Local Enterprises in the Global Economy: Issues of Governance and Upgrading [illustrated edition] 1843760991, 9781843760993, 9781843769743

This book opens a fresh chapter in the debate on local enterprise clusters and their strategies for upgrading in the glo

350 103 1MB

English Pages 416 [405] Year 2004

Report DMCA / Copyright


Table of contents :
Preliminaries......Page 1
Contents......Page 6
List of figures......Page 8
List of tables......Page 9
List of contributors......Page 11
Acknowledgements......Page 13
1 Globalized localities: introduction......Page 14
2 Regions in the ‘world economic triangle’......Page 33
3 Making sense of global standards......Page 66
4 Governance in global value chains......Page 108
5 The underground revolution in the Sinos Valley......Page 123
6 How globalization affects Italian industrial districts......Page 153
7 Upgrading in the tile industry of Italy, Spain and Brazil......Page 187
8 Local upgrading strategies in response to global challenges......Page 213
9 Clustering and upgrading in global value chains......Page 246
10 Global quality standards and technological upgrading in the Brazilian auto-Components industry......Page 278
11 The effect of global standards on local producers......Page 310
12 Paradoxes and ironies of locational policy in the new global economy......Page 339
13 Chain governance and upgrading......Page 362
Index......Page 396
Recommend Papers

Local Enterprises in the Global Economy: Issues of Governance and Upgrading [illustrated edition]
 1843760991, 9781843760993, 9781843769743

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

Local Enterprises in the Global Economy

For Klara

Local Enterprises in the Global Economy Issues of Governance and Upgrading

Edited by

Hubert Schmitz Professorial Fellow, Institute of Development Studies, University of Sussex, UK

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Hubert Schmitz 2004 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited Glensanda House Montpellier Parade Cheltenham Glos GL50 1UA UK Edward Elgar Publishing, Inc. 136 West Street Suite 202 Northampton Massachusetts 01060 USA The views expressed in this publication are the responsibility of the authors, and not the Volkswagen Foundation or the Institute of Development Studies. A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication Data Local enterprises in the global economy : issues of governance and upgrading / edited by Hubert Schmitz. p. cm. Includes bibliographical references and index. 1. Small business. 2. Small business–Technological innovations. 3. Business enterprises–Technological innovations. 4. Corporate governance. 5. Industrial location. 6. Industrial concentration. 7. Globalization–Economic aspects. I. Schmitz, Hubert. HD2341.L62 2004 338.6⬘42–dc22


ISBN 1 84376 099 1 (cased) Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents vii viii x xii

List of figures List of tables List of contributors Acknowledgements 1

Globalized localities: introduction Hubert Schmitz


Regions in the ‘world economic triangle’ Dirk Messner



Making sense of global standards Khalid Nadvi and Frank Wältring



Governance in global value chains John Humphrey and Hubert Schmitz



The underground revolution in the Sinos Valley: a comparison of upgrading in global and national value chains Luiza Bazan and Lizbeth Navas-Alemán


How globalization affects Italian industrial districts: the case of Brenta Roberta Rabellotti


Upgrading in the tile industry of Italy, Spain and Brazil: insights from cluster and value chain analysis Jörg Meyer-Stamer, Claudio Maggi and Silene Seibel


Local upgrading strategies in response to global challenges: the surgical instrument cluster of Tuttlingen, Germany Gerhard Halder


Clustering and upgrading in global value chains: the Taiwanese personal computer industry Chikashi Kishimoto


Global quality standards and technological upgrading in the Brazilian auto-components industry Ruy Quadros













Local enterprises in the global economy

The effect of global standards on local producers: a Pakistani case study Khalid Nadvi


Paradoxes and ironies of locational policy in the new global economy Jörg Meyer-Stamer


Chain governance and upgrading: taking stock John Humphrey and Hubert Schmitz




Figures 2.1 3.1 3.2 3.3 5.1 5.2 5.3 6.1 7.1 8.1 8.2 9.1 12.1

The world economic triangle The dynamic of international quality management standards: from generic approaches to further diversification The dynamic of international social standards: from diverse to generic approaches The dynamic of environmental standards and its growing diversity US and European buyers’ demands on footwear producers’ performance Domestic buyers’ demands on footwear producers’ performance Footwear producers – investments in process and product upgrading from 1996 to 1999 Total sales in Brenta (1995 prices, euro millions) Key actors and interaction in the tile industry Products of the Tuttlingen cluster Model of main distribution channels Functional upgrading in the PC value chain Types of product upgrading


23 86 88 88 118 119 120 144 195 203 207 246 332

Tables 2.1 3.1 3.2 3.3 3.4 3.5 3.6 4.1 5.1 5.2 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12 8.1

Chain governance and scopes of action in regions Types of actors engaged in defining and implementing standards Typologies for global standards Different generations of global quality management standards ILO core labour conventions and ratification Different generations of global social and environmental standards Overview of key standards Combinations of parameter setting and enforcement Number of firms and workers in the footwear industry of Rio Grande do Sul (RGS) and the Sinos Valley Governance in the footwear value chain and implications for local upgrading of footwear producers The district in 2000 Size distribution of the sample Performance indicators (% of sample firms) Process and product innovation (number of firms) The world market for footwear (US$ billion) The value of sales of some top luxury-fashion companies operating in the footwear industry in 2000 Production for high fashion companies among sample firms (%) Internal functions in sample firms working as subcontractors to high fashion companies Kendall correlation coefficients Performance among sample firms (number and % of firms recording an increase of performance indicators during the last 5 years) Process and product innovations among sample firms Importance of linkages according to sample firms Medical engineering in Tuttlingen: number of firms and workers engaged in manufacturing in 1999 (by firm size) viii

43 59 61 64 72 74 85 103 113 124 145 145 146 147 148 150 152 152 155 156 157 165 204


8.2 8.3 9.1 9.2 9.3 9.4 9.5 10.1 10.2 10.3 10.4 10.5 10.6 11.1 11.2 11.3 11.4 11.5 11.6 11.7 12.1 12.2

Initiatives established and mediated by public and public–private joint action Upgrading strategies of Tuttlingen firms Basic data of main sub-products in the Taiwanese PC industry (2000) Leading sub-products of the Taiwanese PC industry (every three years) The rate of offshore production of main sub-products (volume: %) The rate of OEM/ODM (%) of four main sub-products Upgrading of the Taiwanese PC cluster (the 1980s–the 1990s) Sindipeças members: quality standards certification by type of certificate (August 2000) Status of firms in terms of quality standards certification (August 2000) Sample firms, five largest customers and respective share of sales (1999/00) Sample firms, status of quality standards certification Rank of sources of technical support and information in preparation for certification Sample firms: types of collaboration received from customers Quality assurance certification by sampled surgical instruments firms Average costs, and time taken, for ISO 9000 certification Sources of know-how on ISO 9000 quality assurance certification Changes in internal practices and firm efficiency Changes in competitiveness and market access Changes in performance over the past three years Firm perceptions of buyer’s key priorities Requirements on locations across the industry life cycle A neo-Schumpeterian model of industrial development


216 225 236 238 240 249 258 270 271 277 281 283 289 302 303 304 305 306 307 309 335 336

Contributors Luiza Bazan is a sociologist and project manager at Social Services for Industry (SESI), Federation of Industries of Rio Grande do Sul, Brazil. Gerhard Halder is a geographer and manager of the research and teaching programme in economic and social geography at the University of Stuttgart, Germany. John Humphrey is a Professorial Fellow at the Institute of Development Studies at the University of Sussex and Director of the ‘Globalisation and Poverty Programme’ funded by the Department for International Development. Chikashi Kishimoto is a political scientist and researcher at the Foundation for Advancement of International Science, Tsukuba, Japan. He specializes in the industrial development of East Asian countries. Claudio Maggi is a Research Fellow at the Institute for Development and Peace at the University of Duisburg and Director of a Regional Development Programme in Chile. Dirk Messner is the Director of the German Development Institute, Bonn and former Director of the Institute for Development and Peace at the University of Duisburg, Germany. Jörg Meyer-Stamer was a Research Fellow at the Institute for Development and Peace at the University of Duisburg, Germany, and is now a freelance consultant. He is the pioneer of participatory appraisal of competitive advantage. Khalid Nadvi is a Fellow at the Institute of Development Studies at the University of Sussex where he specializes in the implications of global standards for local development. Lizbeth Navas-Alemán is a researcher and consultant working on issues of governance and industrial upgrading in Latin America. She is currently based at the Institute of Development Studies at the University of Sussex. x



Ruy Quadros is Professor at the Department of Science and Technology Policy, University of Campinas (UNICAMP), Brazil, and specializes in issues of innovation and industrial development. Roberta Rabellotti is a Professor at the Department of Economics at the University of Piemonte Orientale, Italy, and specializes in issues of regional development. Hubert Schmitz is a Professorial Fellow at the Institute of Development Studies at the University of Sussex. He was coordinator of the research project on which this book is based. Silene Seibel was Project Manager at the Federation of Industries specializing in benchmarking of best practices in production management, in Santa Catarina, Brazil. She is now Industrial Director of a large apparel company (Marisol S.A). Frank Wältring was a research officer at the Institute for Development and Peace, University of Duisburg. He works now in Central America for the German Ministry for Cooperation and Development.

Acknowledgements This book is a product of the research project ‘The interaction of global and local governance: implications for industrial upgrading’. The project was a joint initiative of the Institute of Development Studies (IDS) at the University of Sussex and of the Institute for Development and Peace (INEF) at the University of Duisburg. It involved contributors from many parts of the world. In the name of all collaborators I wish to thank the Volkswagen Foundation for funding this project. Complementary funding was received from other organizations, acknowledged in the chapters where this applies. I enjoyed coordinating this project and thank my collaborators for the enthusiasm and energy they invested in this project. Even though we had a great research team, progress was at times difficult. In the research business, the proverbial ‘no pain – no gain’ remains unfortunately true, especially when it comes to distilling the key findings and bringing out the value added. My deepest thanks go to all contributors for providing the nth revision of their chapters. There are three colleagues to whom I am particularly indebted. Dirk Messner, Director of INEF, shared with me the intellectual and practical coordination of the project. John Humphrey, my close collaborator for many years, helped to find a way out of several awkward corners (financial and analytical) and played a key role in developing the conceptual framework. Simone Field helped to edit all the chapters and made sure that the book is accessible and can be enjoyed by all readers irrespective of their disciplinary background.



Globalized localities: introduction Hubert Schmitz

Participating in the global economy brings both opportunities and dangers. The main danger is getting locked into a race to the bottom; in other words, competing by paying low wages, disregarding labour and environmental standards, cutting corners in the production process and avoiding taxation. Many enterprises, particularly in poor countries, use this strategy with knock-on effects for producers in other parts of the world. Witness the relentless price competition in global markets, especially for labour-intensive products; witness the falling terms of trade for developing country manufactured exports (Kaplinsky, 2000; Wood, 1997). Many producers, especially those of small and medium size, find that participating in and gaining from the global economy do not always go together. In order to achieve both export growth and rising incomes, it seems essential for local enterprises to ‘upgrade’ – to make better products, make them more efficiently, or move into more skilled activities. Policy makers in many parts of the world are looking for ways of helping their enterprises to achieve this. Many approaches are being tried. Particularly influential is the idea that the local sources of competitiveness need to be strengthened. The buzzwords are synergy, economies of clustering, systemic competitiveness, collective efficiency or local innovation systems. Studies carried out in the 1980s and 1990s showed many unexpected success stories of local enterprise clusters breaking into global markets. These stories continue to fuel excitement amongst local policy makers and their external backers. Simultaneously, there is a globalization debate which centres on the new rulers of the global economy: the global companies that set the terms under which local export producers operate. The core competence of these global companies is seen to lie in research and development, design, branding and the coordination of suppliers in different parts of the world. If these companies pull the strings, how feasible is it to develop local strategies for avoiding the low road and embarking on the high road to competitiveness? In what circumstances are local upgrading strategies possible? Can local policy networks make a difference – or do global forces undermine them? Another set of questions is concerned with the impact of global quality and labour standards. Do global standards reinforce or undermine local 1


Local enterprises in the global economy

strategies to compete on the high road. This issue is particularly important for developing country clusters. Do the global standards marginalize them or help them to upgrade? These are the questions addressed in this book. The contributors reject the naive optimism that prevails in much of the local development debate. Similarly they reject the pessimism of the globalization critics who fear that local strategies are rendered irrelevant by global forces. Moving away from these unhelpful positions requires new conceptual work: of how the global economy is governed, of how the links between global companies and local enterprises are structured, and of how upgrading can be achieved or thwarted. Starting from this new conceptual basis, the contributors to this book dissect real cases from the developed and developing world: clusters from Latin America and South Asia which are struggling to escape the race to the bottom; an East Asian cluster with the most impressive upgrading record in recent history; and, the transformation of European clusters reflecting both upgrading and downgrading experiences. This introductory chapter guides the reader through the book. It starts by recalling the common point of departure for all contributors; then it sets out the axes along which they advance the debate, it highlights the general findings and brings out the specific contribution of each chapter. The explanation is organized by theme rather than the sequence in which the chapters appear in the book.

1. THE LOCALITY MATTERS: A SURPRISING CONVERGENCE OF VIEWS A paradoxical feature of the recent debate on competitive advantage in globalized markets is the importance given to locality. Influential authors from a range of different disciplines have emphasized the local determinants of upgrading. Mainstream economics re-discovered the importance of locality, prompted by Paul Krugman’s (1991) Geography and Trade. Krugman builds on Alfred Marshall’s (1920) Principles of Economics in which he set out how clustering small enterprises can compete in distant markets due to local external economies. Marshall continues to be the key reference for all writers in this field, but for many of these writers the impact of clustering is not confined to agglomeration economies. The competitive advantage of locality also arises from the combination of rivalry and cooperation between local enterprises and from the partnership of public agencies and private organizations in supporting local enterprises. The importance of local synergy and rivalry has for some time been

Globalized localities


stressed in Michael Porter’s (1990) work on industrial clusters and has been emphatically restated in ‘Regions and the new economics of competition’ (Porter, 2001). Similarly, the literature on industrial districts has for some time hailed the importance of local relationships for competing in the global economy. The Pyke and Sengenberger (1992) volume Industrial Districts and Local Economic Regeneration was particularly influential in a debate which continues to be pursued with vigour in both the academic and policy arena. Take for example, the recent issue of European Planning Studies in which Becattini (2002), Bellandi (2002) and Dei Ottati (2002) reassert and refine their earlier analyses of Italian industrial districts. Likewise, take the recent EU Observatory of European SMEs which is devoted to policies for ‘Regional clusters in Europe’ (EU, 2002). A parallel line of work, concerned with local innovation systems, produced similar messages, even though its origins were different. In the 1990s, the literature on technological development moved from a focus on the individual firm and a strong distinction between innovation and diffusion towards a greater concern with learning-by-interaction (Lundvall, 1998). This led on to the studies of innovation systems, first at the national and then increasingly at the regional and local level (Freeman, 1995; Edquist, 1997; Braczyk et al., 1998; Strambach, 2002). Even though these writers come from different disciplines, most of them share an optimistic view of the scope for local upgrading strategies. In most works there is a strong emphasis on local governance, as for example in Cooke and Morgan’s (1998) The Associational Economy or Scott’s (2000) Regions and the World Economy. The central message is that effective local upgrading strategies build on strong linkages between local enterprises and institutions; and that local policy networks can help local enterprises to reposition themselves in the global economy. A very similar message came from the 1990s research on clusters, local innovation systems, and systemic competitiveness in developing countries. Contributors to this book were amongst the leading researchers in this field. Dirk Messner and Jörg Meyer-Stamer were co-authors of Systemic Competitiveness: New Governance Patterns for Industrial Development (Esser et al., 1996) which stresses the critical role of meso-level policy in building competitive advantage. Messner (1997) then extended the theoretical analysis of the political determinants which enhance or block policy networks concerned with competitive strategies. Meyer-Stamer (2000) opted for the more practical route of making local competitive strategies work. He pioneered the ‘Participatory Appraisal of Competitive Advantage’ (PACA) in which local stakeholders undertake their own analysis of constraints and opportunities and then jointly implement their ideas for repositioning the local economy.


Local enterprises in the global economy

These authors drew their inspiration from a number of sources; they include the research on industrial clusters in developing countries by Khalid Nadvi (1999), Roberta Rabelotti (1997) and Hubert Schmitz (1995). This research helped to overcome the earlier pessimism concerning the export prospects of small firms in developing countries. Conceptually this research was organized around the idea of collective efficiency defined as the competitive advantage derived from local external economies and joint action. The practical implications of this work were synthesized by John Humphrey and Hubert Schmitz (1996) as the triple C of local industrial policy. Customer-oriented, collective and cumulative were proposed as key features of effective approaches for fostering local development. A wealth of further studies on industrial clusters and local innovation systems in developing countries has emerged since then (Cassiolato and Lastres, 1999; Nadvi and Schmitz, 1999) and made a major impact on the policy debate – in developing countries and in the donor agencies concerned with fostering economic development.1 The conclusions and problems often mirror those found in developed countries. However, they tend to differ in one respect. The clusters in developed countries are often global leaders and they play a decisive role in innovation and product design. In contrast, developing country clusters tend to work to specifications that come from outside. A similar divide arises with respect to global standards: the developing country clusters tend to be ‘standard takers’ and not ‘standard makers’. The observation of such differences led us to embark on a new journey to find out in more detail: who defines and monitors the parameters for local producers? How do these producers respond? How do power relationships change over time? What is the connection between upgrading and different global–local power configurations?

2. TAKING INTO ACCOUNT THE NEW GLOBAL LINKAGES While the old debate on clusters and local innovation systems was mainly concerned with local linkages, this book examines in detail the global linkages and how they affect local relationships. This is not just an add-on; it transforms the understanding of local upgrading strategies and options. It asks who sets the parameters for local producers. These producers do not just export into an anonymous global market; often they feed into chains that are governed by powerful global players. Chain governance means that some enterprises set and/or enforce the parameters under which others in the chain operate. Other parameters are set from outside the chains, notably the new quality and labour standards. In this book we ask whether and how

Globalized localities


local enterprises manage to comply with these standards, whether and how the standards enhance or constrain the scope for local policy networks, and whether and how the standards affect different upgrading strategies. Indeed, running through all the chapters is a distinction between different types of upgrading. The common question is how they are affected by different constellations of global and local governance. Answering this question is an ambitious undertaking. It requires bringing together three bodies of literature: ● ● ●

the above-mentioned work on local and regional development the literature on global value chains the disparate work on global standards.

The book provides new assessments of these bodies of work but more importantly it seeks to bring them together, both theoretically and empirically. The building blocks are of two types: new conceptual work and a range of case studies from developed and developing countries. The objective is to provide a new understanding of how global and local governance interact. The contributors show new forms of power and inequality in global chains but they also identify scope for local action. This provides a platform for engaging in the debate on winners and losers from globalization. Who loses and gains from globalization is one of the key questions of our time, central not just to research but to the street protests surrounding recent political summits. In the coming years, the question will become more, not less acute. By showing how and why insertion in global chains enhances or blocks local upgrading, the book can provide much needed analysis and evidence relevant to the academic and political debate on globalization. The remainder of this opening chapter highlights how the contributors to this book have moved the debate forward. During the course of this undertaking, the subsequent chapters are introduced by theme rather than chronologically.



In order to understand the interaction between local and global governance one needs a clear understanding of how the global economy is governed. Most of the current academic literature and policy advice is informed by one of the following two views: a) the neoliberal view of a free global market economy in which local advance depends on local efforts and the market friendliness of government; b) the view which emphasizes the need for


Local enterprises in the global economy

regulation and intervention and puts forward a stratified public governance model, that is intergovernmental organizations at the global level, national governments, local governments. Neither view is helpful for us. Local enterprises and policy makers are not operating in a free global market economy, nor in a global economy which is governed in a stratified way. The contributors to this book depart from these views with a different conceptualization: the network-based global economy. This is spelt out by Dirk Messner in Chapter 2 of this book. He stresses that the global economy is not just governed by a combination of reliance on free markets and rules of inter-governmental organizations. Export oriented regions experience this most directly as they are tied into global value chains often governed by global buyers and they are faced with global standards that are defined by global policy networks. This leads Messner to develop the idea of a governance triangle comprising: a) a local policy network, b) lead firms of global chains, and c) a global policy network concerned with the setting and monitoring of standards. His key point is that local enterprises and policy makers need to interact with these sector-specific global governance structures and that the options and limits for local action arise from this interaction. Thus the purpose of his chapter is to provide a framework for working out the scope for local action: instead of starting from an analysis of local relationships it starts from the global relationships. This perspective is maintained throughout the book, with some chapters focusing on the insertion of local actors in global value chains and others more concerned with the responses to new global standards. Chapter 4 by Humphrey and Schmitz explains how and why global value chains are governed. Central to value chain analysis is the observation that there are lead firms, notably global buyers, which set and enforce the parameters under which other firms in the chain operate. The question is why would these lead firms go to the trouble and expense of setting up and supervising supply chains? No firm will incur the expense of developing arrangements with specific suppliers in order to purchase products that the market freely provides. The authors suggest that there are two reasons why the global buyers do not rely on coordination through the market and seek to govern their chains: ●

Product definition. The more the buyers pursue a strategy of product differentiation, for example, through design and branding, the greater the need to provide suppliers with precise product specification and to monitor that these specifications are met. Risk of supplier failure. The increasing importance of non-price competition based on factors such as quality, response time and reliability of delivery, together with increasing concerns about safety

Globalized localities


and standards, means that buyers have become more vulnerable to shortcomings in the performance of suppliers. Understanding chain governance is not an objective in itself. The contributors to this book study chain governance because it structures the upgrading opportunities of local producers. At least this was the general proposition at the outset. In order to test this proposition, a set of tools was developed (Humphrey and Schmitz, 2000; 2002) which included a distinction between four types of value chains: ●

Arm’s length market relations. Buyer and supplier do not need to develop close relationships because the product is standard or easily customized. A range of firms can meet the buyer’s requirements and the switching costs are low. Networks. Firms develop information-intensive relationships, frequently dividing essential competences between them. The interaction is coordinated and the relationship is characterized by reciprocal dependence. The buyer may specify certain product performance standards or process standards to be attained, but would be confident that the supplier can meet them. Quasi hierarchy. One firm exercises a high degree of control over other firms in the chain, frequently specifying the characteristics of the product to be produced, and sometimes specifying the processes to be followed and the control mechanisms to be enforced. This level of control can arise not only from the lead firm’s role in defining the product, but also from the buyer’s perceived risk of losses from the suppliers’ performance failures. In other words, there are some doubts about the competence of the supply chain. Hierarchy. The lead firm takes direct ownership of some operations in the chain. The case of the intra-firm trade between a transnational company and its subsidiaries falls into this category.

With these distinctions one can then ask whether some types of chains offer local producers better upgrading prospects than others. Such questioning can be further refined if one distinguishes between different types of upgrading: ●

Process upgrading: transforming inputs into outputs more efficiently by re-organizing the production system or introducing superior technology. Product upgrading: moving into more sophisticated product lines (which can be defined in terms of increased unit values).


Local enterprises in the global economy ●

Functional upgrading: acquiring new functions in the chain (or abandoning existing functions) to increase the overall skill content of activities. Inter-sectoral upgrading: using the knowledge acquired in particular chain functions to move into different sectors.

Equipped with these typologies one can then investigate the hypothesis that the upgrading opportunities of local producers vary with the type of chain governance. The conclusions of this investigation are presented in Chapter 13, which draws together the evidence from the case studies in this book. What follows is a brief outline of these case study chapters.



The Sinos Valley in the south of Brazil is an ideal location for testing the above hypothesis because its enterprises feed into chains which are governed in different ways. This makes it possible to conduct comparative analysis which Luiza Bazan and Lizbeth Navas-Alemán provide in Chapter 5. The Sinos Valley is Latin America’s most significant footwear cluster. In order of importance its main markets are the US, Brazil, the rest of Latin America and Europe. The relationships between the producers and customers in these four markets vary significantly. The US chain is a clear case of quasi-hierarchy, that is the combination of high pressure and technical assistance from the US buyers led to fast process and product upgrading but little advance in functional upgrading. Relationships between European buyers and local producers were not quite so close and seemed to provide more space for functional upgrading. The relationships in the Latin American and Brazilian chains are more even and local producers tend to use their own design and own brands. This research by Bazan and Navas-Alemán is the most thorough case study to date of the connection between chain governance and upgrading in that it provides both comparative and historical analysis. The analysis of the US and European chains underlines the enormous gains which local producers can derive from integrating into global chains. It enabled them to become world class producers. The downside was that they remained in a captive relationship with their buyers for a long time. According to the authors, the key point is that this is changing. Their explanations of this change are of general interest even though it is difficult to establish a clear hierarchy of factors. The trigger seems to have been a shift in the main US buyer’s commitment to source from Brazil which brought into focus a long recognized but long neglected strategic issue, that is the need to diversify

Globalized localities


marketing channels and destinations. Some firms found this more difficult than others. Not surprisingly, the most successful firms were those that had begun to operate in several chains simultaneously for some time. More of this was now required. Particularly successful in exploring new markets were a number of firms that had previously prioritized the domestic market. Their key advantage was that they had developed their own design and marketing capabilities. These were essential in particular for diversifying into the Latin American market where relationships with customers tend to be market-based. These results question the widely-held view that exporting provides superior learning opportunities for local firms. While this view holds for process and product upgrading, it does not seem to apply to functional upgrading. While Chapter 5 provides a case of delayed functional upgrading in Brazil, Chapter 6 tells a story of functional downgrading in Italy. When Roberta Rabellotti first presented this finding at a workshop it came as a big surprise because she was studying a world class Italian industrial district. Is this district a casualty of globalization? Not quite. Many enterprises in the footwear district of Brenta continue to export products based on their own designs and using their own brands. However, an important new line of exports has emerged: the production of luxury shoes based on the design and brand of powerful top fashion companies. To be part of the chain, Brenta’s producers accept a functional downgrading, abandoning design and marketing and focusing on production. Rabellotti suggests that in the short term this does not represent a step backwards because the producers inserted in this chain have higher returns than those exporting their own designs and brands. However, if one takes the long-term view one could come to a different conclusion. By abandoning their design competence these Italian producers risk giving up the key advantage which they have over their competitors in developing countries. Whether this will happen is hard to tell. It depends on whether producers dedicate themselves entirely to producing for these top brand companies or whether they engage simultaneously in other chains in which own design and branding remain important. But the space for independence seems to be shrinking. A general lesson from this case is that the growth of global buyers, design and fashion houses changes the market conditions for all producers, even sophisticated ones. The already observable trend towards explicit chain governance arising from global branding and retailing will impact on developed country clusters as well. The shift in upgrading strategy from production to marketing is also observed in Chapter 7 which analyses the changes in the global tile industry and the changing position of tile producing clusters in Italy, Spain and Brazil. The retail market for tiles is however more diffused and global


Local enterprises in the global economy

buyers and chain coordinators have not emerged. Tile producers have taken the initiative and are trying to set the parameters for the retailers and traders. If they succeed, this would be a very interesting reversal of the situation found in other industries. At present, it is too early to come to a judgement on these initiatives. A particular merit of this chapter by Jörg Meyer-Stamer, Claudio Maggi and Silene Seibel is that they also investigate the relationship between manufacturers and the suppliers of inputs, equipment and materials. This relationship, which has attracted a lot of attention in the technological change literature, has been neglected in the value chain literature. Both the Italian and Spanish manufacturers have network based relationships with their suppliers, based on cooperation and the sharing of competences. These relationships have contributed decisively to their upgrading of products and processes and becoming world leaders in this industry. However, such relationships are not without tension. The Italian and Spanish suppliers sell their equipment and materials to tile producers in other parts of the world and thus help those producers to upgrade their products and processes. But they remain followers, as shown with reference to the Brazilian cluster. In contrast, the Italian and Spanish clusters remain leaders in the global tile industry by virtue of the interactive innovation with key input and equipment suppliers. The surgical instrument cluster of Tuttlingen in Germany, analysed in Chapter 8, is also a world leader, relying on interactive innovation to maintain its position. Competition from low wage countries, especially the Pakistani cluster of Sialkot, has eroded Tuttlingen’s traditional strength, which is the manufacture of hand held instruments. The cluster has responded by developing new products and services. Both require substantial investment and active management of the relationships with users. The cluster’s large firms have this capability whereas most of the small firms are confined to the more traditional products in which they are struggling to survive. As shown by Gerhard Halder, some of the small firms have moved from manufacture to trade in surgical instruments – sourced from Pakistan. Large firms also source traditional instruments from low wage countries and maintain (quasi-) hierarchical relationships with their suppliers in Malaysia, Pakistan, Poland and Hungary. Their in-house manufacture concentrates on new products such as surgical implants or instruments for minimally invasive surgery. In order to develop these products, the firms maintain network-based relationships, characterized by the sharing of competences and close cooperation with leading surgeons and hospitals. Large firms have also taken on additional functions by offering hospitals sterilization services and just-in-time delivery of instrument or implant kits. Gerhard Halder concludes that the cluster has become more differentiated,

Globalized localities


that most small firms are losing out, and that large firms capable of combining internal and external knowledge flows do well. As a result of this increasing differentiation, cluster-wide upgrading initiatives have become more difficult and the management of relationships with specific external holders of complementary competence has become strategic. There is a general lesson which emerges from the analyses of the European clusters in Chapters 6, 7 and 8. A leap in understanding resulted from combining the cluster and value chain approaches. At the outset of the research, this combination of approaches was thought to be important mainly for the analysis of developing country clusters. In the end, it was found that bringing together the two perspectives was also critical for the developed country cases.2 Why not go the whole way and rely entirely on the global value chain approach? Four prominent clusters analysed in the book are outsourcing an increasing part of their production. Enterprises from the South Brazilian footwear cluster in the Sinos Valley have set up subsidiaries in the north-east of the country where wages are lower (Chapter 5); enterprises in the Italian cluster of Brenta are farming out the most labour-intensive operations to Eastern Europe (Chapter 6); the German cluster of Tuttlingen acquires simple surgical instruments from suppliers in Pakistan, Malaysia, Poland and Hungary (Chapter 8); and the most rapid increase in outsourcing has occurred in the Taiwanese computer cluster (Chapter 9). Does this signify a decreasing importance in the relevance of clustering for international competitiveness? Kishimoto, in his analysis of the Taiwanese case, answers this question with an emphatic NO. Following Bell and Albu (1999), he distinguishes between the production and knowledge system and shows that although the importance of clustering diminishes in the former, it increases in the latter. Shedding repetitive activities and concentrating on knowledge-intensive activities is seen as a sign of functional upgrading. The analysis of the Taiwanese computer cluster is important for other reasons. It has become the world’s most significant industrial cluster outside OECD countries and accounts for about half of the world’s output in computer hardware. Substantial progress has been achieved in all categories of upgrading and the main question is how this was achieved. Kishimoto argues that the key lies in the interaction of local and global linkages. The global linkages are analysed, as in previous chapters, by using the global value chain approach. Initially local producers operated in quasi-hierarchical chains in which they were able to upgrade processes and products very fast with the assistance of their buyers. Functional upgrading was also achieved but not all the way. Many local firms acquired design capabilities but few established their own brands. Those firms that succeeded in marketing their own design and brand achieved this by maintaining the contract manufacturing for the


Local enterprises in the global economy

main buyers and simultaneously experimenting with own design and brands in other smaller markets. Such leveraging of competences across chains (Lee and Chen, 2000) has been central to the functional upgrading achieved by some of the Taiwanese computer producers. Chapter 13 by Humphrey and Schmitz pulls together the findings on local upgrading in global chains. Clearly power and inequality in global chains need explicit attention in the assessment of cluster upgrading. However, accepting a subordinate role can bring significant advantages. The case material shows that accepting foster parents offers a fast track to product and process upgrading. But integrating into quasi-hierarchical relationships is a double-edged sword. On the one hand it facilitates inclusion and rapid enhancement of product and process capabilities. On the other hand, local producers become tied into relationships that prevent functional upgrading and leave them dependent on a small number of global buyers. The question then is how local firms can break out of such a lock-in. The chapter discusses ways in which this can be achieved and concludes that the chain approach, while providing many new insights, is not always sufficient to explain the upgrading prospects of local producers.



