to control how the variations and weights in the individual indices affect the CIP index. Initially, a different weight wj was assigned to each performance indicator Ij,i. Stability tests confirmed that the weights did not significantly affect ranks, however, so equal weights were allocated to the four performance indicators. With w1 = w2 = w3 = w4 = 1, the general formula then became the following:

Industrial Performance Scoreboard

UNIDOs Industrial Performance Scoreboard was developed in four stages. In the first stage a database of industrial indicators (both output and input factors) for 1985 and 1998 was created for as many countries as possible. Indicators were chosen on the basis of the availability of cross-country data. Four performance indicatorsMVA per capita, manufactured exports per capita, the share of medium- and high-tech activities in manufacturing production and the share of mediumand high-tech products in manufactured exportswere chosen for the competitive industrial performance (CIP) index. In the second stage individual indices of performance Ij,i were standardized according to the general formula

To further simplify, = 1 was chosen, and the result is the simple arithmetic mean of I1,i, I2,i, I3,i and I4,i. Thus,

Cluster analysis
where Xj,i is the i th country value of the j th performance variable. Therefore the highest country in the ranking has a score of 1 and the lowest a score of 0. The third stage consisted of testing the feasibility of computing a composite index based on the four performance indicators selected. Positive and statistically significant correlations between the four performance variables confirmed that a composite index could be constructed as a proxy for overall industrial performance. The CIP index was constructed using the standardized values of the four performance indicators, according to this general formula: Cluster analysis is a statistical technique for identifying relatively homogeneous groups of cases according to their quantitative features. The version used for the report is Kmeans cluster analysis, which is used to cluster large numbers of observations, with squared Euclidean distance (the sum of the squared differences over all the variables) employed to identify a specified number of clusters. The algorithm used for determining the membership of clusters is based on nearest centroid sorting. The values obtained for each cluster are simply the standardized average values of the variables for cases in the clusters. However, data presented in the report have been de-standardized to show averages of real values.

148

Industrial Development Report 2002/2003

Statistical annex

Table A2.1

Manufacturing value added by income level and region, 1985 and 1998

1985 Developing Country group, Value World economies Per income level (billions of shares shares capita or region dollars) (percent) (percent) (dollars) World 2,480.0 100 na 619 Industrialized economies 2,003.3 80.8 na 2,579 Transition economies .. .. na .. Developing economies 476.6 19.2 100 147 High and upper-middle income 222.9 9.0 46.8 578 Lower-middle income 92.4 3.7 19.4 176 Low income 161.3 6.5 33.8 70 Low income (without China and India) 22.2 0.9 4.7 44 Least developed countriesb 5.6 0.2 0.7 31 East Asia 203.7 8.2 42.7 145 East Asia (without China) 98.0 3.9 20.6 278 South Asia 42.0 1.7 8.8 42 Latin America and the Caribbean 171.1 6.9 35.9 462 Latin America and the Caribbean (without Mexico) 133.9 5.4 28.1 454 Sub-Saharan Africa 24.1 1.0 5.1 83 Sub-Saharan Africa (without South Africa) 12.7 0.5 2.7 49 Middle East and North Africa and Turkey 35.8 1.4 7.5 202

1998a Developing Value World economies (billions of shares shares dollars) (percent) (percent) 5,636.1 100 na 4,240.8 75.2 na 169.5 3.0 na 1,225.8 21.7 100 560.2 210.4 455.2 35.8 12.1 649.8 294.3 83.6 360.0 9.9 3.7 8.1 0.6 0.2 11.5 5.2 1.5 6.4 45.7 17.2 37.1 2.9 0.6 53.0 24.0 6.8 29.4

Per capita (dollars) 1,094 5,040 725 300 1,161 311 156 51 35 387 668 65 771

Growth rate 19851998 Per Total capita 6.5 4.5 5.9 5.3 .. .. 7.5 5.6 7.3 6.5 8.3 3.8 6.2 9.3 8.8 5.5 5.9 5.5 4.5 6.4 1.2 1.2 7.8 7.0 3.4 4.0

278.1 38.2 15.1 94.1

4.9 0.7 0.3 1.7

22.7 3.1 1.2 7.7

750 92 40 392

5.8 3.6 1.4 7.7

3.9 0.8 1.4 5.3

Source: UNIDO Scoreboard database (see technical annex). Note: Data on manufacturing value added cover only the 87 economies in the Scoreboard sample: they are expected to account for a very high proportion of the world total. a. The 1998 data reflect sharp drops in production in many economies, particularly in East Asia. b. Includes only 12 of 49 least developed countries.

149

Table A2.2

Manufactured exports by income level and region, 1985 and 1998

1985 Developing Country group, Value World economies Per income level (billions of shares shares capita or region dollars) (percent) (percent) (dollars) World 1,239.2 100 na 292.5 Industrialized economies 1,045.0 84.3 na 1,345.2 Transition economies .. .. na .. Developing economies 194 15.7 100 60.2 High and upper-middle income 143.0 11.5 73.6 371.0 Lower-middle income 33.8 2.7 17.4 64.2 Low income 17.5 1.4 9.0 7.6 Low income (without China and India) 5.3 0.4 2.7 10.5 Least developed countriesa 1.4 0.1 0.7 7.0 East Asia 118 9.5 60.6 84 East Asia (without China) 112 9.0 57.5 317 South Asia 9 0.8 4.9 9 Latin America and the Caribbean 43 3.5 22.1 116 Latin America and the Caribbean (without Mexico) 35 2.8 17.8 117 Sub-Saharan Africa 7 0.6 3.7 25 Sub-Saharan Africa (without South Africa) 2 0.2 1.1 8 Middle East and North Africa and Turkey 17 1.4 8.8 96

1998 Developing Value World economies (billions of shares shares dollars) (percent) (percent) 4,230.0 100 na 3,125.5 73.9 na 117.1 2.8 na 987.4 23.3 100 614.5 159.8 213.2 19.7 6.0 686 519 41 188 14.5 3.8 5.0 0.5 0.1 16.0 12.1 1.0 4.4 62.2 16.2 21.6 2.0 0.6 65.9 49.8 4.0 18.1

Per capita (dollars) 821.0 3,714.4 500.7 242.2 1,273.5 236.2 73.1 28.1 17.5 409 1,178 32 404

Growth rate 19851998 Per Total capita 9.9 8.3 8.8 8.1 .. .. 13.3 11.3 11.9 12.7 21.2 10.7 12.0 14.5 12.5 12.4 12.0 10.0 10.5 19.1 7.9 7.3 12.9 10.6 10.2 10.1

85 19 5 53

2.0 0.4 0.1 1.2

8.1 1.8 0.5 5.1

229 45 14 220

7.1 8.0 7.3 9.1

5.3 4.6 4.4 6.6

Source: UNIDO Scoreboard database (see technical annex). Note: Export data are for all economies in the world, not only the 87 economies in the Scoreboard. a. Includes only 12 of 49 least developed countries.

150

Industrial Development Report 2002/2003

Table A2.3

Technological structure of industrial activity by income level and region, 1985 and 1998 (percent)

Manufacturing value added Change in medium and high tech, 1985 to 1998 Medium and (percentage high tech point) 58.7 1.9 61.2 1.9 42.2 .. 48.7 6.2 53.4 7.8 35.4 9.3 49.9 3.0 25.4 1.0 24.0 1.1 54.4 10.1 56.2 14.9 52.7 2.9 39.7 2.8 43.7 37.6 24.2 36.8 0.2 1.0 2.3 6.1

Country group, income level Resource or region based World 27.1 Industrialized economies 25.5 Transition economies .. Developing economies 37.1 High and upper-middle income 33.9 Lower-middle income 54.0 Low income 32.5 Low income (without China and India) 52.4 Least developed countriesa 52.1 East Asia 31.9 East Asia (without China) 33.1 South Asia 30.3 Latin America and the Caribbean 39.6 Latin America and the Caribbean (without Mexico) 38.7 Sub-Saharan Africa 42.7 Sub-Saharan Africa (without South Africa) 51.8 Middle East and North Africa and Turkey 48.6

1985 Low tech 16.2 15.3 .. 20.4 20.5 19.9 20.6 21.3 25.0 23.8 25.6 19.9 18.0 17.8 18.7 21.6 20.7

Medium and high tech 56.8 59.3 .. 42.5 45.6 26.1 46.9 26.4 22.9 44.3 41.3 49.8 42.5 43.5 38.6 26.5 30.7

Resource based 27.1 25.5 45.1 33.7 30.5 43.9 31.7 47.6 44.4 28.0 26.4 27.6 44.6 42.2 43.6 55.3 41.4

1998 Low tech 14.1 13.3 12.7 17.6 16.1 20.7 18.4 27.0 31.6 17.6 17.4 19.7 15.7 14.2 18.8 20.5 21.8

Manufactured exports Change in medium and high tech, 1985 to 1998 Medium (percentand high age tech point) 63.8 6.1 67.8 4.8 46.9 .. 53.8 20.3 64.9 42.7 35.0 9.2 3.4 59.7 66.2 15.8 56.9 24.4 28.4 20.9 0.9 1.1 20.6 25.9 3.8 25.1

1985 Country group, income level Resource or region based World 23.7 Industrialized economies 21.0 Transition economies .. Developing economies 34.1 High and upper-middle income 26.8 Lower-middle income 62.3 Low income 39.1 Low income (without China and India) 37.8 Least developed countriesa 39.6 East Asia 22.7 East Asia (without China) 21.9 South Asia 32.3 Latin America and the Caribbean 51.3 Latin America and the Caribbean (without Mexico) 58.6 Sub-Saharan Africa 57.9 Sub-Saharan Africa (without South Africa) 67.8 Middle East and North Africa and Turkey 59.9 Low tech 18.6 16.1 .. 32.2 32.7 23.4 46.8 52.0 55.9 38.2 37.9 55.8 16.9 Medium tech 40.9 44.7 .. 21.9 25.7 12.0 10.6 9.3 3.8 23.3 23.9 9.2 24.8 High tech 16.8 18.2 .. 11.6 14.8 2.3 3.5 0.8 0.7 15.8 16.4 2.8 7.0 Medium and high tech 57.7 62.9 .. 33.5 40.5 14.3 14.1 10.1 4.5 39.1 40.3 12.0 31.8 Resource based 17.4 16.8 26.4 18.2 15.8 27.3 13.1 17.8 12.6 12.1 12.8 21.4 24.9 Low tech 18.8 15.5 26.7 28.0 19.3 30.1 51.9 72.9 84.0 28.1 21.1 62.8 18.2

1998 Medium tech 38.7 43.3 34.9 25.6 30.2 17.1 18.4 8.2 3.1 23.6 24.8 11.4 37.2 High tech 25.1 24.5 12.0 28.2 34.8 25.6 16.6 1.0 0.3 36.1 41.4 4.4 19.7

17.7 17.3 19.2 24.5

20.4 18.2 11.6 14.1

3.3 6.6 1.4 1.5

23.7 24.8 13.0 15.6

47.1 45.8 50.5 39.9

17.0 23.3 36.8 37.6

28.9 25.5 11.4 18.8

7.0 5.3 1.3 3.7

35.9 30.8 12.7 22.5

12.2 6.0 0.3 6.9

Source: UNIDO Scoreboard database (see technical annex). a. Includes only 12 of 49 least developed countries.

Statistical annex

151

Table A2.4

Ranking by concentration of manufacturing value added and exports in selected economies, 1985 and 1998

1985 World All shares Rank economies (percent) Manufacturing value added 1 United States 32.4 2 Japan 16.0 3 Germany 7.9 4 France 4.6 5 Italy 4.3 Top 5 China United Kingdom 65.2 4.3 4.0 Developing economies shares (percent) 22.2 14.4 7.8 7.0 5.7

1998 World shares (percent) 25.4 15.9 8.5 6.3 5.0 61.1 4.4 4.2 2.7 1.9 1.8 76.1 0.5 12.8 12.8 8.6 6.2 5.7 44.7 5.3 3.9 3.8 3.6 3.3 64.6 0.3 Developing economies shares (percent) 29.0 12.3 8.0 6.7 6.0 62.0 5.2 4.3 3.6 2.9 1.9 80.0 2.4 17.0 15.0

Developing economies

All Rank economies 1 2 3 4 5 United States Japan Germany China France Top 5 United Kingdom Italy Brazil Canada Spain Top 10 Bottom 30 United States Germany Japan France United Kingdom Top 5 Italy China Canada Belgium Netherlands Top 10 Bottom 30

Developing economies

China Brazil Mexico India Korea, Republic of

6 7 8 9 10

Brazil 2.8 Canada 2.4 Spain 1.8 Top 10 80.4 Bottom 30 0.8 Manufactured exports 1 Japan 13.3 2 3 4 5 6 7 8 9 10 Germany United States 11.4 12.6

Top 5 57.2 Argentina 5.5 Taiwan Province of China 4.9 Indonesia 2.9 Venezuela 2.8 Turkey 2.6 Top 10 75.9 Bottom 30 4.2 Taiwan Province of China 15.0 Korea, Republic of 12.2 Singapore 9.8 Brazil Hong Kong SAR Top 5 Malaysia Mexico Venezuela India China Top 10 Bottom 30 9.1 8.2 57.1 4.4 4.3 3.6 3.2 3.1 75.8 2.2

6 7 8 9 10

China Brazil Korea, Republic of Mexico Taiwan Province of China Top 5 India Argentina Turkey Thailand Indonesia Top 10 Bottom 30 China Korea, Republic of Taiwan Province of China Mexico Singapore Top 5 Malaysia Thailand Brazil Philippines Indonesia Top 10 Bottom 30

1 2 3 4 5 6 7 8 9 10

France 6.5 Italy 5.8 Top 5 51.0 United Kingdom 5.7 Canada 4.9 Netherlands 4.0 Belgium 3.4 Taiwan Province of China 2.3 Top 10 71.2 Bottom 30 0.3

10.7 10.5 10.5 61.1 6.7 4.5 3.9 2.8 2.7 81.6 1.3

Source: UNIDO Scoreboard database (see technical annex).

152

Industrial Development Report 2002/2003

Table A2.5

Tertiary enrolments, total and technical, by income level and region, 1987 and 19951998

1987 Country group, income level Number or region (thousands) Total enrolment Industrialized economies 26,630 Transition economies .. Developing economies 20,473 High and upper-middle income 5,998 Lower-middle Income 6,741 Low income 7,734 Low income (without China and India) 1,198 Least developed countriesb 634 East Asia 6,388 East Asia (without China) 4,323 South Asia 5,087 Latin America and the Caribbean 6,142 Latin America and the Caribbean (without Mexico) 4,897 Sub-Saharan Africa .. Sub-Saharan Africa (without South Africa) 434 Middle East and North Africa and Turkey 2,419 Technical enrolmentc World 9,323 Industrialized economies 4,508 Transition economies .. Developing economies 4,814 High and upper-middle income 1,400 Lower-middle Income 1,136 Low income 2,278 Low income (without China and India) 223 Least developed countriesd 135 East Asia 1,726 East Asia (without China) 905 South Asia 1,384 Latin America and the Caribbean 1,341 Latin America and the Caribbean (without Mexico) 965 Sub-Saharan Africa .. Sub-Saharan Africa (without South Africa) 58 Middle East and North Africa and Turkey 306 Developing World economies shares shares (percent) (percent) 56.5 .. 43.5 12.7 14.3 16.4 2.5 1.3 14.0 9.0 11.0 13.0 na na 100 29.3 32.9 37.8 5.9 3.1 31.2 21.1 24.9 30.0 Number per 1,000 population 34.3 .. 6.3 15.6 12.8 3.3 2.4 2.5 4.6 12.3 5.1 16.6

19951998a Developing World economies Number shares shares (thousands) (percent) (percent) 33,775 6,157 35,346 8,849 12,443 14,053 2,644 785 15,007 9,181 6,545 7,677 44.9 8.2 47.0 11.8 16.5 18.7 3.5 1.0 21.1 12.9 9.2 10.8 na na 100 25.0 35.2 39.8 7.5 2.2 42.5 26.0 18.5 21.7

Number per 1,000 population 40.1 26.3 8.7 18.3 18.4 4.8 3.8 2.3 9.2 21.9 5.4 17.3

Growth rate (percent) 1.8 .. 4.3 3.0 4.8 4.7 6.3 1.7 6.8 6.0 2.0 1.7

10.0 .. 1.0 5.0 100 48.4 .. 51.6 15.0 12.2 24.4 2.4 1.4 18.5 9.7 14.9 14.4

23.9 .. 2.1 11.8 na na na 100 29.1 23.6 47.3 4.6 2.8 35.9 18.8 28.8 27.8

16.6 .. 1.7 13.6 2.2 5.8 .. 1.5 3.6 2.2 1.0 0.4 0.5 1.2 2.6 1.4 3.6

6,257 1,542 924 4,571 14,611 5,850 2,090 6,670 2,100 1,937 2,633 325 138 3,198 1,977 1,271 1,497

8.8 2.2 1.3 6.4 100 40.0 14.3 45.7 14.4 13.3 18.0 2.2 0.9 21.5 13.3 8.6 10.1

17.7 4.4 2.6 12.9 na na na 100 31.5 29.0 39.5 4.9 2.1 46.3 28.6 18.4 21.7

17.7 4.0 2.7 20.5 2.8 7.0 8.9 1.6 4.4 2.9 0.9 0.5 0.4 2.0 4.7 1.0 3.4

1.9 .. 6.0 5.0 3.5 2.0 .. 2.5 3.2 4.2 1.1 3.0 0.1 4.9 6.2 0.7 0.9

10.3 .. 0.6 3.3

20.0 .. 1.2 6.3

3.3 .. 0.2 1.7

1,097 185 117 519

7.4 1.2 0.8 3.5

15.9 2.7 1.7 7.5

3.1 0.5 0.3 2.3

1.0 .. 5.5 4.1

Source: UNIDO Scoreboard database (see technical annex). a. Period for which the most recent enrolment data could be obtained (data were not always available for 1998). b. Includes only 12 of 49 least developed countries. Bangladesh accounts for 64.8 percent of enrolment in this group in 1985 and 58.7 percent in 1998. c. Covers engineering, mathematics, computer science and natural sciences. d. Includes only 12 of 49 least developed countries. Bangladesh accounts for 72.4 percent of enrolments in this group in 1985 and 65.3 percent in 1998.