A central objective of this book is to understand how local firms interact with the global economy, given that this global economy is not a free market economy but governed by a multitude of private and public actors who (try to) define its operational parameters. Global quality and labour standards have become an increasingly important set of parameters. They are seen as an essential instrument for blocking the race to the bottom and nudging enterprises from the low to the high road of competitiveness. The hope pinned on these global standards is not matched by evidence on whether and how they are working. In Chapter 3, Khalid Nadvi and Frank Wältring ask why this is so. Their main argument is that the proliferation of standards in recent years has made it difficult to engage in an empirically informed debate. They then set about reducing the confusion and complexity. They introduce typologies of global standards in quality assurance, environment and employment conditions; they identify the networks of actors engaged in the formulation and implementation of standards; and they distinguish between different generations of standards which helps to understand the bewildering array of current standards and their evolution over time. The categories put forward help researchers and policy makers alike concerned with the formulation, implementation and monitoring of standards.

Globalized localities


Two chapters in this book are specifically concerned with the question of implementation, one in Latin America and one in South Asia. In Chapter 10, Ruy Quadros presents some disturbing findings from Brazil where he investigated the adoption of global quality standards by local auto parts producers. His main question is whether the adoption of such standards has enhanced their upgrading prospects. Since he examines this question in a value chain framework, he is particularly concerned with the technical ties between the auto parts suppliers and their customers, which are transnational producers of automobiles or subsystems. He concludes that the implementation of standards has not improved the weak technical ties between local suppliers and the transnational customers. The latter have neither made substantial efforts to help local suppliers conform to the quality standards nor has the compliance led to the development of technical cooperation in areas such as product and process design. The adoption of standards was achieved mainly with the assistance of local consultants but the customers did not trust the certification process. As a result, the suppliers continue to endure the customers’ own monitoring of quality. Quadros concludes that the expected reduction in transaction costs resulting from the diffusion of global standards has not materialized, that certification has induced local suppliers to upgrade quality processes, that the improvement was small and has been achieved at a high cost, and that certification has not provided a platform for moving into product development. Nadvi provides a more positive assessment of the certification experience in Pakistan but shares some of Quadros’ concerns. In Chapter 11, he examines the responses of the surgical instrument producers of Sialkot to the imposition of global standards in quality assurance. Overall, the research does not support fears that global standards would exclude small and medium sized enterprises from competing in the global economy. Adoption has been fast and widespread. Has this certification enhanced the position of local enterprises in the global economy? Like Quadros, Nadvi concludes that obtaining the certificate merely provides a certificate for entering (or staying in) the international market. The widespread adoption of standards by suppliers has increased choices for buyers. This has contributed to falling prices even though the overall volume of sales has increased. The impact of global quality standards has varied with the tier of subcontractor. The processes for assuring quality have improved amongst the first tier suppliers, but amongst their subcontractors there has been little change. Given the importance and depth of the subcontracting system in Sialkot, this raises questions about the value of the whole certification process. Nadvi also reflects on how global standards have affected the governance of the cluster. He emphasizes that in the early stage, when Sialkot was first


Local enterprises in the global economy

confronted with global quality standards, the local business association played an important role in formulating a positive response, mobilizing support and bringing in technical expertise on quality management. Once joint action had set this process in motion, a wide array of private service providers came to Sialkot. When demands for complying with further quality standards arose, the local enterprises were able to respond rapidly with the assistance of these private consultants. The market was providing the required technical and audit services and local institutions ceased to be important for the implementation of global quality standards. Nadvi provides a contrast in terms of the implementation of global labour standards which was achieved in a different way. The comparison is instructive because both standards were externally imposed, but the certification and auditing have been carried out in different ways. The labour standards, which are focused on the elimination of child labour, are implemented through a network of local and global institutions. This policy network is an example of how local institutions can extend their activities and strengthen their effectiveness by working together with global institutions.



Having examined how global standards and the organization of global value chains affect local firms, what are the implications for local policy aimed at upgrading? What is the scope for local development strategies in a world where trading relationships are increasingly global? In Chapter 2, Messner (2002) argues against the pessimism of those capitulating in the face of the new global governance structures; equally, he argues against the unguarded optimism reigning in many agencies concerned with fostering local development. He calls for a new type of analysis which locates the local economy in the global networked economy and helps to identify where the scope for local action is expanding and where it is shrinking. Key factors to consider are: the specific governance pattern in the relevant global value chain; the core competence of the global lead firms; and the manner in which global standards are implemented. Messner’s key message is that there are not only restrictions but also new opportunities for local action through coalitions with global actors. This analysis builds on his earlier work (Messner 1997) which shows that, in order to be effective, local industrial policy requires building a coalition of the key actors in the public and private sector. He refers to this as building policy networks across the public–private divide and shows the context within which they are likely to work. His new work (Messner, 2002; Chapter 2 of this volume) emphasizes that such policy networks increas-

Globalized localities


ingly need to be formed along a local–global axis. More importantly, it provides a new framework for analysing the scope of action for such policy networks. Collective actors such as business associations are essential players in such policy networks, both because of their sector-specific expertise and their ability to mobilize political and financial support. Individual actors can also play a major role, notably the lead firms of hub-and-spoke clusters. The more clusters are integrated into global markets, the more heterogeneous they become and the more they move towards a hub-and-spoke organization in which the lead firms become the gatekeepers of both material and knowledge flows. The surgical instrument cluster of Tuttlingen (Chapter 8) is perhaps the clearest example. Halder shows that institutions concerned with strengthening existing competencies, for example the highly specialized local training institute (BBT), have the support of all local enterprises. In contrast, an institution concerned with developing new competencies (Competence Centre) does not have the backing of the local lead firm; the fear is that through participating in the new Competence Centre other firms could avail themselves of the fruits of its (the lead firm’s) R&D investment. In other words, the scope for building policy networks depends on the type of upgrading they seek to promote. This applies even more to clusters in developing countries which operate in global chains coordinated by external actors. The issue is straightforward where the policy aim is the strengthening of the existing position of a manufacturing cluster in a global chain. This requires process and product upgrading that is usually incremental in nature. The insertion in a global chain ensures that a great deal of learning occurs in the course of making products defined by external buyers (provided that the local manufacturers make the corresponding investment in people and equipment). Clustering facilitates the rapid diffusion of the knowledge thus acquired. Local industrial policy has an important role to play in expanding infrastructure and strengthening training, testing and certification facilities. Such local industrial policy can usually count on the support of all actors including the buyers because its aim is to strengthen and reinforce the position that the cluster occupies in the global chain. Upgrading aimed at repositioning the cluster requires a more active search and risky investment in capabilities aimed at reaching new markets or reaching old markets in new ways. What scope is there for local industrial policy to foster the radical product upgrading or functional upgrading? A key issue is whether the local lead firms support the repositioning of the cluster. Which is stronger, their allegiance to the local policy network or the relationship to their customer(s) in the global chain? The answer


Local enterprises in the global economy

depends on the way in which the chain is organized. If local firms sell to few buyers to whom they are in a quasi-hierarchy relationship, their risk of supporting a project of repositioning is particularly high. The risk lies not only in the investment of exploring new markets but also in jeopardizing sales to existing markets. It is precisely for this reason that a local policy network in the Sinos Valley collapsed in the mid-1990s. As pointed out by Bazan and Navas-Alemán in Chapter 5, the local policy network became stronger again a few years later when the connection between the main global buyer and local lead firms had loosened. Marcia Leite (2002) also examines the relevance of the value chain for the formation of policy networks. Her study, carried out in collaboration with the contributors to this book, focuses on the auto and plastics sector in the ABC region of São Paulo. These two sectors differ substantially in their involvement in the regional industrial chamber, the most important public–private upgrading initiative launched in recent years. Leite shows how and why the active participation of the plastics firms contrasts with the passive attitude of the auto and auto-parts producers. A reason for this lack of interest can be found in the relationship of the auto-parts producers with their clients who are few in number and pursue global sourcing strategies. Local producers prioritize this relationship over all others. In contrast, the plastics producers have many clients operating in many different chains and are therefore less dependent on specific vertical relationships. This has contributed to their greater reliance on collective efforts and greater involvement in the regional chamber. Further supportive evidence is provided by Meyer-Stamer, Maggi and Seibel in Chapter 7 of this book. Their study of tile clusters in Italy, Spain and Brazil suggests that symmetrical relationships with external customers have facilitated the emergence and functioning of local policy networks. However, change is underway. In two of the three clusters, the local policy networks have weakened. The authors link this weakening to the upgrading strategies of the main local producers. Their competitive strategy has shifted from prioritizing excellence in production to controlling marketing channels and thus increasing their market share. According to the authors, this strategy has intensified competition amongst local firms and has weakened their interest in the local policy network. This conclusion applies to both the Italian and Brazilian cluster. In contrast, the Spanish tile producers, who are not pursuing such marketing strategies, continue to participate actively in collective initiatives. The circumstances in which local policy networks can make a difference are discussed further in Chapter 12 ‘Paradoxes and ironies of locational policy in the new global economy’. Jörg Meyer-Stamer highlights the enormous current interest in local upgrading strategies and then assesses the

Globalized localities


likelihood of these strategies becoming successful. He does this by drawing attention to a number of paradoxes which have not been given sufficient attention in recent debate. First, there is the ‘life cycle paradox’: pursuing active local policy is crucial in the early stage of economic development but effective local policy networks are more likely at a late stage. Then there is what one could call the ‘integrationist paradox’: local policy networks which seek to achieve close relationships between local producers and global producers will be marginalized if they succeed. Finally, the ‘location paradox’: firms are increasingly demanding when it comes to locational quality but show a decreasing propensity to invest in local policy networks. The main message of this insightful chapter is not one of pessimism but realism. It provides many insights which help to define the circumstances in which local upgrading strategies can be expected to succeed – or fail. Reading across the chapters one finds increasing similarities between developed and developing countries. The conflicts, trade-offs and paradoxes are often the same for local enterprises and policy networks in developed and developing countries. In a way this is not surprising. They operate in the same global economy, and the old division of labour has changed. Export manufacturing is no longer concentrated in the developed world; Asian and Latin American developing countries account for a significant (and increasing) part of it. As stressed in Chapter 13, the co-ordination of global value chains by global buyers has contributed considerably to this increase in manufactured exports from the developing world. The build-up of manufacturing capabilities has been phenomenal. However, progress in other capabilities has been much slower. R&D, design, branding and systems integration remain largely concentrated in the developed countries. In this sense, there remains a deep knowledge divide in the global economy. Without narrowing this divide it will be difficult for the gains from globalization to spread more evenly.

NOTES 1. For example, in Brazil, the Instituto Metas produces Clusters – Revista Brasileira de Competitividade, a journal for the many policy makers and practitioners concerned with strengthening local production systems. In India, UNIDO has supported a programme for strengthening the export competitiveness of clusters. In Indonesia, JICA (Japan International Co-operation Agency) supports a project for strengthening the capacity of SME clusters. 2. One of the best examples of using this combination of approaches is the comparative analysis of German and Pakistani clusters by Nadvi and Halder (2002). They show that the two clusters occupy different stages in the same value chain and that the relationships between them are both conflictual and complementary.


Local enterprises in the global economy

REFERENCES Becattini, G. (2002), ‘Industrial sectors and industrial districts: tools for industrial analysis’, European Planning Studies, 10 (4), 483–93. Bell, M. and M. Albu (1999), ‘Knowledge systems and technological dynamism in industrial clusters in developing countries’, World Development, 27 (9), 1715–34. Bellandi, M. (2002), ‘Italian industrial districts: an industrial economics interpretation’, European Planning Studies, 10 (4), 425–37. Braczyk, H.-J., P. Cooke and M. Heidenreich (eds) (1998), Regional Innovation Systems. London: UCL Press. Cassiolato, J. and H. Lastres (1999), Globalização e Inovação Localizada – Experiências de Sistemas Locais no Mercosul, Brasília: IBICT. Cooke, P. and K. Morgan (1998), The Associational Economy: Firms, Regions and Innovation, Oxford: Oxford University Press. Dei Ottati, G. (2002), ‘Social concertation and local development: the case of industrial districts’, European Planning Studies, 10 (4), 451–66. Edquist, C. (1997), Systems of Innovation: Technologies, Institutions and Organizations, London and Washington: Pinter. Esser, K., W. Hillebrand, D. Messner and J. Meyer-Stamer (1996), Systemic Competitiveness: New Governance Patterns for Industrial Development, London: Frank Cass. EU (2002), Observatory of European SMEs No 3, Regional Clusters in Europe, Luxemburg: Office for Official Publications of the European Union. Freeman, C. (1995), ‘The national system of innovation in historical perspective’, Cambridge Journal of Economics, 19 (1), 5–24. Humphrey, J. and H. Schmitz (1996), ‘The triple C approach to local industrial policy’, World Development, 24 (12), 1859–77. Humphrey, J. and H. Schmitz (2000), ‘Governance and upgrading: linking industrial cluster and global value chain research’, IDS Working Paper, No 120, Brighton: Institute of Development Studies. Humphrey, J. and H. Schmitz (2002), ‘How does insertion in global value chains affect upgrading in industrial clusters?’, Regional Studies, 36 (9), 1017–27. Kaplinsky, R. (2000), ‘Is globalisation all it is cracked up to be?’, Journal of International Political Economy, 7 (5), 1–21. Krugman, P. (1991), Geography and Trade, Cambridge, MA: MIT Press. Lee, J-R. and J-S. Chen (2000), ‘Dynamic synergy creation with multiple business activities: toward a competence-based growth model for contract manufacturers’, in R. Sanchez and A. Heene (eds), Theory Development for Competence-Based Management, Advances in Applied Business Strategy, Stanford, CT: JAI Press, pp. 209–28. Leite, M.D.P. (2002), ‘The struggle to develop regional industry policy: the role of the plastics and auto sectors in the regional chamber of ABC, São Paulo’, IDS Working Paper, No 154, Brighton: Institute of Development Studies. Lundvall, B.-A. (1998), ‘Innovation as an interactive process: from user-producer interaction to the national system of innovation’, in G. Dosi, C. Freeman, R. Nelson, G. Silverberg and L. Soete (eds), Technical Change and Economic Theory, London: Pinter, pp. 349–69. Marshall, A. (1920), Principles of Economics, 8th edn., London: Macmillan. Messner, D. (1997), The Network Society: Economic Development and International

Globalized localities


Competitiveness as Problems of Social Governance, London and Portland: Frank Cass. Messner, D. (2002), ‘The concept of the “world economic triangle”: global governance patterns and options for regions’, IDS Working Paper, No 173, Brighton: Institute of Development Studies. Meyer-Stamer, J. (2000), ‘Participatory appraisal of competitive advantage: a methodology to support local economic development initiatives’, mimeo, www.meyer-stamer.de. Nadvi, K. (1999), ‘The cutting edge: collective efficiency and international competitiveness in Pakistan’, Oxford Development Studies 27 (1), 81–107. Nadvi, K. and G. Halder (2002), ‘Local clusters in global value chains: exploring dynamic linkages between Germany and Pakistan’, IDS Working Paper, No. 152, Brighton: Institute of Development Studies. Nadvi, K. and H. Schmitz (eds) (1999), ‘Industrial clusters in developing countries’, World Development, 27 (9) (Special issue). Porter, M. (1990), The Competitive Advantage of Nations, London: Macmillan. Porter, M. (2001), ‘Regions and the new economics of competition’, in A. Scott (ed.), Global City-Regions: Trends, Theory and Policy, Oxford and New York: Oxford University Press, pp. 139–57. Pyke, F. and W. Sengenberger (eds) (1992), Industrial Districts and Local Economic Regeneration, Geneva: International Institute for Labour Studies, ILO. Rabellotti, R. (1997), External Economies and Cooperation in Industrial Districts. A Comparison of Italy and Mexico, London: Macmillan. Schmitz, H. (1995), ‘Collective efficiency: growth path for small-scale industry’, Journal of Development Studies, 31 (4), 529–66. Scott, A. (2000), Regions and the World Economy: The Coming Shape of Global Production Competition, and Political Order, Oxford: Oxford University Press. Strambach, S. (2002), ‘Change in the innovation process: new knowledge production and competitive cities; the case of Stuttgart’, European Planning Studies, 10 (2), 215–31. Wood, A. (1997), ‘Openness and wage inequality in developing countries; the Latin American challenge to East Asian conventional wisdom’, World Bank Economic Review, 11 (1), 33–57.

2. Regions in the ‘world economic triangle’ Dirk Messner 1.


The geography of the world economy is changing. The world economy of yesteryear was mainly viewed as the sum total of national economies and conceived in the categories of periphery and centre. The new world economy is marked by competition between local clusters (Nadvi and Schmitz, 1999), global cities (Sassen, 2000), global city regions (Scott, 2001) and global value chains (Gereffi, 1999) that no longer know national boundaries. The economy is in part breaking its links with territorially and politically constituted entities and creating functional and agglomeration spaces of its own. The reach of national governments ends at their external borders, which have largely ceased to constitute crucial boundaries to the transfer of money, goods, technology and knowledge. Along with its geography, the world economy’s governance patterns are likewise in the midst of a process of change. They go beyond classical international organizations like the IMF, for example, and they are growing in significance and shaping the dynamics of the global economy. Such governance patterns include global regimes like the WTO; global clubs like the International Stability Forum; globally operating firms that organize transnational production and trade networks; and internationally active NGOs (non-governmental organizations) that negotiate with multinational corporations over social and ecological standards. Against this background of increasingly dense global interdependencies and transnational interactions in the world economy we are forced to readdress the issue of whether and to what extent economic development can be formulated and shaped by political means. This chapter centres on the question of the scopes of action open to regions (that is local firms, public organizations and policy makers) in the new world economy. What global governance structures are relevant for local actors? How do global governance mechanisms determine local development? Do local actors have the autonomy and the resources they need to 20

Regions in the ‘world economic triangle’


deal actively with the new demands placed by the global economy, to build specific competitive advantages, and to influence and shape the level of their region’s prosperity? Or are local and regional actors losing their action potentials and becoming passive or reactive adapters to global framework conditions in the world economy?1 There are various views on the role of regions in the world economy. These include research on clusters (Porter, 1998; Nadvi and Schmitz, 1999), systemic competitiveness (Esser et al., 1996; Messner, 1997), innovation systems (Lundvall, 1992; Braczyk, et al., 1998), territorial development (Storper, 1997). From these perspectives one can derive a relatively optimistic view of the scopes of action open to local actors (firms and local policy networks). Regional theorists underline two important trends of globalization: ●

In the global economy, the international competitiveness of firms and the economic efficiency of regions are increasingly based on regional proximity and regional competitive advantages. Globalization does not devalue or level out local and regional specifics. Indeed, it increases the value of them. The geography of the new world economy increasingly centres on regions. Because geographic proximity and specific institutional and business landscapes are growing in significance, regions have (again, in contrast to the critics of globalization) considerable latitude to shape processes of economic development. This implies that globalization does not lead to the disempowerment of politics, in other words regional governance matters.

The key variable of these approaches is the quality of local linkages. Regions whose local actors, by building business networks and developing policy networks in their business environment, have succeeded in optimizing their inter-cluster relationships in the direction of ‘systemic competitiveness’ (Esser et al., 1996) and ‘collective efficiency’ (Schmitz, 1999) are able to develop ‘specific, geographically defined competitive advantages’ (Porter, 1990; 1998). In this way, they can actively influence and improve their position in the world economy. Regions that lack the collective capacity to develop specific competitive advantages will find themselves among the losers in the global economy. Seen in this way, the key to the development dynamics of regions must be sought at the local level. It is from this viewpoint that the World Bank, UNDP, the Interamerican Development Bank, or the German ‘Gesellschaft für technische Zusammenarbeit’ (GTZ) now base their strategies that are directed at strengthening competitiveness and supporting the private sector in developing countries.


Local enterprises in the global economy

However, the regional theorists and the policy recommendations derived from them neglect the specific demands made on concrete regions by the world market. The frame of reference defined by the world market is perceived in terms of framework conditions that are beyond influence. The world economy is treated as if it were a ‘black box’. Regional theorists overlook the significance of global governance structures for the options available to industrial locations and regions. They tend to overrate local action potentials and the ‘internal sovereignty’ of local actors and ignore the specific demands of concrete world market contexts in which regions are integrated. This chapter is an attempt to rectify this deficit. Its point of departure is the idea that regions are tied into specific global market segments and global governance systems that significantly influence the options of local actors and the demands placed on their strategic capabilities. The study looks at the impact of global governance structures on local development strategies. In other words, the focus is on the interplay between local and global governance in the world economy. Section 2 discusses the two established discourses on the world economy in order to shed some light on the ‘black box’ of the world economy: the neoliberal view of the world economy and the inter-governmentalist perspective on global governance in the world economy. Section 3 looks at the world economy from the bottom-up perspective of local clusters. The analysis is based on the empirical studies presented in this book. What emerges is a picture that deviates from both the neoliberal and the inter-governmentalist interpretation of global governance in the world economy. This is captured in the concept of the ‘world economic triangle’ that is based on the process of interaction between regions, global value chains and global networks dedicated to setting standards (see Figure 2.1). Compared with the established discourses on the global economy, the ‘triangle view’ gives rise to new insights about challenges for regions in the process of globalization. Section 4 focuses on the options and limitations of local actors and local strategies in the context of the world economic triangle. What new demands do we see emerging in the world’s regions? The triangle concept makes clear that the efficiency of regions depends not only on intra-cluster relationships but also and above all on transnational interactions and network structures in the world economic triangle. Cluster research has primarily been concerned with the development of ‘systemic competitiveness’ in a given region, while the triangle perspective clearly indicates that for many firms the ‘relevant system’ in which systemic competitiveness must be developed and safeguarded is the world economic triangle. Systemic competitiveness in situ and systemic competitiveness in the world economic tri-

Regions in the ‘world economic triangle’

Intergovernmental global governance

Global buyers, Private and global lead public–private firms global governance

Trade and financial and social architecture (WTO, IMF, ILO) • Global standards can reduce chain governance costs


iona snat




Figure 2.1

l ne

a al v



• Market access • Accelerated learning processes • Core competences of lead firms: limits to local upgrading • Connection between chain governance and local scopes of action

Global standard setting policy networks (ISO 9000/14000, social/environmental standards)


• Environmental and social standards require cooperation with local clusters

Local and regional governance


Local clusters Local policy networks

• Global standards as ‘tickets’ to the chains • High demands on local governance: scanning, monitoring, influence • Global actors ‘intervene in situ’ • Proliferation of standards precisely in labour and resource intensive industry • Standards influence organization of labour in situ

The world economic triangle

angle are linked together in a tense relationship that opens up new scopes of action for local actors and at the same time introduces new challenges and limitations for local firms and policy makers. The analysis demonstrates that we can still say ‘Regions matter!’. But we also see that local development strategies are decisively influenced by global governance mechanisms in the triangle and the interactions of local and global governance (Messner, 2002).

2. THE ESTABLISHED DISCOURSES ON GLOBAL GOVERNANCE IN THE WORLD ECONOMY This section outlines the two central discourses on the world economy as a means of reconstructing the state of the discussion on main structures,


Local enterprises in the global economy

actors and governance patterns in the global economy. It remains to be seen whether these approaches will prove useful in overcoming the ‘blind spot’ that marks the ‘systemic competitiveness’, the ‘collective efficiency’ and other similar approaches that have a tendency to neglect world economic contexts. 2.1

The Neoliberal Perspective

Neoliberal authors’ system of choice is one involving a worldwide economic policy which sets the stage for firms, as well as states, to square-off against one another in a locational competition that is covered by a minimal regulative framework. From this perspective, the lowest possible level of political intervention in global financial, goods and labour markets is the approach best suited to ensure high levels of economic dynamics in both the global economy and its subsystems. Although global governance, international cooperation and coordination of economic policies are seen as necessary here, they are not required to shape global markets and correct their dynamics in social or ecological terms. Instead, they are seen as a means of anchoring the world economy in rule systems that guarantee property rights, safeguard free trade, ensure free movement of capital and minimize state intervention. Neoliberalism’s concept of world economic order thus provides for largely open and unregulated global markets, based on a ‘weak’ multilateral regulatory framework, developed by international organizations or coordination between states. The ‘Washington Consensus’ sums up the core elements of this model (Williamson, 1997). Neoliberal authors are fully aware that competition in the world economy involves not only business enterprises but countries as well, with their specific institutional and tax systems. Neoliberals think it possible to transfer the advantages stemming from competition between different enterprises to the competition between different systems of government regulation in an increasingly networked world economy (Gerken and Lambsdorff, 2001; Siebert, 1999). The core idea is as follows: individual industrial locations offer different packages of taxes and services. Economic actors that want high levels of government services will be prepared to choose locations with high taxes, while actors that prefer low levels of public services will opt for locations with low taxes. Given perfect and no-cost mobility, global competition will tend towards a pareto-optimal spatial distribution of economic activities in the world economy. Largely free world markets and unhindered global competition not only provide for an optimal level of private economic dynamics and growth, they also contribute to the development of efficiency-oriented states at the same time. The real world economy diverges in many areas from the neoliberal

Regions in the ‘world economic triangle’


model because a variety of market barriers continue to exist and because mobility is neither perfect nor no-cost. It also differs hugely for individual factors of production and income groups. The money capital is more mobile than real capital, and the latter in turn is more mobile in the long run than labour, where high mobility is only found in the upper range of the income scale. Subsequently, these different mobilities translate into socioeconomic effects that remain unconsidered in the neoliberal perspective, that is income distribution effects, reallocation of power between mobile and immobile actors, social and ecological races to the bottom. Yet, liberal economic theory and reasons bound up with welfare theory are cited to justify the model as realistic and so it continues to be pursued. Some important globalization critics foresee the prevalence of a ‘neoliberal world economy’ and they do not consider that it will prove possible to shape global market dynamics politically (Bello, 2001; Mittelmann, 2000). To this extent these authors’ views concur with neoliberal views in their analysis of the central development trends of economic globalization. In contrast, globalization critics reject this perspective. They point to a neoliberal neglect of the subsequent impacts of largely untrammelled competition, for example, income-distribution effects, reallocation of power in favour of mobile actors, democracies in the corset of global competition and indications of system-imminent instabilities, that is on the part of the international financial markets. Central to our discussion is the fact that neoliberal authors perceive regions to be tied into global, decentrally organized, anonymous markets to which they are forced to adjust. They are also faced with international regimes designed to guarantee worldwide competition, free trade, property rights and the lowest possible level of government intervention. These global governance structures are created by nation states; thus, for regions they constitute another external set of framework data that are beyond their influence and to which they have to adapt. This perspective also fails to address the question of interactions between local and global governance as well as the scopes open for active regional locational policies. Thomas Friedman, for instance, argues that untrammelled competition between states for global mobile investment will entail a growing convergence of economic policy designs (monetary stability, low taxes for companies and owners of capital, flexible labour legislation, deregulation, privatization, lean government, that is that politics will soon only be in a position to act out the constraints imposed by the world market). The image in which he visualizes this development is the golden straitjacket. As your country puts on the Golden Straitjacket . . . two things tend to happen: your economy grows and your politics shrink . . . [The] Golden Straitjacket


Local enterprises in the global economy narrows the political and economic policy choices of those in power to relatively tight parameters. That is why it is increasingly difficult these days to find any real differences between ruling and opposition parties in those countries that have put on the Golden Straitjacket, its political choices get reduced to Pepsi or Coke – to slight nuances of tastes, slight nuances of policy, slight alterations in design to account for local traditions, some loosening here or there, but never any major deviation from the core golden rules (Friedman, 1999, 87).


The Inter-governmentalist Perspective

In a departure from neoliberalism and the a priori anti-market, anti-worldmarket, and anti-globalization positions embraced by sceptics, authors like Fred Bergsten (1996), Dani Rodrik (2000; 2001), Joseph Stiglitz (2000), Vincent Cable (1999) and José Ocampo (2002) have tracked down some core elements of a global economic order that would be capable of tempering global market forces. ‘The dilemma that we face as we enter the twentyfirst century is that markets are striving to become global while the institutions needed to support them remain by and large national’ (Rodrik, 2000, 348). Economic processes are increasingly international and can no longer be controlled and shaped by national means, politics must also organize effectively at the international level, and do so via more dense multilateral cooperation and coordination among states or in inter- or supranational organizations (for example IMF, World Bank or the EU). In this view, neither globalization and growing world economic integration nor global competition is the problem. The problem is the lack of adequate global structures of cooperation and organization at the level of globalization. According to Reimut Jochimsen: The joint objective . . . must . . . remain creation of a worldwide market economy geared to responsible social, economic, and ecological aims, one in which, as far as trade, capital, technologies, intellectual property rights, and national currencies are concerned, the actors involved can compete fairly and efficiently in free markets. This means no less than constituting, formulating, the world market. (Reimut Jochimsen, 2000, 36).

There are three patterns of argumentation that run counter to the neoliberal worldview and lead to a call for a global regulative policy: ●

Securing market efficiency: A world regulative economic policy is required to create stability (for example in international financial markets), to learn from the Asia crisis for instance (Eichengreen, 1999; Stiglitz, 2000), and to safeguard competition in the global economy. This is a task that national anti-trust authorities are in many cases no longer up to.

Regions in the ‘world economic triangle’ ●


Preventing social and ecological ‘races to the bottom’: World regulative economic policy must contribute to limiting and/or compensating for unwanted income-distribution effects and unintended trends toward social polarization due to economic globalization (Rodrik, 1997). At the same time, it would be essential to develop worldwide framework conditions geared to preventing the overexploitation of environmental resources (Young, 1999). Creating legitimacy for the institution ‘world economy’: Every institution, even the global market, is in the end forced to legitimize itself in social and political terms. Globalization is creating new power imbalances between worldwide mobile actors and immobile actors, intensifying polarization trends within and between societies (Branko, 1999). As a result, it finds itself faced with legitimacy problems that cannot be resolved in the framework of democracies organized on a national basis (Helleiner, 2001).

The discourses on the formulation of a future architecture for the world economic order are marked by a variety of controversies. However, there is also a large measure of concurrence, that is that the nation state and ‘its’ international organizations and regimes (IMF, WTO, OECD, ILO, and so on as well as possible new organizations like a World Environmental Organization that has been proposed) will be the key actors responsible for global governance and world regulative economic policy. Where nation states reach the limits of their capacity to act, they must delegate competencies to international organizations or regimes. This discourse on the global economy of the 21st century, therefore, centres on the model of an inter-governmentally and multilaterally constituted world economic order. The key to global governance is presented as an ‘external set of framework data’. In accordance with the neoliberal paradigm, interactions between local and global governance are not addressed as an issue. States and ‘their’ international organizations provide for global governance, while the field of action open to regional actors appears to be restricted to their own (territorial) locations.2 2.3 Limits of the Neoliberal and Inter-governmental Perceptions of Governance Patterns in the World Economy The established discourses on the world economy reproduce the classical controversies over the issue of ‘more government or less’ that have occupied the fields of economics, political economy and development studies since their infancy. In essence, we concur with the arguments presented by the inter-governmentalists, who point to the normative and factual significance


Local enterprises in the global economy

of the regulative policies of multilateral organizations and regimes for the functioning of the world economy. But these views are not sufficient properly to understand the world economic action context in which regions are forced to develop their strategies that aim to strengthening local competitive advantages: 1. The narrowed-down view of global market allocation and multilateral regulative structures (in the discourses of neoliberals and intergovernmentalists) overlooks the fact that firms and states are not the only actors in the world economy. 2. This being the case, there are in the world economy, apart from anonymous market coordination and multilateral bargaining systems, other important global governance structures that are of central importance to regions. 3. Neoliberal and inter-governmentalist authors concentrate on universal rule systems that are obviously relevant for regions (multilateral rules and standards). Above and beyond these, however, there are more specific global governance mechanisms that must be processed by local actors. The following section argues that there exist specific global governance structures in the world economy that favour local development, and others that restrict local development options. 4. The dominant economic discourses outlined above subscribe to a stratified model of governance. Local, national and global levels of action are perceived as largely independent of one another. According to this view, regions are concerned with adapting as quickly and prudently as possible to global rules and demands. Global governance patterns in the world economy are perceived here as exogenous factors, and regions are conceived in the sense of quasi-closed containers. 5. The following section looks at the interactions between local and global governance, which are at cross-purposes to stratified models. It demonstrates that global and local governance are closely interwoven and that transnational networks and governance patterns are becoming increasingly important in the world economy.



When we look from specific regions ‘into’ the world economy, our gaze is directed to governance patterns in the world economy that are not adequately considered by the neoliberal strand of theory or the intergovernmental strand of the discussion. The empirical studies presented in this volume make it clear that aside from interaction between firms in

Regions in the ‘world economic triangle’


global anonymous markets (arm’s length relations, that is market coordination) and rules of multilateral organizations (for example the WTO), there exist two other patterns of global governance beyond market and inter-governmentalism that effectively influence the choices open to local clusters. Local and regional industrial locations are: ●

on the one hand, increasingly tied into global value chains that are characterized by forms of ‘private global governance’ beyond pure market coordination; and, on the other hand, increasingly faced with global (technical, social and ecological) standards which are as a rule developed and monitored, and in some cases even sanctioned, by private or public–private global policy networks.

If we take these two global governance dimensions into consideration, we come up with a far more complex picture than we find in the established discourses on the world economy. The interactions between regions, global value chains and ‘the world of standards’ give rise to a system context, the ‘world economic triangle’ (see Figure 2.1).3 3.1

Private Governance in Global Value Chains

A considerable share of world trade is accounted for by cross-border intracompany trade, that is exchange between units of multinational corporations (according to UNCTAD estimates, over 30 per cent). The findings of ‘global value chain’ research (Gereffi, 1999; Humphrey and Schmitz, Chapters 4 and 13, this volume) indicate that a substantial share of world trade is organized within relatively stable networks of corporations legally independent of one another. Exchange in these networks is not effected in anonymous markets; instead, it is coordinated in various ways. There are, accordingly, other forms of private governance beyond global market allocation and inter-governmental governance of the world economy. Michael Porter (1990) uses the term ‘value chain’ to refer to the different sequences of activities (logistics, packaging, marketing, after-sales services) in single firms. Gereffi (1994; 1999) further pointed out that specific sequences of value chains may be located in different firms and different countries (thus, global value chains), and that these chains are as a rule organized and co-ordinated by ‘lead firms’. Various empirical studies show that companies from developing countries (in some cases from OECD countries as well) only find access to the markets of (other) industrialized countries in a variety of sectors if these firms are integrated into global production and trade networks. These studies include: the exports of the


Local enterprises in the global economy

East Asian garment industries to the US (Gereffi, 1999); the trade in horticultural products between Africa and the UK (Dolan and Humphrey, 2000); footwear exports from China, Brazil and Italy to the US and Europe (Schmitz and Knorringa, 2000; Bazan and Navas-Alemán, Chapter 5, this volume; Rabellotti, Chapter 6, this volume); as well as studies on the trade relations between Pakistani manufacturers of medical equipment and importers in the US and Germany (Nadvi and Halder, 2002). The studies suggest two conclusions. First, trade in these products is organized by ‘global buyers’ in industrial countries, who often work for wholesalers of brand-name companies. In other words, the local companies and clusters do not produce for anonymous markets but for a limited number of ‘lead firms’, and they are typically integrated within these lead firms’ trade and production networks for longer periods of time. Second, these studies clearly suggest that the form of production in local clusters, their technoorganizational learning processes and their options for local upgrading strategies, depend on the governance patterns prevalent in global value chains. Accordingly, for local clusters, world-market and export orientation implies not only competition in global markets and integration into intergovernmental regulatory structures of the world economy, it means integration into global private governance structures as well. Much of the global value chain research has focused on these governance issues. The empirical studies have shown that it is often not possible to describe the interactions between companies in global production and trade networks as pure market transactions. Instead, what we are observing here are different governance structures: ‘chain governance structures are the relationships and institutional mechanisms through which nonmarket coordination of the chain is achieved’ (Humphrey and Schmitz, 2002, 7). Therefore, the central concern is the attempt to reconstruct the governance structures in global value chains. John Humphrey and Hubert Schmitz (2002; Chapter 4, this volume) work out what it is that is governed in global value chains by different forms of coordination and control. They note that the lead firms define three types of parameters: ●

What is to be produced. This involves the design of products, both in broad conception and detailed specifications. How is it to be produced? This involves the definition of production processes, which can include elements such as the technology to be used, quality systems, labour standards and environmental standards. Physical product flow. How much is to be produced, when, and how the flow of product along the chain is to be handled (Humphrey and Schmitz, 2002, 6–7).

Regions in the ‘world economic triangle’


The way in which these decisions are made and the activities of different units within and between firms in a chain, as well as the way they are coordinated can be described along a continuum extending from market coordination (arm’s-length market relationships) to vertical integration at the other end (hierarchic governance). Between these two poles we can observe network structures in which companies cooperate by pooling complementary competences and others in which the lead firms (as a rule large global buyers) use power resources that lead to highly asymmetrical relationships (quasi-hierarchy). Humphrey and Schmitz (2002, 26–7) characterize the four patterns of interaction and governance as follows: ●

Arm’s length market relations. Buyer and supplier do not develop close relationships. This implies that the supplier has the capacity to produce the product the buyer wants, and also that the buyer’s requirements (including quality, reliability, etc.) could be met by a range of firms. The product should be standard or easily customised and any process requirements can be met by non-transaction specific standards of the sort verified by independent certification. Networks. Firms co-operate in a more information-intensive relationship, frequently dividing essential value chain competences between them. The relationship is characterised by reciprocal dependence. In this case, the buyer may specify certain product performance standards or process standards to be attained, but should be confident that the supplier can meet them. Quasi-hierarchy. One firm exercises a high degree of control over other firms in the chain, frequently specifying the characteristics of the product to be produced, and sometimes specifying the processes to be followed and the control mechanisms to be enforced. This level of control can arise not only from the lead firm’s role in defining the product, but also from the buyer’s perceived risk of losses from the suppliers’ performance failures. In other words, there are some doubts about the competence of the supply chain. The lead firm in the chain may exercise control not only over its direct suppliers but also further along the chain. Hierarchy. The lead firm takes direct ownership of some operations in the chain (Humphrey and Schmitz, 2002, 26–7).

The existence of network governance and quasi-hierarchy in global value chains is empirically well documented. But why is it that firms are willing to invest in building networks or quasi-hierarchy? ‘Governance’ costs time and money. In a world of perfect information and perfect competition market transactions would be the most cost-effective form of interaction


Local enterprises in the global economy

between firms. Network theory (Powell, 1990) as well as some approaches that combine network theory with transaction-cost theory (Jones et al., 1997) show that market coordination (arm’s-length market relationships) and vertical integration (hierarchic governance) often lead to suboptimal solutions. Jones et al. argue that markets are inefficient when it comes to intercompany exchange relations that are marked by ‘frequent, complex and customized exchanges, time pressure and asset specificity’ (Jones et al., 1997, 916). Like Williamson (1979, 249ff.), Jones et al. show that under these conditions it makes sense for firms to cooperate more closely than they would under purely market conditions as a means of managing mutual dependencies and risks (time pressure, frequent and customized exchange) and complementarities in production processes (asset specificity). Now, why is it that these considerations are relevant to the discussion on regions in the world economy of the 21st century? They are relevant because: First, we observe that in many sectors the challenge facing world-marketoriented companies and local clusters is not to compete in ‘free, anonymous markets’ but to be able to deal with different private governance patterns and rule systems and to be able, in this context, to exploit or extend their own concrete options. In other words, we find highly different requirements and options for local actors (companies, policy makers and intermediate organizations), characterized by specific governance structures, in different value chains. Access to global markets, access to global knowledge (technology, production know-how, design, marketing, and so on) and the distribution of profits and rents between companies are crucially influenced by the specific governance structures in global value chains. Second, streams of world trade, patterns of global production and investment, and the integration of specific local industrial locations into the world economy, or their exclusion from it, are often significantly influenced by private global governance structures and decisions are defined by lead firms of global value chains. These effective and powerful forms of global governance find consideration neither in the neoliberal notions of global market allocation (that can conceptualize private governance only as market coordination) nor in the inter-governmental view of the world economy (in which the perspective of governance remains restricted to governmental actors). 3.2

Global Policy Networks and the ‘World of Standards’

In the global economy we cannot help noting a confusing proliferation of global standards. Their genesis and meaning for the new basic structures of

Regions in the ‘world economic triangle’


the world economy and their impacts on the action options of worldmarket-oriented corporations, local clusters, or developing countries making their way into the world economy have as yet been accorded little systematic attention in the literature.4 The studies published are mainly concerned with specific standards (environmental standards, ISO 9000, and so on). Therefore, they do not offer an overall picture of the role played by global standards in the process of structure-building in the world economy (Barrientos, 2000; Mah, 1997; Clapp, 1998). Khalid Nadvi and Frank Wältring (Chapter 3, this volume) bring order into the proliferating tangle of global standards, setting out a comprehensible panorama for the interested reader. The study, first, illustrates that different types of global standards are gaining increasing importance for companies and industrial locations that are oriented to global markets. Second, the study shows that in the ‘world of global standards’, patterns of world economic governance are emerging which are given systematic attention neither in neoliberal circles nor in the inter-governmental perspective. Proceeding from the work of Nadvi and Wältring (Chapter 3, this volume), this section seeks to categorize the great number of existing standards with an eye to outlining trends relevant for our discussion about latitude for regional policy. An important observation is that private actors increasingly drive the emergence of new global standards, in particular NGOs and corporations. The ILO core labour standards, for instance, have been with us for many decades, although their impacts have been very limited. On the other hand, in recent years a number of industries and business networks have developed industry-specific or company-specific social, labour, child-protection and environmental standards that have in some cases been monitored and certified in an extremely effective fashion. Such social and environmental standards come about in global policy networks, sometimes in conflict, sometimes in cooperation between ‘concerned’ companies and NGOs, labour unions, or consumer groups (Fuchs, 2000; Blowfield, 1999; Hilowitz, 1997). Apart from company- and sector-specific standards, recent developments have also seen the emergence of universal social standards (like SA 8000, Ethical Trade Initiative/ETI), the reach of which is worldwide and cross-sectoral. ETI, which sets social labour standards, is an indication of the potential and the reach of standardization in transnational networks. Following negotiations between British retail corporations and UK and African NGOs, labour unions and the British government, the seven largest UK supermarket chains now apply the ETI standards in their retail and production networks with African partners. These practices are monitored by independent institutions. In future ETI standards are to be verifiably implemented in the African companies involved, which are owned directly


Local enterprises in the global economy

by UK supermarket chains, as well as in supplier companies that produce fresh foods (Barrientos et al., 2001). In the export segments so crucial to African economies that have interwoven local supply chains, it has, despite many difficulties, proven possible to set binding social and labour standards that are verified by independent institutions. Aside from these global standards that essentially come about due to pressure exerted by NGOs (sometimes backed by governments, for example in connection with development cooperation), there are other global standards that have been created or actively promoted by (individual, several, or many) corporations operating in their own interests. Four motives can be distinguished here: ●

In sensitive markets (for example, the food industry), international corporations are interested in binding standards (such as hygiene standards) that enable them to secure consumer confidence (motives include credibility and promotion of legitimacy); In global competition, company-specific social and environmental standards are instruments used to distinguish between competing firms (examples of company codes of conduct include the German OTTO Versand and Karstadt, Levi-Strauss, Sainsbury’s, all of which are certified by independent institutions); In global value chains, management systems like ISO 9000 or the ISO 14000 environmental management system, in a sense ‘quality labels’, can contribute to reducing the control costs that lead firms have visà-vis their suppliers and cut the search costs needed to find new suppliers;5 Corporations that are active worldwide and have been pressured by NGOs or other actors into accepting social or ecological standards are interested in seeing these standards (and the costs they entail) established globally and sector-wide. This compensates for competitive disadvantages they might face compared with their direct competitors. This process is initially set in motion politically and selectively (for example, by NGO pressure on individual multinational corporations). It also gives rise to an inherent dynamic working toward self-generalizing standards that result from competition between business enterprises and their interest in rules that are binding for all, that is that do not distort competition.

There are many indications that the essential motor behind the development of global environmental and social standards in the world economy is private policy networks that bring together above all NGOs, labour unions and firms to reach agreements on standards on the model of collec-

Regions in the ‘world economic triangle’


tive bargaining. We can, however, also observe that governments are becoming increasingly active in, or at least initiate or support, global policy networks that develop or monitor global standards. For example, the UK government in the case of the Ethical Trade Initiative or initiatives of German development policy (see Dolan and Humphrey, 2000; Reichert, 2000). Thus, below and beyond the threshold of inter-governmental negotiation systems (such as WTO or ILO), it is also global policy networks that contribute to the setting of standards in the world economy (see Nadvi and Wältring, Chapter 3, this volume). The dynamics of standard formation can be well illustrated with reference to African–British trade relations and production networks in the horticulture industry (Dolan and Humphrey, 2000, 10): Within the last few years, several industry-wide organizations and trade associations in fresh produce have established sectoral codes of practices to reduce their vulnerability to consumer and NGO pressure. Some sectoral codes have their origin in the North, and are being adopted by African suppliers either voluntarily or as a requirement to supply certain buyers. The most significant standard for suppliers of horticulture produce is the EUREPGAP protocol, produced by a network of European retailers to ensure best practice in the production and sourcing of fresh produce. This protocol defines the minimum industry wide standards of technical, environmental and social aspects of production, and has been widely adopted by UK retailers and their suppliers. More recently, 38 supermarket chains worldwide have signed up to a global benchmark standard on food safety, as part of a new Global Food Safety Initiative. Similarly, a variety of sectoral codes have been established through consortia of trade associations and producers in Africa. In Kenya, Zambia, Uganda, and Zimbabwe, associations and exporters, conscious of the need to assure northern buyers of ethical production, moved early to introduce their own benchmark standards as a means of promoting quality assurance in the horticulture sector. . . . More recently, UK retailers have engaged with trade unions, NGOs, and enterprise associations to develop multi-stakeholder social codes and verification systems. Again, these have been increasingly adopted as UK multiples realise that standards developed in concert with public stakeholders enhanced their credibility in global markets.

The global policy networks in which standards are set can be characterized with reference to three core notions: ●

first, there are transnational, multi-actor constellations that bring together private and, increasingly, public actors, at times from wholly different geographic and politically constituted areas (for example African companies, British retail chains, European and African NGOs); second, we can note a pluralism of governance (as a rule cooperative


Local enterprises in the global economy

or conflictual network governance in the standard-setting phase; hierarchic governance, network governance, or market solutions in the certification phase); third, global standard-setting takes place in multilevel governance systems (collaboration between local actors, governments, global private actors, international organizations).

Why is it that these global standards are emerging ‘from the bottom up’, in self-organizing networks, rather than coming from a central institution in the world economy (this would be the logic of ‘intergovernmentalism’)? Three central lines of arguments can be advanced here: 1.



Neoinstitutionalists (North, 1990) argue that firms are often forced to operate with limited information and information-processing capacities. In this perspective, standards, rules and routines are essential to create transparency, in this way lowering transaction costs. In addition, generally accepted standards and rules have the function of creating and safeguarding stable expectations in complex interaction contexts. Stability of expectations is the foundation firms need for their long-term activities (for example for investment decisions). Therefore, when pressure is brought to bear by consumers or NGOs, it is more advisable to reach agreements on global standards that are binding on competitors as well, than it is to accept a situation marked by uncertainty, a lack of rules or constantly changing standards. March and Olson (1989) point out that standards (beyond purely technical rules) always have an ‘orientational’ and ‘sense-giving dimension’ – generating social legitimation for markets. Standards are not merely marginal, action-channelling conditions for utility-maximizing actors. They also define a ‘logic of appropriateness’. Seen in these terms, the idea of neoliberalism, that the market order can be reduced to defining property rights and safeguarding competition, is simply naive. Just as in the age of national capitalism it was national labour unions and other actors that brought about the normative framework in which the market is embedded, globally oriented actors are now acting to come up with a normative framework to tame the global market.

It becomes evident against the background of this line of argument that the successive development and generalization of global standards results not only from abstract idealism (for example on the part of global NGOs) but from the concrete interests of multinational corporations in reducing their transaction costs, in increasing the stability of their expectations, and in enhancing their social legitimacy in the social context of the global economy.

Regions in the ‘world economic triangle’


There are good theoretical reasons for developing standards in the world economy that may be termed ‘system-functional’. The world economy is in possession of a favourable incentive system working toward the ‘spontaneous’ development of global standards in network-based governance structures that are borne by the actors concerned. This is because: (a) there are no central institutions at present that could assume the task of setting hierarchic standards; (b) large-sized inter-governmental bargaining systems (like the WTO or the ILO) operate slowly for structural reasons and are geared to coming up with agreements based on minimal consensuses; and, (c) governments in inter-governmental negotiation systems are a priori overburdened by sectoral, highly specific standardization problems (problems of information and complexity – this is true above all for environmental problems). Why is the emergence of global policy networks in which a variety of different actors develop a universe of standards, relevant for our discussion on the global economy of the 21st century? First, the variegated ‘world of global standards’ is of central importance for world-market-oriented clusters and local industrial locations. Building competitiveness no longer means only acting on the variables ‘price’, ‘ontimeness’, and ‘product quality’ under control, it also increasingly means having to meet (or even influence) diverse global standards that intervene profoundly in the production processes and local social conditions. Second, together with the policy networks in which global standards are emerging, the world economy is experiencing the development of effective and powerful governance patterns that are not sufficiently perceived by the established economic discourses. There is much indication that the global policy networks outlined here are rapidly giving rise to multiple environmental and social standards in the global economy. Similar attempts are being made to develop further the ecological and social rules established by the WTO but, with little success. Third, the governance structures of global value chains are closely interlinked with those of the ‘world of global standards’. On the one hand, we can observe that the existence of global standards forces lead firms in global value chains to ensure compliance with these standards among their suppliers, some of whom are active worldwide, as well as to monitor suppliers’ activities and offer suppliers their support in meeting standards. In other words, global standards call for ‘chain governance’. On the other hand, it is also true that international chain governance structures may become superfluous if relevant standards are increasingly monitored and certified by external actors (NGO monitoring systems, private certification companies).


Local enterprises in the global economy

3.3 The Triangle Perspective in the Context of the Established Discourses on the World Economy Now we bring together the strands of the presented arguments in five key points. First, there are two dimensions of global governance (coordination of global value chains and global standard-setting policy networks) that are neglected in the established economic discourses. When we observe these two neglected patterns of global governance, we are forced to perceive a complex 21st century world economy that can not be adequately understood by the categories provided by market theories or on the basis of the concepts of world economic order advocated by inter-governmentalism. The neglected global governance mechanisms work in very specific ways and are of particular importance to sectorally specialized regions. The triangle concept was introduced to open up these new perspectives. Second, against this background, globalization can not be described as a unilinear process of universal ‘market-economization’ (as described by neoliberals and many critics of globalization). Instead, we observe that, in parallel to processes of deregulation and liberalization, new, non-market coordination patterns are emerging in the global economy. Neoliberal theorists should note that private actors are advancing these new forms of governance beyond the market. Global production and trade structures are increasingly organized in global value chains in which market coordination is supplemented by private network governance or quasi-hierarchic governance. Global technical, but also social and ecological, standards come about in multi-actor constellations which are marked by cooperation and collaboration among firms, NGOs, labour unions, scientists, and (as a rule in subsidiary roles) governments and international organizations. The global economy is not a giant anonymous market, but embedded in significant private governance structures. This means coordination distinct from arm’s length market relations. World-market-oriented companies and regions must be familiar with these governance patterns and their modes of operation if they are to be capable of actively building viable competitive advantages. Third, many advocates of a more regulated world economy likewise neglect the new global governance patterns, which are marked above all by interaction with and between private actors. They tend to remain within an inter-governmental frame of reference in which nation states and their international organizations represent the central actors involved in shaping the world economic order. The controversy between neoliberals and intergovernmentalists is concerned with the interplay and the distribution of power between the ‘world of the economy’ and the ‘world of states’, and continuing with the old controversy over ‘more market’ versus ‘more state’. A glance at our world economic triangle reveals that ‘the world of society’

Regions in the ‘world economic triangle’


(Czempiel, 1993) is incessantly growing in significance, that is the basic structures of the world economy and the approaches needed to shape them can be understood adequately only when we cease to view in isolation the ‘worlds of’ the economy, states and society. Fourth, the world economic triangle offers regions a different view of how they are affected by, and how they can participate in, the global economy. Compared with the inter-governmental perspective, it leads, on the one hand, to a consideration of new actors such as NGOs and companies, local and global business networks, and local and global policy networks. On the other hand, this approach focuses our attention on the interplay between different levels of action (for example local British and African NGOs, that is translocal alliances, supported by British development cooperation, enter into negotiations with global lead firms on labour standards; the result is interaction between local and global governance). In contrast, the inter-governmental perspective tends to focus above all on creating and strengthening international organizations. Fifth, when inter-governmental discourses on the world economy are viewed together with the triangle perspective, the following governance mechanisms become visible in the global economy: ●

First, international organizations and regimes that have been created and are controlled by nation states are of great significance. Therefore, the attempt to shape globalization is associated with a shift of state competences and sovereignties to higher-level organizations, and thus linked to a ‘centralization of politics’. Two features characteristic of the governance type ‘international organization’ are ‘inter-governmental negotiation systems’ and ‘quasi-hierarchic governance’ (for example of the WTO by clubs made up of advanced industrialized countries). Second, global market coordination is modified by a great variety of forms of private governance in global value chains. The governance patterns in global value chains shape global investment flows, technology transfers, learning processes and the links between industrial locations and the world economy, or the way in which such locations are marginalized in global competition. Third, global policy networks are an important factor involved in the setting of norms and standards in the world economy. This involves marked interplay between a great variety of private and public actors; structure-building takes place in cross-border value chains, sectors, or subsectors (such as the forestry or food industries); multilevel structures and network governance play an instrumental role in the ‘world of global standards’.


Local enterprises in the global economy ●

Fourth, the interactions between local and global governance mechanisms and between local and global actors in the world economic triangle shows that the architecture of the global economy cannot be depicted adequately by a stratified model but is best represented in the form of an interwoven multilevel system. Five, at least, while inter-governmentalist (and neoliberal) economists focus on the universal rules governing the world economy, the concept of the world economic triangle indicates that regions are also integrated in highly specific global governance rule systems. This gives rise to the question of whether it is possible, in the context of the triangle, to distinguish global governance constellations that tend to encourage, or to block, local developments.

To conclude, inter-governmental economists are right by not wanting to leave the world economy to the markets and calling for global regulative policies and international organizations as institutions of stabilization and frameworks for embedding economic globalization in social and ecological terms. The triangle view, however, indicates that it is not lone states and their international organizations that have the power to shape world markets. Rather, patterns of private governance in global value chains, the interplay between private and public actors from different societies in the ‘world of global standards’, and complex interactions in the triangle (which will be discussed in more detail below) are important building blocks of the architecture of the world economy of the 21st century.

4. LOCAL DEVELOPMENT STRATEGIES IN THE WORLD ECONOMIC TRIANGLE By focusing on the interaction of local and global governance, the triangle perspective opens up a new perception on determinants of international competitiveness and local development strategies. Because the triangle view takes into account the specific form in which regions are integrated in concrete global value chains and ‘worlds of global standards’, it paves the way to a more precise understanding of the options and limits of local industrial upgrading processes, to the scopes of action for local actors to design and implement their development strategies, and for new demands placed on local development policies.

Regions in the ‘world economic triangle’


4.1 The Significance of Global Value Chains for Local Development Policy Chapters 5 and 6 on the Italian and Brazilian footwear clusters and Chapter 7 on the tile clusters in Italy, Spain and Brazil indicate specific correlations between the governance structures in global value chains, the core competences of lead firms, and local scopes for independent cluster and development strategies: ●

The core competences of lead firms define certain limits (although these limits can be overcome) on local upgrading processes. Local firms or clusters that attempt to advance into core competence fields of global buyers are endangering their position and their existence in the global value chain. This is demonstrated in the case of the Brazilian footwear producers in the second half of the 1990s and the Italian footwear cluster over the last few years. The Brazilian footwear producers of the Sinos Valley are integrated in global value chains whose governance patterns are described as quasi-hierarchic. The relationships between the Brazilian companies and the global lead firms may be characterized as ‘asymmetric interdependencies’. As soon as conflicts of interest develop between local actors and the global lead firms, both the scopes open for local strategies and the bargaining potentials of local actors turn out to be relatively small. Recently the Sinos Valley cluster successively pursued a strategy of diversifying their ties to global value chains. Groups of local footwear companies are now increasingly integrated in European and Latin American value chains, characterized by network and market coordination. In this way, they have significantly increased the options available for Brazilian producers to build up competences in marketing and design. This example shows that it is possible to open up new local scopes of action even in the context of global value chains. It also underlines the fact that different structures in value chains define the framework conditions essential to local development strategies. For a long time, Brenta, as a ‘world-class location’ in the footwear industry, was integrated in market-based value chains in which the lead firms hardly set any parameters ‘from outside’. In this framework both the local cluster and local policy networks have larger scopes of action. The governance structures in the ‘top brand global value chains’ in which Brenta has been integrated since the mid-1990s are described as


Local enterprises in the global economy

‘somewhere in between network and quasi-hierarchy’ (Rabellotti, 2001, 27). On the one hand, the local producers have specific and firstclass production know-how that cannot simply be replaced by other suppliers. This seems to indicate balanced relationships between lead firms and local suppliers. On the other hand, the lead firms are in a position to dictate to local firms’ parameters in strategically relevant fields that offer chances of good potential returns (design and marketing). This appears to indicate quasi-asymmetrical relationships between the local and global actors concerned. In this context local scopes of action are smaller than they were under the previous conditions of market governance in the global value chain, though they are presumably greater than in the case of the Brazilian footwear cluster. The tile clusters are integrated in network-like value chains. These relationship patterns can be described as symmetrical interdependencies. In these cases the options open to local firms and policy networks in shaping their locations are numerous and these firms and networks can rely far more on their own local efficiency and effectiveness (and are far less dependent on external influences). However, the research also shows that individual tile producers’ attempts to control the external marketing operations makes their participation in local policy networks less likely.