Statistical annex

153

Table A2.6

Ranking by concentration of tertiary enrolments, total and technical, in selected economies, 1987 and 19951998

1987 World shares (percent) 29.1 9.5 5.3 4.4 3.9 52.2 3.5 3.2 3.1 2.6 2.6 67.2 0.5 19.6 13.2 8.8 5.4 4.9 51.8 4.0 3.4 3.3 3.3 2.9 68.8 0.9 Developing economies shares (percent) 21.8 10.1 7.4 7.2 6.1 52.6 5.9 4.8 3.7 2.9 2.6 72.4 1.1 25.6 17.1 7.8 6.7 5.6 62.8 4.7 4.4 2.9

19951998a World shares (percent) 19.2 7.7 7.4 5.9 5.2 45.4 3.1 3.0 2.8 2.8 2.7 59.7 0.6 12.3 12.0 8.4 7.4 5.5 45.6 5.1 4.3 3.0 3.0 2.7 63.7 1.3 Developing economies shares (percent) 16.5 15.8 6.5 6.3 5.2 50.3 4.9 4.0 3.5 3.3 3.0 69.0 2.1 18.3 16.3 11.1 6.6 6.0 58.3 5.8 4.3 3.4 3.0 2.4 77.2 2.8

All Rank economies Total enrolment 1 United States 2 India 3 Japan 4 China 5 France Top 5 6 Germany 7 Korea, Republic of 8 Brazil 9 Mexico 10 Italy Top 10 Bottom 30 Technical enrolment 1 United States 2 India 3 China 4 Japan 5 Germany Top 5 6 Mexico 7 Korea, Republic of 8 France 9 10 United Kingdom Philippines Top 10 Bottom 30

Developing economies

All Rank economies 1 2 3 4 5 6 7 8 9 10

Developing economies

India China Korea, Republic of Brazil Mexico Top 5 Philippines Indonesia Argentina Egypt Turkey Top 10 Bottom 30 India China Mexico Korea, Republic of Philippines Top 5 Brazil Argentina Indonesia

United States China India Russian Federation Japan Top 5 Indonesia Korea, Republic of Germany France Canada Top 10 Bottom 30 United States Russian Federation China India Japan Top 5 Korea, Republic of Germany United Kingdom Indonesia Mexico Top 10 Bottom 30

China India Indonesia Korea, Republic of Philippines Top 5 Brazil Mexico Thailand Turkey Argentina Top 10 Bottom 30 China India Korea, Republic of Indonesia Mexico Top 5 Philippines Brazil Taiwan Province of China Colombia Argentina Top 10 Bottom 30

1 2 3 4 5 6 7 8 9 10

Colombia 2.4 Taiwan Province of China 2.4 Top 10 79.5 Bottom 30 1.7

Source: UNIDO Scoreboard database (see technical annex). a. Period for which the most recent enrolment data could be obtained (data were not always available for 1998).

154

Industrial Development Report 2002/2003

Table A2.7

R&D financed by enterprises by income level and region, 1985 and 19951998

19951998a As a As a share of share of Growth manumanurate Value Developing facturing Value Developing facturing 1985 to Country group, (millions World economies Per value (millions World economies Per value 1995 income level of shares shares capita added of shares shares capita added 1998 or region dollars) (percent) (percent) (dollars) (percent) dollars) (percent) (percent) (dollars) (percent) (percent) World 97,133.6 100 na 22.9 3.9 353,288.9 100 na 71.4 6.3 10.4 Industrialized economies 95,034.1 97.8 na 122.3 4.7 333,088.6 94.3 na 402.4 7.9 10.1 Transition economies -.. na .. .. 2,078.7 0.6 na 8.8 1.6 .. Developing economies 2,099.6 2.2 100 0.6 0.4 18,121.7 5.1 100 4.6 1.5 18.0 High and upper-middle income 1,698 1.7 80.9 4.4 .. 16,057 4.5 88.6 33.3 .. 18.9 Lower-middle income 98 0.1 4.7 0.2 .. 569 0.2 3.1 0.8 .. 14.5 Low income 303 0.3 14.5 0.1 .. 1,496 0.4 8.3 0.5 .. 13.1 Low income (without China and India) 1 17.7 Least developed countriesb East Asia .. .. .. .. .. 14,125.8 4.0 77.9 8.7 2.2 .. East Asia (without China) 1,115.1 1.1 53.1 3.2 1.1 13,028.4 3.7 71.9 31.0 4.4 20.8 South Asia 303.3 0.3 14.4 0.3 0.7 397.6 0.1 2.2 0.3 0.5 2.1 Latin America and the Caribbean 423.1 0.4 20.2 1.1 0.2 2,783.7 0.8 15.4 6.3 0.8 15.6 Latin America and the Caribbean (without Mexico) 170.9 0.2 8.1 0.6 0.1 2,647.2 0.7 14.6 7.5 1.0 23.5 Sub-Saharan Africa 183.7 0.2 8.7 0.6 0.7 501.2 0.1 2.8 1.3 1.3 8.0 Sub-Saharan Africa (without South Africa) 0.1 0.4 11.3 Middle East and North Africa and Turkey 74.3 0.1 3.5 0.4 0.2 313.3 0.1 1.7 1.4 0.3 11.7 1985
Source: UNIDO Scoreboard database (see technical annex). a. Period for which the most recent R&D data could be obtained (data were not always available for 1998). b. Includes only 12 of 49 least developed countries.

Statistical annex

155

Table A2.8

Ranking by concentration of R&D financed by enterprises in selected economies, 1985 and 19951998

1985 World shares (percent) 37.6 23.9 20.4 3.6 3.1 88.6 1.8 1.4 1.2 1.1 1.0 95.0 Developing economies shares (percent)

19951998a World shares (percent) 34.7 30.5 Developing economies Developing shares economies (percent) Korea, Republic of 52.4 Taiwan Province of China Brazil China Singapore Top 5 South Africa India Argentina Turkey Indonesia Top 10 Bottom 30

All Rank economies 1 United States 2 3 4 5 6 7 8 9 10 Japan Germany France United Kingdom Top 5 Canada Switzerland Netherlands Italy Sweden Top 10 Bottom 30

Developing economies Taiwan Province of China 29.3 Korea, Republic of 21.0 India Mexico South Africa Top 5 Brazil Turkey Singapore Chile Indonesia Top 10 Bottom 30

All Rank economies 1 United States 2 3 4 5 6 7 8 9 10 Japan

14.4 12.0 8.7 85.6 7.0 3.2 1.7 0.9 0.4 98.8

Germany 9.7 France 4.9 United Kingdom 2.9 Top 5 82.6 Korea, Republic of 2.7 Switzerland 1.7 Sweden 1.6 Italy 1.5 Canada 1.2 Top 10 91.3 Bottom 30

14.4 12.1 6.1 3.3 88.2 2.8 2.2 1.6 1.6 0.8 97.2

Source: UNIDO Scoreboard database (see technical annex). a. Period for which the most recent R&D data could be obtained (data were not always available for 1998).

156

Industrial Development Report 2002/2003

Table A2.9

Foreign direct investment inflows by income level and region, 19811985 and 19931998

Average Country group, valuea income level (millions or region of dollars) World 56,375.4 Industrialized economies 42,541.8 Transition economies .. Developing economies 13,833.6 High and uppermiddle income 9,676.4 Lower-middle Income 2,505.2 Low income 1,652.0 Low income (without China and India) 657.7 Least developed countriesb 41.1 East Asia 6,038.5 East Asia (without China) 5,104.3 South Asia 196.6 Latin America and the Caribbean 4,091.1 Latin America and the Caribbean (without Mexico) 3,194.2 Sub-Saharan Africa 508.7 Sub-Saharan Africa (without South Africa) 501.7 Middle East and North Africa and Turkey 2,998.7

19811985 Developing World economies shares shares (percent) (percent) 100.0 na 75.5 na .. na 24.5 100.0 17.2 4.4 2.9 1.2 0.1 10.7 9.1 0.3 7.3 69.9 18.1 11.9 4.8 0.3 43.7 36.9 1.4 29.6

Per capita (dollars) 13.3 54.8 .. 4.3 25.1 4.8 0.7 1.3 0.2 4.3 14.5 0.2 11.1

Average valuea (millions of dollars) 314,045.6 199,982.5 9,597.5 104,465.6 43,785.2 18,280.0 42,400.4 2,945.5 547.6 64,377.9 26,565.0 2,522.9 31,291.1

19931998 Developing World economies shares shares (percent) (percent) 100.0 na 63.7 na 3.1 na 33.3 100.0 13.9 5.8 13.5 0.9 0.2 20.5 8.5 0.8 10.0 41.9 17.5 40.6 2.8 0.5 61.6 25.4 2.4 29.9

Per capita (dollars) 63.4 241.6 40.8 26.9 95.0 28.5 15.2 4.5 1.6 39.7 63.3 2.1 70.4

Growth rate (percent) 15.4 13.8 .. 18.4 13.4 18.0 31.1 13.3 22.0 21.8 14.7 23.7 18.5

5.7 0.9 0.9 5.3

23.1 3.7 3.6 21.7

10.8 1.7 1.9 16.9

24,485.0 3,155.0 1,822.4 3,196.6

7.8 1.0 0.6 1.0

23.4 3.0 1.7 3.1

69.3 8.2 5.3 14.1

18.5 16.4 11.3 0.5

Source: UNIDO Scoreboard database (see technical annex). a. Annual averages calculated for the periods 19811985 and 19931998. b. Includes only 12 of 49 least developed countries.

Statistical annex

157

Table A2.10

Technology licence payments abroad by income level and region, 1985 and 1998

Country group, Value income level (billions or region of dollars) World 11,091.8 Industrialized economies 9,286.9 Transition economies .. Developing economies 1,804.9 High and upper-middle income 1,230.1 Lower-middle income 539.9 Low income 34.9 Low income (without China and India) 9.7 Least developed countriesb 0.2 East Asia .. East Asia (without China) 942.3 South Asia 25.1 Latin America and the Caribbean 696.9 Latin America and the Caribbean (without Mexico) 554.9 Sub-Saharan Africa 127.8 Sub-Saharan Africa (without South Africa) 7.5 Middle East and North Africa and Turkey 12.8

1985a Developing World economies shares shares (percent) (percent) 100 na 83.7 na .. na 16.3 100 11.1 4.9 0.3 0.1 .. 8.5 0.2 6.3 68.2 29.9 1.9 0.5 .. 52.2 1.4 38.6

Per capita value (dollars) 2.6 12 .. 0.6 3.2 1.0 .. 2.7 0 1.9

Value (billions of dollars) 70,471.0 54,825.4 583.8 15,061.8 11,409.7 2,937.8 714.3 93.5 21.8 11,568.3 11,148.3 225.6 2,348.8

1998a Developing World economies shares shares (percent) (percent) 100 na 77.8 na 0.8 na 21.4 100 16.2 4.2 1.0 0.1 16.4 15.8 0.3 3.3 75.8 19.5 4.7 0.6 0.1 76.8 74.0 1.5 15.6

Per capita value (dollars) 14.2 66.2 2.5 3.9 23.6 4.3 0.2 0.1 0.1 7.1 26.6 0.2 5.3

Growth rate (percent) 16.7 15.9 .. 19.3 18.7 13.9 26.1 19.0 42.3 .. 22.9 20.1 10.7

5.0 1.2 0.1 0.1

30.7 7.1 0.4 0.7

1.9 0.4 0.1

1,847.8 229.0 63.6 690.1

2.6 0.3 0.1 1.0

12.3 1.5 0.4 4.6

5.2 0.6 0.2 3.0

10.5 5.0 19.5 39.4

Source: UNIDO Scoreboard database (see technical annex). a. When data for 1985 or 1998 were not available, data for closest years were used. b. Includes only 12 of 49 least developed countries.

158

Industrial Development Report 2002/2003

Table A2.11

Ranking by concentration in technology licence payments abroad in selected economies, 1985 and 1998

1985 World shares (percent) 20.5 11.0 10.5 8.9 7.3 58.1 6.6 3.8 3.8 3.5 3.5 79.3 Developing economies Developing shares economies (percent) Indonesia 21.3 Korea, Republic of 17.9 Taiwan Province of China 9.5 Mexico 7.9 South Africa Top 5 Thailand Ecuador Brazil India Chile Top 10 Bottom 30 6.7 63.3 2.5 2.3 1.7 1.4 1.3 72.5

1998 World shares (percent) 16.0 12.7 8.8 8.7 Developing economies Developing shares economies (percent) Malaysia 15.9 Korea, Republic of 15.7 Singapore 11.7 Taiwan Province of China Hong Kong SAR Top 5 Brazil Indonesia Thailand Mexico Argentina Top 10 Bottom 30

Rank 1 2 3 4 5 6 7 8 9 10

All economies Japan Germany United States France United Kingdom Top 5 Netherlands Argentina Belgium Australia Indonesia Top 10 Bottom 30

All Rank economies 1 United States 2 Japan 3 Ireland 4 5 6 7 8 9 10 United Kingdom

Germany 6.9 Top 5 53.2 Netherlands 4.2 France 3.9 Malaysia 3.4 Korea, Republic of 3.4 Canada 2.9 Top 10 71.0 Bottom 30 0.1

9.4 8.2 61.0 7.1 6.7 5.3 3.3 2.8 86.2 0.3

Source: UNIDO Scoreboard database (see technical annex).

Statistical annex

159

Table A2.12

Information and communication technologies infrastructure by income level and region, 1998 and 2001

Country group, World income level Number shares or region (thousands) (percent) Personal computers, 1998 World 334.3 100 Industrialized economies 266.3 79.6 Transition economies 10.0 3.0 Developing economies 58.1 17.4 High and upper-middle income 32.6 9.7 Lower-middle Income 9.9 3.0 Low income 15.6 4.7 Low income (without China and India) 2.0 0.6 Least developed countriesa 0.3 0.1 East Asia 32.4 9.7 East Asia (without China) 21.4 6.4 South Asia 3.3 1.0 Latin America and the Caribbean 15.5 4.6 Latin America and the Caribbean (without Mexico) 11.0 3.3 Sub-Saharan Africa 3.2 1.0 Sub-Saharan Africa (without South Africa) 1.3 0.4 Middle East and North Africa and Turkey 3.6 1.1 Telephone mainlines, 1998 World 785.4 100 Industrialized economies 480.5 61.2 Transition economies 49.3 6.3 Developing economies 255.6 32.5 High and upper-middle income 88.6 11.3 Lower-middle Income 53.1 6.8 Low income 113.9 14.5 Low income (without China and India) 6.0 0.8 Least developed countriesa 1.5 0.2 East Asia 138.8 17.7 East Asia (without China) 52.5 6.7 South Asia 25.2 3.2 Latin America and the Caribbean 57.1 7.3 Latin America and the Caribbean (without Mexico) 47.2 6.0 Sub-Saharan Africa 6.9 0.9 Sub-Saharan Africa (without South Africa) 2.1 0.3 Middle East and North Africa and Turkey 27.6 3.5
Source: UNIDO Scoreboard database (see technical annex). a. Includes only 12 of 49 least developed countries.

Developing economies shares Per 1,000 (percent) people na na na 100 56.1 17.0 26.9 3.2 0.5 55.9 36.8 5.7 26.8 19.0 5.6 2.2 6.1 na na na 100 34.7 20.8 44.6 2.4 0.6 54.3 20.6 9.9 22.3 18.5 2.7 0.8 10.8 64.9 316.5 43.4 14.2 67.5 14.6 5.4 2.7 0.8 19.3 48.6 2.6 33.3 29.8 7.8 3.4 14.8 152.5 571.1 214.0 62.6 183.6 78.5 39.0 8.8 4.4 82.7 119.3 19.7 122.3 127.2 16.5 5.7 115.0

Developing World economies Number shares shares Per 1,000 (thousands) (percent) (percent) people Internet hosts, 2001 106.1 100 na 20.6 99.7 94.0 na 118.5 1.1 1.0 na 4.8 5.3 5.0 100 1.3 4.6 4.3 87.0 9.4 0.5 0.5 9.1 0.7 0.2 0.2 3.9 0.1 3.0 2.9 0.1 1.6 1.2 0.3 0 2.9 2.7 0.1 1.5 1.1 0.2 0.3 57.7 55.0 1.1 31.1 22.8 5.0 0.3 5.1 na na na 100 52.2 17.2 30.6 0.9 0.3 64.1 35.8 2.0 25.4 21.3 3.2 0.3 5.4 1.8 6.6 3.5 3.2 0.6 1.1 60.6 265.8 23.9 20.4 90.0 21.1 8.7 1.1 0.6 31.8 67.7 1.3 45.2 47.9 6.3 0.8 18.7

0.3 0.3 Mobile telephones, 1998 312.3 100 223.6 71.6 5.5 1.8 83.2 26.6 43.4 13.9 14.3 4.6 25.5 8.2 1.0 0.2 53.3 29.8 1.6 21.1 17.8 2.6 0.3 4.5 0.2 0.1 17.1 9.5 0.5 6.8 5.7 0.8 0.1 1.4

160

Industrial Development Report 2002/2003

Table A2.13

Comparison of main industrial performance and capability indicators by income level and region, 19851998, selected years

Manufacturing Country valued added group, per capita income level (dollars) or region 1985 1998 World 619 1,094 Industrial economies 2,579 5,040 Transition economies .. 725 Developing economies 147 300 High and upper-middle 578 1,161 Lower-middle 176 311 Low 70 156 Low (without China and India) 44 51 Least developed countriesb 31 35 East Asia 145 387 East Asia (without China) 278 668 South Asia 42 65 Latin America and the Caribbean 462 771 Latin America and the Caribbean (without Mexico) 454 750 Sub-Saharan Africa 83 92 Sub-Saharan Africa (without South Africa) 49 40 Middle East and North Africa and Turkey 202 392 Share, top 10 75.9 80.0 Share, bottom 30 4.2 2.4 Share, least developed countriesb 0.7 0.6

Exports per capita (dollars) 1985 1998 293 821 1,345 .. 60 371 64 8 3,714 501 242 1,274 236 73

Tertiary technical enrolment per 1,000 people 1987 1995 2.2 2.8 5.8 .. 1.5 3.6 2.2 1.0 7.0 8.9 1.6 4.4 2.9 0.9

R&D financed by enterprises per capita (dollars) 1985 1998 22.9 71.4 122.3 .. 0.7 4.4 0.2 0.1 402.4 8.8 4.6 33.3 0.8 0.5

Foreign direct investment inflows per capita (dollars) 1981 1993 1985 1997 13.3 63.4 54.8 .. 4.3 25.1 4.8 0.7 241.6 40.8 26.9 95.0 28.5 15.2

Royalties per capitaa (dollars) 1985 1998 2.6 14.2 12 .. 0.6 3.2 1.0 66.2 2.5 3.9 23.6 4.3 0.2

11 7 84

28 18 409

0.4 0.5 1.2

0.5 0.4 2.0

..

8.7

1.3 0.2 4.3

4.5 1.6 39.7

na

0.1 0.1 7.1

317 9

1,178 32

2.6 1.4

4.7 1.0

3.2 0.3

31.0 0.3

14.5 0.2

63.3 2.1

2.7

26.6 0.2

116

404

3.6

3.4

1.1

6.3

11.1

70.4

1.9

5.3

117 25

229 45

3.3 ..

3.1 0.5

0.6 0.6

7.5 1.3

10.8 1.7

69.3 8.2

1.9 0.4

5.2 0.6

14

0.2

0.3

1.9

5.3

0.2

96 75.8 2.2

220 81.6 1.3

1.7 79.5 1.7

2.3 77.2 2.8

0.4 98.8

1.4 97.2

16.9 78.5 2.4

14.1 78.8 1.8

0.1 72.5

3.0 86.2 0.3

0.7

0.6

2.8

2.1

0.3

0.5

0.1

Source: UNIDO Scoreboard database (see technical annex). a. Technology licence payments. b. Includes only 12 of 49 least developed countries.

Statistical annex

161

Table A2.14

Ranking by manufacturing value added, 1985 and 1998

Rank by manufacturing value added per capita 1998 1985 Economy 1 1 Switzerland 2 3 Japan 3 19 Ireland 4 16 Singapore 5 5 Germany 6 6 Finland 7 2 United States 8 4 Sweden 9 9 Austria 10 11 Denmark 11 8 France 12 13 Belgium 13 14 United Kingdom 14 12 Italy 15 17 Netherlands 16 10 Norway 17 7 Canada 18 21 Taiwan Province of China 19 28 Portugal 20 22 Spain 21 18 New Zealand 22 23 Israel 23 15 Australia 24 .. Slovenia 25 29 Korea, Republic of 26 .. Czech Republic 27 25 Bahrain 28 26 Argentina 29 20 Hong Kong SAR 30 36 Uruguay 31 30 Hungary 32 37 Malaysia 33 32 Greece 34 34 Brazil 35 35 Mexico 36 31 Poland 37 38 39 40 41 42 43 44 47 51 43 .. 27 33 45 55 Chile Mauritius Turkey Russian Federation Venezuela Saudi Arabia Peru Thailand

ManuManufacturing facturing value added value added per capita (millions (dollars) of dollars) 1998 1985 1998 1985 8,314.7 3,861.2 59,084 24,982 7,083.5 3,286.8 895,425 396,890 7,042.9 1,337.1 26,094 4,733 6,178.4 1,681.1 19,545 4,174 5,866.3 2,530.6 481,315 196,622 5,557.4 2,493.6 28,637 12,224 5,300.8 3,372.3 1,432,800 802,347 5,295.4 2,599.4 46,874 21,705 5,191.2 2,048.3 41,935 15,475 4,776.3 1,873.9 25,319 9,583 4,761.9 2,084.1 280,223 114,982 4,445.9 1,769.7 45,366 17,445 4,179.0 1,738.7 246,789 98,558 4,082.1 1,873.5 235,087 106,027 3,953.4 1,562.3 62,061 22,641 3,803.2 1,898.3 16,856 7,884 3,489.2 2,263.1 105,725 58,710 3,351.2 2,631.2 2,620.9 2,611.2 2,598.8 2,488.3 2,365.0 2,107.8 1,612.4 1,577.4 1,475.2 1,411.0 1,125.0 947.1 936.6 927.5 912.0 854.6 779.2 748.8 738.9 695.1 662.7 607.3 605.1 584.9 584.5 1,260.3 708.2 1,153.2 1,498.8 971.5 1,692.5 .. 668.1 .. 869.8 862.2 1,322.0 463.2 653.3 368.1 548.4 507.6 494.8 626.6 214.5 182.5 244.1 .. 792.0 546.4 224.7 166.7 73,183 26,228 103,186 9,902 15,497 46,658 4,687 97,866 16,600 1,014 53,293 9,435 3,700 9,579 20,774 9,753 151,274 81,912 30,129 11,099 857 44,106 97,357 14,114 12,550 14,505 35,771 23,316 7,090 44,293 4,903 4,112 26,670 .. 27,264 .. 370 26,130 7,213 1,394 6,911 5,771 5,448 68,640 37,342 23,311 2,584 185 12,274 .. 13,573 6,764 4,380 8,528

Rank by manufacturing value added per capita 1998 1985 Economy 45 38 South Africa 46 40 Costa Rica 47 24 Romania 48 49 El Salvador 49 54 Tunisia 50 52 Jamaica 51 39 Ecuador 52 58 Egypt 53 48 Colombia 54 56 Oman 55 63 China 56 41 Panama 57 53 Paraguay 58 44 Guatemala 59 62 Morocco 60 57 Philippines 61 50 Jordan 62 .. Albania 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 59 42 61 70 65 69 60 71 46 72 64 66 73 67 74 .. 77 76 79 .. 78 80 75 68 .. Bolivia Algeria Honduras Sri Lanka Indonesia Senegal Zimbabwe Pakistan Nicaragua India Cameroon Nigeria Bangladesh Zambia Kenya Yemen Madagascar Central African Republic Uganda Mozambique Malawi Nepal United Republic of Tanzania Ghana Ethiopia

Manufacturing value added per capita (dollars) 1998 1985 556.9 366.5 556.6 327.9 466.3 872.9 425.9 200.7 390.0 174.8 371.9 179.1 354.4 332.2 326.1 132.6 322.2 213.8 293.3 163.5 287.0 100.6 271.3 298.6 246.6 178.9 237.4 228.9 219.3 110.3 189.7 141.4 188.6 196.5 183.8 .. 177.9 153.8 138.3 124.5 115.0 81.7 77.4 72.6 67.0 65.2 64.6 62.2 59.6 39.6 36.6 34.1 26.5 26.0 24.3 22.1 20.6 17.6 15.8 9.2 7.9 132.3 264.5 111.7 50.8 85.6 51.8 123.1 45.0 217.9 43.7 94.3 83.9 33.5 76.8 31.7 .. 28.3 30.0 10.0 .. 20.2 8.3 30.9 57.7 ..

Manufacturing value added (millions of dollars) 1998 1985 23,056 11,476 1,963 866 10,494 19,838 2,580 957 3,641 1,269 958 414 4,315 3,023 20,020 6,166 13,148 6,769 675 228 355,540 105,698 750 647 1,287 646 2,564 1,771 6,091 2,388 14,260 7,731 860 519 614 .. 1,414 4,602 851 2,338 23,418 739 905 9,557 321 63,860 923 7,514 7,489 383 1,072 565 387 90 507 375 217 403 509 169 484 780 5,788 468 804 13,960 330 1,024 4,264 742 33,471 940 6,979 3,280 515 631 .. 287 78 142 .. 145 138 672 728 ..

Source: UNIDO Scoreboard database (see technical annex).

162

Industrial Development Report 2002/2003

Table A2.15

Ranking by manufactured exports, 1985 and 1998

Rank by manufactured exports per capita 1998 1985 1 1 2 10 3 2 4 3 5 4 6 5 7 7 8 8 9 12 10 11 11 9 12 14 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 15 .. 19 17 18 6 13 24 16 20 .. 25 26 21 46 22 29 27 41 43 31 45 30 23 33 40 36 32 42 48 38 51

Economy Singapore Ireland Belgium Switzerland Netherlands Sweden Finland Denmark Austria Germany Canada Taiwan Province of China France Slovenia United kingdom Italy Israel Hong Kong SAR Norway Malaysia Japan Korea, Republic of Czech Republic Spain Portugal United States Hungary New Zealand Mauritius Australia Mexico Costa Rica Greece Thailand Saudi Arabia Bahrain Poland Tunisia Uruguay Jamaica Chile Oman Argentina Philippines

Manufactured exports per capita (dollars) 1998 1985 32,713 7,657.7 15,659 2,323.9 15,050 4,480.1 10,512 3,983.0 8,894.6 3,514.6 8,396.4 3,452.0 7,917.7 2,561.6 6,850.2 2,433.3 6,615.2 2,106.2 5,939.0 2,125.1 5,383.4 2,431.4 4,833.5 4,486.3 4,274.6 4,099.8 3,958.2 3,701.7 3,459.9 3,432.3 2,973.0 2,929.8 2,599.6 2,566.6 2,375.0 2,336.0 2,034.9 2,017.0 1,625.7 1,433.7 1,151.3 1,081.8 970.7 758.0 731.4 701.8 687.9 628.5 554.1 471.8 445.9 443.1 406.4 390.5 374.0 1,572.6 1,507.3 .. 1,304.4 1,313.6 1,309.1 2,928.8 1,876.0 550.3 1,422.6 711.3 .. 544.7 544.3 681.9 71.3 632.0 400.2 467.8 110.5 102.3 334.9 71.5 340.0 628.6 186.2 115.1 150.3 209.6 102.5 64.1 122.2 44.4

Total manufactured exports (millions of dollars) 1998 1985 103,489 19,014 58,018 8,227 153,572 44,165 74,702 25,770 139,628 50,933 74,323 28,824 40,800 12,557 36,313 12,444 53,438 15,912 487,273 165,117 163,121 63,074 105,554 29,092 264,005 83,157 8,472 .. 242,113 73,937 227,949 74,343 22,073 5,542 23,137 15,979 15,212 7,791 65,941 8,626 370,360 171,785 120,700 29,025 26,423 .. 93,505 20,921 23,285 5,449 550,043 162,244 20,400 754 6,165 2,068 1,602 407 21,589 7,371 103,681 8,336 3,423 270 7,970 3,327 44,760 3,658 14,554 442 24,302 5,173 1,552 1,149 6,568 935 14,108 28,119 4,209 267 6,926 836 452 484 1,234 90 3,703 2,429

Rank by manufactured exports per capita 1998 1985 45 39 46 .. 47 28 48 35 49 37 50 .. 51 55 52 74 53 60 54 59 55 62 56 49 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 56 47 34 44 61 53 54 52 67 63 .. 50 71 68 58 57 72 65 70 69 76 64 66 75 73 77 78 .. 79 .. 80

Economy Turkey Romania Venezuela South Africa Brazil Russian Federation Sri Lanka China El Salvador Indonesia Guatemala Morocco Colombia Jordan Algeria Peru Bolivia Panama Ecuador Zimbabwe Paraguay Pakistan Albania Honduras Bangladesh Egypt Senegal Cameroon Nicaragua Kenya India Ghana Nepal Central African Republic Zambia Madagascar Malawi Mozambique United Republic of Tanzania Yemen Nigeria Ethiopia Uganda

Total Manufactured manufactured exports exports per capita (millions (dollars) of dollars) 1998 1985 1998 1985 360.7 115.1 22,885 5,790 339.0 .. 7,630 .. 337.3 409.8 7,841 7,023 322.1 158.5 13,334 4,962 234.4 130.3 38,882 17,617 201.9 .. 29,659 .. 162.1 36.7 3,043 582 135.4 5.8 167,681 6,049 134.0 22.8 812 109 132.0 23.7 26,895 3,856 128.6 21.7 1,389 168 111.9 55.4 3,108 1,200 103.9 103.0 95.2 90.9 80.6 80.3 77.8 74.7 66.4 56.4 52.8 48.2 37.3 36.5 34.5 34.0 29.9 28.3 26.4 21.5 16.3 15.4 11.0 8.5 5.8 4.2 2.9 2.0 1.5 1.4 0.9 33.9 70.7 184.9 79.1 22.6 42.1 40.2 43.3 11.2 18.7 .. 47.9 8.1 9.8 24.2 27.0 6.1 17.1 8.1 9.4 5.1 18.2 11.9 5.1 5.9 4.6 2.6 .. 2.6 .. 0.2 4,241 470 2,848 2,254 641 222 948 874 347 7,428 176 297 4,691 2,242 300 477 143 829 25,855 396 373 53 107 124 61 60 93 34 177 85 19 1,073 187 4,045 1,541 133 91 366 360 40 1,776 .. 201 793 458 154 269 21 341 6,209 119 85 48 80 52 43 62 57 .. 216 .. 2

Source: UNIDO Scoreboard database (see technical annex).

Statistical annex

163

Table A2.16

Ranking by technological structure of manufacturing value added, 1985 and 1998 (percent)

Rank by share of mediumand hightech in manufacturing value added 1998 1985 Economy 1 1 Singapore 2 2 Japan 3 13 Ireland 4 3 Germany 5 7 Switzerland 6 4 United States 7 5 United Kingdom 8 8 Sweden 9 25 Korea, Republic of 10 11 Netherlands 11 24 Malaysia 12 9 India 13 12 Brazil 14 31 Taiwan Province of China 15 14 Saudi Arabia 16 15 Israel 17 18 France 18 26 Finland 19 33 Hong Kong SAR 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 10 22 19 21 27 17 16 23 .. 28 20 6 29 30 .. 40 55 44 66 43 38 39 61 35 46 Italy Canada China Denmark Australia Belgium Norway Austria Slovenia Spain Czech Republic Hungary Poland South Africa Russian Federation New Zealand Indonesia Egypt Thailand Turkey Nigeria Argentina Philippines Mexico Colombia

Share of high- and medium-tech industry 1998 1985 80 67 66 64 65 53 64 64 63 57 63 62 62 58 61 57 60 47 60 56 60 47 59 56 58 54 57 54 54 53 53 52 52 51 51 51 51 51 50 50 50 49 48 46 45 44 41 40 40 39 39 38 38 37 36 36 35 43 52 52 50 46 38 56 48 49 48 46 50 51 48 .. 44 48 58 44 44 .. 34 25 31 18 32 34 34 22 37 30

Share of low-tech and resource-based industry 1998 1985 20 33 34 36 35 47 36 36 37 43 37 38 38 42 39 43 40 53 40 44 40 53 41 44 42 46 43 46 46 47 47 48 48 49 49 49 49 49 50 50 50 51 52 54 55 56 59 60 60 61 61 62 62 63 64 64 65 57 48 48 50 54 62 44 52 51 52 54 50 49 52 .. 56 52 42 56 56 .. 66 75 69 82 68 66 66 78 63 70

Rank by share of mediumand hightech in manufacturing value added 1998 1985 Economy 45 45 Guatemala 46 36 Pakistan 47 32 Romania 48 62 Senegal 49 51 Venezuela 50 47 Greece 51 41 Portugal 52 70 Jordan 53 64 Costa Rica 54 42 Algeria 55 34 Malawi 56 49 Bangladesh 57 48 El Salvador 58 37 Zimbabwe 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 65 52 54 59 58 57 50 .. 63 79 69 .. 56 .. 75 60 78 77 72 .. 71 .. 73 80 67 74 53 68 76 Chile Jamaica Morocco Peru United Republic of Tanzania Zambia Kenya Bahrain Uruguay Oman Central African Republic Yemen Tunisia Albania Ghana Panama Sri Lanka Uganda Nicaragua Nepal Honduras Mozambique Mauritius Bolivia Cameroon Paraguay Ecuador Madagascar Ethiopia

Share of high- and medium-tech industry 1998 1985 35 30 34 36 34 41 34 21 32 28 31 29 31 33 31 14 30 21 29 32 29 38 28 28 28 29 27 34 26 25 25 25 25 24 24 22 21 20 20 20 19 19 17 16 16 15 15 15 12 12 12 11 11 11 11 10 9 20 28 25 24 24 25 28 .. 21 10 14 .. 25 .. 12 23 10 11 14 .. 14 .. 12 7 17 12 27 16 11

Share of low-tech and resource-based industry 1998 1985 65 70 66 64 66 59 66 79 68 72 69 71 69 67 69 86 70 79 71 68 71 62 72 72 72 71 73 66 74 75 75 75 75 76 76 78 79 80 80 80 81 81 83 84 84 85 85 85 88 88 88 89 89 89 89 90 91 80 72 75 76 76 75 72 . 79 90 86 .. 75 .. 88 77 90 89 86 .. 86 .. 88 93 83 88 73 84 89

Source: UNIDO Scoreboard database (see technical annex).

164

Industrial Development Report 2002/2003

Table A2.17

Ranking by technological structure of manufactured exports, 1985 and 1998 (percent)

Rank by share of mediumand hightech in manufactured exports 1998 1985 Economy 1 1 Japan 2 34 Philippines 3 13 Singapore 4 24 Mexico 5 3 United States 6 26 Malaysia 7 2 Germany 8 4 Switzerland 9 9 United Kingdom 10 7 Korea, Republic of 11 17 Taiwan Province of China 12 46 Hungary 13 6 France 14 5 Sweden 15 16 Spain 16 .. Czech Republic 17 12 Ireland 18 11 Italy 19 22 Netherlands 20 20 Finland 21 8 Austria 22 10 Canada 23 14 Belgium 24 19 Israel 25 33 Thailand 26 23 Portugal 27 18 Denmark 28 21 Hong Kong SAR 29 48 China 30 15 Poland 31 25 Brazil 32 39 Costa Rica 33 .. Slovenia 34 31 South Africa 35 .. Romania 36 29 Turkey 37 36 Argentina 38 27 Norway 39 35 Greece 40 37 India 41 .. Russian Federation 42 60 Indonesia 43 32 Tunisia 44 28 Zimbabwe 45 50 Guatemala 46 43 Uruguay 47 44 Australia 48 42 New Zealand 49 38 Morocco 50 49 El Salvador 51 51 Venezuela 52 40 Pakistan 53 47 Colombia 54 69 Egypt

Complex exports 1998 Medium and Medium high tech tech 51.5 81.1 10.4 74.7 17.6 74.3 38.9 65.5 34.4 65.4 18.2 65.1 47.7 64.8 39.7 62.9 34.7 62.9 35.1 62.3 26.3 38.1 36.8 33.5 43.1 40.4 12.0 40.8 25.7 25.4 36.9 36.0 37.1 17.8 16.6 33.5 23.5 12.4 18.4 27.7 28.1 8.7 15.9 21.2 20.6 18.1 20.8 15.3 13.4 11.4 13.0 10.2 11.2 14.4 11.0 12.6 9.9 10.3 12.2 5.8 9.7 8.6 6.5 6.6 61.3 58.8 58.4 58.2 52.5 51.9 51.2 50.9 50.0 49.8 49.1 47.1 46.9 46.1 44.9 39.7 39.5 36.8 36.6 35.7 34.3 32.6 27.8 25.9 23.6 23.5 23.3 21.0 17.9 16.6 16.3 15.5 15.5 15.3 15.0 14.6 14.6 14.5 12.4 11.5 10.3 9.2 8.9 8.8 1985 Medium and Medium high tech tech 59.2 80.0 4.7 10.5 19.5 39.9 16.5 25.1 33.6 59.4 6.3 21.1 48.0 61.3 40.0 57.0 28.0 45.6 35.7 47.9 20.0 1.7 34.1 38.8 31.1 .. 14.1 34.3 20.7 26.9 37.0 36.9 32.1 16.9 11.4 18.7 24.1 18.4 2.9 28.2 20.5 5.9 .. 11.6 .. 17.1 8.4 17.9 9.1 7.0 .. 1.3 14.0 18.2 3.1 5.5 5.0 5.3 8.5 2.3 3.5 7.8 4.8 0.4 35.4 5.8 48.7 52.2 37.1 .. 39.9 43.8 30.9 32.6 46.2 44.1 38.5 33.9 13.8 25.2 34.9 32.6 4.1 37.8 23.9 8.1 .. 16.6 .. 18.2 10.3 20.9 10.5 9.8 .. 1.9 15.0 18.7 3.8 6.4 6.3 6.5 8.9 3.9 3.5 7.9 5.6 0.7 1998