Linking the cluster perspective with the global value chain approach proves useful as an analytic frame of reference. The development dynamics and paths outlined for the clusters analysed could not be explained from a purely ‘local perspective’, that is on the basis of the classic industrial district approach. Furthermore, the global value chain perspective opens up a more precise understanding of the limits and potentials of locational policy at the local level. This is clearly illustrated in the case of the Sinos Valley. The reason why an apparently reasonable upgrading strategy (development of local design and marketing competences) failed was neither the inability of intermediary local actors nor the project’s lack of economic feasibility. It was because a strategy of this kind would have affected the core competences of the lead firms and was therefore blocked by major local exporters. Upgrading processes were thus blocked by the governance structures specific to the global value chain (quasi-hierarchical governance) and asymmetrical power structures, both within the global value chain and between the actors at the local level. It only became possible to operate an active and promising local locational policy in the Sinos Valley cluster once it gradually proved possible to pursue a strategy involving diversification of the value chain. In contrast, the examples in the tile cluster show that the scopes for local locational policies and upgrading processes are large in the

Regions in the ‘world economic triangle’

Table 2.1


Chain governance and scopes of action in regions

Arm’s length market relations



High autonomy for local actors to design and implement local development strategies. Absence of interventions of global firms restricting local scopes of action coincides with the challenge to develop local strategies without the support from global firms and without access to their strategic knowledge. (Examples: Sinos Valley cluster integrated in the ‘Latin American Chain’; Brenta cluster exporting to different European buyers) High autonomy for local actors based on symmetrical interdependences between local and global actors. Very favourable conditions for fast learning processes in the cluster, combining local and global resources (knowledge, information flows, technological capabilities) and local and global competitive advantages. (Examples: Brazilian, Italian, Spanish tile clusters; Sinos Valley cluster integrated in the ‘European Chain’) Restricted autonomy for local actors to define their own development strategies. Global firms determine the parameters for local actors; local actors have to adjust. Building up local development strategies is likely only if the interests of global firms (core competences) are not touched. Tensions between local and global interests are transformed in interest tensions within the cluster, thus complicating local collective action. (Examples: Sinos Valley cluster integrated in the ‘US-Chain’; Brenta cluster integrated in the ‘Top Brand Chain’)

context of value chains that are organized in networks. Whether this scope is then used depends on other factors. The significance of global value chains for local development strategy is illustrated in Table 2.1. Linking the cluster perspective with the global value chain concept enables us to see the following demands facing local development policy that are neglected in the context of an exclusively local frame of reference. ●

First, local policy makers (in public or private organizations) should be very familiar with how the global value chains in their locations are integrated if they are to be able to realistically assess the specific demands facing locational policies. Second, it becomes clear here that local development policy should not only be geared to focus on local forces but must also seek to network local competitive advantages and global potentials actively (in the value chain).


Local enterprises in the global economy ●

Third, policy makers should realize that local competitive advantage (of clusters) and global competitive advantage (in the chain) are potential competitors (as is shown in particular by the case of Brenta). Fourth, local actors must learn to seek integration in different global value chains to strengthen their bargaining power vis-à-vis global lead firms (as the case of Sinos Valley particularly highlights). Fifth, local policy makers should realize that it is precisely in dynamic clusters that more and more relevant actors find global networking more important and occasionally even more cost-efficient than investments in ‘local collective efficiency’.

These demands on local policy makers have rarely been considered in the context of the established cluster strategies. The challenges are considerable given that local policy networks are increasingly reliant on the know-how of global contexts and need a capacity to interact with global actors. These factors are the sine qua non of successful local development policies. This means that linking the local cluster perspective and the global value chain approach is a good guard against voluntarist recommendations on local development that may come about when the relevant actors are blind to specific structures in concrete global value chains, which tend to limit local scopes of action. Conversely, it can also help to identify and assess options that had not been considered before. Do regions matter in the triangle perspective of the world economy? The answer is, Yes, . . . but . . .! The considerations developed until this point indicate that there continue to be geographically bound competitive advantages and that locational policy can help strengthen these advantages. Yet, in order to build up systemic competitiveness, local development strategies must first be viewed in the context of their specific global value chains. Second, the specific needs, options and limits of locational policy come better into focus here: regions matter, but they form part of a larger, more complex and intertwined transnational economic context. 4.2

The Significance of the ‘World of Global Standards’ for Regions

The growing significance of the ‘world of global standards’ in the world economy was addressed at length above. The issue here is what relevance do global policy networks that develop standards have for the development dynamics of industrial locations, and what demands do they entail for local firms and policy makers? The studies presented in this volume as well as other investigations permit us to draw five important conclusions.

Regions in the ‘world economic triangle’


First, access by local suppliers to global value chains is increasingly bound up with international technical standards (for example safety standards in the toy industry) and global quality management standards (for example ISO 9000, ISO 14000). These standards provide for (technical) compatibility in the world economy and constitute for lead firms an instrument that can be used to check the efficiency of potential suppliers in a costeffective way. In many industries it is the management quality standards in particular that constitute an initial filter in the process in which global lead firms select their suppliers (as shown by Quadros in Chapter 10, this volume, for automotive suppliers in Brazil; Dolan and Humphrey, 2000 for fruit production in Africa; Nadvi and Kazmi 2001 in terms of Pakistani producers of medical equipment). Second, the demands on local firms and policy makers are rising against the background of a proliferation of different global standards (Quadros, Chapter 10, this volume). Competitiveness does not only mean the capacity to strengthen technological competence, it also requires local actors to keep an eye on, and to comply with, the changing and highly complex tangle of global standards if they are not to lose market access and continue developing new markets. The permanent task of scanning and monitoring global standards is a major challenge for both local firms and local policy networks (see Dolan and Humphrey, 2000; Nadvi and Kazmi, 2001; Barrientos, 2000). These demands are especially high when the task is not only to adopt global standards but also to take a hand in shaping them in the context of global networks. Most world-market-oriented firms from advanced countries are concerned to be present in the global networks responsible for developing and setting standards relevant to their own operations. Only in this way is it possible not to fall into the role of the passive ‘rule taker’ and to ensure that one’s own interests are not left out of consideration in the process of standard-setting. Companies, their organizations, and policy makers from developing countries, should be highly interested in bringing their influence to bear in the making of global standards, for example, in preventing such standards from taking on the character instruments of a quasi-protectionism.6 As in the analysis of global value chains, we see here that the demands placed on the governance capacities of local actors are growing at an enormous rate. However, the new challenges facing local actors can also be met with the aid of new alliances, for example, local and global NGOs, sometimes together with bilateral or multilateral institutions of international development that join forces in transnational networks with an eye to gaining social concessions from global lead firms or even local producer clusters. From the perspective of industrial cluster or innovation system approaches, local actors move above all on a local or sometimes a national playing field. Whereas, seen in terms of the triangle,


Local enterprises in the global economy

local actors are forced to move at once in both local and global arenas. ‘Think global and act local’ is no longer a viable model in the framework of the world economic triangle. Instead, it is essential to think local and global and to act at the local and global level in networked multilevel systems. Third, the study by Nadvi and Kazmi (2001) on Pakistani instrument producers and the Dolan and Humphrey (2000) study on African fruit producers indicate that global actors (multinational corporations, NGOs and international organizations) are increasingly present in local industrial sites to monitor and certify global standards, to provide help in implementing them, or to work toward their acceptance (Clapp, 1998; Glaser, 1999). Through this process of interaction between local and global governance ‘local networks’ become transnational networks in situ in which completely new alliances and political forces may arise. As stressed in section 1 of this chapter, the world economy can no longer be conceived in terms of a ‘stratification model’ in which local, national and international dimensions and action spaces are ‘piled’ one on top of the other and whereby actors largely operate independently from one another. Instead, transnational functional spaces and ‘cross-border activities’ of actors are gaining in significance. They comprise global actors who influence economic and political dynamics in situ ‘from the outside’, local actors seeking to influence who must undertake efforts to influence and shape standards discussed in global networks, and local and global actors interacting in regions. Fourth, the number of global social and environmental standards is growing rapidly in sensitive sectors (for example labour-intensive industries, industries close to raw materials, food industries). These are the sectors in which social and ecological problems and health-relevant impacts frequently occur and are highly visible to the public, the consumers, and to NGOs in advanced countries that are the driving forces behind the proliferation of social and ecological standards. In other words, it is precisely in industries with low levels of technological complexity (which include industries in developing countries that have ‘natural competitive advantages’), that global standards and the high demands which they imply for the global governance capacities of local actors are assuming ever greater significance. Thus, building competitiveness is often no longer dependent only on compliance with the classic parameters of competition (time, price and quality of products and services) but also requires the capacity to orient products and production processes to global social and environmental standards. Even on the ‘low roads’ of the world economy, knowledge-based competitive advantages and governance capacities of local actors are gaining in importance. Fifth, global standards can have direct impacts on the forms in which

Regions in the ‘world economic triangle’


labour is organized in local industrial locations. Nadvi and Kazmi (2001) document that the establishment of global social standards for producers of sports equipment in Pakistan has led to a situation in which global buyers have basically restructured their supplier structures in Pakistan. To lower costs for monitoring compliance with global standards and to minimize risks from many small suppliers and many potential actors who violate standards, they have reduced the number of their suppliers in Pakistan and now prefer close cooperation with larger companies. Since the 1980s, the big sports equipment buyers have markedly decentralized their supplier structures, and smaller companies have grown into global value chains via complex supplier networks in producer countries and with an eye to reducing costs. In contrast, global social standards are inducing a reorganization of the local clusters integrated in global value chains that favour larger firms and show a tendency toward centralized supplier structures. No matter whether we view this trend in normative terms7 or in economic terms, from the perspective of developing regions and small companies, one factor that cannot be ignored is the crucial forces of social and environmental standards ‘in situ’.

5. REGIONS IN THE WORLD ECONOMIC TRIANGLE – CONCLUSIONS Compared with the emphasis placed by Gary Gereffi (1994; 1999) that local development options are primarily determined by the specific structures of global value chains, our empirical studies arrive at a more differentiated assessment. In the context of the triangle, the ability or inability of local actors to deal with world economic challenges, to build independent techno-organizational competences and global governance capacities prove to be important influencing factors for development successes or failures of local industrial locations in the world economy. Therefore, we can continue to say: ‘Regions matter!’. But the empirical studies also point to the limitations of concepts of industrial clusters, innovation systems or systemic competitiveness, all of which concentrated on local relationships without taking adequate consideration of the specific global contexts in which localities or regions are integrated. Depending on the governance structures in specific global value chains, the triangle perspective shows us that there exist ‘windows of opportunity’ or ‘dead ends’ for local development strategies aimed at strengthening competitiveness. ‘Regions matter, but . . .’ the scopes and limits for local governance are influenced by the following global forces highlighted by the world economic triangle:


Local enterprises in the global economy ● ● ●

the specific governance patterns in global value chains; the specific core competences of global lead firms in value chains; the specific governance structures in global networks involved in the setting of standards; and, the concrete rules agreed on in standard-setting networks and the manner in which these rules are implemented and sanctioned, as well as the impacts they unfold in regions.

Beyond that, the triangle concept is useful to show that if regions are to strengthen their competitiveness, it is not enough to use locational policy focused on local forces (intra-cluster relationships). It is essential at the same time: ●

to use the analysis of global governance structures to assess the scope for local strategies and to develop realistic strategies, compatible with the dynamics in the triangle, as a means of avoiding any voluntarist efforts; to build up local governance capabilities in order to play a role in shaping global governance structures (for example global social and ecological standards); and, to link local competences prudently with global resources (for example local technological potentials with technological nodes in global value chains); to use the presence of global actors in local policy networks (for example NGOs, lead firms and international organizations involved in the monitoring and implementation of global standards on the ground) to shape locational factors favourably.

The ‘playing field’ of local actors is thus expanding, above all in complexity (multilevel policy; multi-actor constellations). Furthermore, local actors are confronted with a paradox: the diversity of options is growing (for example the possibility of diversification of sales channels; networking of strengths and global competence pools; coalitions with global actors, aimed for instance at strengthening the social and ecological dimensions in situ). Yet, at the same time dense interaction between local and global governance gives rise to restrictions on local action (for example the power of global lead firms and the growing number of global standards). Whether and how development blockades will emerge is a question that can only be answered empirically.

Regions in the ‘world economic triangle’


NOTES 1. By ‘region’ we mean subnational units. The trend has been for locational policies to be formulated at the subnational (regional and local) level. This trend is of course a function of the size of a given economy. In small countries like Uruguay or Costa Rica, locational policy continues to be formulated at the national level. On the other hand, in medium-size countries (like Chile) or in larger countries (like Brazil or Germany) subnational regions are gaining in relevance as spaces of active locational policy. 2. The literature seldom systematically links these inter-governmentalist world economic discourses with approaches focusing on regions in the world economy. But they are complementary in nature. As a rule, the ‘intergovernmentalists’ are not concerned with the question of regional scopes of action and local governance in the world economy. Where they do turn their attention to the issue, they tend to sympathize with concepts that are used to argue that competitiveness comes about on the basis of the interplay between markets, state and private governance ‘on the ground’ (Rodrik, 1997; 2001). The authors concerned with the question of regions in the world economy often have a reciprocal approach: wherever they address structures of the world economy (for example Esser et al., 1996; Messner, 1997), which are for them in essence ‘black boxes’, they tend to refer to the publications of inter-governmentalist theorists of the world economic order. 3. The triangle perspective at first leaves inter-governmental governance structures (like the WTO) out of consideration, but without underestimating their significance. The intergovernmental regulative patterns in the world economy constitute a kind of ‘global macropolicy’ which the triangle approach views as a set of external data. The triangle investigates the specific global governance contexts and world market structures that are tied into the specific locations. 4. In all developing-country business locations covered by the IDS–INEF project (with the exception of Brazilian automotive suppliers), social and environmental standards play an increasingly important role. In nearly all of the case study locations, ISO standards have assumed growing importance. Evidence from the IDS–INEF projects indicates that global standards are gaining importance in the world economy, and in particular for export-oriented developing countries. Thus far, however, no studies have appeared that provide exact data on the broad significance of global standards in world trade or segments of world trade. There is also a lack of studies that look into how different types of global standards affect local firms, regions and local governance structures. The studies that the IDS–INEF projects have prepared in this area provide some first points of departure for this area (Quadros, 2002; Nadvi and Kazmi, 2002; Navas-Alemán and Bazan, 2002). There is, in other words, need for research in both fields. The following considerations on the significance and impact of global standards in the world economy require additional research efforts to deepen and verify them. 5. In his study on Brazilian automotive suppliers, Quadros (Chapter 10, this volume) shows that ISO standards are seen as a necessary condition to qualify as a partner of the global players. But in this cluster the ISO standards have not contributed to lowering the ‘total transaction costs’ in the value chain. Since there is some doubt as to the reliability and credibility of the Brazilian and the international certifiers, global auto makers are insisting on compliance with additional standards defined and monitored by the auto makers themselves. The result for the Brazilian firms is additional costs for ISO certification, but no corresponding benefits. 6. Dolan and Humphrey (2000) point out that the Kenyan fruit-importing industry has succeeded in developing particularly stringent (sanitary and environmental) standards of its own and that these have become current in various global sales channels. 7. In normative terms there might be disagreement on how to judge the rise in social standards in export-oriented companies due to global standards at the expense of the exclusion of small, employment-intensive companies in developing countries from global value chains.


Local enterprises in the global economy

REFERENCES Barrientos, S. (2000), ‘Globalization and ethical trade. Assessing the implications for development’, Journal of International Development, 21, 412–27. Barrientos, S., C. Dolan and A. Tallontire (2001), ‘The gender dilemma in ethical trade’, Natural Resource Institute Working Paper, Chatham: Natural Resources Institute. Bello, W. (2001), The Future in the Balance. Essays on Globalization and Resistance, Oakland: Food First Books. Bergsten, F.C. (1996), Global Economic Leadership and the Group of Seven, Washington, DC: Institute for International Economics. Blowfield, M. (1999), ‘Ethical trade: a review of developments and issues’, Third World Quarterly, 20 (4), 753–70. Braczyk, H.J., P. Cooke and M. Heidenreich (1998) (eds), Regional Innovation Systems, London: UCL Press. Branko, M. (1999), ‘The true world income distribution’, Policy Research Working Paper 2244, Development Research Group, Washington, DC: World Bank. Cable, V. (1999), Globalization and Global Governance, London: The Royal Institute of International Affairs. Clapp, J. (1998), ‘The privatisation of global environmental governance. ISO 14000 and the developing world’, Global Governance, 4 (3), 295–316. Czempiel, E. (1993), Weltpolitik im Umbruch, Munich: Beck Verlag. Dolan, C. and J. Humphrey (2000), ‘Governance and trade in fresh vegetables: the impact of UK supermarkets on the African horticulture industry’, Journal of Development Studies, 37 (2), 147–76. Eichengreen, B. (1999), Toward A New International Financial Architecture. A Practical Post-Asia Agenda, Washington, DC: Institute for International Economics. Esser, K., W. Hillebrand, D. Messner and J. Meyer-Stamer (1996), New Governance Patterns for Industrial Development, London: Frank Cass. Friedman, T.L. (1999), The Lexus of the Olive Tree: Understanding Globalization, New York: Harper Collins. Fuchs, P. (2000), ‘Codes of conduct – neue Handlungsoptionen zur Regulierung transnationaler Konzerne “von unten”?’ in Ch. Dörrenbacher and D. Plehwe, (eds), Grenzenlose Kontrolle? Organisatorischer Wandel und politische Macht multinationaler Unternehmen, Berlin: Edition Sigma. Gereffi, G. (1994), ‘The organization of buyer-driven global commodity chains. How U.S. retailers shape overseas production networks’, in G. Gereffi and M. Korzeniewicz (eds), Commodity Chains and Global Capitalism, Westport: Greenwood Press, pp. 95–122. Gereffi, G. 1999, ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Gerken, L. and O.G. Lambsdorf (2001), Ordnungspolitik in der Weltwirtschaft, Baden-Baden: Nomos Verlag. Glaser, N. (1999), ‘GTZ unterstützt Entwicklung von Soziallabels’, Entwicklungspolitik, No 23/24, 11–13. Helleiner, G.K. (2001), ‘Markets, politics and globalization: can the global economy be civilized’, Global Governance, 7 (3), 243–63. Hilowitz, J. (1997), ‘Social labelling to combat child labour: some considerations’, International Labour Review, 136 (2), 215–32. Humphrey, J. and H. Schmitz (2002), ‘Developing country firms in the world

Regions in the ‘world economic triangle’


economy. Governance and upgrading in global value chains’, INEF-Report No 61, Duisburg: Institute for Development and Peace. Jochimsen, R. (ed.) (2000), Globaler Wettbewerb und weltwirtschaftliche Ordnungspolitik, Bonn: Dietz Verlag. Jones, C., W. Hesterly and S. Borgatti (1997), ‘A general theory of network governance. Exchange conditions and social mechanisms’, Academy of Management Review, 22 (4), 911–45. Lundvall, B.A. (1992), National Systems of Innovation, London: Continuum. Mah, J.S. (1997) ‘Core labour standards and export performance in developing countries’, The World Economy, 20 (6) 773–85. March, J.G. and J.P. Olson (1989), Rediscovering Institutions, The Organizational Basis of Politics, New York: Free Press. Messner, D. (1997), The Network Society. Economic Development and International Competitiveness as Problems of Social Governance, London: Frank Cass. Messner, D. (2002), ‘The concept of the “world economic triangle”. Governance patterns and options for regions’, IDS Working Paper 173, Brighton: Institute of Development Studies. Mittelman, J.H. (2000), The Globalization Syndrome. Transformation and Resistance, Princeton: Princeton University Press. Nadvi, K. and G. Halder (2002), ‘Local clusters in global value chains. Exploring dynamic linkages between Germany and Pakistan’, IDS Working Paper 152, Brighton: Institute of Development Studies. Nadvi, K. and S. Kazmi (2001), ‘Global standards and local responses. Case study from Pakistan’, mimeo, Brighton: Institute of Development Studies. Nadvi, K. and H. Schmitz, (eds) (1999), ‘Industrial clusters in developing countries’, Special Issue of World Development, 27 (9). Navas-Alemán, L. and L. Bazan (2002), ‘Value chain governance and local implementation of quality, labour and environmental standards: opportunities for upgrading activities in the footwear industry’, unpublished manuscript, Brighton: Institute of Development Studies. North, D. (1990), Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press. Ocampo, J.A. (2002), ‘La reforma financiera internacional, una agenda ampliada’, in C. Maggi and D. Messner (eds), Gobernanza Global desde una Perspectiva Latinoaméricana, Caracas: Nueva Sociedad. Porter, M. (1990), The Competitive Advantages of Nations, London: Macmillan. Porter, M. (1998), ‘Clusters and the new economics of competition’, Harvard Business Review, Nov–Dec: 77–90. Powell, W. (1990), ‘Neither market nor hierarchy. Network forms of organisation’, Research in Organizational Behaviour, 12, 295–336. Quadros, R. (2002), ‘Global quality standards, chain governance and the technological upgrading of Brazilian auto-components producers’, IDS Working Paper No 156, Brighton: Institute of Development Studies. Rabellotti, R. (2001), ‘The effect of globalisation on industrial districts in Italy: the case of Brenta’, IDS Working Paper No 144, Brighton: Institute of Development Studies. Reichert, T. (2000), ‘Vom Beschluß zur Umsetzung – Sozialstandards in der Weltwirtschaft’, Discussion Paper/unpublished, Eschborn: Gesellschaft für technische Zusammenarbeit/GTZ.


Local enterprises in the global economy

Rodrik, D. (1997), Has Globalization Gone too Far? Washington, DC: Institute for International Economics. Rodrik, D. (2000), ‘Governance of economic globalization’, in J. Nye and J. Donahue (eds), Governance in a Globalizing World, Washington, DC: Brookings Institution Press, pp. 347–66. Rodrik, D. (2001), The Global Governance of Trade. As if Development Really Mattered, New York: UNDP. Sassen, S. (2000), Cities in a World Economy, Thousand Oaks: Pine Forge Press. Schmitz, H. (1999), ‘Global competition and global co-operation. Success and failure in the Sinos Valley’, World Development, 27, 1627–50. Schmitz, H. and P. Knorringa (2000), ‘Learning from global buyers’, Journal of Development Studies, 37 (2), 177–205. Scott, A. (2001), Global City-Regions, Oxford: Oxford University Press. Siebert, H. (1999) ‘Disziplinierung der Nationalen Wirtschaftspolitik durch Internationale Kapitalmobilität’, in D. Duwendag (ed) Finanzmärkte im Spannungsfeld von Globalisierung, Regulierung und Geldpolitik, Berlin: Springer Verlag. Stiglitz, J. (2000), ‘Reforming the global economic architecture. Lessons from the recent crisis’, The Journal of Finance, 1, 1508–21. Storper, M. (1997) The Regional World. Territorial Development in a Global Economy, New York: Guilford Press. Williamson, J. (1997), ‘The Washington Consensus revisited’, in E. Emmerij (ed.), Economic and Social Development into the XXIst Century, Washington, DC: Inter-American Development Bank. Williamson, O.E. (1979), ‘Transaction-cost economics. The governance of contractual relations’, Journal of Law and Economics, 22 (2), 233–61. Young, O.R. (1999), Governance in World Affairs, New York: Cornell University Press.


Making sense of global standards Khalid Nadvi and Frank Wältring*



Globalization has heightened interest in global standards. Such standards address a wide range of issues from labour conditions, health and safety norms, quality management procedures, to environmental and social concerns. Various actors take a keen interest in these standards, notably consumers in advanced countries, international NGOs (non-governmental organizations), global buyers and producers, and UN agencies. Their interests and motivations for promoting standards differ widely. Some are concerned with defending or advancing narrow interests. Others are driven by wider concerns such as protecting the vulnerable (people or environment) or halting the ‘race to the bottom’. However, governments and enterprises in developing countries find that, while they are expected to comply with global standards, they have little say in the making of standards. Not surprisingly, global standards feature significantly in key policy debates on the future of the world economy. This is apparent at four levels and revolves around the role of standards in: i) promoting economic efficiency and international trade; ii) reflecting concerns on the social and ecological dimensions of international trade; iii) providing pressure and opportunity to switch from the low to the high road of competitiveness; and, iv) pointing to new forms of global governance. However, the proliferation of standards in recent years has made it very difficult to conduct an orderly analysis of these debates. In this chapter we seek to reduce the confusion and complexity that arises from the proliferation of standards. Our objective is to enable the reader to gain an overview, to categorize standards and actors, and to bring out common trends. The remainder of this introduction briefly outlines the four policy debates relating to global standards. 1.1

Global Standards Improve Efficiency in the World Economy

Standards, by providing a set of common and widely understood benchmarks, have traditionally been seen as an important factor in smoothing trade relations. By efficiently transmitting information, standards reduce 53


Local enterprises in the global economy

transaction costs. The demand for standards has, however, accelerated sharply with the globalization of production and trade. The ever-more complex interrelations to be found between producers, suppliers, retailers and consumers across the world has accentuated the need for harmonization of norms and forms of codification. As Reardon et al. (2001, 6–7) state, standards reassure ‘. . . consumers about credence characteristics such as food safety, worker conditions and location authenticity . . . [which] cannot be known to consumers through sensory inspection or observation in consumption’. In addition, standards enhance business-to-business ties by improving coordination of global production and distribution systems. This is particularly pronounced given the ways in which local producers are integrated into global value chains. Value chains have emerged as a powerful tool in understanding how the distinct functions that turn raw materials into traded end-products are inter-linked through complex arrangements between globally diverse actors (Gereffi, 1999; Sturgeon, 2000; Humphrey and Schmitz, 2001; Kaplinsky, 2000). Common standards promote compatibility between diverse actors within the chain, help organize their linkages, reduce transaction costs associated with chain governance and lower risks for actors in the chain. 1.2 Global Standards Underline the Social and Ecological Dimensions of International Trade The concerns that standards now address have gone beyond technical norms to environmental issues, working conditions, human rights and social and ethical values. The focus is not only what is produced, but how it is produced and delivered. In some cases, such as organic food standards, the ‘what’ and the ‘how’ are closely connected. This underlines the importance attached to the social and ecological dimension of international trade and has prompted a vociferous debate, both inside and outside international forums, between those who argue that such standards pose new forms of non-tariff barriers and those who view compliance as one path for developing countries to avert the pitfalls of globalization. A particular pitfall is the ‘race to the bottom’ where Southern firms are locked into a downward spiral of competition based on lowering wages and the flouting of labour and environmental norms. This raises particular challenges for developing countries with weak social and environmental infrastructure and regulatory institutions.

Making sense of global standards


1.3 Global Standards can be a New Basis for International Competitiveness Standards provide a basis to differentiate markets and create competitive niches. Compliance with ethical, social and environmental standards can be one important way to add value. However, accruing the rents that come with compliance requires new forms of knowledge. Upgrading technology is a well-understood, albeit often difficult, process for firms and for governments. However, global process standards demand upgrading on process and management issues. Small and medium sized enterprises (SMEs) from developing countries are especially vulnerable and ill-equipped for this. 1.4 Global Standards imply New Forms of Global Governance in the World Economy Global process standards create new challenges for private and public governance at both local and global levels. First, they point to the relative erosion of the regulatory powers of the nation state. The growing influence of global standards in global markets is likely to weaken national standards. This is because national standards must increasingly comply with international norms, or risk becoming irrelevant. Consequently, sovereignty over standard setting moves out of the national domain. Second, whereas national standards were largely defined in the public arena, global process standards are increasingly being formulated by private and public–private initiatives. The range of private actors engaged in the process of defining and implementing standards is varied. They include private business, consisting of firms, business associations and chambers of commerce, as well as other, not-for-profit, actors within civil society, such as issue-based NGOs, trade unions and concerned consumer groups. Moreover, many of these private actors, from firms to NGOs, are distinct at the local and global levels. This suggests new institutional arrangements and complex local and global networks of public and private actors. It also indicates potential conflicts between the competing interests of the forprofit business sector and not-for-profit civil society actors. Mediating such conflicts requires new forms of global governance. Third, they raise specific challenges for local government and local policy networks. These include monitoring global standards and developing mechanisms for dialogue with global NGOs, and, more generally, building new alliances between local and global actors that also acknowledge potential differences in the interests of local and global actors. Improving our understanding of the role of standards within these debates is hampered by the increasing number and types of standards.


Local enterprises in the global economy

What limited evidence there is focuses on individual standards, which often misses the bigger picture. This chapter’s primary contribution is to reduce the complexity and confusion in this area by providing a comparative overview across the range of different global standards. It puts forward typologies and distinctions that make comparison easier and draws out common threads. Thus, it distinguishes between different steps in the policy cycle of a standard, between different networks of actors at the local, national and global levels, and the different governance forms engaged in the steps. Finally, it provides a dynamic perspective by illustrating the distinct trajectories in the evolution of different types of standards. The chapter is structured as follows: the next section provides initial distinctions and typologies. Section 3 then uses these distinctions in the field of quality assurance and health and safety standards. Section 4 reviews environmental and social standards. The final section draws together core elements of the typology and compares trajectories of standard evolution.