Simple exports 1985 Low tech and Low Resource resource tech based based 12.4 5.3 17.7 12.8 29.6 42.4 7.2 36.2 43.4 5.0 8.1 13.1 5.6 13.6 19.1 4.4 29.6 34.0 15.3 13.2 28.6 17.3 19.6 36.9 10.9 16.5 27.4 39.7 8.2 47.9 50.2 2.8 15.8 14.9 22.5 .. 12.5 33.3 10.3 16.7 28.5 4.7 18.3 18.6 18.4 41.4 16.8 60.6 10.3 13.2 14.7 13.3 .. 9.0 .. 38.6 7.2 3.8 31.4 31.4 .. 3.2 22.2 8.8 5.0 36.2 2.9 9.3 15.9 5.2 3.6 53.5 6.9 8.8 9.4 0.2 20.6 27.5 26.4 .. 26.7 17.1 33.4 43.5 18.3 23.4 25.6 36.1 19.6 29.3 23.8 3.0 9.2 9.4 30.2 5.9 .. 29.4 .. 15.9 26.6 14.4 31.4 28.1 .. 15.6 14.1 10.2 11.2 10.4 24.1 21.3 30.6 6.4 36.7 4.1 17.7 15.4 59.6 3.0 36.4 42.4 48.9 .. 39.2 50.3 43.7 60.2 46.8 28.0 43.8 54.7 38.0 70.7 40.6 63.7 19.5 22.5 44.9 19.2 .. 38.4 .. 54.5 33.8 18.2 62.8 59.6 .. 18.8 36.4 19.0 16.3 46.6 26.9 30.6 46.5 11.6 40.3 57.6 24.6 24.2 Low tech and Low Resource resource tech based based 8.0 6.3 14.3 13.8 6.9 20.7 6.6 13.3 19.9 16.9 5.9 22.9 10.1 11.1 21.2 9.9 15.0 24.9 13.7 11.2 24.9 15.4 16.4 31.8 12.4 14.2 26.7 19.1 9.8 28.9 29.1 17.4 13.9 12.4 16.0 26.5 10.4 30.8 12.0 9.6 24.2 8.9 16.7 12.8 20.6 36.9 19.3 53.0 45.6 32.2 11.5 20.1 27.7 12.3 50.9 49.2 8.0 4.6 29.7 38.1 5.2 18.2 53.7 11.4 14.4 23.7 5.0 8.2 22.4 24.1 4.1 74.4 11.9 30.9 5.3 12.5 15.6 19.5 17.1 14.9 28.7 12.4 21.3 35.0 14.6 20.0 22.4 35.8 15.7 19.5 17.3 4.2 9.0 18.2 30.2 11.2 38.1 29.9 17.4 12.4 24.4 12.0 26.5 23.6 18.5 21.4 21.0 16.7 24.4 17.8 21.3 28.3 30.0 29.0 31.6 4.5 18.3 30.5 34.3 29.9 29.4 31.9 33.1 41.4 39.1 43.3 33.3 44.6 38.7 28.9 39.1 48.6 36.3 56.5 36.6 57.3 54.6 50.5 41.7 31.3 65.8 42.2 68.3 61.7 32.4 16.7 56.2 61.7 23.7 39.5 74.7 28.1 38.8 41.4 26.3 36.5 52.3 53.1 35.7 78.9 30.2 61.4

High tech 29.6 64.3 56.7 26.6 31.0 46.9 17.1 23.2 28.2 27.2 35.0 20.7 21.6 24.7 9.3 11.5 39.3 10.1 24.3 24.4 12.2 11.1 9.7 28.3 28.3 6.2 16.0 24.5 18.2 8.0 6.2 23.9 11.9 4.7 3.0 5.3 2.5 5.6 4.5 5.2 3.3 5.3 4.3 0.9 4.0 2.0 4.7 4.2 0.3 5.7 0.5 0.6 2.5 2.1

High tech 20.8 5.8 20.4 8.6 25.8 14.8 13.2 17.0 17.6 12.2 15.4 4.1 14.6 13.4 6.0 .. 25.8 9.5 10.2 5.7 9.2 7.1 6.4 17.0 2.4 6.5 10.9 14.2 1.2 9.6 3.4 2.3 .. 4.9 .. 1.2 1.9 3.0 1.5 2.8 .. 0.6 1.1 0.6 0.7 0.9 1.3 1.2 0.4 1.6 0.1 0.2 0.8 0.3

Statistical annex

165

Table A2.17

Ranking by technological structure of manufactured exports, 1985 and 1998 (percent) (continued)

Rank by share of mediumand hightech in manufactured exports High 1998 1985 Economy tech 55 54 Kenya 1.6 56 57 Chile 0.7 57 67 Honduras 0.5 58 30 Oman 1.6 59 41 Bahrain 0.4 60 63 Saudi Arabia 61 61 Jordan 2.6 62 80 Bolivia 1.9 63 55 Peru 0.8 64 73 Ecuador 0.8 65 .. Albania 1.8 66 56 Panama 2.4 67 64 Sri Lanka 1.5 68 68 Nicaragua 0.8 69 65 Mozambique 1.1 70 62 Bangladesh 0.3 71 79 Paraguay 0.5 72 75 Nepal 0.1 73 77 Cameroon 0.1 74 66 Zambia 0.1 75 53 Jamaica 0.1 76 58 United Republic of Tanzania 1.1 77 78 Nigeria 78 52 Mauritius 0.2 79 45 Senegal 0.2 80 72 Malawi 81 59 Madagascar 0.3 82 74 Central African Republic 0.2 83 76 Uganda 0.1 84 71 Algeria 85 70 Ghana 86 .. Yemen 87 .. Ethiopia

Complex exports 1998 Medium and Medium high tech tech 6.0 7.6 5.6 6.3 5.5 6.0 4.3 5.8 5.3 5.7 5.2 5.2 2.4 5.0 3.1 5.0 3.9 4.6 3.4 4.2 2.4 4.2 1.7 4.0 2.5 4.0 3.1 3.9 2.3 3.4 2.6 2.9 1.7 2.2 1.7 1.9 1.7 1.8 1.7 1.8 1.4 1.5 0.4 1.4 1.2 1.2 0.9 0.6 0.6 0.7 0.7 0.1 0.1 1.5 1.5 1.4 1.4 1.0 0.9 0.8 0.8 0.8 0.1 0.1 0.1 1985 Medium and Medium high tech tech 2.4 3.2 2.2 2.4 0.9 1.0 16.0 17.7 3.3 7.3 1.6 1.6 1.3 1.8 2.9 3.1 0.2 0.3 .. .. 1.1 2.7 1.2 1.4 0.5 0.7 0.8 1.1 1.6 1.8 0.2 0.2 0.1 0.2 0.9 1.1 2.5 3.2 1.6 3.3 5.0 0.5 0.8 0.3 0.5 0.5 .. .. 2.3 0.1 3.4 5.9 0.5 2.0 0.3 0.2 0.5 0.6 .. .. 1998

Simple exports 1985 Low tech and Low Resource resource tech based based 4.3 28.1 32.5 0.7 29.7 30.4 4.9 18.7 23.5 3.1 4.9 8.0 1.2 0.9 2.1 0.4 13.7 14.1 14.8 12.3 27.0 0.5 19.4 19.8 8.7 41.5 50.1 0.2 12.1 12.3 .. .. .. 10.5 15.0 25.5 25.8 19.5 45.3 0.5 6.3 6.8 17.1 22.9 40.0 63.2 16.5 79.7 1.7 11.5 13.2 53.9 11.7 65.6 0.5 10.2 10.7 0.2 11.3 11.5 9.5 77.9 87.3 3.9 0.1 40.0 4.7 3.8 6.4 0.2 0.1 0.1 0.2 .. .. 10.7 1.3 46.2 35.2 13.3 9.7 43.6 0.2 39.3 21.7 .. .. 14.6 1.4 86.2 39.9 17.1 16.1 43.8 0.4 39.3 21.8 .. .. Low tech and Low Resource resource tech based based 11.8 23.8 35.6 3.7 34.3 37.9 7.6 16.9 24.5 2.6 5.7 8.3 4.2 4.1 8.2 1.5 18.0 19.5 17.6 25.2 42.8 9.5 34.0 43.5 13.1 22.0 35.1 3.5 14.8 18.3 61.0 19.7 80.7 10.8 16.6 27.4 59.0 13.2 72.2 2.8 19.2 22.0 2.7 19.1 21.8 87.4 2.5 89.9 8.9 17.3 26.2 74.1 2.3 76.4 2.6 20.3 23.0 3.7 6.2 9.9 17.7 70.1 87.8 3.2 1.0 64.8 7.6 5.8 21.1 0.1 1.8 0.2 0.5 0.5 10.8 11.9 0.1 28.0 63.2 7.3 18.7 44.3 0.6 17.7 22.1 1.1 3.6 15.1 1.1 92.8 70.8 13.1 39.9 44.4 2.4 18.0 22.5 1.6 14.4

High tech 0.7 0.1 0.2 1.8 4.0 0.5 0.2 0.1 .. 1.6 0.2 0.2 0.3 0.2 0.0 0.1 0.1 0.7 0.6 0.1 0.9 1.2 0.2 0.2 .. ..

Source: UNIDO Scoreboard database (see technical annex).

166

Industrial Development Report 2002/2003

Table A2.18

Ranking by Harbison-Myers index of skills

Rank 1998 1985 1 2 2 18 3 1 4 4 5 7 6 9 7 12 8 8 9 28 10 6 11 16 12 10 13 15 14 11 15 19 16 14 17 .. 18 17 19 22 20 25 21 20 22 13 23 21 24 41 25 5 26 .. 27 24 28 32 29 37 30 30 31 3 32 23 33 36 34 34 35 27 36 .. 37 35 38 31 39 39 40 29 41 38 42 46 43 42 44 33

Economy Canada Australia United States Finland New Zealand Belgium Norway Netherlands United Kingdom Korea, Republic of France Spain Sweden Denmark Austria Germany Russian Federation Japan Ireland Italy Greece Israel Taiwan Province of China Portugal Argentina Slovenia Switzerland Poland Singapore Peru Uruguay Philippines Chile Costa Rica Panama Bahrain Czech Republic Jordan Hong Kong SAR Venezuela Hungary South Africa Romania Egypt

Harbison-Myers Index 1998 1985 62.05 39.50 50.55 24.20 50.25 39.60 45.05 29.40 40.80 26.70 38.95 26.25 38.85 25.85 38.35 26.60 37.55 19.70 36.10 26.80 35.90 24.65 34.85 26.25 34.45 24.70 34.30 25.95 32.80 24.00 31.65 24.95 30.75 .. 30.05 24.35 29.90 22.30 29.10 20.45 28.55 22.95 28.35 25.40 27.80 22.48 27.20 12.90 26.75 27.80 25.05 .. 25.00 20.95 23.30 16.90 23.05 14.80 22.55 18.40 21.85 30.40 21.60 21.25 21.00 15.90 20.95 16.50 20.20 20.10 20 .. 20.00 16.30 18.55 18.00 18.45 14.00 17.75 18.65 17.65 14.60 17.05 11.50 16.95 12.80 16.45 16.80

Rank 1998 1985 45 48 46 43 47 26 48 44 49 50 50 47 51 40 52 61 53 45 54 49 55 51 56 57 57 52 58 53 59 67 60 55 61 54 62 56 63 64 64 59 65 65 66 .. 67 58 68 62 69 60 70 63 71 66 72 73 73 74 74 68 75 70 76 72 77 69 78 77 79 76 80 75 81 71 82 80 83 79 84 78 85 81 86 82 87 83

Economy Thailand Colombia Ecuador Bolivia Turkey Saudi Arabia Mexico Tunisia El Salvador Algeria Malaysia Indonesia Brazil Sri Lanka China Jamaica Morocco Nicaragua Mauritius Paraguay Oman Albania Honduras Zimbabwe India Guatemala Nepal Nigeria Yemen Ghana Cameroon Bangladesh Pakistan Zambia Senegal Kenya Madagascar Uganda Central African Republic Ethiopia Malawi Mozambique United Republic of Tanzania

Harbison-Myers Index 1998 1985 15.55 10.75 15.30 12.55 15.00 20.25 14.80 12.00 14.70 9.80 13.45 11.10 12.95 13.15 12.55 6.85 12.05 11.75 11.65 10.20 11.10 9.20 10.35 8.30 10.15 9.10 10.05 9.10 9.75 5.15 9.60 8.50 9.55 8.60 9.40 8.50 9.35 5.75 8.95 7.30 8.95 5.45 8.30 .. 8.20 7.60 8.15 6.70 8.10 7.10 6.55 6.40 6.40 5.40 5.05 3.85 4.45 3.85 4.40 4.75 4.35 4.05 4.30 3.95 4.10 4.40 3.35 2.85 3.30 2.95 3.20 3.05 3.10 4.00 1.95 1.75 1.70 1.80 1.45 1.95 1.00 0.70 0.90 0.60 0.75 0.55

Source: UNIDO Scoreboard database (see technical annex). Note: The Harbison-Myers Index is the average of the percentage of the relevant age groups enroled in secondary and tertiary education, with tertiary enrolments given a weight of five.

Statistical annex

167

Table A2.19

Ranking by tertiary enrolments in technical subjects, 1985 and 1998

Rank by tertiary technical enrolment as share of population 1998 1985 1 1 2 2 3 .. 4 21 5 10 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 16 12 23 11 17 40 8 18 26 9 3 13 30 20 32 33 15 19 7 39 28 .. 34 25 31 24 .. 4 5 .. 29 14 27 22 35 55 43 37 36

Share of population (percent) Economy 1998a 1985 Korea, Republic of 1.65 0.78 Finland 1.33 1.00 Russian Federation 1.18 .. Australia 1.17 0.50 Taiwan Province of China 1.06 0.59 Spain 0.97 0.54 Ireland 0.91 0.58 Austria 0.78 0.50 Germany 0.77 0.58 United Kingdom 0.75 0.54 Portugal 0.73 0.26 Sweden 0.73 0.60 Chile 0.73 0.52 Greece 0.72 0.48 Canada 0.69 0.60 United States 0.68 0.75 New Zealand 0.68 0.57 Israel 0.68 0.45 Norway 0.67 0.51 Japan 0.64 0.41 Italy 0.64 0.40 France 0.61 0.56 Denmark 0.60 0.52 Panama 0.59 0.60 Netherlands 0.56 0.26 Philippines 0.55 0.47 Bahrain 0.52 .. Colombia 0.51 0.36 Switzerland 0.51 0.49 Romania 0.49 0.42 Hong Kong SAR 0.49 0.49 Slovenia 0.49 .. Singapore 0.47 0.71 Argentina 0.47 0.68 Czech Republic 0.46 .. Peru 0.46 0.47 Venezuela 0.45 0.56 Mexico 0.44 0.48 Belgium 0.43 0.50 Jordan Algeria Poland Costa Rica Bolivia 0.42 0.41 0.39 0.34 0.34 0.35 0.13 0.23 0.31 0.35

Number (thousands of people) 1998a 1985 742.5 320.7 68.0 49.2 1,749.2 .. 212.0 81.5 226.8 379.7 32.6 63.0 631.1 439.1 72.6 64.5 103.1 75.0 203.2 1,792.9 24.8 37.4 29.3 808.2 364.0 355.1 31.4 15.6 86.6 387.3 3.0 197.1 36.0 111.2 30.2 9.7 14.1 162.3 47.9 108.2 97.9 400.1 43.6 17.5 115.1 151.9 11.5 25.4 115.7 209.7 20.7 37.5 454.1 305.7 26.2 50.8 65.2 48.3 158.3 1,822.6 18.8 19.8 21.2 501.6 224.9 311.1 26.6 13.6 38.7 271.5 .. 115.8 31.8 95.8 27.5 .. 18.1 210.9 .. 94.9 102.0 375.7 49.4 10.0 29.8 85.9 8.5 21.6

Rank by tertiary technical enrolment as share of population 1998 1985 45 45 46 42 47 6 48 38 49 41 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 46 63 53 47 49 48 44 56 57 62 54 52 51 50 60 na 64 78 61 58 66 76 59 70 67 69 68 65 73 .. 71 80 72 81 74 77 75 79

Economy Turkey Uruguay Ecuador El Salvador Morocco Tunisia Indonesia Nicaragua Honduras Thailand Brazil South Africa Guatemala Hungary Malaysia Saudi Arabia Egypt India Paraguay Jamaica Albania China Zimbabwe Sri Lanka Bangladesh Nepal Cameroon Madagascar Nigeria Senegal Pakistan Mauritius Oman Zambia Yemen Kenya Mozambique Uganda Central African Republic Ghana United Republic of Tanzania Ethiopia Malawi

Share of population (percent) 1998a 1985 0.33 0.22 0.29 0.25 0.29 0.66 0.26 0.28 0.25 0.25 0.24 0.23 0.22 0.20 0.19 0.18 0.17 0.17 0.16 0.13 0.12 0.12 0.12 0.11 0.11 0.11 0.10 0.09 0.08 0.08 0.08 0.06 0.06 0.06 0.05 0.05 0.04 0.04 0.03 0.02 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.19 0.08 0.14 0.19 0.16 0.16 0.22 0.12 0.11 0.08 0.13 0.15 0.15 0.16 0.09 0.20 0.08 0.01 0.08 0.09 0.06 0.01 0.09 0.03 0.05 0.03 0.04 0.07 0.02 0.01 0.03 0.01 0.02 0.01 0.01 0.01 0.01 0.01

Number (thousands of people) 1998a 1985 198.3 114.1 9.3 7.6 32.7 63.2 15.0 13.8 66.7 56.8 21.4 13.8 439.1 137.3 9.7 4.9 11.3 8.7 110.5 81.8 289.3 225.9 68.1 68.9 17.0 9.3 16.7 11.6 26.7 13.8 23.4 18.5 69.6 75.0 1,086.3 1,233.8 5.5 6.1 2.9 2.1 3.6 6.1 1,221.0 821.5 9.5 0.9 15.4 13.8 90.0 97.9 16.0 10.5 8.4 10.0 8.2 9.9 63.3 23.5 4.4 3.6 63.4 28.5 0.5 0.4 0.9 1.0 2.7 1.1 3.2 1.3 4.6 5.5 2.1 0.8 2.5 2.7 0.4 2.1 3.6 6.5 0.8 0.2 1.9 0.9 5.9 0.5

Source: UNIDO Scoreboard database (see technical annex). Note: Ranking is based on tertiary technical enrolment as a percentage of the population. Technical subjects include pure science, mathematics and computing and engineering. a 1998 or latest year available.

168

Industrial Development Report 2002/2003

Table A2.20

Ranking by productive enterprisefinanced research and development, 1985 and 1998

Rank by per capita R&D 1998 1985 1 2 2 3 3 5 4 4 5 1 6 12 7 6 8 10 9 7 10 15 11 8 12 11 13 23 14 19 15 14 16 21 17 22 18 9 19 13 20 16 21 18 22 23 24 24 17 25 26 33 27 34 28 27 29 25 30 20 31 46 32 26 33 46 34 38 35 46 36 31 37 32 38 29 39 46 40 46 41 28 42 46 43 46 44 46 45 41 46 36 47 48 39 49 50 51 30 52 53 54 55 56 57 40 58 59 60 61

Economy Switzerland Japan Sweden United States Germany Finland Denmark France Norway Belgium Netherlands Austria Korea, Republic of Singapore United Kingdom Ireland Australia Canada Israel Taiwan Province of China Italy Slovenia Spain New Zealand Czech Republic Portugal Brazil Greece South Africa Hungary Argentina Poland Russian Federation Malaysia Costa Rica Chile Turkey Romania Venezuela Hong Kong SAR Mexico Panama Uruguay China Indonesia India Mauritius Thailand Egypt Colombia Jordan Guatemala Algeria Saudi Arabia Peru Morocco Philippines Honduras Nicaragua Sri Lanka Yemen

Per capita (dollars) 1998a 1985 859.9 203.3 858.4 192.3 653.9 113.0 465.9 153.7 418.1 255.5 413.4 53.8 328.4 80.0 297.6 63.3 275.5 79.0 272.7 36.8 258.8 78.3 214.4 58.8 211.2 10.8 198.4 14.5 174.5 52.5 152.8 14.1 148.0 12.0 143.7 66.9 134.0 52.9 122.5 33.3 90.1 18.7 73.3 55.2 8.8 50.7 25.8 32.3 14.1 1.1 13.7 1.1 13.5 4.1 12.8 5.9 11.3 14.1 8.5 8.3 4.6 7.5 6.7 0.2 5.5 5.3 1.5 4.8 1.3 2.5 3.1 2.3 1.8 1.5 3.3 1.4 1.1 0.9 0.8 0.1 0.4 0.4 0.3 0.3 0.1 0.2 0.2 0.2 1.9 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1

Total value (billions of dollars) 1998a 1985 6.05 1.32 107.68 23.22 5.78 0.94 122.44 36.57 34.13 19.85 2.11 0.26 1.72 0.41 17.30 3.49 1.20 0.33 2.76 0.36 4.00 1.13 1.72 0.44 9.50 0.44 0.59 0.04 10.22 2.98 0.55 0.05 2.67 0.19 4.22 1.73 0.74 0.22 2.61 0.62 5.16 1.06 0.15 2.16 0.34 0.19 0.08 0.33 0.14 0.01 2.19 0.15 0.14 0.04 0.50 0.18 0.12 0.15 0.29 0.32 0.17 1.11 0.14 0.01 0.02 0.07 0.02 0.29 0.07 0.06 0.07 0.05 0.01 0.14 0.25 1.10 0.15 0.01 0.40 0.30 0.01 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

Share of GNP (percent) 1998a 1985 1.85 1.29 2.08 1.72 2.61 0.96 1.74 0.90 1.42 1.56 1.75 0.50 1.02 0.74 1.13 0.67 0.83 0.53 1.02 0.46 1.01 0.89 0.75 0.67 2.10 0.48 0.69 0.20 0.93 0.65 0.99 0.28 0.79 0.12 0.78 0.51 0.82 0.90 0.99 0.60 0.48 0.25 0.77 0.39 0.21 0.33 0.40 0.71 0.14 0.05 0.32 0.07 0.12 0.10 0.38 0.35 0.27 0.75 0.11 0.27 0.25 0.32 0.17 0.01 0.20 0.12 0.13 0.17 0.10 0.16 0.17 0.07 0.01 0.05 0.14 0.05 0.02 0.16 0.08 0.01 0.12 0.14 0.01 0.01 0.02 0.02 0.01 0.01 0.10 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

Statistical annex

169

Table A2.20

Ranking by productive enterprisefinanced research and development, 1985 and 1998 (continued)

Rank by per capita R&D 1998 1985 61 37 61 44 61 43 61 42 61 45 61 35 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61

Economy Tunisia Malawi Madagascar Kenya Jamaica Ecuador Albania Bahrain Bangladesh Bolivia Cameroon Central African Republic El Salvador Ethiopia Ghana Mozambique Nepal Nigeria Oman Pakistan Paraguay Senegal United Republic of Tanzania Uganda Zambia Zimbabwe

Per capita (dollars) 1998a 1985 0.2 0.5

Total value (billions of dollars) 1998a 1985

Share of GNP (percent) 1998a 1985 0.03

Source: UNIDO Scoreboard database (see technical annex). Note: All economies with negligible values are given the same rank. a Data for 1998 or latest available year.

170

Industrial Development Report 2002/2003

Table A2.21

Ranking by foreign direct investment inflows, 19811985 and 19931997

Rank by per capita foreign direct investment inflows 1998 1985 1 1 2 5 3 18 4 2 5 8 6 7 7 16 8 34 9 3 10 13 11 4 12 10 13 17 14 21 15 12 16 9 17 28 18 .. 19 11 20 24 21 23 22 35 23 15 24 30 25 26 26 .. 27 22 28 31 29 14 30 .. 31 61 32 40 33 67 34 37 35 36 36 25 37 27 38 79 39 33 40 45 41 41 42 48 43 19 44 44 45 42 46 6 47 51 48 49 49 65 50 47 51 .. 52 62 53 59 54 .. 55 54 56 .. 57 20

Economy Singapore Belgium Sweden New Zealand Hong Kong SAR Netherlands Norway Denmark Switzerland Ireland Australia United Kingdom France Austria Canada United States Finland Hungary Malaysia Chile Israel Panama Spain Argentina Portugal Czech Republic Costa Rica Mexico Greece Slovenia Peru Venezuela Poland Germany Taiwan Province of China Italy Colombia Jamaica Brazil Bolivia Ecuador Uruguay Tunisia Paraguay Thailand Oman South Africa Korea, Republic of China Mauritius Romania Philippines Indonesia Albania Morocco Nicaragua Jordan

Per capita (dollars) 1993 1981 1997 1985 2,536.0 563.4 1,116.2 120.7 922.5 34.7 735.0 203.9 727.7 103.3 711.6 105.9 589.3 40.9 551.8 15.2 529.8 140.0 484.2 51.5 376.9 132.4 367.6 76.6 362.1 39.4 304.6 27.5 292.8 69.2 271.3 79.5 260.2 17.8 236.1 .. 229.5 73.6 229.4 19.3 191.1 19.8 189.0 14.1 182.3 46.5 149.1 16.6 149.0 18.8 132.1 .. 110.4 22.8 102.4 16.5 96.7 47.3 92.9 .. 91.1 1.4 88.4 7.7 86.3 0.5 77.1 9.7 74.5 10.0 63.0 18.8 62.2 18.6 58.7 3.9 49.6 15.4 49.5 4.7 46.3 6.1 42.0 3.3 41.2 30.6 40.6 4.8 38.0 5.6 37.3 111.2 37.1 2.8 36.8 2.9 30.1 0.8 25.7 3.4 20.6 .. 20.1 1.2 19.8 1.5 19.7 .. 19.4 2.4 18.8 .. 16.1 28.3

Total value (billions of dollars) 1993 1981 1997 1985 8.20 1.53 10.58 1.84 8.10 0.28 2.69 0.62 2.75 1.34 11.92 1.45 2.62 0.17 2.99 0.07 4.47 0.56 1.47 0.16 6.35 1.87 20.91 4.13 22.89 2.10 2.65 0.20 8.06 1.78 70.00 21.83 1.46 0.09 2.39 .. 4.63 1.10 3.38 0.13 1.11 0.08 0.46 0.03 7.65 1.71 5.39 0.48 1.53 0.17 1.30 .. 0.37 0.07 6.81 0.90 1.08 0.44 0.21 .. 2.20 0.02 1.89 0.10 3.13 0.02 6.81 1.52 1.74 0.45 3.55 1.09 1.98 0.44 0.14 0.01 7.28 1.74 0.30 0.02 0.51 0.06 0.14 0.38 0.20 0.20 0.01 2.45 0.28 0.07 0.15 1.33 0.01 1.61 0.13 37.81 0.93 0.03 0.51 .. 1.54 0.05 3.66 0.22 0.08 .. 0.51 0.04 0.07 .. 0.07 0.07

Share of gross domestic investment (percent) 1993 1981 1997 1985 26.54 17.91 24.16 7.60 25.25 1.60 22.31 10.95 10.24 6.90 15.50 5.91 7.73 1.00 9.60 0.69 6.60 2.24 15.11 3.91 8.82 4.75 12.07 5.60 8.59 2.01 4.80 1.30 8.08 2.56 5.67 2.66 7.57 0.68 23.57 .. 14.10 10.91 20.23 6.74 5.08 1.53 20.74 3.46 6.77 5.07 10.34 2.88 6.32 2.91 8.58 .. 15.94 7.04 11.04 2.19 4.81 4.39 4.88 .. 16.91 0.14 15.05 0.75 13.27 0.11 1.32 0.60 2.78 1.50 1.90 1.11 11.29 6.19 10.59 1.37 5.06 4.33 30.89 5.59 15.75 1.92 6.10 0.41 8.39 7.72 9.93 1.08 4.07 2.49 3.43 6.40 6.28 0.10 0.99 0.47 13.54 0.87 2.65 1.46 6.21 .. 8.46 0.67 6.16 1.00 20.24 .. 7.72 1.34 16.79 .. 3.84 3.99

Share of GDP (percent) 1993 1981 1997 1985 9.57 8.36 3.91 .. 3.66 0.29 4.79 2.92 1.96 .. 3.01 1.13 1.81 0.28 1.78 0.13 1.37 0.55 2.64 0.93 1.88 1.15 1.90 0.90 1.49 0.40 1.15 0.31 1.49 0.53 0.99 0.54 1.21 0.17 5.58 .. 5.73 3.81 5.26 0.92 1.22 0.33 6.13 0.59 1.38 1.04 1.94 0.58 1.54 0.76 2.77 .. 4.18 1.80 2.49 0.51 0.93 1.08 1.09 .. 3.85 0.09 2.53 0.16 2.65 0.03 0.28 .. 0.66 .. 0.33 0.26 2.54 1.33 3.63 0.29 1.08 0.83 5.22 0.90 3.04 0.38 0.81 0.09 2.22 2.50 2.27 0.29 1.48 0.72 0.63 1.72 1.01 .. 0.36 0.14 5.51 0.31 0.74 0.32 1.44 .. 2.01 0.18 1.90 0.27 3.15 .. 1.63 0.35 4.50 .. 1.01 1.42

Statistical annex

171

Table A2.21

Ranking by foreign direct investment inflows, 19811985 and 19931997 (continued)

Rank by per capita foreign direct investment inflows 1998 1985 58 .. 59 39 60 43 61 32 62 58 63 46 64 53 65 38 66 66 67 57 68 52 69 50 70 60 71 77 72 64 73 73 74 68 75 72 76 56 77 71 78 80 79 29 80 69 81 75 82 63 83 55 84 78 85 76 86 70 87 74

Per capita (dollars) 1993 1981 Economy 1997 1985 Russian Federation 15.4 .. Saudi Arabia 13.8 7.8 Nigeria 13.5 5.2 Egypt 13.3 15.5 Turkey 12.0 1.7 Honduras 11.2 3.9 Sri Lanka 10.6 2.7 Guatemala 9.0 9.6 Ghana 7.9 0.7 Yemen 7.3 1.9 Japan 7.1 2.8 Zambia 6.7 2.9 Senegal 6.6 1.5 Uganda 5.8 Pakistan 5.1 0.9 Zimbabwe 4.2 United Republic of Tanzania 3.3 0.4 Mozambique 3.1 El Salvador 2.1 2.0 India 2.1 0.1 Bahrain 1.7 36.8 Cameroon 1.2 16.7 Madagascar 0.8 0.3 Nepal 0.6 Kenya 0.5 0.9 Central African Republic 0.4 2.2 Algeria 0.4 0.4 Bangladesh 0.3 Malawi 0.1 0.1 Ethiopia 0.1

Total value (billions of dollars) 1993 1981 1997 1985 1.98 .. 0.42 1.63 1.23 0.28 0.78 0.75 0.74 0.09 0.06 0.02 0.19 0.05 0.09 0.07 0.13 0.01 0.14 0.01 1.07 0.37 0.06 0.01 0.06 0.01 0.12 0.65 0.09 0.04 0.09 0.02 0.01 0.01 1.64 0.06 0.01 0.08 0.01 0.16 0.01 0.01 0.01 0.02 0.01 0.01 0.01 0.03 0.01

Share of gross domestic investment (percent) 1993 1981 1997 1985 2.52 .. 1.00 0.20 30.72 8.23 7.83 8.43 1.76 0.75 4.92 3.12 5.91 2.99 4.20 5.68 9.73 4.79 12.03 1.90 0.07 0.10 12.18 5.27 7.58 2.53 13.80 5.66 1.38 3.06 0.01 9.20 0.30 10.24 0.23 0.71 2.22 2.16 0.12 0.76 6.90 1.13 8.05 2.81 0.98 1.18 0.03 0.92 1.30 3.02 8.50 0.07 0.05 0.44 0.34 0.33 0.58 0.08

Share of GDP (percent) 1993 1981 1997 1985 0.56 .. 0.33 .. 5.36 1.03 1.32 2.38 0.43 0.13 1.57 0.48 1.49 0.83 0.64 0.78 2.19 0.20 2.11 .. 0.02 0.03 1.75 0.75 1.34 0.34 2.16 1.06 0.26 0.61 1.77 .. 1.88 0.01 0.14 0.25 0.51 0.03 0.14 .. 0.18 2.06 0.32 0.08 0.28 0.01 0.15 0.25 0.20 0.78 0.02 0.02 0.09 0.06 0.07 0.09 0.01

Source: UNIDO Scoreboard database (see technical annex). Note: Annual averages calculated for available figures within the periods 19811985 and 19931997.

172

Industrial Development Report 2002/2003

Table A2.22

Ranking by royalty and licence payments abroad, 1985 and 1998

Rank by payments per capita 1998 1985 1 11 2 1 3 3 4 2 5 4 6 33 7 5 8 7 9 17 10 16 11 10 12 13 13 12 14 6 15 8 16 19 17 15 18 9 19 20 20 21 21 14 22 25 23 22 24 30 25 .. 26 23 27 .. 28 37 29 18 30 39 31 .. 32 31 33 47 34 27 35 24 36 43 37 29 38 26 39 40 40 36 41 38 42 34 43 28 44 35 45 49 46 45 47 50 48 32 49 53 50 48 51 46 52 55 53 42 54 61 55 44 56 41 57 64 58 51 59 57 60 58 61 54

Economy Ireland Singapore Netherlands Hong Kong SAR Switzerland Malaysia Belgium Sweden United Kingdom Austria Finland Norway Japan New Zealand Canada Taiwan Province of China Germany Australia Korea, Republic of Spain France United States Israel Portugal Hungary Italy Slovenia Thailand Argentina Jamaica Czech Republic Denmark Brazil Egypt Panama Morocco Costa Rica Ecuador Greece Mexico Poland Indonesia South Africa Chile Peru Philippines Turkey Uruguay Kenya Colombia El Salvador Romania Honduras Madagascar Bolivia Zimbabwe China Tunisia Senegal India Pakistan

Per capita (dollars) 1998 1985 1,683.1 21.3 559.2 191.1 188.8 50.6 184.7 101.2 151.7 49.8 107.8 2.6 107.7 42.5 106.0 34.2 103.7 14.2 100.4 15.2 79.8 23.7 76.9 18.5 70.8 18.8 70.4 36.6 68.4 31.3 65.0 9.3 59.6 15.7 53.8 24.9 51.0 7.9 47.4 6.0 46.2 17.8 41.8 4.9 35.2 5.9 29.1 3.3 21.2 .. 20.1 5.9 19.5 .. 13.1 0.9 11.7 13.9 11.6 0.8 10.9 .. 8.5 3.3 6.5 0.2 6.4 4.1 6.4 5.2 6.2 0.5 6.1 3.6 5.6 4.6 5.5 0.8 5.2 1.9 5.0 0.9 4.9 2.4 4.0 3.8 3.8 1.9 3.2 0.2 2.1 0.3 1.9 0.2 1.8 3.0 1.3 0.1 1.3 0.2 1.1 0.3 0.9 0.16 0.8 0.5 0.6 0.6 0.4 0.5 0.6 0.3 0.2 0.2 0.2 0.2 0.1 0.1

Total value (millions of dollars) 1998 1985 6,235.8 75.6 1,769.0 474.5 2,964.5 733.7 1,235.0 552.4 1,078.2 322.2 2,392.0 41.2 1,099.2 419.3 938.5 285.5 6,122.7 807.3 810.9 114.8 411.4 116.1 341.0 77.0 8,947.3 2,270 266.9 119.7 2,073.2 812.1 1,419.0 172.0 4,893.4 1,216 1,009.7 392.6 2,369.3 322.8 1,866.3 229.2 2,716.7 981.9 11,292 1,170 209.6 24.9 290.0 33.2 214.6 .. 1,154.9 331.8 38.6 . 804.0 45.5 422.0 420.0 30.0 1.9 112.6 .. 45.3 16.8 1,075.0 30.0 392.0 189.4 17.6 11.2 171.5 11.6 21.5 9.6 68.0 42.0 58.0 8.0 501.0 142.0 195.0 32.4 1,002.0 385.0 165.4 120.3 56.0 23.0 80.0 4.0 158.0 17.0 124.0 9.7 6.0 8.9 39.9 1.8 54.0 7.0 6.9 1.4 21.0 1.4 5.1 2.3 9.8 0.1 5.2 2.5 6.0 4.8 420.0 11.0 2.6 1.2 2.2 0.2 200.8 25.1 19.7 6.7

Share of GNP (percent) 1998 1985 8.998 0.432 1.852 2.589 0.762 0.573 0.781 1.584 0.380 0.316 2.942 0.142 0.424 0.529 0.414 0.290 0.484 0.176 0.374 0.174 0.329 0.221 0.224 0.124 0.219 0.168 0.482 0.567 0.357 0.241 0.527 0.168 0.224 0.096 0.261 0.241 0.594 0.354 0.336 0.140 0.185 0.189 0.143 0.029 0.217 0.100 0.273 0.150 0.470 .. 0.100 0.079 0.199 .. 0.610 0.119 0.145 0.502 0.667 0.104 0.213 .. 0.026 0.030 0.140 0.014 0.495 0.603 0.212 0.215 0.498 0.096 0.219 0.264 0.370 0.284 0.047 0.020 0.136 0.081 0.129 0.047 0.767 0.465 0.121 0.229 0.076 0.160 0.132 0.023 0.200 0.057 0.062 0.014 0.030 0.204 0.391 0.031 0.054 0.021 0.061 0.038 0.069 0.003 0.111 0.065 0.264 0.005 0.065 0.092 0.084 0.110 0.045 0.004 0.014 0.015 0.047 0.009 0.047 0.012 0.032 0.020

Statistical annex

173

Table A2.22

Ranking by royalty and licence payments abroad, 1985 and 1998 (continued)

Rank by payments per capita 1998 1985 62 .. 63 62 64 52 65 63 65 .. 65 .. 65 56 65 .. 65 65 65 60 65 65 65 65 65 65 65 65 59 65 65 65 65 65 65 65

Economy United Republic of Tanzania Paraguay Cameroon Bangladesh Russian Federation Albania Algeria Bahrain Central African Republic Ethiopia Ghana Guatemala Jordan Malawi Mauritius Mozambique Nepal Nicaragua Nigeria Oman Saudi Arabia Sri Lanka Uganda Venezuela Yemen Zambia

Per capita (dollars) 1998 1985 0.1 .. 0.1 0.1 0.2 .. .. .. 0.03 ..

Total value (millions of dollars) 1998 1985 4.7 .. 0.5 1.0 1.6 5.1 1.2 2.0 .. .. .. 1.2 0.2 2.2 ..

Share of GNP (percent) 1998 1985 0.065 .. 0.006 0.001 0.012 0.021 0.012 0.007 0.001 .. .. .. 0.142 0.004 0.002 ..

Source: UNIDO Scoreboard database (see technical annex). Note: Some values are interpolated (see technical annex). All economies with negligible data are given the same rank.