Standards are agreed criteria, or as Hawkins (1995, 1) states ‘external points of reference’, by which a product or a service’s performance, its technical and physical characteristics, and/or the process and conditions under which it has been produced or delivered, can be assessed. David (1995, 22) argues that ‘. . . having dependable standards . . . [makes] it simpler for all parties to a deal to recognize what is being dealt in’. Compliance criteria need to be measurable with well-defined procedures. In addition, standards require a degree of authority that ensures that they are legally binding or voluntarily enforced, with sanctions for non-compliance. 2.1

Product and Process Standards

It is now common to distinguish between product and process standards. Traditionally, standards focused on the characteristics of a product. This included, for example, size, composition, function and health and safety impact. Therefore, product standards were sector-specific and technical in nature. They were generated by private business as well as by government.1 Product standards began to be internationalized from the 1950s onward. This involved coordination in regional forums (such as the European Union – EU) and in international arenas (such as the International Electrotechnical Commission, the International Telecommunications Union, and most prominently, the International Organization for Standardization – ISO). Harmonization of national product standards

Making sense of global standards


facilitated international trade and reduced potential ‘market failures’ that could arise from distinct national standards.2 Since the mid-1980s there has been growing interest in process standards. These address specific aspects of the production process. This implies that, unlike product standards, process standards can be generic, sector- or firmspecific. In some cases, process standards include clearly defined and measurable benchmarks, allowing firms to gauge how well they perform in reaching particular targets. In other cases, benchmarks can be contentious, especially in areas where ethical, social and environmental values are not universally held. Such differences in values cause friction because, like product standards, the formulation of process standards has moved from the national to the international arena. Moreover, a wide range of actors, both public and private, are involved in the formulation, implementation and monitoring of these standards. The distinction between product and process standards, while widely used, is becoming hazy as producers seek to reduce the range of applicable standards by incorporating process concerns (such as levels of pesticide residues in food crops) into product features. Within the universe of standards there are a number of sub-categories. This adds to the confusion, especially where the boundaries between, and within, these sub-categories are unclear. Thus, there is ambiguity concerning the distinction between standards, codes and labels. We consider labels and codes of conduct to be a distinct sub-category of standards. Labels provide consumers with a simple way to rapidly and easily acquire information about product characteristics (like the woolmark label, which shows that a garment is made from pure wool), or about conditions of production (such as the fair trade label). Labels tend to be sector-specific and concentrate on particular themes. In contrast, codes of conduct are usually firmspecific. They stipulate the criteria of accepted practices adopted by a company and transmitted to its employees, its suppliers and its wider stakeholders, including its clients and shareholders. These practices can range from employment conditions, social and environmental norms, to the firm’s role in the community (van Liemt, 1998b). A further point to note is the distinction between global process standards that are universal in nature and those that, while global, are adjusted to national circumstances. Often company codes of conduct, or certain social and labour standards, incorporate specific International Labour Organization (ILO) conventions or require that a firm complies with national regulations regarding work or safety practices. Clearly, legal stipulations on issues like the minimum wage, working hours and social security benefits vary from country to country. In contrast, some global standards have been adapted as specific national standards. For example,


Local enterprises in the global economy

there are various national versions of the ISO 9000 quality assurance standard, that is Brazil’s ABNT9000, Korea’s KS9000, Pakistan’s PS9000. Irrespective of these differences, all standards provide a codified basis for conveying information and this information provision can be critical. As well as facilitating transactions, such information can also facilitate coordination between interdependent agents. This can serve to reduce costs and ensure efficient use of resources within the supply chain. Such coordination functions are especially significant where uniformity is of importance, or where complex decisions require detailed information of products and production processes. This is especially important in technologically sophisticated production systems. Here, quality assurance (and technical product) standards allow firms to maintain complex supply chains and to engage in joint R&D with diverse and distant suppliers. To be of use, and to provide confidence to consumers and firms, standards have to have legitimacy. This is tied to the manner in which monitoring and certification takes place, as well as the type of actors engaged in defining the standard. This takes us to the distinct policy steps associated with the formulation and implementation of standards. 2.2

The Policy Cycle: Four Steps

The policy cycle for standards comprises four distinct steps: standard setting; standard monitoring; assistance on achieving standard compliance; and, sanctions for non-compliance. Each of these steps involve diverse actors. The credibility of a standard is largely related to the types of actors engaged in setting the standard and monitoring compliance. With compliance, for example, there are three distinct alternatives. First party certification relies solely on self-monitoring. In terms of public legitimacy, this usually results in the least degree of credibility. Second party certification shifts monitoring to the user of the product or services, or alternatively to trade bodies who monitor on behalf of their members. While this can enhance the credibility of the standard, there can be conflicts of interest. Third party certification transfers monitoring to neutral and independent auditors. The credibility of the certification is directly linked to the credibility of the auditor. Auditors can include accredited firms who provide market-based certification services, or NGOs and civil society groups who uphold the values associated with the specific standard. The range of actors engaged in these four distinct steps can be extensive. Table 3.1 summarizes the main categories of such actors who range from private business, NGOs and trade unions, to the public sector. Moreover, such actors can operate at local, national and global levels, and be engaged in the distinct functions of formulating and monitoring the implementa-


Making sense of global standards

Table 3.1

Types of actors engaged in defining and implementing standards

Types of Actors




Local or national firms, trade associations and certification firms

TNCs, Global trade associations, global certification firms


Local or national NGOs, consumer groups and trade unions

Global NGOs, international trade union federations

Local and national government and standards organizations

International and regional organizations



tion of standards. To understand how standards are set and assessed we need to have an understanding of networks. 2.3


The networks required to define complex standards often come about because the resources needed to formulate the standard and make it credible are distributed amongst a variety of actors (Messner and MeyerStamer, 2000). Moreover, there is an element of interdependence amongst such actors within the network (Messner, 1997). Not only do different actors come together because they have specific core competencies, they also need each other in order to make a standard reliable, transparent, efficient and legitimate. For example, in defining environmental standards global NGOs can provide a core competence in determining the criteria against which compliance is measured. This can also enhance the standard’s legitimacy because consumers are more likely to attach credence to the claims of standards formulated in such partnerships, than standards that evolve from business alone. At the same time, such NGOs need businesses to implement standards in order to achieve their objectives. Hence, the pressure to work with business when defining a meaningful standard. Bringing such diverse agents together is a complicated task. The various parties have to agree on common rules. This requires communication and a modicum of trust (Messner, 1997). Without the latter, each actor would seek to promote its own objective with little regard for collective concerns. Apart from different interests, power structures are involved in different network constellations. An actor’s influence and centrality increases in relation to the importance that other actors ascribe to the resources controlled


Local enterprises in the global economy

by him or her, and their core competencies. Core competencies could be specific expert knowledge, control of information and communication resources, reliability and legitimacy resources, and control over financial resources (Messner, 1997). The role and power of local and global actors in shaping standards differ. National governments and national standards organizations often lack the necessary capacity to define and implement standards, while local firms and trade associations can be weak in formulating commonly agreed norms. This is especially so where local actors are closely tied into global production through value chains in which power rests with external lead firms. Similarly, local NGOs may monitor globally-defined standards; however, their ability to shape such standards and influence global NGOs is often limited. Thus, the relative influence of global and local/private and public actors in defining and monitoring standards has clear consequences for the nature of governance. 2.4

Typology for Global Standards

Making sense of the diversity of global standards is essential for researchers and policy makers. The first step is to construct a typology to map the distinct standards. Initially, standards can be distinguished according to the following criteria: ● ● ● ●

● ● ●

scope – process, product standards geographical reach – national, regional, international function – social, labour, environmental, quality, safety, ethical key drivers – public, private (business and not-for-profit organizations) public–private forms – management standards, company codes, labels coverage – generic, sector specific, firm/value chain specific regulatory implications – legally mandatory, necessary for competition, voluntary

On the basis of some of these distinctions, a framework for reviewing global standards is outlined in Table 3.2. The typology in Table 3.2 is insufficient for providing a sense of the trends in standard development. Thus, we use the notion of ‘generations’ of standards to highlight the chronological stages of development of different standards and their changing influence. This should not be viewed as a sequence of superseding ‘generations’. Rather, different ‘generations’ of standards can, and do, co-exist at the same time. We turn now to applying this typology to the distinct bodies of global standards.


Quality assurance Environmental Health Labour Social Ethical


Key drivers International business International NGOs International trade unions ● International • organizations

Codes of conduct ● Firm/value● Label • chain specific ● Standard ● Sector specific ● ● Generic •


Typologies for global standards

Field of application

Table 3.2

Regulatory implication

First party ● Legally mandatory Second party ● Market competition ● Third party • requirement ● Private sector ● Voluntary • auditors ● NGOs ● Government ●

Certification process



Local enterprises in the global economy


Globalization of production has accelerated demand for greater control over quality assurance in production processes. Thus, quality assurance standards have become directly linked with supply chain management. They potentially influence production, outsourcing and the increasingly complex interrelations that exist between producers, suppliers, distributors and retailers. As Table 3.3 shows, these standards can be distinguished according to distinct ‘generations’. This distinguishes between international quality management standards that are generic, sector-specific, and more recently, firm-specific. We discuss each of these separately. In addition, this section briefly reviews leading international health and safety standards relating to food production value chains. Although they are not quality assurance standards, they are closely related in terms of their function and their consequences for management practices within the production process. 3.1 Generic Standards – ISO 9000 (International Organization for Standardization) The ISO 9000 standard provides assurance that a product or service conforms to established and specified requirements, and that the firm or service provider has appropriate quality management procedures in place. The standard is seen as promoting better and more assured control of quality within international supply chains, improving market transparency of suppliers and reducing transaction costs related to quality management. We view ISO 9000 as the first generation of global quality management standards. The standard is generic and can be applied to manufacturing, service, and public sectors. It is the most widely held international standard. It constituted the first foray by the ISO in the area of process, as opposed to product, standards. Launched in 1987, over 340 000 ISO 9000 certificates had been issued worldwide by the end of 1999, with certification levels rising annually by over 26 per cent (ISO, 2000). More than half of these certificates were issued in the EU, although the most rapid growth was seen in Australia, the USA, Japan and China. The main driver behind ISO 9000 is private business, but its roots lie in the public sector. It is based on the British Standards Institution’s quality management standard, BS5750. Developed in 1979, BS5750 emerged from public standards designed for the UK defence industry. It was primarily adopted by public sector enterprises, and by the early 1980s was being promoted in the UK as a tool to enhance private sector competitiveness (Seddon, 2000).3

Making sense of global standards


Although this is a voluntary standard, the popularity of ISO 9000 stems from both public and private pressures. Public regulators have made it a mandatory requirement in many markets. For example, the EU has adopted the standard as part of its ‘Global Approach to Testing and Certification’, which states guiding principles for EU policy on conformity assessment (Wilson, 1999). Within the private sector, many companies use the standard as a filtering mechanism to assess the process competencies of their suppliers. While they do not necessarily rely on the standard, those without ISO 9000 certification are often excluded from the supply chain in various sectors and markets. Thus, the standard is seen by many developing country firms as key to obtaining access and enhancing competitiveness in global markets. Before reviewing the standard itself, it is worth briefly considering the structure of the ISO. The ISO is an international non-profit and nongovernmental federation of national standards organizations. Constituted in 1947, and based in Geneva, it now has 138 national standards organizations as members. It is, however, a somewhat opaque body. As national standards organizations, its members have the regulatory task of setting and defining national standards. In some countries these are purely publicsector bodies. However, in most industrialized countries such organizations involve participation by the private sector in standard formulation. In many cases, these are private organizations (for example, DIN, Deutsches Institut für Normung, the German Standards Institute) wherein government has a limited role, and business and sectoral associations are the main drivers. Thus, the ISO is often referred to as a ‘hybrid private–public regime’ (Clapp, 1998, 295). Standard formulation in the ISO is a long, complex and decentralized process. Detailed negotiations over the exact content of standards are undertaken in 187 technical committees, over 500 sub-committees and some 2000 working groups. Each deal with specific standards or sectors, with representation from industry, research centres, government, consumers, and international organizations. In addition, strategic advisory groups discuss what role the ISO might take in new arenas (Clapp, 1998). Developed country bodies and institutions, especially standards organizations, are most active in the various forums and committees. Many developing country members have rarely, and in some cases never, been part of these deliberations. The ISO itself admits that eight countries provide 80 per cent of all the secretariats of the technical committees, sub-committees and working groups of the organization (ISO, 2001). Haufler (2000, 6) also argues that large enterprises from industrialized countries are especially influential, and that despite the public–private nature of the organization, the ISO is effectively a ‘corporate private regime’.



b) HACCP: Health and safety standards

2nd generation a) AS 9000, SECTOR QS 9000 SPECIFIC

ISO 9000


a) Large TNCs, sector business associations, accredited certification bodies b) International public institutions (e.g. WHO, FAO), national control institutions with public duties, governmental representatives

The (ISO) represented through national standardization bodies, large business actors mainly from industrialized countries and accredited certification bodies

Actors involved

Influence in international trade

a) TNCs, lead firms Increasing influence in in the chain technically complex sectors where specialized quality assurance codes are required b) National Increasing influence in governments, international pharmaceutespecially in ical and food-based trade industrialised with growing concerns countries relating to process management in the international food chain

Industry (trade Voluntary, but increasingly associations, TNCs, becoming mandatory in certification bodies) some European markets. Also gaining influence in the USA and Japan

Key drivers

Different generations of global quality management standards

1st generation GENERIC


Table 3.3

b) 3rd-party certification through public–private institutions with public duties

a) 3rd-party marketbased auditors

3rd-party marketbased auditors

Certification process


3rd generation COMPANY BASED

DaimlerChrysler Supermarket Codes (Tesco/ Sainsbury)

c) EUREPGAP: Food Quality, and Crop Management standards

Powerful TNCs with a dominant position in the world market and a leading role in their supply chain

Extremely prominent in European fresh produce value chain, adopted by all leading UK supermarkets and food importers

TNCs, lead firms in Increasing influence due to the chain technological based complexity in know-how intensive sectors, and also in the food products sector.

c) Food retailers, c) Private Sector importers and suppliers Industry

1st- and 3rd-party monitoring

c) 3rd-party marketbased auditors


Local enterprises in the global economy

The ISO 9000 standard contains guidelines for setting up a quality management system within a firm or organization. It verifies that the certified firm has in place a well-documented quality system, including the ability to trace purchases from suppliers. The supplier does not have to be certified, although there is growing pressure on first-tier suppliers also to be compliant with the standard. The standard does not address, nor is it a guarantee of, product quality. Rather, it provides assurance that the quality management procedures of the firm are independently certified as conforming to accepted norms, and that the firm has in place a mechanism for responding to the needs and quality concerns of its customers. One needs to distinguish between quality management and product quality. It is possible for a certified firm to manufacture products that might be considered of poor quality. Nevertheless, it would be expected that a proper implementation of the standard would result in quality improvements as quality concerns of consumers are fed back to the firm. However, Seddon (2000) argues that the emphasis on maintaining proper documentation results in a paper trail that falls far short of a true quality management system. Despite this criticism, compliance can lead to improvements in quality practices and be a tool in upgrading process management. Among entrepreneurs it is common to hear: ‘It is not difficult to get the ISO certificate, but it is difficult to keep it.’ This underscores the ongoing nature of monitoring and certification. Once certified, the firm has to undergo regular sixmonthly audits and a re-certification every three years. During this time it has to demonstrate an improvement in its quality management practices. The ISO itself does not monitor compliance or issue certificates. Instead, certification is undertaken by independent auditors who offer marketbased services. Leading international certification firms include SGS (Société Générale de Surveillance), DNV (Det Norske Veritas), Lloyd’s Register, Moodys, and BVQI (Bureau Veritas Quality International), as well as national standards organizations such as the British Standards Institution. Certification costs can be high. In general, costs depend largely on the nature and scope of the certification, and on competition between certification bodies. Civil society actors are not involved in the monitoring process. As outlined earlier, lead firms within the value chain increasingly demand compliance from their first-tier suppliers. However, compliance does not necessarily require certification of quality assurance practices adopted by second and third-tier suppliers further down the chain.4 Certification costs are pushed down the supply chain, as they are the responsibility of the certified supplier and not the lead firm. Criticisms have been made that, due to the costs of certification and the management changes required, ISO 9000 is skewed against small firms (UNIDO, 1996).

Making sense of global standards


Assistance on compliance is limited. In some countries, firms have been provided financial subsidies to offset costs of compliance, and lead firms in global supply chains do assist local suppliers to incorporate the standard (Nadvi, 1999). The ISO itself is engaged in promoting standards in developing countries through the committee on developing countries (DEVCO) (ISO, 2001). Nevertheless, technical support is primarily obtained through specialist service providers and technical consultants. Sanctions on non-compliance are largely market enforced. A certificate can be withdrawn by the certification agent.5 Usually, however, the firm and the certification body share a common interest to help the firm improve on its compliance. Yet this can lead to complications, especially where potential conflicts of interest arise from close ties between auditors and technical consultants engaged in helping firms implement the standard. Ultimately, it is the reputation of the certification agent that is at stake. Where an end user finds that a certified firm does not comply in its procedures with the stipulations of the standard, it can raise the matter with the national accreditation body with which the auditor is accredited. This is rare. It points to the potential weakness of the sanctions for non-compliance within the code. It also underlines the earlier observation that, for many supply chains, compliance to ISO 9000 is only considered an entry requirement and not a guarantee of a particular level of process competence. In part, it is these considerations that have resulted in moves to more specialized quality assurance standards. 3.2

Sector- and Firm-specific Standards

The second and third generation of quality assurance standards were developed in part on the basis of ISO 9000. Their distinguishing feature is that they move from generic to sector-specific (second generation) and firmspecific standards (third generation). This shift reflects the increasing technical complexity of production, supply chain management in particular sectors, and particular firm value chains. Leading examples of international sector-specific quality assurance standards include the AS 9000 and QS 9000 standards that apply to the aerospace and automobile industries respectively. They contain ISO 9000 in its entirety, but have additional requirements specific to the sector to which they apply. They adopt the same documentation and monitoring principles of the ISO standards and constitute the generally accepted quality assurance norm within their respective industries. QS 9000, for example, is commonly required of first-tier suppliers in the automobile industry. Both standards emerged as a result of coordinated actions by lead firms within their specific sectors to increase transparency and quality assurance


Local enterprises in the global economy

along their own supply chain. QS 9000 was introduced by Ford, Chrysler and General Motors. AS 9000 was officially released in 1997 by the Society of Automotive Engineers (SAE) with General Electric Aircraft Engines playing a leading role.6 The dominance of US firms and their supply chain practices in the formulation of QS 9000 has resulted in many leading European and Japanese manufacturers requiring their global suppliers to comply with their own national standards for quality assurance in the auto sector. Quadros (2002) observes this trend in the case of Brazil, where suppliers to firms like Peugeot, Renault and Toyota were no longer relying solely on QS 9000 standards (Quadros, 2002). Some lead firms were demanding that suppliers comply with specific national standards, such as the German VDA (Verband der Automobilindustrie) by Volkswagen and the French EAQF (Evaluation d’Aptitude Qualité Fournisseurs) by Peugeot. Finally, an emerging trend is the development of firm-specific standards. This is pronounced in technically complex industries, and where lead firms manage extensive and complex international supply chains. Some large international manufacturers with diverse and globally distributed supply chains, such as the newly-merged Daimler-Chrysler corporation, require more detailed standards from their suppliers than the sector-specific standards. There are also signs of such firm-specific standards in the food products sector to which we now turn. 3.3.

Health and Food Safety Standards

There has been a rapid concentration of food retailing and, consequently, of food production and packaging in the developed world during the 1990s (Dolan and Humphrey, 2000). This is in line with the spectacular rise of international supermarket chains and large-scale processors. Such concentration has resulted in complex contracts between global food producers and retailers, and food producers and suppliers in the developing world. Coordinating these value chains has increasingly involved compliance with various food standards. Food safety standards are one specific form of sector standards (second generation). To reassure consumers, numerous standards and labels have emerged dealing with: food safety and quality; characteristics such as organic produce, environmental and ethical considerations (animal welfare standards); and, the regional authenticity of farm products (British farm standard label or Kenyan flower council standards). They have emerged as strategic tools in creating brand identity; facilitating product differentiation and market segmentation (Reardon et al., 2001). Some of these new food standards are promoted by public bodies or governed by regional regulations. Others are the result of private initiatives,

Making sense of global standards


with supermarkets and suppliers in the food sector becoming standard setters. Here, we review a leading public food standard, Hazard Analysis and Critical Control Point (HACCP), and a private food standard, the European Retailers Representative Group’s standards on Good Agricultural Practices (EUREP-GAP). Both have gained wide currency in recent years. Food safety codes have been an important area of intervention by the World Trade Organization (WTO) through the agreement on Sanitary and Phytosanitary Standards (SPS). The SPS agreement covers sanitary (human and animal) and phytosanitary (plant health) measures to protect human or animal health from food-borne risks, plant-carried diseases or pests. These measures can take many forms, such as requiring specific treatment or processing of products, setting maximum levels of pesticide residues, or restricting use of certain additives in food. They apply to domestically produced food, livestock and plants, as well as to imported products.7 The leading global initiative within the framework of the SPS agreement is the HACCP. This is a food safety management standard that concentrates on prevention strategies for known hazards. It also aims to minimize the risks of such hazards occurring at specific and critical points in the food chain. It was developed by a network of public actors on the global and local level. The UN’s Food and Agriculture Organization (FAO) and the World Health Organization (WHO) developed the ‘Codex Alimentarius’, and adopted national guidelines for the application of HACCP within the member countries in 1993. It is required for food products such as fish, seafood, meat, dairy products, fruit and vegetables (FAO, 1998). The application of the HACCP system is compatible with the implementation of quality management systems, such as the ISO 9000 series, and HACCP is the system of choice for the management of food safety (FAO, 1998). As an international standard in the food sector, HACCP has become a mandatory requirement in most industrialized countries. In the US approximately 38 states have made HACCP mandatory. The EU also introduced HACCP as a mandatory standard in 1993 (UNIDO, 1999), while many other governments have integrated the standard into law. This is a reflection of the interest of public bodies to improve transparency and safeguard the health of the population, but sometimes it is also mandated by governments to strategically position domestic exporters. Agribusiness firms targeting export markets, or newly deregulated domestic markets, have begun to adopt HACCP for strategic and competitive reasons (Reardon et al., 2001). In addition to public standards, there are important private food safety standards motivated by both business and NGOs. A prominent example is


Local enterprises in the global economy

EUREP-GAP. Starting in the late 1990s, EUREP-GAP has very rapidly gained wide circulation in the European fresh produce retail sector. EUREP has over 100 members, including prominent retailers and suppliers. It ‘has authorised 20 certification bodies to carry out its audits in over 25 different countries’ (FPJ, 2001). In the UK, for example, the standard has been adopted by the five leading supermarket chains that collectively account for 70 per cent of total UK food retailing. The standard is, in effect, an industry-wide response at formulating a single code that can offset the numerous country and firm-specific standards. EUREP-GAP covers a range of issues, with a particular focus on integrated crop and pest management. This includes proper documentation of quality and plant health from the seed treatment and nursery stages of commercial farming, through to the use of fertilizers and pesticides and finally to harvesting and packaging stages in production. Certification is undertaken by independent auditors accredited with EUREP (such as AFAQ, Bureau Veritas, and SGS). The EUREP-GAP protocol, while specific to the farm produce sector, is in many ways procedurally similar to the ISO 9000 standard, that is the standard places greatest emphasis on issues that directly impact on food hygiene, food quality and human welfare. The standard also goes beyond ‘traditional’ quality assurance codes in that it explicitly recognizes environmental and social considerations (see EUREP, 2001). To conclude, quality management standards are the most popular of global process standards. They emphasize issues of traceability, documentation and stage-wise quality assurance. The key drivers behind such standards are private business. These standards have gained wide popularity due to their reliance on independent auditing, and the reduction of transaction costs associated with organizing complex global value chains. In technically complex sectors, as well as industries where customer credence is critical, we observe a move away from generic to sector-specific standards. Many standards, especially food safety standards, have been defined by public interests or through public–private partnerships (Reardon et al., 2001). These factors point to public–private network forms of global governance in standard formulation, with business as the key player. In terms of certification and monitoring, quality management standards have well developed independent auditing procedures, indicating arm’s-length governance. Sanctions for non-compliance are either enforced by the market, or applied by national and regional regulatory bodies. However, it is unclear as to how effective such sanctions are, especially through regulatory bodies. Finally, the evolution of quality assurance standards indicates a distinct move from generic to sector- and firm-specific standards. Civil society actors are by and large absent. The next section reviews social and

Making sense of global standards


environmental standards. Here, we observe quite different trends and networks in standard formulation and implementation.



Social and environmental concerns lie at the heart of the new ‘rules’ on international trade. This has resulted in a rapid proliferation of global standards in these areas. Many of these standards have been influenced by multilateral initiatives and leading international institutions. The 1992 UN Earth Summit (UNCED) and the 1995 UN Social Summit provided a dynamic push in promoting environmental sustainability and social development. They also influenced private initiatives on environmental and social standards (Fues, 2000). ILO conferences led to the ILO ‘Declaration on Fundamental Principles and Rights at Work’. In June 1998 a follow-up conference resulted in a consensus on ‘core labour standards’. Since 1995, ratification of individual ILO conventions has steadily increased (Table 3.4). In recent years, discussion has arisen on how to make labour standards more effective, and how to integrate social issues into the WTO’s rule framework. This marked the origin of a continuing debate between industrialized and developing countries, incorporating NGOs and unions from the North and the South. Whereas advocates see the social clause as providing a basis for promoting basic international social rights, opponents fear it as a protectionist instrument. Additionally, exposure to pressure from consumers and organized NGO campaigns has fuelled initiatives by private business to collaborate with NGOs in formulating effective and legitimate environmental and social standards. This has resulted in an increasing number of voluntary codes, labels and standards in the private sector.8 Such developments can be observed in environmental resource-intensive and labour-intensive sectors, particularly those marked by highly globalized production. These include the petro-chemicals, mining, agriculture, forestry, chemicals, textiles, carpets, clothing and footwear industries. Furthermore, such standards are increasingly significant for value chains in sectors where consumer perceptions on ethical, social and environmental norms are a core element of competition. These factors, at the level of global institutions and individual firms, point to similarities in the evolution and trajectory of environmental and social standards. In sharp contrast to most quality assurance standards, environmental and social standards are increasingly formulated in networks that include civil society, public and private actors (Diller, 1999). While such standards clearly differ in scope and focus, these similarities underline the



Freedom of association and collective bargaining




Worst forms of child labour

All 7 conventions


Minimum age


29/1930 105/1957

Forced labour



From 1–7

8) The prohibition and immediate action for the elimination of the worst forms of child labour

7) Minimum age for admission to employment (not less than 15 years)

5) Equal remuneration for men and women workers for work of equal value 6) Discrimination in respect of employment and occupation

3) Forced or compulsory labour 4) Abolition of forced labour

1) Freedom of association and protection of the right to organize (No.87/1948) 2) Application of the principles of the right to organize

Name of convention

ILO core labour conventions and ratification


Table 3.4




139 115









86 83



% of ILO members in 2001

Introduced in November 2000

85 (⫹37)

142 (⫹20)

144 (⫹17)

152 (⫹13) 146 (⫹31)

146 (⫹18)

128 (⫹14)


Number of states ratified (or in process of ratification)

Making sense of global standards


need to consider social and environmental standards together. Their trajectories ‘move’ quite differently to that observed for quality assurance standards. We discern five distinct ‘generations’ of standards (Table 3.5). These include: company-specific codes of conduct; sector-specific codes and labels defined by business; generic standards defined by business; sector-specific codes and labels defined by business–NGO partnerships; and finally, generic standards defined by business–NGO–government partnerships. In these five generations there are three distinct trends in how standards have been defined: (i) by business alone (the first, second and third generation standards); (ii) by business and civil society together (fourth generation); and, (iii) through tri-partite arrangements that link business, civil society and the state (fifth generation). The distinct trajectories of social and environmental standards demonstrate similarities and differences. In social standards we observe a process of convergence through generic standards that involve joint action by industry, NGOs, and the public sector. In contrast, environmental standards are becoming more divergent, responding to the distinct needs of specific sectors. This underlines the difficulties of defining a common framework for minimum global environmental standards. This section is structured around the three trends mentioned above. 4.1

Business defined Standards

The first generation of standards were company codes of conduct. In a recent study, the Organisation for Economic Co-operation and Development (OECD) listed 233 existing company codes of conduct (Diller, 1999). The use of codes of conduct is most widespread in the US. In 1990, 85 per cent of the 100 largest US corporations had a company code. In the UK, this figure was 42 per cent, while in the Netherlands the total was 22 per cent (van Liemt, 1998a). Traditionally, company codes of conduct focused on a firm’s relationship with its employees. More recently, especially in sectors where production is marked by an extensive international division of labour and where consumers are ethically aware, firms have been forced to pay closer attention to social conditions in their supply chains. Thus, some of the more prominent company codes of conduct are found in buyer-driven sectors, such as garments, food, toys and sports goods, and amongst well-known brandname retailers, such as Nike, Reebok, Levi-Strauss, C&A, Mattel, and supermarket chains like the German Karstadt or UK’s Sainsbury (FEER, 2000). In most cases, the code’s guidelines are set by the lead firms and they are internally monitored. Although firms do not have to collaborate with other actors in setting company codes, this can be a difficult process. There are also high transaction costs incurred in monitoring the code. Moreover, the

74 ISO, national standardisation bodies, business mainly from industrialised countries


ISO 14000: Environmental management standards (using the model of ISO 9000)

Not necessary, but gets increasing influence especially in natural resource intensive sector

Sporadic, but with more comprehensive influence according to the sector approach

3rd generation Businessdefined international standards

Enterprise associations, Chambers, suppliers

Enterprise associations

Influence in international trade

ICC, Eco-tex, AVE: Sector specific codes and labels formulated and implemented by enterprise associations

Key drivers

2nd generation Businessdefined sector codes and labels

Actors involved TNCs as lead Existence of a large number of firms of supply firm codes, focused on some chains brand name companies in consumer near sectors and in buyer-driven chains

Examples / Contents

Different generations of global social and environmental standards

1st generation e.g. Nike, Reebok, TNCs and their Company codes Karstadt, etc.: suppliers of conduct Self-obligations of TNCs on the firm and supplier level, internal formulation and implementation


Table 3.5

3rd party monitoring through market based certification bodies, setting more difficult. Legitimacy high

2nd party monitoring through associated sector association: setting quite easy, still weak legitimacy

1st party self monitoring; setting process easy, legitimacy weak



Transfair, FSC, Rugmark, etc.: NGO fostered sector specific codes and labels, formulated and implemented mainly through NGO and business partnership with independent monitoring procedures and civil society participation

SA 8000, FLA, ETI: Tripartite social minimum standards to harmonize the diverse numbers of codes and to increase legitimacy, transparency and traceability (existence of divergent approaches)

4th generation Business and NGO defined sector-specific codes and labels

5th generation Tripartite defined generic social standards

Social NGOs, unions, TNCs (buyers and producers), certification bodies, governments

Public Sector, NGOs

NGOs, religious NGOs associations, solidarity groups, minority groups, unions, large retailers

Increasing influence despite disagreements between special actors involved in the formulation of the standards

Gain increasing importance according to new strategies of NGOs and retailers

3rd party monitoring through certification bodies

3rd party monitoring through certification bodies or NGOs (setting difficult, keeping legitimacy requires constant negotiation)