174

Industrial Development Report 2002/2003

Table A2.23

Ranking by modern physical infrastructure, 1985 and 1998 (number of telephone mainlines)

Rank by lines per 1,000 people 1998 1985 1 2 2 1 3 4 4 7 5 3 6 5 7 10 8 8 9 9 10 16 11 17 12 14 13 6 14 18 15 12 16 13 17 19 18 15 19 11 20 21 21 20 22 24 23 26 24 23 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 22 27 28 29 38 47 31 25 40 48 46 32 41 30 42 34 49 33 35 36

Economy Switzerland Sweden United States Norway Denmark Canada Netherlands France Germany Singapore Hong Kong SAR United Kingdom Finland Greece Australia Japan Belgium Austria New Zealand Israel Italy Ireland Korea, Republic of Taiwan Province of China Spain Portugal Slovenia Czech Republic Hungary Turkey Uruguay Bahrain Poland Mauritius Chile Argentina Malaysia Russian Federation Colombia Costa Rica Jamaica Romania Panama Saudi Arabia

Per 1,000 people 1998 1985 675.4 501.6 673.7 627.8 661.3 486.4 660.1 423.2 659.7 497.3 633.9 481.1 593.1 401.8 569.7 416.6 566.8 416.1 562.0 324.3 557.7 323.4 556.9 374.0 553.9 446.6 522.2 313.8 512.1 391.9 502.7 375.2 500.3 307.6 491.0 361.1 479.1 395.8 471.1 278.8 450.7 304.5 434.7 198.6 432.7 159.7 420.1 413.7 413.5 374.8 363.9 335.9 254.1 250.4 245.5 227.6 213.7 205.5 202.7 197.6 196.6 173.5 171.8 165.7 162.4 151.3 142.6 228.5 242.6 145.3 145.2 129.3 69.8 44.4 95.8 167.4 66.9 39.0 44.5 89.9 61.5 102.6 57.3 79.5 33.2 87.6 77.6 71.6

Total number (thousands) 1998 1985 4,799.3 3,245.4 5,963.3 5,242.5 178,751 115,721 2,925.7 1,757.7 3,497.0 2,543.3 19,206 12,481 9,310.6 5,823.5 33,524 22,983 46,505 32,330 1,777.9 805.3 3,729.2 1,764.4 32,889 21,200 2,854.5 2,189.0 5,491.1 3,116.8 9,601.4 6,175.1 63,540 45,300 5,104.6 3,032.4 3,966.1 2,728.3 1,816.8 1,294.8 2,809.1 1,180.0 25,954 17,233 1,610.4 703.0 20,088 6,517.5 9,174.8 16,288 4,121.4 742.9 3,746.2 3,396.8 16,125 823.5 157.8 8,800.4 247.8 3,045.8 7,323.6 4,383.7 4,228.0 9,317.0 1,454.4 286.5 1,336.2 738.8 2,231.1 288.1 71.1 2,487.5 39.6 535.8 2,723.0 963.6

Rank by lines per 1,000 people 1998 1985 45 44 46 37 47 39 48 45 49 50 50 43 51 62 52 52 53 57 54 51 55 77 56 53 57 55 58 58 59 56 60 64 61 54 62 59 63 65 64 66 65 60 66 63 67 70 68 73 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 72 71 61 74 67 69 68 86 76 75 78 80 81 79 85 83 84 82 87

Economy Brazil Venezuela South Africa Mexico Oman Jordan Thailand Tunisia El Salvador Ecuador China Bolivia Peru Egypt Paraguay Morocco Algeria Guatemala Honduras Philippines Nicaragua Albania Sri Lanka Indonesia

Per 1,000 people 1998 1985 120.5 53.3 116.7 70.8 114.6 68.4 103.6 49.5 92.3 29.6 85.5 55.9 83.5 12.3 80.6 26.4 80.0 19.1 78.3 29.5 69.6 3.0 68.8 26.2 66.7 21.2 60.2 18.5 55.3 20.9 54.4 11.0 53.2 24.5 40.8 16.1 38.1 11.0 37.0 9.3 31.3 13.4 30.5 11.2 28.4 5.4 27.0 3.7 4.1 4.6 12.5 3.5 7.0 5.9 6.4 1.2 3.0 3.0 2.8 2.5 2.4 2.7 1.5 2.2 1.7 2.3 1.0

Total number (thousands) 1998 1985 19,989 7,211.2 2,712 1,213.5 4,743 2,141.2 9,928.7 3,739.1 212.6 41.3 390.2 147.9 5,112.8 630.8 752.2 191.6 484.7 91.1 953.0 268.5 86,230 3,120.0 547.1 154.4 1,654.8 412.3 3,696.1 860.4 288.4 75.4 1,509.9 238.6 1,591.5 536.9 441.1 124.6 234.8 45.9 2,782.6 510.3 150.3 45.7 101.9 33.2 532.7 86.2 5,499.9 598.9 21,538 3,174.7 2,549.8 440.2 201.6 103.7 140.1 22.2 221.9 70.9 269.9 117.6 85.5 42.7 194.0 20.1 138.9 37.6 77.2 29.9 67.6 37.4 462.1 204.3 121.9 36.6 380.6 42.1 57.9 168.6 9.5 52.0 19.7 150.6 22.5 24.4 100.7 2.6

28,879 14,758 7,078.7 1,812.8 605.9 209.9 426.8 76.7 3,653.4 1,990.2 418.3 168.1 2,957.8 886.6

India 22.0 Pakistan 19.4 Zimbabwe 17.3 Senegal 15.5 Yemen 13.4 Kenya 9.2 Zambia 8.8 Nepal 8.5 Ghana 7.5 Cameroon 5.4 Mozambique 4.0 Nigeria 3.8 United Republic of Tanzania 3.8 Malawi 3.5 Bangladesh 3.0 Madagascar 2.9 Uganda 2.8 Ethiopia 2.8 Central African Republic 2.7

Source: UNIDO Scoreboard database (see technical annex).

Statistical annex

175

Table A2.24

Ranking by traditional physical infrastructure, 1985 and 1998 (kilograms of oil equivalent)

Rank 1998 1985 1 .. 2 18 3 1 4 2 5 4 6 3 7 9 8 5 9 6 10 11 11 10 12 16 13 7 14 13 15 20 16 .. 17 12 18 8 19 14 20 29 21 15 22 19 23 24 24 27 25 .. 26 28 27 32 28 25 29 31 30 17 31 23 32 26 33 21 34 30 35 38 36 34 37 36 38 22 39 35 40 42 41 41 42 33 43 56 44 40

Economy Bahrain Singapore United States Canada Finland Sweden Belgium Norway Australia Saudi Arabia Netherlands New Zealand Germany France Japan Russian Federation Denmark Czech Republic United Kingdom Korea, Republic of Switzerland Austria Ireland Taiwan Province of China Slovenia Israel Oman Italy Spain Poland South Africa Venezuela Hungary Greece Malaysia Hong Kong SAR Portugal Romania Argentina Chile Jamaica Mexico Thailand Turkey

Commercial energy use per capita 1998 1985 13,688 .. 8,660.5 3,147.8 8,075.6 7,448.8 7,929.9 7,447.7 6,435.0 5,338.2 5,868.6 5,700.0 5,610.9 4,532.7 5,500.8 4,894.3 5,483.8 4,690.4 4,906.3 4,217.5 4,799.8 4,251.9 4,434.6 3,467.5 4,231.4 4,627.8 4,223.6 3,628.7 4,083.5 3,039.5 4,018.8 .. 3,994.3 3,895.0 3,937.8 4,613.0 3,863.4 3,581.7 3,834.5 1,913.5 3,698.9 3,544.5 3,439.1 3,071.9 3,411.9 2,519.2 3,251.4 2,145.2 3,212.6 .. 3,014.2 1,938.6 3,003.1 1,736.6 2,839.1 2,394.4 2,729.4 1,868.5 2,720.7 3,400.7 2,636.3 2,579.5 2,525.8 2,149.8 2,492.5 2,809.1 2,434.6 1,868.9 2,237.2 970.7 2,171.8 1,373.0 2,051.3 1,140.0 1,956.9 2,791.9 1,729.9 1,297.3 1,573.8 611.2 1,551.7 699.7 1,501.1 1,487.2 1,319.5 304.5 1,140.2 773.1

Rank 1998 1985 45 37 46 43 47 50 48 39 49 48 50 51 51 49 52 63 53 53 54 44 55 67 56 47 57 45 58 62 59 61 60 46 61 59 62 52 63 57 64 55 65 66 66 68 67 58 68 71 69 65 70 77 71 78 72 64 73 54 74 69 75 72 76 73 77 60 78 82 79 .. 80 70 81 81 82 .. 83 76 84 75 85 74 86 79 87 80

Economy Jordan Brazil China Algeria Uruguay Zimbabwe Panama Paraguay Costa Rica Colombia Nigeria Tunisia Ecuador Indonesia El Salvador Egypt Zambia Peru Nicaragua Bolivia Guatemala Honduras Philippines Kenya India Mozambique United Republic of Tanzania Pakistan Mauritius Cameroon Sri Lanka Ghana Morocco Nepal Albania Senegal Ethiopia Yemen Bangladesh Madagascar Malawi Central African Republic Uganda

Commercial energy use per capita 1998 1985 1,080.7 1,067.3 1,051.0 605.9 907.0 491.9 904.3 868.2 882.8 504.3 865.5 391.4 856.2 501.2 824.2 173.3 768.8 347.1 761.2 544.7 753.3 148.2 738.5 515.3 713.2 529.5 692.5 220.5 690.8 234.2 655.9 527.0 634.0 236.6 620.7 380.7 550.9 295.3 547.7 312.5 535.6 151.6 531.6 145.1 520.2 259.2 494.1 100.6 479.1 169.8 460.9 37.7 455.3 35.2 442.3 173.1 427.0 345.5 413.4 126.2 385.9 88.7 383.4 77.8 339.6 236.6 320.8 14.5 317.2 .. 315.0 123.3 286.7 16.7 207.9 .. 196.8 40.3 56.2 40.5 46.3 42.2 42.1 33.1 27.6 25.0

Source: UNIDO Scoreboard database (see technical annex).

176

Industrial Development Report 2002/2003

Table A3.1 Ranking of economies by basic indicators of industrial performance and by composite index of competitive industrial performance, 1998

(b)+ Share of medium- and (a)+ high-tech Manufacturing Manufactured activities in value added exports manufacturing per capita per capita value added index index index Rank Economy (a) Economy (b) Economy (c) 1 Switzerland 1 Singapore 0.871 Singapore 0.872 2 Japan 0.852 Ireland 0.663 Ireland 0.775 3 Ireland 0.847 Switzerland 0.661 Switzerland 0.743 4 Singapore 0.743 Belgium 0.497 Japan 0.595 5 Germany 0.705 Japan 0.471 Germany 0.576 6 Finland 0.668 Finland 0.455 Finland 0.513 7 United States 0.637 Sweden 0.447 Sweden 0.510 8 Sweden 0.637 Germany 0.443 United States 0.483 9 Austria 0.624 Austria 0.413 Belgium 0.467 10 Denmark 0.574 Denmark 0.392 Denmark 0.428 11 France 0.572 Netherlands 0.373 Austria 0.402 12 Belgium 0.534 France 0.355 France 0.380 13 United Kingdom 0.502 United States 0.350 United Kingdom 0.372 14 Italy 0.490 United Kingdom 0.314 Netherlands 0.366 15 Netherlands 0.475 Italy 0.306 Canada 0.350 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Norway Canada Taiwan Province of China Portugal Spain New Zealand Israel Australia Slovenia Korea, Republic of Czech Republic Bahrain Argentina Hong Kong SAR Uruguay Hungary Malaysia Greece Brazil Mexico Poland Chile Mauritius Turkey Russian Federation Venezuela Saudi Arabia Peru Thailand South Africa Costa Rica Romania El Salvador Tunisia Jamaica Ecuador Egypt Colombia 0.457 0.419 0.402 0.316 0.315 0.313 0.312 0.299 0.284 0.253 0.193 0.189 0.177 0.169 0.134 0.113 0.112 0.111 0.109 0.102 0.093 0.089 0.088 0.083 0.079 0.072 0.072 0.069 0.069 0.066 0.066 0.055 0.050 0.046 0.044 0.042 0.038 0.038 Canada Norway Taiwan Province of China Israel Slovenia Portugal Spain New Zealand Australia Korea, Republic of Hong Kong SAR Czech Republic Bahrain Malaysia Argentina Hungary Uruguay Mexico Greece Mauritius Brazil Poland Chile Costa Rica Turkey Saudi Arabia Thailand Russian Federation Venezuela South Africa Peru Romania Tunisia Jamaica El Salvador Oman Ecuador Colombia 0.292 0.281 0.275 0.213 0.207 0.194 0.194 0.182 0.167 0.166 0.137 0.136 0.105 0.101 0.094 0.087 0.074 0.067 0.067 0.065 0.058 0.056 0.051 0.048 0.047 0.047 0.046 0.042 0.041 0.038 0.036 0.033 0.031 0.029 0.027 0.023 0.022 0.020 Norway Italy Taiwan Province of China Korea, Republic of Australia Israel Spain New Zealand Slovenia Portugal Hong Kong SAR Czech Republic Malaysia Bahrain Argentina Hungary Greece Mexico Brazil Uruguay Chile Mauritius Turkey Thailand Poland Saudi Arabia Costa Rica Venezuela South Africa Russian Federation Romania Peru Tunisia Jamaica El Salvador Oman Colombia China 0.315 0.303 0.298 0.237 0.222 0.212 0.210 0.189 0.181 0.157 0.121 0.110 0.103 0.095 0.090 0.077 0.063 0.058 0.058 0.056 0.048 0.048 0.048 0.045 0.044 0.041 0.038 0.038 0.037 0.036 0.030 0.027 0.027 0.024 0.021 0.019 0.018 0.017

(c)+ Share of medium-tech and high-tech products in manufactured exports final index Economy (d) Singapore 0.883 Switzerland 0.751 Ireland 0.739 Japan 0.696 Germany 0.632 United States 0.564 Sweden 0.562 Finland 0.538 Belgium 0.495 United Kingdom 0.473 France 0.465 Austria 0.453 Denmark 0.443 Netherlands 0.429 Taiwan Province of China 0.412 Canada 0.407 Italy 0.384 Korea, Republic of 0.370 Spain Israel Norway Malaysia Mexico Czech Republic Philippines Portugal Hungary Slovenia Australia Hong Kong SAR New Zealand Thailand Brazil Poland Argentina Costa Rica China Turkey South Africa Greece Romania Bahrain Uruguay Russian Federation Tunisia Venezuela Chile Guatemala Indonesia India Zimbabwe El Salvador Morocco 0.319 0.301 0.301 0.278 0.246 0.243 0.241 0.240 0.239 0.221 0.211 0.204 0.186 0.172 0.149 0.143 0.140 0.129 0.126 0.108 0.108 0.102 0.095 0.089 0.087 0.077 0.068 0.060 0.056 0.056 0.054 0.054 0.052 0.051 0.048

Statistical annex

177

Table A3.1 Ranking of economies by basic indicators of industrial performance and by composite index of competitive industrial performance, 1998 (continued)

Rank 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87

Economy Oman China Panama Paraguay Guatemala Morocco Philippines Jordan Albania Bolivia Algeria Honduras Sri Lanka Indonesia Senegal Zimbabwe Pakistan Nicaragua India Cameroon Nigeria Bangladesh Zambia Kenya Yemen Madagascar Central African Republic Uganda Mozambique Malawi Nepal United Republic of Tanzania Ghana Ethiopia

Manufacturing value added per capita index (a) 0.034 0.034 0.032 0.029 0.028 0.025 0.022 0.022 0.021 0.020 0.018 0.016 0.014 0.013 0.009 0.008 0.008 0.007 0.007 0.007 0.007 0.006 0.004 0.003 0.003 0.002

Economy Egypt China Panama Philippines Guatemala Paraguay Morocco Jordan Bolivia Albania Algeria Sri Lanka Honduras Indonesia Zimbabwe Senegal Pakistan Nicaragua Cameroon India Bangladesh Nigeria Kenya Zambia Yemen Central African Republic Madagascar Uganda Mozambique Malawi Nepal United Republic of Tanzania Ghana Ethiopia

(a)+ Manufactured exports per capita index (b) 0.020 0.019 0.017 0.017 0.016 0.015 0.014 0.012 0.011 0.011 0.010 0.009 0.009 0.008 0.005 0.005 0.005 0.004 0.004 0.004 0.004 0.003 0.002 0.002 0.002 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.000 0.000

Economy Ecuador Egypt Philippines Panama Guatemala Morocco Jordan Albania Paraguay Algeria Indonesia Bolivia Sri Lanka Nicaragua Honduras Zimbabwe Senegal India Pakistan Mozambique Cameroon Bangladesh Nigeria Zambia Kenya Yemen Central African Republic Madagascar Malawi Uganda Ghana Nepal United Republic of Tanzania Ethiopia

(b)+ Share of medium- and high-tech activities in manufacturing value added index (c) 0.016 0.015 0.015 0.013 0.013 0.012 0.012 0.011 0.011 0.009 0.009 0.008 0.007 0.007 0.006 0.006 0.004 0.004 0.004 0.004 0.003 0.003 0.002 0.002 0.002 0.002

Economy Saudi Arabia Colombia Mauritius Egypt Peru Oman Pakistan Ecuador Kenya Jordan Honduras Jamaica Panama Bolivia Albania Sri Lanka Nicaragua Paraguay Mozambique Bangladesh Algeria Cameroon Senegal Zambia Nigeria Nepal United Republic of Tanzania Malawi Madagascar Central African Republic Uganda Yemen Ghana

(c)+ Share of medium-tech and high-tech products in manufactured exports final index (d) 0.047 0.041 0.041 0.038 0.035 0.032 0.031 0.025 0.025 0.024 0.023 0.022 0.022 0.021 0.021 0.017 0.017 0.015 0.013 0.011 0.009 0.008 0.008 0.007 0.006 0.006

0.002 0.002 0.002 0.002 0.001 0.001 0.000 0.000

0.001 0.001 0.001 0.001 0.001 0.001

0.005 0.003 0.003 0.003 0.003 0.001 0.001 0.000

0.000 0.000

Ethiopia

Source: UNIDO Scoreboard data set (see technical annex). Note: Ranking is based on individual indicators contained in detailed annex tables for chapter 2. Column b is the average of a and b, c is the average of b and c and d is the average of all individual indices (the final CIP index).