Local enterprises in the global economy

absence of independent verification raises doubts of legitimacy. In some cases, weaknesses in self-monitored codes became apparent after NGOs and the media attacked particular firms or their subcontractors for code infringements (van Liemt, 1998a). The plethora of codes can be a disadvantage, causing greater uncertainty and confusion for both suppliers and consumers regarding the content of codes, their reliability and their legitimacy. These concerns have motivated the second generation of social and environmental standards, namely sector-specific codes and labels. Various industry-wide organizations have begun, or have been specifically set up, to promote voluntary codes of conducts. The leading examples include the International Chamber of Commerce (ICC), Responsible Care, the foreign trade association of the German retail sector, Aussenhandelsvereinigung des deutschen Einzelhandels (AVE), and the Eco-Tex label (Reichert, 2000; Robins and Roberts, 2000, Chahoud, 1998). The Paris-based ICC launched a Business Charter for Sustainable Development in 1991 to help businesses around the world improve their environmental management and performance. One of the best known sector-specific organizations is the chemical industry’s Responsible Care programme, launched after the Bhopal disaster in the mid-1980s by US and Canadian chemical producers to promote high standards of pollution prevention, product stewardship and community awareness. Existing enterprise associations in industrialized countries have started to develop harmonized codes as a preventive strategy to counter consumer and NGO pressure. For example, in 2000 AVE developed a unified code of conduct for its members and suppliers. Another example of an environmental label developed by industry is the Eco-tex label. Founded in 1991 by an association of 130 textile entrepreneurs, Eco-tex certifies sustainable products. It focuses on the production process as well as on the end product. Formulated by the sector organizations itself, it has an audit system that monitors levels of formaldehyde and pesticides in the production process (Chahoud, 1998). The EUREP-GAP standards in the food products sector, as discussed earlier, also fall within this framework. As with AVE and Eco-Tex, most sector-specific codes and labels are formulated by a network of relatively homogeneous actors. This makes it relatively easy to come to an agreement on the criteria for standard setting. Such sector-oriented codes harmonize different firm-specific approaches, increase legitimacy and reliability and reduce transaction costs. Business-defined sector-specific codes are often monitored through second party certification. In firm-specific codes, the lead firm determines the implementation principles to which suppliers must conform. Large vertically integrated buyers or business associations employ different forms of

Making sense of global standards


monitoring. In some cases, buyers or trade associations directly monitor the production processes of suppliers. Sometimes monitors are integrated into the supplier’s firm. In other cases, the buyer or association contracts a service agency or accredits their own certification bodies to audit compliance to the codes using guidelines developed by the buyer or trade association (FEER, 2000). Very rarely, there are codes of conduct involving third party certification. These sometimes involve the participation of NGOs or local civil society groups. High profile brand-name companies, such as Nike and Reebok, delegate monitoring to local NGOs or neutral service agencies to increase the reliability and transparency of the code and raise public confidence in the brand. The third generation of environmental standards are developed in large part by business but, unlike the first and second generation, they are generic as opposed to firm- or sector-specific. The most prominent example is the ISO 14000 environmental management standard. The number of new ISO 14000 certificates issued worldwide in 1999 was 14106, an increase of 78.9 per cent since 1998. Moreover, some countries are beginning to introduce the standard in national regulatory systems (Haufler, 2000). The scope of coverage of ISO 14000 is somewhat different from ISO 9000. Formulated in 1996, it is based on the earlier British standard BS 7750. Its emergence is seen as a response by industry to the growing environmental consciousness of the 1980s and 1990s, and demands by NGOs and multilateral bodies for environmentally sustainable practices in production. It does not apply to the whole supply chain. Instead, it is concerned with environmental management practices of the certified unit alone. The influence of ISO 14000 along the value chain is weak. Consequently, many producers favour ISO 14000 as a ‘defensive’ measure against pressures to meet more regulated environmental standards. Nevertheless, it certifies that the enterprise has adopted management practices that demonstrate environmental responsibility and reduce environmental costs within the firm. As with ISO 9000 and overall product quality, ISO 14000 compliance does not suggest that a product or service is environmentally sound. Certification is through internal monitoring and independent third party audits, using procedures and service providers similar to those adopted in ISO 9000. 4.2

NGO–Business defined Standards

From standards defined by business, we turn to standards and labels that emerge through networks that bring together business and NGO interests. These address environmental, social, and increasingly ethical concerns (Blowfield, 1999). These are the fourth generation of standards that are


Local enterprises in the global economy

sector-specific in coverage. They are often supported by governments and international institutions keen to encourage independent codes and labels that involve civil society actors. Prominent examples of such sector-specific codes and labels include: Transfair (or Fairtrade), Rugmark, Forestry Stewardship Council (FSC), Marine Stewardship Council (MSC) and the Clean Clothes Campaign (CCC). While their objectives often differ, they have in common the involvement of a wide range of social and economic actors, with NGOs as the main drivers. ●

The Transfair (or Fairtrade) label started through initiatives of solidarity groups in different industrial countries. Various labels emerged from different fair trade initiatives concerned with the dependence of developing country farmers on volatile international commodity trade and the lack of access to fair and reasonable loans. During the 1990s, products carrying the Fairtrade mark, like coffee or cocoa, became a common sight in European supermarkets. However, their market share is minimal and has declined of late (Robins and Roberts, 2000). The RUGMARK Foundation concentrates on child labour in the carpet industry. Supported by United Nations Children’s Fund (UNICEF) and ILO, it recruits carpet producers and importers to make or sell carpets that are free of child labour. Producers have the right to place the RUGMARK label on their carpets. At present RUGMARK’s efforts are mainly focused on India, Nepal and Pakistan. The label accounts for a significant share of the European and US market, with over 30 per cent of carpet sales in Germany (Reichert, 2000). FSC is a non-profit association founded in 1993. Set up in Canada it has an international agenda to promote sustainable forest management. The main driver has been environmental NGOs, particularly the Worldwide Fund for Nature (WWF) that campaigned against the international tropical timber trade (Kiekens, 2000). Since its foundation, FSC has gained increasing influence in the international timberproducts trade. It is especially important in the European market. The volume of FSC-certified timber has grown from 1 million hectares in 1995 to almost 17 million hectares at the end of 1999 (Kiekens, 1999).9 More recently, leading retailers, especially in the Western European home furnishing products (DIY) sector, are using the FSC label as an advertising tool to promote environmentally and socially conscious timber demand. In 1995, WWF set up buyers’ groups. By 1999, there were 15 buyers’ groups chiefly in Europe and North America. The participation of DIY retailers is important

Making sense of global standards


because they often entail a large share of the market. In the UK, for example, DIY stores account for almost 25 per cent of the timber trade (Kiekens, 1999). Although NGOs were the main driver of FSC, the standard itself was formulated through a network of NGOs and business. Formulation procedures within the advisory board involve social, economic and ecological interest groups with equal voting rights. This implies that, while it is an environmental forest management standard, it also incorporates social principles. National groups are set up within countries where forest-owners want to implement the FSC label. These national groups are organized along the similar tripartite membership and voting system seen at the global level. National groups adjust the ten global principles of the standard to local circumstances and formulate country-specific requirements. The Clean Clothes Campaign, like FSC, also pays special attention to the participation of local civil society. In contrast to FSC it focuses on social issues. It is organized through an international NGO network that tries to build linkages with retailers and companies. The goal is to improve working conditions in the global garment industry. The network comprises the widest variety of organizations in the standard debate including trade unions, consumer organizations, researchers and solidarity groups. Their guiding codes draw on ILO conventions and include those on child labour, minimum wage, the right to collective bargaining and freedom of association. Although not a label, CCC issues ‘stickers’ that retailers can use in marketing. In several countries, including Sweden, France, the UK and the Netherlands, projects have been set up involving companies and participating organizations to develop independent monitoring systems. The FSC and the Clean Clothes Campaign are of special interest for different reasons. First, they are both a mix of label and code. Second, they emphasize a role for local civil society in implementation. Third, FSC incorporates environmental and social concerns with professional and independent certification procedures. Fourth, FSC operates a comprehensive scheme of certification along the whole value chain. Local civil society is involved in the implementation and monitoring process, although accredited certification bodies from the FSC are responsible for third party certification. The FSC system differentiates between two certification and monitoring schemes: one for the forest; the other for the value-added companies engaged in the chain of custody of FSC-certified timber. Both require third party certification. However, the sustainable principles of the FSC focus only on forest management. They do not apply to the subsequent value-added processes or the manner in which


Local enterprises in the global economy

products are made from FSC-certified timber (Kiekens, 1999; 2000). Enterprises in the value chain only have to comply with documentation, transport and storage requirements to insure that FSC-labelled products are indeed from FSC-certified forests. With this system, every FSC-certified product can be traced back to the forest. This procedure demonstrates that the real costs of implementing the standard rests on the downstream supplier, the forest or plantation owner, while the retailer accrues the credibility benefits of FSC at little cost. Such initiatives involving the collaboration of NGOs and business in defining standards demonstrate more complex network arrangements than firm and association codes. They are a relatively new phenomenon and they have to be analysed with the evolving strategies of NGOs and business in the last decade. Large retailers and producers in certain sectors recognize that compliance with independently monitored standards, developed in partnership with NGOs, can enhance legitimacy and reduce vulnerability to consumer campaigns. Moreover, they are aware of the advantages such labels provide as part of marketing strategies, and in differentiating product niches. NGOs, for their part, have learnt that cooperation can be more effective for achieving environmental and social goals. This has enabled many NGOs to expand their activities and enter new arenas. Despite this change of attitude by different interest groups, the management of such networks remains a difficult task. There are often network failures as actors with different aims and powers try to set standards collectively. As Messner (1997) notes, there are four core problems in network governance. First, the greater the number of actors, the higher the risk of veto positions. Second, the search for a consensus between different interest groups often leads to agreement only on the smallest common denominator. Third, networks often prioritize short-term interests over long-term objectives. Fourth, networks tend to externalize costs at the expense of the network environment due to intended or unintended effects (Messner, 1997). To prevent such failures, coordination between public and private (both for-profit and not-for-profit) actors is important. Attempts at developing generic global standards often fail because of network problems. This is more likely in the environmental than in the social arena. Social standards usually have a framework of reference. In most cases they refer either to a single ILO convention, the eight ILO core standards, or the Declaration of Universal Human Rights. Such reference points do not, as yet, exist in the environmental field. Instead, there are a variety of environmental labels and codes, with sector- and process-specific environmental criteria. These include energy, ecological and biological

Making sense of global standards


standards, standards for recycling, forestry, cars, and so on. Forming a consensus on a generic agreement is difficult. It is compounded by the large number of actors involved in the negotiation process. Standards like FSC, therefore, demonstrate a new trend that combines social and environmental issues to shape a sustainable approach, albeit at the sector level. 4.3

Generic Public–Private Social Minimum Standards

During the last five years new forms of generic social standards, formulated through various public–private networks have emerged on the international arena. We view them as the fifth generation of social standards. Based on NGO–business partnerships, either with public support or directly initiated by government, they seek to harmonize the diverse firm- and sector-specific codes and develop a global social minimum standard. Examples include the social management standard Social Accountability 8000 (SA 8000), the Fair Labour Association (FLA) (both based in the US), and the Ethical Trading Initiative (ETI) in the UK. A common feature amongst these initiatives is their reference to the core standards of the ILO. ●

Since 1999, the New York based ‘Council on Economic Priorities Accreditation Agency’ (CEPAA – recently renamed as Social Accountability International, SAI) developed a global minimum standard with the objective of harmonizing the diverse social standards in international trade. SA 8000 tries to transfer the experiences of established quality assurance standards, like ISO 9000, to social management. Certification bodies, unions, companies and NGOs have participated in formulating the standard. The standard itself includes the ILO core labour standards and issues like a ‘living wage’, hours of work and freedom of association. Like the ISO approach, it is based on a model of factory certification by professional independent auditors. Officially, local NGOs can be accredited as SA 8000 auditors. However, in reality, local NGOs are yet to undertake such a task. Consequently, SA 8000 has been criticized for excluding local NGOs from an active role in code certification (Jeffcott and Yanz, 2000), and shifting certification costs to suppliers. Furthermore, professional auditing services are considered less experienced in the detection of workplace violations and less independent due to pre-existing contractual relationships with enterprise management (Diller, 1999). In response to the criticism that CEPAA has faced in recent years, it has increased social audit training for certification bodies and integrated local NGOs and unions more widely in gathering information and making appeals (SAI, 2000). Criticisms from


Local enterprises in the global economy

the business perspective question the strength and attractiveness of the SA standard. Nevertheless, international organizations and businesses participate in the SA 8000 Advisory Board.10 They support SA 8000 because its independent market-based certification procedure, and the engagement of unions and NGOs, add credibility. However, the use of SA 8000 has yet to become widespread, and many lead firms, especially in the US, focus on their own firm or sector-specific codes. Nevertheless, SA 8000 is currently at the forefront of the agenda for harmonization of the diverse codes of conduct and the development of an international social minimum standard. In the US market, the SA 8000 competes with the FLA. In 1998, the FLA grew out of the Apparel Industry Partnership (AIP) which was initiated by the US government in August 1996 to work towards eliminating sweatshops. The AIP brought together apparel and footwear companies, human rights groups, labour unions, religious organizations, consumer advocates and universities to work on an industrywide international code of conduct.11 The FLA’s workplace code includes the ILO core standard; but it does not include a living wage. FLA members are moving towards implementing a monitoring and certification system which, in contrast to the SA 8000 factory certification model, will certify Northern brands based on a sample monitoring of 30 per cent of the company’s suppliers. As with SA 8000, companies hire external monitors from a list of auditing firms including FLA-accredited NGOs. While the FLA also provides a mechanism for third parties to register complaints, it is not yet clear how much information on steps taken to eliminate abuses will be made publicly available. However, it seems that the FLA (in some pilot projects) is putting more emphasis on involving local NGOs in the monitoring process than other code initiatives (Jeffcott and Yanz, 2000). A leading publicly supported national initiative in the UK is the ETI (Barrientos, 2000). This is a cooperative programme of NGOs, unions, universities and TNCs aimed at improving the working conditions between TNCs and their suppliers. It is governed by a board of directors, which includes the Department for International Development from the British government. ETI, unlike SA 8000 and FLA, is not a factory or brand certification programme. Instead, the ETI members aim to ‘identify and promote good practice in the implementation of codes of labour practices, including the monitoring and independent verification of the observance of code provisions’ (Mabott, 2000, 16). According to the ETI, members follow the principle of ‘learning by doing’ in the sense that member companies commit themselves to

Making sense of global standards


bringing their own codes into conformance with the ETI Base Code, which is based on the ILO core standards and on the living wage issue. The learning-approach is evident in ETI’s focus on pilot projects, in which companies, NGOs and unions are experimenting with different models of code monitoring and verification. The companies have to map and assess labour practices in their supply chains and identify major problems encountered by their suppliers. They then develop an internal monitoring programme and plans for independent certification along the supply chain. The independent certification bodies can either be professional audit companies, as in SA 8000, or local NGOs and unions.12 Thus, ETI does not develop a single form of certification and monitoring procedure. It tries to implement third party certification through professional auditing firms, while certification rules are defined in a tripartite fashion with the involvement of local actors. The objective is to demonstrate a joint implementation structure (professional services in the market and local inspectors) that is recognized as operating independently of management control and enlisting local participation.13 At present, the ETI has not only the widest range of actors engaged in any fifth generation social standards, it is also the most far-reaching of such standards and the one that has attracted most attention from leading retailers (Reichert, 2000; van Liemt, 1998a; Barrientos, 2000).14 In all the three approaches above, local suppliers bear the costs of certification and implementation. While SA 8000, FLA and ETI try to give ILO conventions more policy impact on the firm and supply chain level, they differ widely in how they seek to reach this aim. SA 8000 aims to raise transparency and credibility for large companies in their search for suppliers, thereby reducing risks and transaction costs for the lead firm by pushing them down the supply chain. In contrast, FLA is mainly based on informing consumers about the social responsibility of Northern brand firms and their suppliers. ETI, with its Basic Code, has to be seen as an institution that wants to encourage the search for better implementation procedures with reference to local circumstances and the participation of local workers and NGOs. In terms of the negotiation costs and the danger of network failures, the ETI approach involves a continuing process of negotiation with the different local stakeholders in the standard setting procedure. The ETI was initiated by national government which pressured British TNCs and NGOs to develop code guidelines. Therefore, the government can be seen as the lead actor, although the process of standard setting is based on a participatory approach. In contrast, SA 8000 works along clearly defined rules and guidelines, although


Local enterprises in the global economy

supplier difficulties in standard compliance gets no attention. SA 8000 may gain importance as a global minimum standard because of its marketbased, and relatively easy, implementation process. But, given its voluntary character it is not clear whether it would be widely accepted in the private sector. Despite debates in global public forums, private, social and environmental standards lack a hierarchical institution that can enforce such standards. This raises doubts on their future influence.

5. COMPARING TYPOLOGIES AND TRAJECTORIES OF STANDARDS As seen in sections 3 and 4, standards have evolved along distinct paths. In this section we summarize this discussion by considering first the typology of standards, and then turn to a comparison of the distinct trajectories. We end by considering the nature of links between public and private agendas that influence the ways in which standards have evolved. 5.1

Typologies of Standards

Earlier we set out a typology for global standards. This sought to capture the key elements of each standard, including coverage, major drivers, certification process, regulatory and governance implications. This framework is used in Table 3.6 below to summarize the evidence from the previous sections. What is apparent is the degree of similarity across diverse standards. These include, for example, the ways in which standards, codes and labels sit side by side. Take the case of ISO 9000. While not a label, firms compliant with the standard often use it as a marketing tool. The ISO 9000 logo suggests, often inaccurately, to consumers and the wider public that the firm meets accepted international norms on quality assurance and, by association, quality management. Similarly, while the forestry stewardship council has a clearly defined standard, it is also a powerful, and in many markets a clearly recognizable, label suggesting sustainable forestry management. Consumers are rarely aware of the specific requirements and procedures. This points to a gap in public understanding of what standards and labels are, while recognizing that they can radically influence consumer behaviour. Another area of similarity brought out by the comparison of the range of standards is the growing importance attached to independent third party monitoring. This adds credibility to the standard, while lowering monitoring costs to firms. In many cases, standards are certified through specialist



Codes and labels


Eco-Tex, AVE

Company codes



Standard Sector-specific NGOs, unions, and codes and labels business

FairTrade, FSC, Rugmark


Business associations



State, business and NGOs

International business

Standard (and label)


ISO 14000


Firm QA codes

Sector-specific International organization and government




Sector-specific International business

International business

Key drivers

Social and Environmental Standards SA 8000, ETI, FLA Standard and code


QS 9000/AS 9000 EUREP-GAP

Quality Assurance and Food Safety Standards ISO9000 Standard Generic (and label)


Overview of key standards

Field of application

Table 3.6

Voluntary and sector requirement

Voluntary. Market requirement and legally mandatory in some markets

Regulatory implication





1st and 3rd party firm and NGOs

Mandatory for all suppliers

1st and 2nd party Voluntary business associations

3rd party NGOs

3rd party private auditors

3rd party private auditors and NGOs

1st and 3rd party

3rd party public and Increasingly legally public–private bodies mandatory

3rd party private auditors

3rd party private auditors

Auditing process


Local enterprises in the global economy

service providers. In some, this involves direct monitoring by civil society actors. 5.2

Trajectories of Public and Private Standards

sa ran ed t ge -b as wl ed kn o of se cre a In

Technological complexity in sectors and supply chains

ic cif s pe ent es m or ire f m equ eo er as nc cre lia In mp co




The explosive growth of global standards during the 1990s has been driven by globalization. There is a greater need for regulation of quality, safety, as well as social and environmental issues to prevent a ‘race to the bottom’ and to improve coordination of complex global value chains. This has involved global public effort. WTO rules have raised the importance of national standards in relation to safety and quality issues. The ILO’s Declaration on Core Labour Standards has led to a new dynamic within the organization and increased the number of ratifications to ILO conventions. More importantly, despite the often weak enforcement of the core labour standards, they have become a model for private social standards. Finally, different UN summits have sensitized governments on sustainable and ethical development. They have influenced, and been influenced by, the development of private standards. They have collectively brought the interdependent relations of trade rules, social order and sustainable development, at least rhetorically, onto the agenda. Despite these developments, publicly defined standards are limited. In contrast, private standards have gained influence in trade relations. Different trajectories are observable as Figure 3.1 shows. In this context, the ISO 9000 standard can be regarded as a minimum global requirement

Company based QM-Standards

Sector-specific QM and Food Safety Standards

ISO 9000 – Generic International QM-Standards National QM standards e.g. BS 5750

Figure 3.1 The dynamic of international quality management standards: from generic approaches to further diversification

Making sense of global standards


and a base line. It provides a set of codified rules that enjoy a high degree of legitimacy. This legitimacy is a function of independent certification and the active role of business in the formulation of the standards. Its effect is to raise transparency in the market, allowing lead firms to select suppliers with greater confidence based on their production capabilities and quality assurance procedures. Support for compliance to ISO 9000 is uneven, and concerns exist regarding how sanctions on non-compliance are enforced. It is also apparent that more sophisticated quality assurance procedures are required as sectors and value chains become more complex in terms of product technology, logistics or chain coordination. In such cases, the base ISO 9000 standard may prove insufficient to provide the required level of quality assurance to the lead firm. Thus, we observe the emergence of more specialized, sector-specific quality management standards, such as EUREP-GAP, AS 9000 and QS 9000. The emergence of company based quality assurance standards also reflects various pressures. For example, mergers of large firms with distinct supply chains and their own codes, rules and ‘languages’ of supply chain management call for greater harmonization of quality management systems. In addition, as one moves along the trajectory from simple production and supply chain systems towards technically complex production systems, demands for company-based quality assurance standards within the chain are likely to increase. Compliance with the generic ISO 9000 standards, or even sector-specific standards, is insufficient. Such complexity may be related to the technological frontiers of the particular value chain, or the importance attached to particular needs of lead firms. For example, the importance of quality assurance in food hygiene drives Nestlé’s own quality assurance code to which its suppliers must comply (Reardon et al., 2001). In such situations, it is possible that company-based quality management standards will be of greatest importance. Thus, we observe the ‘pyramid trend’ shown in Figure 3.1. Here the influence of quality management standards is increasing, while the requirements of these quality management standards is associated with the nature of technological complexity of the sector and the need for more specialized, knowledgebased codification. In contrast to the pattern seen above, our discussion of the various types of social and environmental standards indicates a very different trajectory of standard evolution (see Figures 3.2 and 3.3). The impact of social and environmental standards on value chains will differ according to sectors. Such standards are likely to gain importance in particular types of value chains that have a wide international division of labour based on labour cost and resource-based international competition. The diversity of the


Local enterprises in the global economy


ex pl


g sin

ne of ity ts ex n pl e m em co ng g rra sin a ea cr In

Firm-specific codes/labels Generic minimum standards defined in a Public–Private Partnership NGO + business defined sector-specific codes/labels Generic business defined standards

ea re se o qu f ire co m mp en li ts an ce


c De


? future development





Vulnerability of lead firms to consumer and NGO pressure

Firm specific codes/labels


Business sector specific codes/labels

Figure 3.2 The dynamic of international social standards: from diverse to generic approaches

NGO + business defined sector-specific codes/labels

rea s req e of uir com em p en lian ts ce

fn yo xit ts ple en om em g c ang sin arr



Generic business defined standards



Vulnerability of lead firms to consumer and NGO pressure

Firm-specific codes/labels


Business sector specific codes/labels

Figure 3.3 The dynamic of environmental standards and its growing diversity

Making sense of global standards


existing standards will still continue in the future, although the creation of global codes of conduct in the form of a social minimum standard will potentially reduce the heterogeneous approaches. With rising consumer consciousness, lead firms are being pressured to take on greater social and environmental responsibility for their value chains. To reduce their vulnerability to such pressures, such firms rely more heavily on standards, especially those with a high level of public legitimacy. Hence, key actors and procedures, like independent monitoring, that raise legitimacy have to be integrated into standard negotiations. This makes the networks more complex. The different ‘generations’ of social and environmental standards demonstrate a much more conflictive constellation than quality management standards. In order to gain legitimacy, these standards have to involve a larger number of actors in the network. At the same time social and environmental standards are much more difficult to codify than quality management standards. They require a more complex negotiation process, as well as the participation of different actors with different core competencies and legitimacy resources. The different standard approaches, according to their legitimacy requirements, depend on the integration of different interest groups and independent monitoring procedures. Thus we observe the ‘inverse-pyramid trend’ shown in Figures 3.2 and 3.3. Here, the influence of social and environmental standards is increasing in the context of growing consumer consciousness. At the same time, the demands that this places, and the need to give the standard legitimacy, requires the integration of a rising number of actors. Social standards tend to be generic approaches along the guideline of the ILO core labour standards. Once these global minimum standards are framed, further specialized standards, like firm- or sector-specific codes, could follow. Already some of the SA 8000 certified firms, like the German retailer Otto, see it as a minimum requirement surpassed by their own firm-specific social standards (Merck, 1998). Hence, with social standards one could see a future development along lines similar to quality assurance standards (the top pyramid of Figure 3.2), with moves towards sector- and firm-specific standards based on the generic codes within the ILO core labour standards. Figure 3.3 demonstrates that the trajectory in environmental standards differs. Here a tendency to generic standards is not observed. Although NGO–Business codes and labels are increasingly significant, they are also becoming more diverse; differentiated by sectors, areas and countries. This results in a complex array of environmental codes. Without a ‘least common denominator’, there is no trend towards defining minimum global environmental standards. Thus, it is unclear whether the trajectory to be observed with environmental standards will be similar to that for social


Local enterprises in the global economy

standards, and it is possible that the evolution of standards in this area will need to incorporate more complex networks of local and global actors.



Despite liberalization, the global economy continues to be governed by ‘rules’. But the rules are changing, and international standards point to one such set of changes. Concerns about quality assurance, health and safety, as well as ethical, social and environmental aspects of production are now central to the global agenda on trade. In some markets, compliance with particular standards constitutes entry criteria. In others, it is a basis for defining market niches and creating competitive advantages. As a result, firms in developing countries have come to realize that their capacity to compete internationally is often linked to their ability to comply with global standards. While these standards represent new challenges, there remain fears that standards are the new barriers to trade: fears that developing countries lack adequate technical infrastructure to engage in standard formulation, or promote compliance; fears that small firms, short of the technical and financial resources needed for compliance, are the most disadvantaged in meeting global standards. These fears reflect anxieties that standards, far from averting the ‘race to the bottom’, may effectively marginalize particular producers. These preoccupations underline the importance of making sense of standards, both for policy makers and academic researchers. The proliferation of standards, however, makes this an especially difficult task. This is a particular concern for firms forced to implement diverse standards, and for consumers confronted with a confusing array of labels and standards on which to make informed choices. Faced by this assortment, few studies have gone beyond individual standards. Our view is that there is a value added to be had from a typology and a comparative perspective. Thus, as the first step in reducing the complexity, we set out to make sense of the leading global standards. We have shown how such standards have evolved; identified the main drivers behind their development; and outlined their monitoring and certification procedures. We have paid particular attention to the trajectory of standards, and shown through a comparative typology how trajectories of some standards are diverse while others follow a similar path. This categorization of standards is thought to be useful for most of the development debates in which global standards play a critical role, be they concerned with issues of efficiency, equity or management of the new global economy. However, as pointed out in the introduction, the application of these categories to such debates was not the objective of this chapter.

Making sense of global standards


In summary, this chapter underlines the benefits that a comparative perspective provides to our understanding of global standards. The typology proposed in this review shows the similarities and differences across standards. Similarly, the discussion on the evolution of standards, according to their distinct generations, shows how trends differ between quality assurance standards on the one hand, and social and environmental standards on the other, as well as how social and environmental standards are being pulled along different paths. It remains unclear what shape these bodies of standards are going to take in the future. Clearly, there are benefits to be had from the harmonization of standards, and the concomitant reduction of the multitude of competing standards. In some areas this has happened. In others, the technical nature of the standard and the specific needs of each sector may require diverse approaches that limit the possibilities, and the desirability, of such harmonization.


1. 2.


4. 5. 6. 7. 8. 9.

Institute of Development Studies (IDS), University of Sussex, UK and Institute of Development and Peace (INEF), University of Duisburg, Germany. Financial support from the Volkswagen Stiftung, Germany and the Department for International Development, UK is gratefully acknowledged. The authors thank Stephanie Barrientos, Afonso Fleury, Gerhard Halder, Claudio Maggi, Andreas Mertens, Jorg Meyer-Stamer, Judith Tendler, Bernd Wahlbrinck and especially Hubert Schmitz, Dirk Messner and John Humphrey for comments and suggestions on previous drafts. An earlier version of this paper, with an extended bibliography, was published as INEF Report 58/2002, INEF, Duisburg. The usual disclaimers apply. Health and safety were of particular concern (for example in shaping product safety standards for children’s toys, minimum pesticide residues in food products, fire safety codes for various household items). One ‘market failure’, for example, is the distinct national standards for colour television technology in the US (NTSC) and Europe (PAL and SECAM). In contrast, common technical standards were critical to the development of the global information and communications technology industry (David, 1995; Steinmuller, 1995; Tassey, 1995). A 1982 white paper on ‘Standards, Quality and International Competitiveness’, issued by the UK Department of Trade and Industry (DTI), saw BS 5750 as a basis for raising quality, improving economic performance and enhancing the reputation of British industry (see Seddon, 2000). The constant improvement of ISO standards (for example, ISO 9001:2000) and rising competition in the world market are likely to deepen the influence of ISO 9000 along the value chain and challenge subcontractors to give quality assurance more attention. In 1997, for example, 4233 certificates were withdrawn for not meeting the targets set in the first certification process (ISO, 1998). See Internet: http://www.us.tuv.com. It is recognized that SPS standards can constitute unjustified barriers to trade (see UNIDO, 1999). Since the implementation of SPS there has been a number of complaints raised with the WTO Dispute Settlement Body. See van Liemt (1998a, 1998b), Murray (1997), Caldwell (1998), Robins and Roberts (2000). Although this increase is impressive, it constitutes only 0.5 per cent of the world’s forests.

92 10.


12. 13.


Local enterprises in the global economy These are representatives of different Workers and Social Federations, Amnesty International, and firms like Avon, Toys ‘R’ Us, Body Shop, Sainsbury (UK), OttoVersand (Germany), Grupo M (Dominican Republic), Eileen Fisher (US), and others (see website) . Members of the FLA include Nike, Reebok, Levi Strauss, Liz Claiborne and PhillipsVan Heusen, as well as 131 universities whose licensed apparel is produced by US manufacturers. NGO members include the International Labour Rights Fund and others (see FLA homepage). By mid-2000 ETI had initiated four pilot monitoring projects in four different sectors and countries: clothes in China; horticulture in Zimbabwe; wine in South Africa; and bananas in Costa Rica. Evaluation along the supply chain and compliance to the Code is still weak. Out of 4556 suppliers of the ETI member companies in the year 2000, only 32 per cent were evaluated and only 20 per cent of the total complied to the company code or the ETI base code. (Mabott, 2000, 18) Members of the ETI include: Sainsbury, Levi Strauss, Littlewoods, Marks & Spencer, Safeway, The Body Shop, as well as a number of British NGOs and three national and international unions (Mabott, 2000).

REFERENCES Barrientos, S. (2000), ‘Globalisation and ethical trade: assessing the implications for development’, Journal of International Development, 12, 559–70. Blowfield, M. (1999), ‘Ethical trade: a review of developments and issues’, Third World Quarterly, 20 (4), 753–70. Caldwell, D.J. (1998), ‘Ecolabeling and regulatory framework – a survey of domestic and international fora’, in: http://www.consumerscouncil.org. Chahoud, T. (1998), Handel und Umwelt: Förderung umweltfreundlicher Prozeß- und Produktionsverfahren in Entwicklungsländern, Berlin: DIE. Clapp, J. (1998), ‘The privatisation of global environmental governance: ISO 14000 and the developing world’, Global Governance, 4 (3), 295–316. David, P. (1995), ‘Standardisation policies for network technologies: the flux between freedom and order revisited’, in R. Hawkins, R. Mansell and J. Skea (eds), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Aldershot, UK and Brookfield, US: Edward Elgar. Diller, J. (1999), ‘A global conscience in the global marketplace, labour dimensions of codes of conduct, social labelling and investor initiatives’, International Labour Review, 138 (2), 99–129. Dolan, C. and J. Humphrey (2000), ‘Governance and trade in fresh vegetables: the impact of UK supermarkets on the African horticulture industry’, Journal of Development Studies, 37 (2), 147–76. EUREP (2001), Eurep-Gap Protocol 2000, http://www.eurep.org. FAO (1998), ‘Food quality and standards’, Service Food and Nutrition Division, Food and Agriculture Organization of the United Nations, Rome. FEER (Far Eastern Economic Review) (2000), ‘The inspector calls, working conditions in Asian factories: Mattel has a proactive approach’, Far Eastern Economic Review, July, and Internet: http://www.feer.com. FPJ (2001), The Fresh Produce Journal, 5 October.

Making sense of global standards


Fues, T. (2000), ‘Auf dem Weg zur Weltsozialordnung? Beiträge zur Debatte über globale Armutsstrategien’, INEF-Report, 44, Duisburg, www.inef.de. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48 (1), June, 37–70. Haufler, V. (2000), ‘Negotiating international standards for environmental management systems: the ISO 14000 standards’, Case study for the UN Vision Project on Global Public Policy networks, in: http://www.globalpublicpolicy.net. Hawkins, R. (1995), ‘Introduction: addressing the problematique of standards and standardisation’, in R. Hawkins, R. Mansell and J. Skea (eds), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Aldershot, UK and Brookfield, US: Edward Elgar. Humphrey, J. and H. Schmitz (2001), ‘Governance in global value chains’, IDS Bulletin, 32 (3), Institute of Development Studies, Brighton, pp. 19–29. ISO (1998), ‘Standards and world trade’, International Standards Organization, Geneva in: www.iso.ch/wtotbt/. ISO (2000), ‘The ISO survey of ISO 9000 and ISO 14000 certificates, eight cycle – 1999’, International Standards Organization, Geneva, in: http://www.iso.ch /9000e/9k14ke.htm. ISO (2001), ISO 9000, International Organization for Standardization, Geneva, in: http://www.iso.ch/iso/en/iso9000-14000/iso9000/iso9000index.html. Jeffcott, B. and L. Yanz (2000), ‘Visions of ethical sourcing’, Financial Times, May, and Internet: http://www.maquilasolidarity.org/resources/codes/fintimes 00.htm Kaplinsky, R. (2000), ‘Spreading the gains from globalisation: what can be learned from value chain analysis?’, IDS Working Paper 110, Institute of Development Studies, Brighton, May. Kiekens, J.-P. (1999), ‘Forest certification, part I: origins, background and recent trends’, Engineered Wood Journal, Autumn. Kiekens, J.-P. (2000), ‘Forest certification, part II: impacts on forestry, trade and consumer information’, in: http://www.apawood.org/news/journal/spring 2000. html. Mabott, F. (2000), ‘Global social standards, social protectionism and best practices principles: the ethical trading initiative’, paper presented at CDG Conference Global Social Standards in Practices: Monitoring of Codes of Conduct, Hannover, 13–15 October. Merck, J. (1998), ‘Sozialverantwortung im Handel: Der SA 8000 als Element des Otto Versand’, in Forum Wirtschaftsethik, No.4, in http://www.kirchen.de/akademie/rs/referate/wsethik Messner, D. (1997), The Network Society, Economic Development and International Competitiveness as Problems of Social Governance, London: Frank Cass. Messner, D. and J. Meyer-Stamer (2000), ‘Governance theories and network concepts. Tools to study the dynamics of clusters and global value chains’, draft for a workshop of the IDS/INEF project ‘The interaction of global and local Governance’, January. Murray, J. (1997), ‘Corporate codes of conduct and fair labour standards’, paper presented at International Workshop on ‘Global Production Systems and Labour Markets, International Institute for Labour Studies’, Geneva, 22–23 May. Nadvi, K. (1999), ‘Collective efficiency and collective failure: the response of the Sialkot surgical instrument cluster to global quality pressures’, World Development, 27 (9), September, 1605–26.


Local enterprises in the global economy

Quadros, R. (2002), ‘Global quality standards, chain governance and the technological upgrading of Brazilian auto-component producers’, IDS Working Paper 156, Institute of Development Studies, Brighton, May. Reardon, T., J-M. Codron, L. Busch, J. Bingen and C. Harris (2001), ‘Global change in agrifood grades and standards: agribusiness strategic responses in developing countries’, International Food and Agribusiness Management Review, 2 (3), 421–35. Reichert, T. (2000), ‘Vom Beschluß zur Umsetzung – Sozialstandards in der Weltwirtschaft’, mimeo, Eschborn: GTZ. Robins, N. and S. Roberts (2000), ‘The workbook, part 1–3: a new deal for trade, environment and development?’, The Business of Trade, Partnerships for the Future, in: http://www.iied.org. SAI (2000), ‘Measuring the social accountability of business in Central and Eastern Europe: A Consultative Workshop on SA 8000’, 4–6 March, Budapest, Hungary, in www.cepaa.org. Seddon, J. (2000), The Case Against ISO 9000, Dublin: Oak Tree Press. Steinmuller, W.E. (1995), ‘The political economy of data communication standards’, in R. Hawkins, R. Mansell and J. Skea (eds), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Aldershot, UK and Brookfield, US: Edward Elgar. Sturgeon, T.J. (2000), ‘How do we define value chains and production networks?’, background paper for Bellagio Value Chains Workshop, Bellagio, 25 September – 1 October. Tassey, G. (1995), ‘The roles of standards as technology infrastructure’, in R. Hawkins, R. Mansell and J. Skea (eds), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Aldershot, UK and Brookfield, US: Edward Elgar. UNIDO (1996), ‘Trade implications of international standards for quality and environmental management systems, survey results’, UNIDO, Expert Group Meeting, Vienna, October, unedited version, ISED.9 (SPEC.). UNIDO (1999), ‘Trade implications of standards and conformity assessment procedures’, July, Vienna: UNIDO. van Liemt, G. (1998a), ‘Codes of conduct and international subcontracting: a “private” road towards ensuring minimum labour standards in export industries’, draft version, ILO, Geneva, http://www.ilo.org. van Liemt, G. (1998b), ‘The social policy implications of codes of conduct, with particular reference to the relations between companies adopting such codes and their suppliers and subcontractors’, paper presented at international workshop on ‘Global Production and Local Jobs: New Perspectives on Enterprise Networks, Employment and Local Development Policy’, International Institute of Labour Studies, Geneva, 9–10 March. Wilson, S.R. (1999), ‘The impact of standards on industrial development and trade’, Quality Progress, July, pp. 71–5.


Governance in global value chains John Humphrey and Hubert Schmitz*



Studies of industrial clusters have highlighted the way in which competitive advantage arises from the local organization of firms. Their collective efficiency results from more than agglomeration; the quality of local relationships is also important. Both relationships amongst firms and between firms and support organizations have been a focus of attention. Relationships with agents outside the cluster are given much less attention, or described as predominantly arm’s-length and market-based. Buyers purchase products largely designed and made within the cluster. These buyers appear to have little influence on what products are made, or how they are made. This view of the relationship with the outside world can be questioned even in the context of Italian clusters (see Chapter 6 by Rabellotti, this volume). It would, however, be particularly difficult to sustain this view for developing country clusters which are the subject of several chapters in this book. Analysis of trade in labour-intensive products has highlighted the fact that trade in these products is increasingly organized by global buyers, working or acting on behalf of major retailers or brand-name companies. This has been shown to be the case in, for example, the trade of garments between East Asian countries and the US (Gereffi, 1999), the trade in horticultural products between Africa and the UK (Dolan and Humphrey, 2000) and the trade in footwear from China and Brazil to the US and Europe (Schmitz and Knorringa, 2000). Furthermore, trade in these and other products is also subject to an increasing range of standards, codes of conduct and labelling schemes (see Chapter 3 by Nadvi and Wältring, this volume), which specify the product characteristics and the way in which products are made. Such requirements have extended beyond issues of technical and safety standards to cover issues such as labour conditions and environmental impact. In other words, governance occurs not only within clusters, but also in international trade. Much of this trade is coordinated by the lead firms of global value chains and through global standards. The way in which this 95


Local enterprises in the global economy

global governance is exercised has a significant impact on the organization of clusters and their upgrading opportunities. These issues are discussed in many of the chapters in this book. This chapter is particularly concerned with the global governance structures themselves, that is with global value chains, global standards and their interaction. The objective is to understand this public–private institutional context for international trade. To this end the chapter deals with four questions: What is chain governance? Why does it arise? How do chain governance and standards interact? What are the likely trends in the global governance of value chains?



A significant amount of trade in the global economy is carried out in the form of transactions between subsidiaries of transnational companies. This has long been recognized. Of more recent origin is the attention given to trade that is organized through networks of legally independent firms using a variety of transactional relationships. Global value chain research in particular seeks to understand the nature of these relationships and their implications for development. The concept of ‘governance’ is central to the global value chain approach. We use the term to express that some firms in the chain set and/or enforce the parameters under which others in the chain operate. A chain without governance would just be a string of market relations. Governance can be exercised in different ways, and different parts of the same chain can be governed in different ways. Other chapters in this book show why these differences matter for the upgrading prospects of local producers (for a synthesis see Chapter 13). This chapter seeks to deepen our understanding of chain governance. It is easy to point to instances of governance in global value chains. One clear example would be the way in which UK supermarkets exercise control over their fresh vegetable supply chains.1 Not only do they work directly with importers and developing countries’ firms to specify the type of products to be supplied (including varieties, processing and packaging), but they also specify quality, environmental and labour requirements. These requirements are enforced through a system of auditing and inspection and, ultimately, through the decision to keep or discard a supplier. Clearly, governance in value chains has something to do with the exercise of coordination and control along the chain. At any point in the chain, the activities performed are defined by a set of parameters. The three key parameters are:

Governance in global value chains

1. 2. 3.


What is to be produced: product design and specifications. How it is to be produced. This involves the definition of production processes, which can include elements such as the technology to be used, quality systems, labour standards and environmental standards. How much is to be produced, and when: production scheduling and logistics.

The question of governance arises when some firms in the chain work to parameters set by others. When this happens, governance structures are required to transmit information about parameters and enforce compliance. In short, governance refers to the inter-firm relationships and institutional mechanisms through which non-market, or ‘explicit’, coordination of activities in the chain is achieved.2 Governance, in the sense of arrangements that make possible the nonmarket coordination of activities, does not always occur in value chains. Many goods are traded in markets through a series of arm’s-length market relationships between firms. The parameters are defined solely by each firm at its point in the chain. For example, a firm might make a product according to its own estimations of market demand (‘make to forecast’), using a design that has no reference to any particular customer (that is either a completely standard product, or a product developed in-house) and using production processes it has selected or developed. A ready-made and readyto-buy product is then placed on the market. In such market situations, intermediaries frequently facilitate flows of information about supplier capabilities and innovations and about buyer requirements (Spulber, 1996), but neither suppliers nor buyers adjust their activities to meet the demands of specific customers. These types of relationships arise not only in markets with many buyers and sellers, but also in markets where large suppliers provide standard products used by many smaller buyers. There are various ways in which inter-firm relationships can differ from this pattern. For example, when production is scheduled according to ‘make-to-order’ rather than ‘make-to-forecast’, the decisions about ‘when’ and ‘how much’ will be made jointly by producer and buyer. This is typical when products have many possible variants, which renders make-to-forecast uneconomic. Even here, however, the supplier is responsible for the overall design and the range of customized options available to the buyer. Value chain governance arises when one firm coordinates the product, process and scheduling/logistics parameters followed by another firm. This often happens when buyers coordinate the value chain.3 The parameters can be specified at varying levels of detail. In the case of product definition, the buyer might set a design challenge for the producer, which the producer then resolves by providing its own technology and design. Alternatively, the


Local enterprises in the global economy

buyer might provide a design or even detailed drawings for the producer to follow. Similarly, process parameters can be specified at different levels of detail. In some cases, the buyer may merely refer to the process standards to be attained. In other cases, the buyer will specify precisely how particular standards should be attained by requiring and perhaps helping to introduce particular production processes and monitoring procedures. When the buyer sets the parameters followed by firms in the chain, we refer to it as the ‘lead firm’ in the chain. The fact that this lead role can be played by a variety of firms led Gereffi to distinguish between producerdriven and buyer-driven global value chains (Gereffi, 1994). In producerdriven chains, the key parameters are set by firms which control key product and process technologies, for example in the car industry. In buyerdriven chains, the key parameters are set by retailers and brand-name firms that focus on design and marketing, not necessarily possessing any production facilities. Product and process parameters can also be set by agents external to chains, as has been argued by Kaplinsky (2000, 125). Government agencies and international organizations regulate product design and manufacture, not only with a view to consumer safety, but also in order to create transparent markets (for example, by defining standard weights and sizes or technical norms). Examples of such parameter setting by agents external to the chain include food safety standards, norms with regard to the safety of products such as children’s toys, electrical equipment and motor vehicles and control of hazardous substances in a wide range of products. Once again, these norms can refer to the product (are its physical characteristics and design in conformance with requirements?) or to the process (is it being produced in ways which conform to labour or environmental standards?). External agencies may also define process parameters. In some cases, the definition of process standards is used as a means of achieving product standards (for example, hygienic food preparation systems are designed to produce safe food) and in others because of the intrinsic value of particular types of processes (for example, animal welfare requirements). Governments may set standards which are compulsory and have legal force. Standards may also be set by non-legal agreements (for example, code of conduct) and by a variety of unofficial agencies, such as NGOs, which pressure for compliance with labour and environmental standards.4 The ways in which these externally-defined standards interact with chain governance is discussed in section 4.

Governance in global value chains




The setting and/or enforcement of parameters along the chain is costly for the buyer, since it requires asset-specific investments in relationships with particular suppliers. Such investment also increases the rigidity of supply chains by raising the costs of switching suppliers. Why are companies prepared to incur such costs? This section provides an answer in two steps. First it addresses the question of what drives parameter setting and enforcement in general. Second, it asks why such practices are more likely to occur when buyers are sourcing from developing country producers. The role of lead firms in setting product parameters is particularly influenced by product differentiation strategies. Standard products can be bought through arm’s-length market relationships, because designs and product specifications are well established. However, buyers may focus on differentiating their products from those of their competitors, or they may believe they have a better understanding of market requirements than suppliers. In these situations, the buyers’ own designers are more likely to provide the overall design brief, if not the detailed design. If a particular product is part of a broader product differentiation strategy (such as the overall statement of a fashion house for a particular season), which involves matching a variety of accessories, then the design of any one product becomes dependent on the broader product strategy. Therefore, the buyer has to coordinate the activities of the different product suppliers. The extent of the chain governance that arises from buyer design specifications depends upon the characteristics of the product. First, specification of design parameters along the chain is more likely to occur with integral product architectures than with modular product architectures. As summed up by Novak (2000), an integral product architecture is one in which there are many, tightly-connected interfaces between parts. Changing one part of the product has many implications for other parts of the product. A modular product architecture is one which has few, welldefined interfaces between parts. Changes in the design of one part do not necessarily lead to changes in design of other parts of the product. With integral product architectures, the buyer has to coordinate the design decisions of various suppliers, knowing that a change in one part might have implications for others.5 Second, the need for value chain governance increases when design changes have implications for activities earlier in the value chain. Some types of product differentiation may be achieved through combining and transforming standard products purchased from suppliers. Therefore, a design change has no implications for suppliers. However, in other sectors, a decision about product design may mean that changes have to be made much earlier in the production process. For


Local enterprises in the global economy

example, introducing new footwear designs might have implications not only for footwear producers, but also for component suppliers and tanning factories. In this case, coordination along the value chain would be required. Specification and enforcement of process parameters arise for two reasons. First, process parameters can be a source of product differentiation. Products may be differentiated according to a range of factors, including claims about how they are made, their constituent components, and so on. In the food industry, in particular, this leads to the emergence of ‘credence goods’, which in the context of agricultural produce have been described as follows: A credence good is a complex, new product with quality and/or safety aspects that cannot be known to consumers through sensory inspection or observationin-consumption . . . The quality and safety characteristics that constitute credence attributes include the following: (1) food safety; (2) healthier, more nutritional foods (low-fat, low-salt, and so on); (3) authenticity; (4) production processes that promote a safe environment and sustainable agriculture; (5) ‘fair trade’ attributes (for example working conditions) (Reardon et al., 2001: 424–5).

A claim about such a good requires that it be made in a way that ensures the credence characteristics are actually present. Control of production processes becomes particularly important because some credence characteristics cannot be verified by inspection of the product. Therefore, the buyer has to be provided with some assurances about the process in order to establish the claim. A second reason for monitoring processes is concern over quality. One of the features of the management of inter-firm relationships in the past 20 years has been an increasing concern with product quality and the use of process controls to provide assurances about quality. The control of quality at source through developing quality procedures has been developed within enterprises in the form of total quality management and this has also been applied along the value chain through supply chain management. Suppliers may be required to institute particular quality management procedures as the price of entry into their customers’ supply chains. In other words, buyers specify process parameters relating to quality. This process has been well-documented in the case of the automotive industry (see, for example, Sako, 1992; Lamming, 1993; and Quadros, Chapter 10, this volume) and for the surgical instrument industry (Nadvi, Chapter 11, this volume). Value chain governance may arise from two distinct tendencies. First, the trend towards increasing outsourcing of activities previously carried out inhouse means that managerial coordination within the firm is replaced by value chain governance. Second, product differentiation and concern for

Governance in global value chains


process standards coordination leads to increasing coordination of activities previously carried out at arm’s-length. But why would such governance be a particularly marked characteristic of the sourcing by global buyers from developing country producers? Parameter setting and enforcement seems particularly important in global value chains that bring together global buyers and developing country producers. Developing country firms frequently have the characteristic of what Hobday has termed the ‘latecomer’ firm. A latecomer firm is ‘behind technologically [and] . . . dislocated from the mainstream international markets it wishes to supply’ (1995, 34). A similar point has been made by Keesing and Lall (1992). They argue that producers in developing countries are expected to meet requirements that frequently do not (yet) apply to their domestic markets. This creates a gap between the knowledge and capabilities required for the domestic market and those required for the export market. Therefore, buyers may have to supply both product and process specifications to their suppliers and to introduce systems which ensure that products and processes meet the required standards. If the gap in capabilities has to be closed quickly, buyers may need to invest in a few selected suppliers and help them to upgrade. Shortfalls in suppliers’ knowledge of market demands may be particularly likely to arise in fast-moving markets characterized by innovation and product differentiation. They are also likely to arise when developing country suppliers are integrated into global value chains and exposed to the demands of more sophisticated markets. Shortfalls in capabilities are likely to arise as firms engage in non-price competition. Some of this non-price competition may place direct challenges to supply capabilities through increasing quality, more rapid supply response, and so on. In addition, to the extent that suppliers make non-standard products, the buyer becomes more reliant on the performance of particular suppliers, whose failures would impose significant cost. Increasing concerns about product safety, labour standards and environmental standards expose buyers (both retailers and manufacturers) in developed countries to the risks of loss of reputation if shortcomings are found at their suppliers. This further increases the need for value chain governance It follows from this that the need for parameter setting along the chain may decrease as the capabilities of developing country suppliers improve and diffuse. At the initial stages of a supply relationship, buyers may feel the need to provide detailed instructions and undertake close monitoring of supplier performance. As the suppliers become more experienced, and as they are able to demonstrate their reliability to the customer, the latter may indicate the standards to be met, but leave it to the supplier to work out how to meet them. An important corollary of this point is that the


Local enterprises in the global economy

extent to which product and process parameters are set by the buyer does not depend upon the intrinsic characteristics of the product, such as its complexity or its closeness to the technology frontier, but rather derives from the capabilities of the suppliers and the risks faced by the buyer.

4. THE INTERACTION OF STANDARDS AND CHAIN GOVERNANCE Value chain governance is also influenced by agents external to the chain that regulate product design and manufacture, not only with a view to consumer safety, but also in order to create transparent markets (for example, by defining standard weights and sizes or technical norms). The interaction between standards and chain governance can be seen particularly clearly in the case of process standards. Parameters set from outside the chain can have two quite opposite impacts on the need for chain governance. On the one hand, where clearly defined standards and systems for enforcement (for example, certification schemes) are developed, the need for intervention by lead firms is reduced. Instead of monitoring performance along the value chain, buyers can rely upon external monitoring and verification systems to guarantee product and process standards. Having the right certification becomes a necessary condition for entering the value chain, and the supplier has to take the initiative (and bear the cost). On the other hand, agents external to the value chain may impose new requirements on lead firms. The UK Food Safety Act, for example, places upon food retailers a requirement for ‘due diligence’ with respect to the manufacture, transport, storage and preparation of food. This increases their responsibilities and the need for value chain governance. Similarly, to the extent that firms in the public eye are subject to pressures on labour and environmental issues that are not necessarily translated into certification and labelling schemes with external monitoring and verification, the firm has to take direct responsibility for what happens in the value chain. Table 4.1 is a simplified table which sets out and exemplifies the options. Ignoring more complex situations, such as joint parameter setting between firms and external agencies and the role of collective institutions (sectoral or regional) in setting and monitoring standards, the table highlights the contrast between parameter setting and enforcement by firms in the chain (or by agents specifically contracted for the purpose by these firms) and the role of external agents in setting standards and enforcing compliance. There are some reasons to expect that parameters developed or adopted by lead firms within the chain will be enforced by the lead firms, or by


Parameter adoption External agents

3. Firms are expected not to use suppliers that employ child labour, but this expectation is not accompanied by any system for enforcing the ban. The firms have to develop their own enforcement systems.

1. Specification of quality systems and enforcement through audit, either directly by the lead firm itself or through an agent acting directly on its instructions. For example, requirement for labour standards above the legally-required minimum, verified by the lead firm or its agents.

Lead firm

External agents

4. The EU requires that surgical instrument manufacturers exporting to the European market must be ISO 9000 certified. The certification is carried out by independent certification agencies (Nadvi, 2001).

2. Lead firm requires suppliers to conform to a process standard or code of practice for which an independent monitoring or certification system exists. Examples would include ISO 9000, ISO 14000 and SA 8000 certification.6

Parameter monitoring and enforcement

Combinations of parameter setting and enforcement

Parameter adoption Lead firm

Table 4.1


Local enterprises in the global economy

agents contracted by them. Conversely, parameters set by agents external to the chain will also be enforced by agents external to the chain. These are the two situations described in boxes 1 and 4 in the table. In the case of box 1, the greater the extent to which the lead firm specifies non-standard parameters, the greater is the likelihood that it will also have to arrange for enforcement, carrying out this activity directly, or contracting others to do it. These ‘others’ might be other agents within the chain (for example, UK supermarkets requiring their importers to monitor the quality systems of horticultural producers and exporters), or a third party specifically hired for the task; this occurs when NGOs or independent monitors are hired by companies to verify labour standards at the suppliers. The key point here is not whether the firm or an agent does this work, but that the firm defines the parameters to be met and arranges for compliance to be monitored. In contrast, box 4 describes cases where the parameters are specified by agents external to the chain (in the case described, by a government agency) and the monitoring processes are also in the hands of agents external to the chain. In this case, no individual firm in the chain takes responsibility for defining or enforcing the parameters. They apply to all the firms in the chain. However, the table also shows that parameter setting and enforcement may be split. Box 2 outlines the case where lead firms require the suppliers to adhere to certain general process standards. The decision to insist on a standard is made by the lead firm, but if the standard is widely known and adopted, then it is likely that organizations (standards agencies, consultancy firms, and so on) exist for both certifying companies and helping firms meet the specified standard. Box 3 describes cases where the parameters are imposed by external agents (by governments or by NGOs), but the lead firm is responsible for specifying and monitoring the processes which are meant to lead to the required outcome. In these cases, the lead firm has a particular requirement imposed on it, but it has to make the necessary arrangements to ensure compliance along the chain. We can hypothesize that there is some incentive for firms to shift parameter setting and enforcement from boxes 1 and 3 to boxes 2 and 4. Such a shift would reduce the cost of direct monitoring and lead to external certification. Generally speaking, the costs of this certification are borne by the supplier, not the buyer. However, for this process to take place it is necessary for the parameters being specified to be widely applicable across different firms and to have credible means of external monitoring and enforcement. It may be the case that in the early stages of the development of new process parameters, such as labour standards, these are initially enforced by lead firms within the chain. As standards become more generalized, then external systems of enforcement develop. These might be

Governance in global value chains


generic, such as the SA 8000 labour standard, but they could also be sectoral or regional. Sectoral standards may be created by firms that collaborate in order to reduce the costs that they incur for value chain governance.7 Regional standards (for example, the standard developed by the Kenyan Flower Council) may arise as a form of product differentiation and branding. To the extent that such external systems of parameter setting and enforcement develop and gain credibility, then the role of process parameters in generating the need for governance by firms within the chain will decline. If it were the case that certification systems demonstrating adherence to a range of process standards, including quality, environmental and labour standards were developed, this might substitute for process controls by lead firms. Direct monitoring and control of suppliers could be substituted by certification processes. Nevertheless, there are reasons to believe that direct parameter setting and enforcement by lead firms will continue to be important in value chains. First, firms might still wish to specify product parameters. Second, it is not clear how effective standards and certification are. Widely applicable process parameters may not be a guarantee of good performance in areas such as quality. Close links with suppliers may remain indispensable. Third, there may be other areas of supplier behaviour, such as reliability of delivery and willingness to develop longterm partnerships that are not captured by certification schemes.

5. WHAT ARE THE LIKELY TRENDS IN CHAIN GOVERNANCE? This chapter has shown why and how global value chains are governed. Other chapters in this book examine the implications for cluster development and local upgrading strategies. We conclude this chapter with reflections on two questions: Will chain governance become more or less dominant in international trade? What are the implications of the analysis presented in this chapter for trends in value chain governance? ●

The general increase in chain governance is connected to the big changes in retailing in the advanced countries. There has been an enormous concentration in retailing, particularly pronounced in the US and the UK, but also evident in Germany, France, and more recently in countries with traditionally very diffuse retail sectors such as Italy and Japan. Concentration in retailing does not necessarily lead to concentration in sourcing but the scenario which is emerging is increasingly clear: an increasing number of developing country


Local enterprises in the global economy

producers engage in contract manufacturing for a declining number of global buyers. One reason for this is that there are economies of scale in value chain governance. The cost of developing procedures and systems does not rise with the number of suppliers being supervised. Therefore, large firms have advantages when developing global production networks. Brands play an increasingly important role in enterprise strategy, particularly in consumer products such as garments and footwear. The enormous investment required to create (or maintain) brands is increasingly made by retailers or other companies which have no (or only limited) production facilities of their own. Product and process definition, however, is a strategic part of their operation. To the extent that luxury segments of markets for products such as clothes and shoes become increasingly dominated by global brands, the companies holding these brands will play an increasing role in structuring global value chains. This tendency is already evident in parts of the Italian footwear industry (see Chapter 6 by Rabellotti, this volume). Because brands stand for high quality or well-defined images, they need to define and enforce product and process parameters. Branding and chain governance thus tend to go together. Chain governance is not however limited to the sourcing of branded products. We have argued that the risk of supplier failure is a key driver of chain governance. Will this risk diminish with time? The risk of suppliers not being able to produce to the required specification is highest in new producer countries. Over the last two decades, many new producer countries have been able to export to advanced country markets under the tutelage of the global buyers. As the competence of these suppliers increases, chain governance through the buyers can be expected to loosen, provided that the increasing competence of suppliers is accompanied by the emergence of local agents who can monitor and enforce the compliance with general or buyer-specific standards. Some of the formerly new producers will become world leaders in producing promptly to the specification of the foreign buyer. To some extent this is already happening, for example in the Taiwanese computer cluster (Chapter 9 by Kishimoto) and the South Brazilian footwear cluster (Chapter 5 by Bazan and Navas-Aleman), both of which are loosening ties with the foreign buyers. There is however a counter-tendency. While non-price factors (quality, brand, speed) have come to play an increasing role for competing in global markets, price competition continues to be unrelenting, leading to a downward pressure on prices, particularly in labour-intensive

Governance in global value chains


products sourced from developing countries. The resulting profit squeeze leads buyers to scout continuously for new producers who offer lower labour costs. This then raises again the risk of supplier failure and the need for chain governance. While this process has probably bottomed out in traditional products such as garments and shoes, the cycle continues to be reproduced for newer products such as computer monitors or all-year-round available fruits and vegetables. There may be a shift to parameter setting and enforcement by agents outside the chain. The more conformance/compliance with parameters can be codified, generalized and credibly applied, the less need there is for governance from within the chain. Given the costs of chain governance, particularly with respect to the enforcement of process standards, there are strong reasons for global buyers to transfer the costs of monitoring to the suppliers through the introduction of certification schemes. Nevertheless, this can only be achieved if standards are both credible and relevant to the needs of buyers.