178

Industrial Development Report 2002/2003

Table A3.2 Ranking of economies by basic indicators of industrial performance and by composite index of competitive industrial performance, 1985

(b)+ Share of medium- and (a)+ high-tech Manufacturing Manufactured activities in value added exports manufacturing per capita per capita value added index index index Rank Economy (a) Economy (b) Economy (c) 1 Switzerland 1 Switzerland 0.760 Switzerland 0.840 2 United States 0.873 Singapore 0.717 Japan 0.633 3 Japan 0.851 Sweden 0.562 Sweden 0.627 4 Sweden 0.673 Belgium 0.521 Singapore 0.616 5 Germany 0.655 Japan 0.518 Germany 0.592 6 Finland 0.645 Finland 0.490 United States 0.551 7 Canada 0.585 United States 0.481 Finland 0.523 8 France 0.539 Germany 0.466 Belgium 0.491 9 Austria 0.529 Canada 0.451 Canada 0.448 10 Norway 0.491 Netherlands 0.431 Denmark 0.420 11 Denmark 0.484 Austria 0.402 Netherlands 0.401 12 Italy 0.484 Denmark 0.401 Austria 0.401 13 Belgium 0.457 France 0.368 France 0.397 14 United Kingdom 0.449 Norway 0.368 United Kingdom 0.378 15 Australia 0.437 Hong Kong SAR 0.362 Norway 0.377 16 Singapore 0.434 Italy 0.328 Ireland 0.340 17 Netherlands 0.403 Ireland 0.324 Italy 0.323 18 New Zealand 0.387 United Kingdom 0.310 Hong Kong SAR 0.291 19 Ireland 0.345 Taiwan Province Australia 0.259 of China 0.265 20 Hong Kong SAR 0.341 Australia 0.249 Israel 0.246 21 Taiwan Province New Zealand 0.235 Taiwan Province of China 0.325 of China 0.243 22 Spain 0.297 Israel 0.210 New Zealand 0.223 23 Israel 0.250 Spain 0.184 Spain 0.190 24 Romania 0.224 Bahrain 0.153 Korea, Republic of 0.130 25 Bahrain 0.224 Korea, Republic of 0.132 Argentina 0.120 26 Argentina 0.222 Venezuela 0.128 Portugal 0.108 27 Venezuela 0.203 Portugal 0.126 Bahrain 0.102 28 Portugal 0.182 Argentina 0.119 Venezuela 0.099 29 Korea, Republic of 0.171 Romania 0.112 Romania 0.096 30 Hungary 0.167 Poland 0.092 Hungary 0.094 31 Poland 0.160 Saudi Arabia 0.092 Brazil 0.088 32 Greece 0.140 Greece 0.092 Greece 0.080 33 Saudi Arabia 0.140 Hungary 0.088 Poland 0.077 34 Brazil 0.130 Malaysia 0.083 Saudi Arabia 0.077 35 Mexico 0.126 Brazil 0.073 Malaysia 0.067 36 Uruguay 0.118 Mexico 0.070 Mexico 0.062 37 Malaysia 0.093 Uruguay 0.069 South Africa 0.059 38 South Africa 0.093 South Africa 0.057 Uruguay 0.057 39 Ecuador 0.084 Mauritius 0.049 Costa Rica 0.037 40 Costa Rica 0.083 Costa Rica 0.048 Algeria 0.037 41 Panama 0.075 Algeria 0.045 Peru 0.036 42 Algeria 0.066 Ecuador 0.045 Mauritius 0.035 43 Turkey 0.061 Panama 0.040 Turkey 0.034 44 Guatemala 0.057 Turkey 0.038 Panama 0.032 45 Peru 0.056 Jamaica 0.036 Ecuador 0.032 46 Nicaragua 0.054 Chile 0.033 Chile 0.030 47 Chile 0.054 Peru 0.033 Jamaica 0.030 48 Colombia 0.053 Guatemala 0.030 Nicaragua 0.024 49 El Salvador 0.050 Tunisia 0.029 Tunisia 0.023 50 Jordan 0.049 Jordan 0.029 Colombia 0.023 51 Mauritius 0.045 Colombia 0.029 Guatemala 0.022 52 Jamaica 0.044 Nicaragua 0.028 Jordan 0.021 53 Paraguay 0.044 El Salvador 0.026 El Salvador 0.020

(c)+ Share of medium-tech and high-tech products in manufactured exports final index Economy (d) Switzerland 0.808 Japan 0.725 Germany 0.635 Sweden 0.633 United States 0.599 Singapore 0.587 Finland 0.494 Belgium 0.489 Canada 0.474 France 0.450 Austria 0.445 United Kingdom 0.426 Denmark 0.424 Netherlands 0.398 Ireland 0.379 Italy 0.379 Norway 0.348 Hong Kong SAR 0.320 Taiwan Province of China 0.292 Israel 0.290 Spain Korea, Republic of Australia New Zealand Poland Portugal Brazil Mexico Argentina Malaysia Bahrain South Africa Greece Hungary Venezuela Turkey Romania Zimbabwe Oman Tunisia Saudi Arabia Uruguay Thailand Costa Rica Philippines Morocco Mauritius Peru Colombia India Panama Jamaica Chile 0.259 0.247 0.214 0.188 0.176 0.159 0.140 0.125 0.122 0.116 0.099 0.096 0.093 0.088 0.085 0.082 0.072 0.071 0.069 0.064 0.063 0.062 0.058 0.053 0.044 0.038 0.037 0.037 0.035 0.034 0.032 0.032 0.030

Statistical annex

179

Table A3.2 Ranking of economies by basic indicators of industrial performance and by composite index of competitive industrial performance, 1985 (continued)

Rank 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

Economy Tunisia Thailand Oman Philippines Egypt Bolivia Zimbabwe Honduras Morocco China Cameroon Indonesia Nigeria Zambia Ghana Senegal Sri Lanka Pakistan India Bangladesh Kenya United Republic of Tanzania Central African Republic Madagascar Malawi Uganda Nepal

Manufacturing value added per capita index (a) 0.043 0.041 0.040 0.035 0.032 0.032 0.030 0.027 0.026 0.024 0.022 0.020 0.020 0.018 0.013 0.011 0.011 0.010 0.009 0.007 0.006

Economy Thailand Oman Paraguay Philippines Zimbabwe Bolivia Morocco Egypt Honduras Cameroon China Indonesia Nigeria Zambia Sri Lanka Senegal Ghana Pakistan India Kenya Central African Republic Bangladesh United Republic of Tanzania Madagascar Malawi Nepal Uganda

(a)+ Manufactured exports per capita index (b) 0.025 0.024 0.023 0.020 0.018 0.018 0.017 0.017 0.017 0.013 0.012 0.012 0.010 0.010 0.008 0.007 0.007 0.006 0.005 0.004 0.004 0.004

Economy Thailand Oman Paraguay Zimbabwe Philippines Morocco Egypt Bolivia Honduras China Cameroon Zambia Indonesia Nigeria Senegal Sri Lanka Ghana Pakistan India Kenya Central African Republic Bangladesh United Republic of Tanzania Madagascar Malawi Nepal Uganda

(b)+ Share of medium- and high-tech activities in manufacturing value added index (c) 0.020 0.018 0.017 0.017 0.015 0.013 0.013 0.012 0.012 0.011 0.010 0.009 0.009 0.007 0.006 0.006 0.005 0.005 0.004 0.004 0.003 0.003

Economy Algeria Pakistan Guatemala El Salvador Ecuador Senegal Jordan China Nicaragua Paraguay Kenya Indonesia Honduras Egypt Zambia Bolivia United Republic of Tanzania Sri Lanka Cameroon Madagascar Bangladesh Nigeria

(c)+ Share of medium-tech and high-tech products in manufactured exports final index (d) 0.029 0.028 0.028 0.027 0.025 0.023 0.022 0.021 0.020 0.013 0.013 0.012 0.012 0.012 0.010 0.009 0.009 0.008 0.008 0.008 0.008 0.006

0.006 0.006 0.005 0.003 0.000 0.000 0.003 0.003 0.002 0.000 0.000 0.002 0.002 0.002 0.000 0.000 Ghana Central African Republic Malawi Nepal Uganda 0.006 0.003 0.003 0.001 0.001

Source: UNIDO Scoreboard data set (see technical annex). Note: Ranking is based on individual indicators contained in detailed annex tables for chapter 2. Column b is the average of a and b, c is the average of b and c and d is the average of all individual indices (the final CIP index).

Table A3.3

Regression results for export structure and growth in manufactured exports

46 large exporters Independent variable Share of medium- and high-tech exports in 1985 Change in share over time Standard coefficient 0.408* 0.873** t-statistic 3.026 6.452 Adjusted R2 = 0.489 Mean 0.63 0.02 Standard coefficient 0.420* 0.364*

34 small exporters t-statistic 2.356 2.013 Adjusted R2 = 0.131 Mean 0.08 0.07

Source: Calculations by UNIDO based on UNIDO Scoreboard database (see technical annex). * Significant at the 5 percent level. ** Significant at the 1 percent level. Note: The dependent variable is annual compound growth rate of manufactured exports, 19851998. The regressions satisfy all statistical tests of heteroskedasticity and collinearity.

180

Industrial Development Report 2002/2003

Bibliography

Background papers
Baxter, A., J. Perkin and M. Mulligan. Information and communication technology: implications for industry in developing countries. Best, M. Globalization and localization of value networks. Chang, H.-J. Technology transfer, intellectual property rights and industrial development in developing countries. Chudnovsky, D. National innovation systems in an FDI-led development process: the Argentine case in the 1990s. Cooke, P. Strategies for regional innovation systems: learning transfer and application. De Bandt, J. Learning processes: requirements and difficulties. Debresson, C., X. Wei, X. Shiqing and P. Mohnen. Strengths and weaknesses of innovative clusters and learning capablities in China before it joins the multilateral trade system. Diab, T. National industrial innovation systemsArab region. Edquist, C. Systems of innovation and development (SID). Gereffi, G. Prospects for industrial upgrading by developing countries in the global apparel commodity chain. Hamann, K. Globalization trends in the food and agro industries: value chains and the future competitive situation. Humphrey, J. Global value chains and local development in the automotive industry. Inklaar, A. Standards regulation and quality. Kaplinsky, R., M. Morris and J. Readman. Globalization and upgrading: innovation and learning in the wood furniture value chain. Khemani, R. S. The role and importance of competition policy in fostering corporate governance and competitiveness. Kumar, N. Nature and determinants on technology upgrading and innovation in the Indian software industry. Mathews, J. A. Catching-up strategies in technology development, with particular reference to East Asia. Metcalfe, S. Technology and economic development: a comparative perspective. Pietrobelli, C. National industrial systems in Africa: the nature and deficiencies of technological effort in African industry. Radosevic, S. Nature and determinants of innovation and technology upgrading in industry in Central and Eastern Europe. Salazar de Buckle, T. The leather global value chain.

Scheel, C., and M. A. Prez G. National innovation systems in Latin America. Teubal, M. The systems perspectives to innovation and technology policy (ITP): theory and application to developing countries in NIEs. Wallace, D. Promoting environmentally sound technologies through innovation in business and policies.

References
Adelman, I., and L. E. Westphal. 1979. Industrial priorities in the Republic of Korea. In Industrial Priorities in Developing Countries: The Selection Process in Brazil, India, Mexico, Republic of Korea, and Turkey. Vienna: United Nations Industrial Development Organization. Amsden, A. H. 1989. Asias Next Giant: South Korea and Late Industrialization. New York: Oxford University Press. . 2001. The Rise of the Rest: Challenges to the West from Late-Industrializing Economies. Oxford: Oxford University Press. Appelbaum, R., and G. Gereffi. 1994. "Power and profits in the apparel commodity chain." In E. Bonacich and others, eds., Global Production: The Apparel Industry in the Pacific Rim. Philadelphia, Pa.: Temple University Press. Barro, R. J., and J. W. Lee. 1993. International comparisons of educational attainment. Journal of Monetary Economics 32: 363394. . 1996. International measures of schooling years and schooling quality. American Economic Review, Papers and Proceedings 86(2): 218223. Barry, F., ed. 1999. Understanding Irelands Economic Growth. Basingstoke: Palgrave. Bell, M., and M. Albu. 1999. Knowledge systems and technological dynamism in industrial clusters in developing countries. World Development 27(9): 17151733. Bell, M., and K. Pavitt. 1993. Technological accumulation and industrial growth: contrasts between developed and developing countries. Industrial and Corporate Change 2(2): 157210. Bell, M., B. Ross-Larson, and L.E. Westphal. 1984. "Assessing the performance of infant industries." Journal of Development Economics 16(September): 101128.

181

Benavente, J. M., G. Crispi, J. M. Katz and G. Stumpo. 1997. New problems and opportunities for industrial development in Latin America. Oxford Development Studies 25: 261278. Best, M. 1990. The New Competition: Institutions of Industrial Restructuring. Cambridge: Polity. . 2001. The New Competitive Advantage: The Renewal of American Industry. Oxford: Oxford University Press. Borrus, M., and J. Zysman. 1997. Wintelism and the changing terms of global competition: prototype of the future? BRIE Working Paper 96B. Berkeley Roundtable on the International Economy, Berkeley, Calif. [http://brie.berkeley.edu]. Cantwell, J., and O. Janne. 1998. Globalisation of innovatory capacity: the structure of competence accumulation in European home and host countries. Discussion Papers in International Investment and Management, no. 253. University of Reading, Department of Economics. Cantwell, J., and G. D. Santangelo. 2000. Capitalism, profits and innovation in the new techno-economic paradigm. Journal of Evolutionary Economics 10: 131157. Carlsson, B. 1995. Technological Systems and Economic Performance: The Case of Factory Automation. Dordrecht: Kluwer Academic. Chandler, A. D. 1990. Scale and Scope: The Dynamics of Industrial Capitalism. Cambridge, Mass.: Belknap. . 1992. Organizational capabilities and the economic history of the industrial enterprise. Journal of Economic Perspectives 6(3): 79100. Chenery, H. B., S. Robinson and M. Syrquin. 1986. Industrialization and Growth: A Comparative Study. Oxford: Oxford University Press for the World Bank. Cheng, T.-J., S. Haggard and D. Kang. 1999. Institutions and growth in Korea and Taiwan: the bureaucracy. In Y. Akyuz, ed., East Asian Development: New Perspectives. London: Cass. Cimoli, M., ed. 2000. Developing Innovation Systems: Mexico in a Global Context. London: Continuum. Cyhn, J. 2001. Technology Transfer and International Production: The Development of the Electronics Industry in Korea. Cheltenham: Edward Elgar. Dahlman, C. J., and J-E. Aubert. 2001. China and the Knowledge Economy. Washington, D.C.: World Bank. Dahlman, C. J., and O. Sananikone. 1990. Technology strategy in Taiwan Province of China: exploiting foreign linkages and investing in local capability. World Bank, Washington, D.C. Dahlman, C. J., B. Ross-Larson and L. E. Westphal. 1987. Managing technological development: lessons from newly industrializing countries. World Development 15(6): 759775. Dicken, P. 1998. Global Shift: Transforming the World Economy. 3d ed. London: Paul Chapman. Dodgson, M., D. M. Gann and A. J. Salter. 2001. The intensification of innovation. SPRU Electronic Working Paper 65. University of Sussex, Science Policy Research Unit. [http://www. sussex.ac.uk/spru]. Dosi, G., D. Teece and J. Chytry, eds. 1998. Technology, Organisation and Competitiveness. Oxford: Oxford University Press.

DTI (U.K. Department of Trade and Industry). 1998. Our Competitive Future: Building the Knowledge-Driven Economy. London. Dunning, J. H. 1993. Multinational Enterprises and the Global Economy. Wokingham: Addison Wesley. . 1997. Alliance Capitalism and Global Business. London: Routledge. Easterly, W. 2001. The Elusive Quest for Growth: Economists Adventures and Misadventures in the Tropics. Cambridge, Mass.: MIT Press. Economist. 1996. World economy survey. 28 September, London. . 2000a. Have factory, will travel. 10 February, London. . 2000b. Untangling e-conomics: survey of the new economy. 23 September, London. Edquist, C., and M. McKelvey, eds. 2001. Systems of Innovation: Growth, Competitiveness and Employment. Cheltenham: Edward Elgar. Enos, J. 1992. The Creation of Technological Capabilities in Developing Countries. London: Pinter. Ernst, D. 1997. From partial to systemic globalization: international production networks in the electronics industry. BRIE Working Paper 98. Berkeley Roundtable on the International Economy, Berkeley, Calif. . 2000. Carriers of cross-border knowledge diffusion: information technology and global production networks. EastWest Center Working Papers, Economics Series, no. 3. East-West Center, Honolulu. [http://www.eastwestcenter.org/ about-dy-detail.asp?staff_ID=141]. Ernst, D., T. Ganiatsos and L. Mytelka, eds. 1995. Technological Capabilities and Export Performance: Lessons from East Asia. Cambridge: Cambridge University Press. European Commission. 2000. European Trend Chart on Innovation: the European Innovation Scoreboard. Directorate General, Enterprises Innovation and SME Programme. [http://trendchart. cordis.lu/Reports/Documents/EuropeanInnovationScoreboard_ backgroundpaper.pdf]. Evans, P. 1999. Transferable lessons? Re-examining the institutional prerequisites for East Asian economic policies. In Y. Akyuz, ed., East Asian Development: New Perspectives. London: Frank Cass. Felipe, J. 1999. Total factor productivity growth in East Asia: a critical survey. Journal of Development Studies 35(4): 141. Figueiredo, P. N. Forthcoming. Technological Learning and Competitive Performance. Cheltenham: Edward Elgar. Frankel, J. A. 2000. Globalization of the Economy. NBER Working Paper 7858. Cambridge, Mass.: National Bureau of Economic Research. Freeman, C., and C. Perez. 1988. Structural crises of adjustment, business cycles and investment behaviour. In G. Dosi, C. Freeman, R. Nelson, G. Silverberg and L. Soete, eds., Technical Change and Economic Theory. London: Pinter. Galenson, W. 1979. Economic Growth and Structural Change in Taiwan. Ithaca: Cornell University Press. Gereffi, G. 1997. "Global shifts, regional response: can North America meet the full-package challenge?" Bobbin 39(3): 1631.

182

Industrial Development Report 2002/2003

. 1999a. "A commodity chains framework for analysing global industries." Institute of Development Studies, Brighton, UK. . 1999b. "International trade and industrial upgrading in the apparel commodity chain." Journal of International Economics 48(1): 3770 [http://www.sciencedirect.com]. . 2000. "The transformation of the North American apparel industry: Is NAFTA a curse or a blessing?" Integration and Trade 4(11): 4795. Gerschenkron, A. 1962. Economic Backwardness in Historical Perspective. Cambridge, Mass.: Belknap. Golub, S., and A. A. Mbaye. 2000. "Obstacles and opportunities for Senegals international competitiveness: case studies of the peanut oil, fishing and textile industries." Research paper. Swarthmore College, Swarthmore, Pa. Gordon, R. 2000. Does the new economy measure up to the great inventions of the past? Journal of Economic Perspectives 14(4): 4974. Granitsas, A. 1998. "Back in fashion: Hong Kongs leading garment makers are going globallearning to add value and high technology." Far Eastern Economic Review May 21: 5254. Guerrieri, P., S. Iammarino and C. Pietrobelli. 2001. The Global Challenge of Industrial Districts: Small and Medium-Sized Enterprises in Italy and Taiwan. Cheltenham: Edward Elgar. Hamel, G., and C.K. Prahalad. 1994. Competing for the future." Harvard Business Review 72(3): 122128. Harbison, F. H., and C. S. Myers. 1964. Education, Manpower and Economic Growth. New York: McGraw-Hill. Hobday, M. G. 1995. Innovation in East Asia: The Challenge to Japan. Cheltenham: Edward Elgar. Hoff, K., and J. E. Stiglitz. 2001. Modern economic theory and development. In G. M. Meier and J. E. Stiglitz, eds., Frontiers of Development Economics: The Future in Perspective. New York: Oxford University Press for the World Bank. Hou, C.-M., and S. Gee. 1993. "National systems supporting technical advance in industry: the case of Taiwan Province of China." In R. R. Nelson, ed., National Innovation Systems: A Comparative Analysis. New York: Oxford University Press. Humphrey, J. 2000. "Assembler-supplier relations in the auto industry: globalization and national development." Competition and Change 4(3): 245271. Humphrey, J., and H. Schmitz. 1998. Trust and inter-firm relations in developing and transition economies. Journal of Development Studies 34(4): 3261. . 2000. Governance and upgrading: linking industrial cluster and global value chain research. Working Paper 120. Institute of Development Studies, Brighton. ILO (International Labour Organization). 1995. Recent Developments in the Clothing Industry. Geneva: International Labour Office. . 1998. World Employment Report 199899: Employability in the Global EconomyHow Training Matters. Geneva: International Labour Office. . 2001. World Employment Report 2001: Life at Work in the Information Economy. Geneva: International Labour Office.

IMD (International Institute for Management Development). 2000. World Competitiveness Yearbook. Lausanne. IMF (International Monetary Fund). 1999. Balance of Payments Statistics Yearbook 1999. Washington, D.C. ITRI (Industrial Technology Research Institute). 1988. "Industrial Technology Research Institute: 15th Anniversary." Chung Hsing, Taiwan Province of China. Kaplinsky, R. 1993. "Export processing zones in the Dominican Republic: transforming manufactures into commodities." World Development 21(11): 18511865. . 1998. "Globalization, industrialization and sustainable growth: the pursuit of the nth rent." IDS Discussion Paper 365. University of Sussex Institute of Development Studies, Brighton, UK. . 2000. "Globalization and unequalization: what can be learned from value chain analysis?" Journal of Development Studies 37(2): 117146. Kaplinsky, R., and J. Readman. 2000. "Globalization and upgrading: what can (and cannot) be learnt from international trade statistics in the wood furniture sector?" Institute of Development Studies and Center for Research in Innovation Management, University of Brighton, UK. Katz, J. M., ed. 1987. Technology Generation in Latin American Manufacturing Industries. London: Macmillan. Kim, L. S. 1998. "Technology policies and strategies for Southeast Asian countries: lessons from the Korean experience." Technology Analysis and Strategic Management 10(3): 311323. . 2000. "The dynamics of public policy, corporate strategy, and technological learning: lessons from the Korea experience." College of Business Administration, Korea University, Seoul. Kim, L., and R. R. Nelson. 2000. "Technology, Learning, and Innovation: Experiences of Newly Industrializing Economies." Cambridge: Cambridge University Press. Kim, S. R. 1998. "The Korean system of innovation and the semiconductor industry: a governance perspective." Industrial and Corporate Change 7(2): 275309. Kline, S., and N. Rosenberg. 1986. An overview of innovation. In R. Landau and N. Rosenberg, eds., The Positive Sum Strategy. Washington, D.C.: National Academy Press. Kolodny, H., B. Stymne, R. Shani, J. R. Figuera and P. Lillrank. 2001. "Design and policy choices for technology extension organizations." Research Policy 30(2): 201225. Krugman, P. R. 1991. Geography and Trade. Cambridge, Mass.: MIT Press. . 1994. Competitiveness: a dangerous obsession. Foreign Affairs 73(2): 2844. Lall, S. 1992. Technological capabilities and industrialization. World Development 20(2): 165186. . 1996. Learning from the Asian Tigers: Studies in Technology and Industrial Policy. London: Macmillan. . 1999a. Competing with labour. Issues in Development Discussion Paper 31. International Labour Organization, Geneva.

Bibliography

183

, ed. 1999b. The Technological Response to Import Liberalization in Sub-Saharan Africa. London: Macmillan. . 2000. The technological structure and performance of developing country manufactured exports, 198598. Oxford Development Studies 28(3): 337369. . 2001a. Competitiveness indices and developing countries: an economic evaluation of the Global Competitiveness Report. World Development 29(9): 15011525. . 2001b. Competitiveness, Technology and Skills. Cheltenham, U.K.: Edward Elgar. Lall, S., and M. Teubal. 1998. Market-stimulating technology policies in developing countries: a framework with examples from East Asia. World Development 26(8): 13691386. Lall, S., and G. Wignaraja. 1998. Mauritius: Dynamizing Export Competitiveness. Economic Paper 33. London: Commonwealth Secretariat. Lall, S., G. B. Barba-Navaretti, S. Teitel and G. Wignaraja. 1994. Technology and Enterprise Development: Ghana under Structural Adjustment. London: Macmillan. Lall, S., G. Wignaraja, M. Selleck and P. Robinson. 1997. Zimbabwe: enhancing export competitiveness. Report for the Zimbabwe Ministry of Industry and Commerce. Sponsored by the Commonwealth Secretariat, London. Lindert, P. H., and J. G. Williamson. 2001. Globalisation and inequality: a long history. Paper presented at the Annual World Bank Conference on Development EconomicsEurope, Barcelona, 2527 June. Loewendahl, H. 2001. A framework for FDI promotion. Transnational Corporations 10(1): 142. Lucas, R. E. 1988. On the mechanics of economic development. Journal of Monetary Economics 22: 342. Lundvall, B.-A., ed. 1992. National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning. London: Pinter. Luthria, M. 2000. Fostering innovation in developing countries. World Bank, East Asia Department, Washington, D.C. Draft. Magarios, C. A. 2000. Marginalization versus prosperity: improving the creation and distribution of gains brought by the process of globalisationreflections on the development agenda. Vienna: United Nations Industrial Development Organization (UNIDO). . 2001. From marginalization to prosperity: how to improve and spread the gains of globalization, In C. A. Magarios and F. C. Sercovich, eds., Gearing Up for a New Development Agenda. Vienna: United Nations Industrial Development Organization. Magarios, C. A., and F. C. Sercovich, eds. 2001. Gearing Up for A New Development Agenda. Papers and Proceedings of the Meeting on Marginalization vs Prosperity: How to Improve and Spread the Gains of Globalization, United Nations Industrial Development Organization, Vienna. Magarios, C. A., G. Assaf, S. Lall, J. D. Martinussen, R. Ricupero and F. Sercovich. 2001. Reforming the UN System: UNIDOs NeedDriven Model. The Hague: Kluwer Law International for United Nations Industrial Development Organization (UNIDO).

Mansell, R., and U. Wehn. 1998. Knowledge Societies: Information Technology for Sustainable Development. Oxford: Oxford University Press for the United Nations Commission on Science and Technology for Development. Martin, B. 1996. Technology foresight: a review of recent government exercises. STI Review 17: 1550. Maskus, K. 2000. Intellectual Property Rights in the Global Economy. Washington, D.C.: Institute for International Economics. [http://www.iie.com/publications/publication.cfm?pub_id=99]. Mathews, J. A. 2002. Dragon Multinational: A New Model for Global Growth. Oxford: Oxford University Press. . Forthcoming. "The origins of dynamics of Taiwan"s R&D consortia." Research Policy. Mathews, J. A., and D. S. Cho. 2000. Tiger Technology: The Creation of a Semiconductor Industry in East Asia. Cambridge: Cambridge University Press. McCulloch, N., L. A. Winters and X. Cirera. 2001. Trade Liberalization and Poverty: A Handbook. London: Centre for Economic Policy Research and U.K. Department for International Development. Metcalfe, J. S. 1995. Technology systems and technology policy in an evolutionary framework. Cambridge Journal of Economics 19(1): 2546. Morris, M. 2000. "Creating value-chain cooperation." IDS Bulletin 32(3): 127136. Mowery, D. C., and N. Rosenberg. 1989. Technology and the Pursuit of Economic Growth. Cambridge: Cambridge University Press. . 1998. Paths of Innovation: Technological Change in 20thCentury America. Cambridge: Cambridge University Press. Nadvi, K. 2001. Industrial Clusters and International Competitiveness. Basingstoke: Palgrave. Narula, R., and J. H. Dunning. 2000. Industrial development, globalisation and multinational enterprises: new realities for developing countries. Oxford Development Studies 28(2): 141168. Nelson, R. R., ed. 1993. National Innovation Systems: A Comparative Analysis. Oxford: Oxford University Press. Nelson, R. R., and H. Pack. 1999. The Asian miracle and modern growth theory. Economic Journal 109(July): 416436. Nelson, R. R., and S. J. Winter. 1982. An Evolutionary Theory of Economic Change. Cambridge, Mass.: Harvard University Press. Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organizational Science 5(1): 1437. NSF (National Science Foundation). 1998. Science and Engineering Indicators 1998. Washington, D.C. . 2000. Science and Engineering Indicators 2000. Washington, D.C. OECD (Organisation for Economic Co-operation and Development). 1992. Technology and the Economy: The Key Relationships. Paris. . 1996a. Globalisation of Industry. Paris. . 1996b. The Knowledge-Based Economy. Paris. . 1996c. Technology and Industrial Performance: Technology Diffusion, Productivity, Employment and Skills, International Competitiveness. Paris.

184

Industrial Development Report 2002/2003

. 1999a. Globalisation of Industrial R&D: Policy Issues. Paris. . 1999b. Science, Technology and Industry Scoreboard 1999. Paris. . 2000a. Differences in Economic Growth across the OECD in the 1990s: The Role of Innovation and Information Technologies. Paris: OECD, Directorate for Science, Technology and Industry. . 2000b. Measuring the ICT Sector. Paris. . 2000c. A New Economy? The Changing Role of Innovation and Information Technology in Growth. Paris. OHearn, D. 1998. Inside the Celtic Tiger. London: Pluto. Pack, H., and L. Westphal. 1986. "Industrial strategy and technological change: theory versus reality." Journal of Development Economics 22(June): 87128. Pavitt, K. 2001. Can the large Penrosian firms cope with the dynamics of technology? SPRU Electronic Working Paper 68. University of Sussex, Science Policy Research Unit. [http:// www.sussex.ac.uk/spru]. Pigato, M. 2001. Information and communication technology, poverty and development in Sub-Saharan Africa and South Asia. World Bank, Africa Department, Washington, D.C. Pohjola, M. 1998. Information technology and economic development: an introduction to the research issues. Working Paper 153. United Nations University, World Institute for Development Economics Research, Helsinki. Porter, M. E. 1990. The Competitive Advantage of Nations. London: Macmillan. Porter, M. E., and S. Stern. 2000. The New Challenge to Americas Prosperity: Findings from the Innovation Index. Washington, D.C.: Council on Competitiveness. Prahalad, C. K., and Hamel, G. 1990. "The core competence of the corporation." Harvard Business Review 68(3): 7991. Prokopenko, J., ed. 1999. "Productivity promotion organizations: evolution and experience." Free Working Paper, PMD/1/E. International Labour Organization Management Development Programme, Geneva. [http://www.ilo.org/public/english/ employment/ent/mandev/publ/publist.htm] Pyke, F., G. Becattini and W. Sengenberger, eds. 1990. Industrial Districts and Inter-Firm Co-operation in Italy. Geneva: International Labour Organization, International Institute for Labour Studies. Quah, D. T. 1999. The weightless economy in growth. Business Economist 30: 4053. Radosevic, S. 1999. International Technology Transfer and Catch-Up in Economic Development. Cheltenham: Edward Elgar. Rasiah, R. 1995. Foreign Capital and Industrialization in Malaysia. London: Macmillan. . 2000. Industrial technology transition in Malaysia: implications for developing economies. Draft prepared for World Bank Institute, Washington, D.C. National University of Malaysia, Kuala Lumpur. Redding, S. 1999. "Dynamic comparative advantage and the welfare effects of trade." Oxford Economic Papers 51(1): 1539.

Reinert, E. 1995. Competitiveness and its predecessors: a 500-year cross-national perspective. Structural Change and Economic Dynamics 6: 2342. Rodrigo, C. G. 2001. Technology, Economic Growth and Crises in East Asia. Cheltenham: Edward Elgar. Rodrik, D. 1996. Coordination failures and government policy: a model with applications to East Asia and Eastern Europe. Journal of International Economics 40(1/2): 122. . 2001. The global governance of trade as if development really mattered. Background paper for United Nations Development Programme, New York. [http://ksghome. harvard.edu/~.drodrik.academic.ksg/papers.html]. Romer, P. 1993. Idea gaps and object gaps in economic development. Journal of Monetary Economics 32: 531555. Schive, C. 1990. The Foreign Factor: The Multinational Corporations Contribution to the Economic Modernization of the Republic of China. Stanford, Ca.: Hoover Institution Press. Schmitz, H. 1995a. "Collective efficiency: growth path for smallscale industry." Journal of Development Studies 31(4): 529 566. . 1995b. "Small shoemakers and Fordist giants: tale of a supercluster." World Development 23(1): 928. . 1997. Collective efficiency and increasing returns. Working Paper 50. Institute of Development Studies, Brighton. . 1999a. From ascribed to earned trust in exporting clusters. Journal of International Economics 48(1): 139150. . 1999b. "Global competition and local co-operation: success and failure in Sinos Valley." World Development 27(9): 16271650. Schmitz, H., and K. Nadvi. 1999. Clustering and industrialization: introduction. World Development 27(9): 15031514. Sengenberger, W., and F. Pyke. 1992. Industrial districts and local economic regeneration: research and policy issues. In F. Pyke and W. Sengenberger, eds., Industrial Districts and Local Economic Regeneration. Geneva: International Labour Office. Sercovich, F. C., with C.-Y. Ahn, C. Frischtak, M. Mrak, H. Muegge, W. Peres and S. Wangwe, eds. 1999. Competition and the World Economy: Comparing Industrial Development Policies in the Developing and Transition Economies. Cheltenham, UK: Edward Elgar. Shapira, P. 1992. "Lessons from Japan: helping small manufacturers." Issues in Science and Technology 8(3): 6672. . 1998. "Manufacturing extension: performance, challenges and policy issues." In L. Branscomb and J. Kellereds, eds., Investing in Innovation. Cambridge: MIT Press. Spar, D. 1998. Attracting High-Technology Investment: Intels Costa Rican Plant. FIAS Occasional Paper 11. Washington, D.C.: Foreign Investment Advisory Service of the International Finance Corporation and the World Bank. Spenser, H., R. Loader and A. Swinbank. 1999. The impact of sanitary and phytosanitary measures on developing country exports of agricultural and food products. Paper presented at the conference on Agriculture and the New Trade Agenda from a

Bibliography

185

Development Perspective: Interests and Options in the WTO 2000 Negotiations, World Bank, Geneva, 12 October. Stiglitz, J. E. 1987. Learning to learn, localized learning and technological progress. In P. Dasgupta and P. Stoneman, eds., Economic Policy and Technological Development. Cambridge: Cambridge University Press. . 1996. Some lessons from the East Asian miracle. World Bank Research Observer 11(2): 151177. . 1999. The World Bank at the millennium. Economic Journal 109: F577597. Stokey, N. 1991. "Human capital, product quality, and growth." Quarterly Journal of Economics 106: 587616. Streeten, P. 2001. Globalisation: Threat or Opportunity? Copenhagen: Copenhagen Business School Press. Sturgeon, T. J. 1997. Turnkey production networks: a new American model of industrial organisation? BRIE Working Paper 92A. Berkeley Roundtable on the International Economy, Berkeley, Calif. [http://brie.berkeley.edu]. Subramanian, A. and D. Roy. 2001. "Who can explain the Mauritian miracle: Meade, Romer, Sachs, or Rodrik?" Working Paper 01/116. International Monetary Fund, Washington, D.C. Swann, G. M. P., M. Preverez and D. Stout. 1998. The Dynamics of Industrial Clustering: International Comparisons of Computing and Biotechnology. Oxford: Oxford University Press. Taiwan Province of China, Council for Economic Planning and Development. 2002. Taiwan Statistical Data Book 1998. Taipei. Teece, D. 1996. Firm organization, industrial structure and technological innovation. Journal of Economic Behaviour and Organization 31(2): 193225. . 2000. Firm capabilities and economic development: implications for the newly industrializing economies. In L. Kim and R. Nelson, eds., Technology, Learning and Innovation: Experiences of Newly Industrializing Economies. Cambridge: Cambridge University Press. Teitel, S. 1984. Technology creation in semi-industrial economies. Journal of Development Economics 16(1): 3961. . 1987. Science and technology indicators, country size and economic development: an international comparison. World Development 15(9): 12251235. Teubal, M. 1996. R&D and technology policy in NICs as learning processes. World Development 24: 449460. . 1997. A catalytic and evolutionary approach to horizontal technology policy. Research Policy 25: 11611188. UNCTAD (United Nations Conference on Trade and Development). 1995. World Investment Report. Geneva. . 1996. The TRIPS Agreement and Developing Countries. Geneva. . 1999. World Investment Report 1999. Geneva. . 2000. The Competitiveness Challenge: Transnational Corporations and Industrial Restructuring in Developing Countries. Geneva. . 2001. World Investment Report 2001: Promoting Linkages. Geneva.

UNDP (United Nations Development Programme). 2001. Human Development Report 2001: Making New Technologies Work for Human Development. New York and Oxford: Oxford University Press. UNESCO (United Nations Educational, Scientific and Cultural Organization). Various years. Statistical Yearbook. Paris. UNIDO (United Nations Industrial Development Organization). 1979. Industrial policies in developing countries: the selection process in Brazil, India, Mexico, Republic of Korea and Turkey. UNIDO-ID/217; UN-E.78.II.B.12. New York. . 2001. Integrating SMEs in Global Value Chains. Vienna. Unger, K., and M. Oloriz. 2000. Globalisation of production and technology. In M. Cimoli, ed., Developing Innovation Systems: Mexico in a Global Context. London: Continuum. USITC (United States International Trade Commission). 1997. Production Sharing: Use of U.S. Components and Materials in Foreign Assembly Operations, 19921995. USITC Publication 3032. Washington, D.C.: USITC. Venables, A. J. 1996. Localization of industry and trade performance. Oxford Review of Economic Policy 12(3): 5260. Wade, R. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton, N.J.: Princeton University Press. . 2001. Winners and losers. Economist, 26 April, London. WEF (World Economic Forum). 1999. The Global Competitiveness Report 1999. Oxford: Oxford University Press. . 2000. The Global Competitiveness Report 2000. Oxford: Oxford University Press. Wells, L. T., and A. G. Wint. 1990. Marketing a Country: Promotion as a Tool for Attracting Foreign Investment. FIAS Occasional Paper 1. Washington, D.C.: Foreign Investment Advisory Service of the International Finance Corporation and the Multilateral Investment Guarantee Agency. Westphal, L. E. Forthcoming. Technology strategies for economic development in a fast changing global economy. Economics of Innovation and New Technology. Wignaraja, G., and G. Ikiara. 1999. "Adjustment, technological capabilities and enterprise dynamics in Kenya." in S. Lall, ed., The Technological Response to Import Liberalization in SubSaharan Africa. London: Macmillan. Wong, P.-K. 1998. Leveraging the global information revolution for economic development: Singapores evolving information industry strategy. Information Systems Research 9(4): 323341. . 1999a. From leveraging multinational corporations to fostering technopreneurship: the changing role of S&T policy in Singapore. National University of Singapore, Centre of Management of Innovation and Technopreneurship. Draft. . 1999b. National innovation systems for rapid technological catch-up by small, late industrializing economies. National University of Singapore, Centre of Management of Innovation and Technopreneurship. Draft. . 2000. From using to creating technology: the evolution of Singapores national innovation system and the changing

186

Industrial Development Report 2002/2003

role of public policy. National University of Singapore, Centre of Management of Innovation and Technopreneurship. Draft. World Bank. 1993. The East Asian Miracle. Oxford: Oxford University Press. . 1999. World Development Report 1998/99: Knowledge for Development. New York: Oxford University Press. . 2000. World Development Indicators 2000. Washington, D.C. . 2001a. Intellectual property: balancing incentives with competitive access. In World Bank, Global Economic Prospects and the Developing Countries. Washington, D.C. . 2001b. World Development Indicators 2001. Washington, D.C. World Development. 1999. Special issue on industrial clusters in developing countries, 27(9).

Bibliography

187