There is no single tendency for value chain governance. Its costs and inflexibilities have to be weighed against the capabilities it offers for product differentiation and risk reduction. Instead of characterizing the issue in terms of the presence or absence of value chain governance, it may be more helpful to think about different forms of value chain governance, ranging from quasi-hierarchy through to modular networks.8 For global buyers, the challenge of value chain governance is to maintain sufficient control over the value chain while minimizing costs. The trade-off is likely to be expressed in different varieties of inter-firm governance in value chains, and these are discussed in the chapters in this book.

NOTES * 1. 2.



This chapter is a substantially revised version of our article in the IDS Bulletin, 32 (3), 2001. These issues are discussed in Dolan and Humphrey (2000). By restricting the term ‘governance’ to non-market coordination of economic activities, we are distinguishing between ‘market coordination’ and ‘coordination through governance mechanisms’. Clemons et al. (1993) use the term ‘explicit coordination’ to refer to this type of inter-firm governance. In many cases, parameter setting goes ‘backwards’ along the chain, from buyer to seller, but this is by no means always the case. Buyer and seller may set parameters jointly if they each have competences relevant to the parameters being set. In a few cases, parameter setting goes ‘forwards’ from seller to buyer. Franchise operations are the clearest example of this. The issue of governance through product and process standards has become increasingly


5. 6. 7. 8.

Local enterprises in the global economy complex, partly due to the proliferation of such standards. For an overview, see Chapter 3 by Nadvi and Wältring (this volume). For a discussion of the impact of modular product design architecture on industry structure, see Galvin and Morkel (2001). For a discussion of ISO 9000 certification, and also some of the possible shortcomings of the certification processes in the surgical instrument trade, see Nadvi (2001) and Nadvi’s Chapter 11 in this book. See, for example, the EUREP standard for good agricultural practice developed by European food importers and retailers, http://www.eurep.org. These issues are discussed in Gereffi et al. (forthcoming).

REFERENCES Clemons, E., S. Reddi and M. Row (1993), ‘The impact of information technology on the organization of economic activity: the “move to the middle” hypothesis’, Journal of Management Information Systems, 10 (2), 9–35. Dolan, C. and J. Humphrey (2000), ‘Governance and trade in fresh vegetables: the impact of UK supermarkets on the African horticulture industry’, Journal of Development Studies, 37 (2), 147–76. Galvin, P. and A. Morkel (2001), ‘The effect of product modularity on industry structure: the case of the world bicycle industry’, Industry and Innovation, 8 (1), 31–47. Gereffi, G. (1994), ‘The organisation of buyer-driven global commodity chains: how US retailers shape overseas production networks’, in G. Gereffi and M. Korzeniewicz (eds), Commodity Chains and Global Capitalism, Westport: Praeger: pp. 95–122. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Gereffi, G., J. Humphrey and T. Sturgeon (forthcoming), ‘The governance of global value chains’, Review of International Political Economy. Hobday, M. (1995), Innovation in East Asia: The Challenge to Japan, Aldershot, UK and Brookfield, US: Edward Elgar. Kaplinsky, R. (2000), ‘Globalisation and unequalisation: what can be learned from value chain analysis?’ Journal of Development Studies, 37 (2), 117–46. Keesing, D. and S. Lall (1992), ‘Marketing manufactured exports from developing countries: learning sequences and public support’, in G. Helleiner (ed.), Trade Policy, Industrialisation and Development, Oxford: Oxford University Press, pp. 176–93. Lamming, R. (1993), Beyond Partnership: Strategies for Innovation and Lean Supply, London: Prentice-Hall. Nadvi, K. (2001), ‘Global standards and local responses’, paper for the workshop on ‘Local upgrading in global chains’, Brighton, Institute of Development Studies, February. Novak, S. (2000), ‘Auto value chain development’, paper for the ‘IMVP Annual Sponsors Meeting’, Boston. Reardon, T., J.-M. Codron, L. Busch, J. Bingen and C. Harris (2001), ‘Global change in agrifood grades and standards: agribusiness strategic responses in developing countries’, International Food and Agribusiness Management Review, 2 (3), 421–35. Sako, M. (1992), Prices, Quality and Trust, Cambridge: Cambridge University Press.

Governance in global value chains


Schmitz, H. and P. Knorringa (2000), ‘Learning from global buyers’, Journal of Development Studies, 37 (2), 177–205. Spulber, D. (1996), ‘Market microstructure and intermediation’, Journal of Economic Perspectives, 10 (3), 135–52.

5. The underground revolution in the Sinos Valley: a comparison of upgrading in global and national value chains Luiza Bazan and Lizbeth Navas-Alemán 1.


The recent debate on global value chains has been concerned with the influence of global buyers on the prospects for manufacturers in developing countries. Gereffi (1999), the pioneer of this literature, has defined value chain governance as the coordination of different and dispersed activities involved in a chain. Since these activities aggregate different value at each stage, being the governor in a value chain ultimately implies deciding who in the chain performs the better remunerated activities as well as the less lucrative ones. Thus, it is clear that the implications of chain governance go far beyond merely organizing activities: governance is intimately linked with upgrading and the distribution of gains along the chain. While the global chain literature emphasizes that the upgrading prospects of local producers are determined from outside the locality, the cluster literature suggests that advances in upgrading depend above all on relationships within the locality. Particular attention is given to cooperation between local producers and to the role of public and private support services (Nadvi, 1997; Rabellotti, 1997; Schmitz, 1995). However, this body of literature has increasingly acknowledged the need to include external linkages in the analysis of upgrading processes (Nadvi and Schmitz, 1999). In this chapter we combine the value chain and cluster approaches to answer the following questions: What type of chain do local firms operate in? Does the type of chain have an influence on their upgrading? What is the scope for local collective strategies if producers operate in chains governed by global buyers? We examine these questions for the case of the Sinos Valley, a footwear cluster in the south of Brazil. Firms in this cluster feed into a number of chains with different characteristics. It therefore 110

The underground revolution in the Sinos Valley


lends itself to comparative analysis of how insertion in global value chains affects upgrading in local firms. In order to conduct this comparative analysis, we use the chain and upgrading distinctions proposed by Humphrey and Schmitz (2000). Chain governance can take three main forms: network (cooperation among – more or less – equals); quasi-hierarchy (buyers have a high degree of control over suppliers); and hierarchy (buyers take direct ownership of some operations). When a chain does not have any type of governance, it is characterized by market-based relationships. Upgrading, in turn, can vary from product (moving into more sophisticated product lines), process (transforming inputs into outputs in more efficient ways) and functional upgrading (acquiring new, higher value added functions in the chain, that is branding, marketing, design). The main market of the Sinos Valley cluster is the USA, whose buyers are the undisputed governors of the value chain. In addition, the cluster produces for the European, Latin American and the Brazilian domestic market. The governance of these chains varies and this is central to our research design: a systematic comparison and analysis of the different types of chain governance and types of upgrading in local firms. The chapter shows that chain governance is tightest and most hierarchical in the US value chain, where process and product upgrading are common among producers but functional upgrading is rare. Governance is somewhat less hierarchical in the European chain and functional upgrading is slightly more common. The national chains, catering for the Brazilian market, are characterized by market-based relationships and functional upgrading is very widespread. The Latin American chain presents hybrid characteristics (quasi-hierarchy and market). The chapter also analyses tensions between local cluster and global chain governance. This is particularly intense in the Sinos Valley, which has been exposed to tight chain governance but its cluster governance has historically been strong, based on many local institutions (see Schmitz, 1995; 1998; Bazan, 1997; Ruas, 1995). This strong local governance helps to explain our findings of collective functional upgrading efforts (which we regard as a discreet ‘underground revolution’) within the cluster, despite the presence of widespread quasi-hierarchical relationships. However, it was the study of changes along the value chain that provided an explanation for the new functional upgrading activities observed in the Sinos Valley cluster. These changes included risk-diversification strategies by global buyers, which in turn loosened governance ties that encouraged diversification moves on the side of local producers. Nevertheless, there are elements endangering this functional upgrading ‘revolution’. They are also to be found both at the cluster and at the chain level via internal rivalries within


Local enterprises in the global economy

the cluster and the stronger commitment that some local producers may still feel to their global buyers. Fieldwork for this research was carried out in 2000. Qualitative data was collected through in-depth interviews with 11 global buyers and nine local business associations and governmental agencies. Quantitative data was obtained through a survey of 81 local firms, covering 40 footwear producers and 41 input suppliers, sampled from the database of one of the leading local business associations. Input suppliers were included in order to deepen our value chain analysis. The samples were stratified according to size of firm and market orientation1. The survey covered the period from 1996 to 1999. Earlier periods are also covered by drawing on secondary sources. The second section of this chapter presents an overview of the Sinos Valley’s governance and upgrading trajectory. The third section investigates the relationship between the patterns of chain governance present in the cluster and the corresponding patterns of upgrading. Section 4 explains why the relationship between chain governance and types of upgrading is not always clear cut; this also justifies our qualified optimism regarding upgrading prospects for the Sinos Valley. Sections 5 and 6 use the value chain perspective to analyse the local responses to global challenges faced by the footwear industry in the Sinos Valley: section 5 focuses on the firm level and section 6 examines the collective initiatives taking place within the cluster. The chapter concludes by summing up its main contributions to the current debate.



In 2000, Brazilian footwear producers had an output of 580 million pairs and exported 163 million. This performance meant that Brazil was ranked in sixth place amongst global exporters (Abicalçados, 2002). All this would seem quite ordinary for a country the size of Brazil except for the fact that 84 per cent of these exports and more than half of this output came from one single cluster – the Sinos Valley, in the southernmost state of Rio Grande do Sul (bordering Argentina and Uruguay). According to records provided by ACI (the commercial association of Novo Hamburgo, the ‘capital’ of the Sinos Valley), there are 693 firms2 and 101000 workers employed directly or indirectly in footwear production in the Sinos Valley (see Table 5.1). In view of the range and depth of input suppliers, service providers and support institutions that had developed around the local footwear industry, Schmitz (1995) referred to the Sinos Valley as a ‘Super Cluster’. Indeed, the cluster is also remarkable for its institutional density. As


The underground revolution in the Sinos Valley

Table 5.1 Number of firms and workers in the footwear industry of Rio Grande do Sul (RGS) and the Sinos Valley Region Sector Footwear producers Tanneries Machinery producers Input suppliers Export agents and transportation firms Total

Rio Grande do Sul

Sinos Valley





367 91 90 222 93

88 083 15 181 5935 23 132 1296

269 63 75 198 88

71386 10755 2536 15598 1258


133 627



Note: In 50 firms (23 input suppliers and 27 export agents) the number of workers is unknown. Therefore, the number of workers corresponds to 813 firms of RGS and 645 in the Sinos Valley. Source: ACI (2000)

other authors have recorded (Schmitz, 1995; Bazan, 1997; Klein, 1991; Ruas, 1995) the Sinos Valley has a rich network of business associations and public–private technology and training centres, most of which were created in the 1960s and 1970s. Therefore, it is not surprising that when US ex-producers of footwear were scouting the developing world for promising sourcing locations, the Sinos Valley stood out for its many advantages. First, it was a wellestablished cluster, with a long experience in serving the huge Brazilian market. Second, it had already consolidated an extensive network of input suppliers. Finally, it had specialized support institutions. The ‘discovery’ of the Sinos Valley in the late 1960s and its subsequent insertion into the global footwear value chain triggered a process of fast growth. In two decades, the cluster transformed itself from a local producer that supplied the national market only, to one of the world’s leading exporters. However, for all its potential and the enviable range of local institutions, the footwear is still a commodity, known for its price range (US$ 10–11 in the US) rather than for its design or brand. This is not to say that the product and process upgrading achieved by the cluster is not impressive, that is quality and speed are its strong points, but the cluster failed to reposition itself as a higher value added producer. At least, this is what Schmitz (1998) concluded. He claimed that this was due to the type of value chain in which the leading cluster firms were inserted.


Local enterprises in the global economy

It seems that most collective attempts to upgrade the cluster functionally were met by resistance from a handful of large local firms who wanted to avoid upsetting their main foreign buyer. This US buyer was acquiring close to 100 per cent of these firms’ output and more than 50 per cent of the entire cluster’s. It was indeed a delicate situation and Schmitz (1998) concluded that leading firms used their clout within local associations to hold back initiatives aimed at greater marketing independence. Functional upgrading seemed to be blocked by the cluster’s insertion into the quasihierarchical US value chain. This conclusion ran contrary to a more optimistic scenario that Gereffi (1994) proposed in his chain analysis in Asia. As the pioneer of recent value chain literature, Gereffi proposes a view of an almost ‘natural’ process of learning and upgrading taking place in the firms that join a global value chain as suppliers. He draws most of his evidence from the Asian garment sector where he depicts a number of success stories regarding firms that managed to improve not only their product and process capabilities but that were also able to develop their design, brand and marketing channels. Therefore, despite the many similarities between the footwear and textile chains (both buyer-driven, low-tech, labour intensive, semi-commodities), conclusions regarding their upgrading patterns were very different. In order to deepen our understanding of the relationship between chain governance and upgrading, this chapter takes a different approach. It compares different types of chains and examines the implications for upgrading.

3. COMPARISON OF GOVERNANCE AND UPGRADING PATTERNS While the US has long been its main market (69 per cent of total Brazilian exports in 2000, according to Abicalçados, 2002) the Sinos Valley also exports to other parts of the world. This makes it possible to conduct comparative analysis of four chains – the domestic, European, Latin American and the US value chains. These four chains present different governance characteristics allowing for an exploration of whether chain governance makes a difference in upgrading at the firm level. This is one of the main questions of the research. The pattern of governance in the US value chain constitutes an excellent example of quasi-hierarchy. The US is the largest importer of footwear in the world, buying more than 23 per cent of all the world’s trade (Abicalçados, 2002) and is also the host location of the majority of branded retailers that source their products from the developing world.

The underground revolution in the Sinos Valley


Being the pioneers of this method of global sourcing, combined with their strong buying power, has made the US capable of imposing a certain type of governance on ‘their’ value chains. The US quality control practices (namely with in-house monitoring technicians), of tight deadlines and ever increasing pressure to lower production costs have been the blueprint for international sourcing in the footwear industry for more than 30 years. The typical characteristics of a quasi-hierarchical value chain are apparent in the Sinos Valley’s US value chain. The buyers are the undisputed leaders in the chain, exerting control over intermediaries, local producers and often input suppliers as well (this latter influence is called ‘convening power’3). This asymmetrical relationship with local producers can be explained by six main factors. First, a small number of agents with close ties to large US retailers controlled a high share of exports. Throughout most of the 1980s and until 1998–99, the main export agent in the cluster (closely linked to one US retailer) accounted for between 40 per cent and 50 per cent of the clusters’ exports. Other agents (10 in total), when added together were responsible for another 30 per cent to 40 per cent and the remainder was in the hands of the smaller agents in the cluster (Interview with a trade association officer, 2000). These figures changed dramatically after 1998–99, when the main US buyer changed ownership, causing a series of knock-oneffects further down the value chain (this will be explained in section 5). Second, buyers have resisted sharing their knowledge on higher value added activities such as design, branding, marketing and chain coordination. Third, buyers had numerous and different sourcing (or ‘exit’) options such as China, Spain and Portugal in the unlikely scenario that local producers did not accept their terms. Fourth, this same degree of choice was not available to local producers operating in this chain due to their lack of original design and marketing channels into the US. As a result, being cut off from this huge market was a major deterrent to dissent. Fifth, even though concerned about quality, the US market tends to be a price-driven value chain, which results in low profits. Buyers bargained with several footwear producers in the Sinos Valley simultaneously in order to keep prices low (particularly when dealing with non-branded footwear). Finally, US final buyers are very large, demanding bigger batches than buyers in other markets, and yet the US market has very few ‘gatekeepers’ (importers). This combination of factors places final buyers, and intermediaries linked with them, in a very comfortable position to exert power over producers. Other buyers from the developed world (mostly in Europe) have copied the US’ sourcing practices, including the division of labour between global buyers (in charge of design, marketing, branding and supply chain organization) and local footwear producers that only deal with the manufacturing


Local enterprises in the global economy

aspects. However, even if there are many similarities between the US and the European value chains, there are also many differences. As a result, the phrase ‘soft quasi-hierarchy’ is more appropriate to describe the European value chain sourcing from the Sinos Valley. Many local producers distinguished European buyers from US buyers because of their higher interest in quality. This emphasis on quality led buyers to have more respectful and closer relationships with producers because they have often had to work together to achieve such quality improvements. European buyers were also reported to be more willing to pay slightly higher prices. This need to appear respectful may be linked to the necessity of finding ways to compete with long established US buyers (and their large orders) in attracting the best Sinos Valley producers. However, this does not allow for this type of chain governance to be called a network since differences in power between buyers and producers are still significant. Indeed, European buyers still have greater control over what needs to be produced and they have more exit options (sourcing from other similarly priced places) than Brazilian footwear producers. Furthermore, the European chain has its gatekeepers as well, that is the big importers/wholesalers. They often try (and succeed) to insert themselves between independent Brazilian agents or Brazilian footwear producers attempting to sell directly to the large chains of shops in Europe. The organization of the domestic value chain is completely different from the global chains outlined above – the relationships between producers and buyers are market-based.4 This is reflected by a number of different indicators such as the low degree of buyer concentration and sales concentration to main clients, the strategic option of selling directly to retailers by using the producer’s own sales representatives and, more importantly, the main strategic activities (design, branding and marketing) are carried out by producers instead of buyers. Sales representatives are hired by producers and this sets them apart from exporting agents (who are hired by buyers) in the quasi-hierarchical chains. Therefore, the main differences are that sales representatives are accountable to the producers, buyers (local wholesalers and retailers) do not control any of the strategic activities mentioned above and these conditions, combined with the atomized local footwear demand make market-based relations the main trend in the chain. The main obstacle facing domestic producers is the fierce competition amongst themselves. According to local firms, a small number of footwear producers (that own famous local brand names) already dominate the domestic market making it difficult for non-branded producers to succeed in it. As stated by the owner of a medium sized firm supplying the US value chain: ‘The domestic market belongs to the big branded firms or to the “fabriquetistas” [small, usually sub-contracted firms]. You have to have

The underground revolution in the Sinos Valley


your brand, and this demands big investments’. In addition, the influx of cheap Chinese footwear in the 1980s made the conquest of new markets more urgent than ever. Brazilians had developed a high quality footwear industry that was not paralleled in neighbouring countries. In fact, many consider Brazil the ‘Italy’ of Latin American footwear making. Therefore, it was domestic-oriented firms who first saw the opportunity to use their competitive advantage and sell to the Latin American market either by means of their own sales representatives or contacting the importers/retailers directly. This is why the majority of Brazilian firms selling to the Latin American market continue to be the branded, (originally) domestic-oriented firms. They have been entering the Latin American market in the same way that they operate within the domestic value chain, that is through market-based relationships, and this had been the pattern of governance that defined the Latin American value chain until the mid-1990s. This picture is evolving, and nowadays, entering the Latin American market has become a market-diversifying strategy for all types of Brazilian footwear producers, even the ones that do not have their own brand for either the US, European or domestic markets. One of the new strategies is that of export agents creating their own brands and hiring Brazilian producers to do the manufacturing (quasi-hierarchy). Another strategy involves producers selling (without a brand) to Latin American importers/shops that put their own brand on the shoes. In this case, since there is no agent and no instructions given by the buyer to the producer we could consider this a type of market relationship, but it is still rare. Some of the export agents interviewed for this study felt that Brazilian branded producers have ‘more vanity’ and are not as flexible regarding sales without brands or other schemes to sell to other countries. Arguably, unbranded producers that are traditionally net exporters do not have anything to lose by being more ‘open to sell without a brand in Latin America’. By contrast, domestic branded companies do have something to lose – the chance to cash in on the hard-won assets of brand and design. Therefore, the pattern of governance in the Latin American value chain depends very much on the channel used by Brazilian producers to enter these markets. The more they use export agents, the more the quasihierarchical US pattern of governance is replicated onto the Latin American value chain. The opposite is also true: the more firms enter such markets through direct sales, the more they take the shape of the domestic value chain – market-based. Since Latin America was not the main market of most firms at the time of this survey, it remains to be seen which pattern of governance becomes more common in the near future, and particularly because the importance of the Latin American markets is increasing.


Local enterprises in the global economy

Having described how the four chains are governed we can now investigate whether the upgrading experiences of local producers differ according to the type of chain governance. One way in which we analysed the influence of chain governance on upgrading was by comparing the type of upgrading efforts that were determined by buyer demands. In the star diagrams in Figures 5.1 and 5.2, we highlight the different demands buyers place on those footwear producers operating in a quasi-hierarchical chain (US or Europe) and those who work in a market-based chain (Brazilian domestic market). Product quality improvements Innovative use of new materials

Innovative modelling

5 4.5 4 3.5 3 2.5 2 1.5 1

Variety of lines/collections


Faster deliveries

Reliable deliveries

Innovative design

US buyers’ expectations European buyers’ expectations

Price reductions

Compliance with international technical norms Increasing flexibility

Author’s survey, 2000. Respondents: 11 buyers

Figure 5.1 US and European buyers’ demands on footwear producers’ performance As discussed above, buyers in quasi-hierarchies have expectations regarding their producers’ performance that clearly reflect a greater concern with product and process upgrading compared to functional upgrading. In the US chain, for instance, buyers place emphasis on price, quality improvements and flexibility rather than in design. During interviews with both buyers and footwear producers in the quasi-hierarchical chains, it became clear that although ‘innovative’ modelling5 and usage of new materials are buyers’ concerns, they are limited to the footwear producer’s ability to cope with the new models and materials requested by the buyers, because it is the latter who provide the original design. A different situation characterizes the market-based chain. Although quality, delivery reliability, innovative modelling and design are also con-


The underground revolution in the Sinos Valley Product quality improvements Innovative use of new materials

Innovative modelling

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0

Price reductions

Faster deliveries

Innovative design

Reliable deliveries

Variety of lines/collections

Compliance with national technical norms Increasing flexibility


Author’s survey, 2000. Respondents: 20 local producers

Figure 5.2

Domestic buyers’ demands on footwear producers’ performance

cerns for these buyers, they reported that ‘innovative’ modelling and design are more important. They expect original designs and independent innovation to be carried out by producers because domestic buyers do not have this expertise. Additionally, domestic buyers expect Brazilian producers to create several design collections a year and to have an increasing number of models and lines of products. 3.1 Upgrading Patterns in the US and European Value Chains (quasihierarchies) These chains are characterized by strong product and process upgrading and weak functional upgrading. Apart from quality programmes, footwear producers that operate mainly in the US value chain have invested heavily in cutting costs by reducing rejection rates. They have also made the highest investments in product and process upgrading, as Figure 5.3 illustrates. This is consistent with buyers’ expectations, that is that the producer excels first and foremost in manufacturing. These footwear suppliers have become highly competent in footwear production, reaching quality levels that are comparable with the ones achieved by Italian manufacturers, thus corroborating Knorringa and Schmitz’s (2000) findings. Local footwear producers operating in the European value chain seem to


Local enterprises in the global economy New machines 5

New raw materials and components



3 2

Waste reduction

Speeding up deliveries

1 0

Total quality programmes

Quality improvement

Data automation

New managerial techniques Improvement of the production process

US Source:



Author’s survey, 2000. Respondents: 40 local producers

Figure 5.3 Footwear producers – investments in process and product upgrading from 1996 to 1999 have invested less in the improvement of their production processes than US footwear suppliers and indeed, invested less than domestic-oriented firms. This can be explained by the fact that European buyers have been particularly selective when choosing suppliers in Brazil and because these producers made most of their investments before 1996. Between 1996 and 1999, few producers in this chain invested in new production methods with the exception of the adoption of CAD-CAM, an innovative tool used nowadays by many firms in the cluster regardless of the chain they operate in. Many sampled producers in the quasi-hierarchical chains stated that their agents were actively supporting them in product and process upgrading strategies beyond quality control. Export agents working within the US and European value chains were reportedly committed to the improvement of every aspect concerning the production process in supplying firms. In other words, we could see a commitment on the side of agents in terms of the product and process upgrading of the producers that work for them, even if most of the support is given in a top-down, hierarchical way. However, as pointed out by Schmitz (1999), this flow of knowledge used to be stronger in the cluster’s earlier days. The Sinos Valley is now a mature cluster with the majority of producers having already learned the most important pieces of knowledge regarding production.

The underground revolution in the Sinos Valley


Furthermore, among the sampled firms that have the US as their main market, approximately half of them had introduced new lines in this market. In most cases, these new lines were of a higher price. Notably, all new lines remained under the brand of the US buyer. Nevertheless, producers were very satisfied with the skills acquired when introducing these new lines. For instance, there were several producers of sandals that had been ‘upgraded’ over a period of time to supply boots (higher-cost and more complex products). Surveyed firms that operate mainly in Europe also reported a high level of process and product upgrading. In fact, 50 per cent of them had produced higher value added lines (under the buyer’s brand) in the last three years. However, progress in functional upgrading was more difficult to find among producers working mainly for the US and European value chains. In order to progress in this aspect, local producers would have to invest heavily in developing their weaker areas, that is branding, marketing, design and R&D. They would have to try and create their own designs, develop a brand, participate more actively in the marketing strategy, and so on. These efforts could even be undertaken in another market, different from those of the US or Europe. This leads us to the question: Are producers in quasi-hierarchical chains taking these steps? According to our results, none of the surveyed producers in the US value chain have launched their own new lines or new brands in new markets. In contrast most producers (around 75 per cent) in the European chain had introduced their own brands in new markets and one of them even designed a new line (distinct from the style it had currently been producing) for that new market. Even though these achievements in functional upgrading are attributed to producers who had developed under a European quasi-hierarchy, they did not take place in a European country but in the Brazilian domestic market. These results hint at a finding that we corroborated during our indepth interviews, that is working in the European rather than in the US value chain does offer better learning opportunities for firms to upgrade functionally in new markets. We argue that this is due to the most notable difference between European and US buyers – while the former tend to be more quality-driven, the latter tend to be more price-driven. The US value chain tends to put great pressure on prices. However, this does not imply that they are not demanding about quality. According to our findings, what US buyers want is a better quality for the lowest price possible and this expectation makes it difficult for them to sustain trustful relationships with their footwear producers. Footwear producers prefer to deal with European buyers because they are perceived as more respectful and reasonable in their demands. It follows that producers want to have longer (and closer) relationships with European buyers, which is


Local enterprises in the global economy

a prerequisite for high quality learning to take place between two links of a value chain. We believe the kind of learning that takes place in the European quasi-hierarchy is more conducive for the functional upgrading of their footwear suppliers than the type of learning that takes place in the US quasi-hierarchy. Indeed, we gathered evidence on the failed attempts of several nonbranded footwear producers (that had been operating in the US quasi-hierarchy for decades) to enter the domestic market by creating their own brands and designs. They were easily swept away by the established branded domestic producers. Large investments in market research, advertisement and hiring of designers were lost in those attempts, and many of those firms decided to compete in the less demanding Latin American market (usually through export agents). The domestic market posed a formidable challenge even for the traditional suppliers to the US value chain that considered themselves part of an ‘elite’ of world class manufacturers. This does not appear to have been a substitute for years of experience in creating original designs, developing marketing strategies and nurturing strong brands, even if it was for the domestic market only. 3.2 Upgrading Patterns in the Brazilian Domestic Market and Latin America (Market-based) These chains are characterized by strong functional upgrading and moderate product and process upgrading. Domestic-oriented producers have devoted more resources to investments in product and process upgrading than firms operating in the European value chains but fewer resources than US-oriented producers. According to our survey, domestic-oriented producers have had to divide their ‘upgrading’ resources between the two types – functional and product/process. The reason for this seemingly even distribution lies in their recent preparations for exporting to new markets (part of what we have called the ‘Underground Revolution’). This issue will be detailed later; here we will deal with the product/process type of upgrading. Domestic-oriented firms, used to making their major investments in internal market research, branding and design, realized that in order to penetrate export markets (beyond Latin America) they needed to match the internationally renowned product and process performance levels of producers exporting to the US or Europe. At the same time, domestic-oriented producers had to guard their local market share, given recent attempts of suppliers to the US trying to compete in the internal market by introducing their own brands in Brazil. The boundaries of competition are blurring, so further investments in process and product upgrading became imperative, but they did not count on the support of export agents. In the case of

The underground revolution in the Sinos Valley


domestic-oriented firms, private consultants have played a key role in offering advice for implementing any kind of upgrading strategies. As for functional upgrading, it has been shown that domestic-oriented firms working within market relationships tend to be responsible for all the strategic tasks of design and branding. At the same time, domestic and Latin American buyers have less knowledge of these two areas (design and branding) than local producers. Domestic buyers specialize in sales and marketing, therefore any over-dependence by producers on (domestic) buyers based on knowledge is unlikely to occur. Despite selling through intermediaries (sales representatives),6 producers in the domestic market do so on different terms to those selling to the US and European markets, that is sales representatives are basically concerned with trading, and they have no links with any other activity other than finding buyers for the footwear producers. Sometimes they can also help producers by facilitating indirect exchange of information between producers and retailers. As outlined earlier, our findings show that in most cases functional upgrading among domestic-oriented firms is linked with product and process upgrading as well. For instance, the development of new lines of products (product and process upgrading) is often done in conjunction with the launching of new brands and the search for new markets (functional upgrading). Data for this chain shows that more than half of the firms have developed new lines at higher priced levels for their traditional market. Approximately 20 per cent of them have done so through a new brand. Moreover, 15 per cent of surveyed firms have introduced new higher value lines in new markets. Finally, some firms have developed new lines at a lower price either for new markets (20 per cent) and many for their traditional market (55 per cent). In most cases, it has been done through the launching of new brands (75 per cent and 54.5 per cent respectively)7. To summarize this section, we can conclude that footwear producers operating in value chains based on market relationships (such as the Brazilian domestic market) are more likely to achieve functional upgrading than producers in quasi-hierarchical value chains such as the US value chain. Softer quasi-hierarchies such as the European value chain have been less inimical to functional upgrading. Product and process upgrading are actively encouraged and supported by quasi-hierarchical buyers. Table 5.2 presents these conclusions in a more schematic way. Our results show that although the Sinos Valley has followed a successful upgrading trajectory from its humble origins before the decade of the export boom (1970s) to its current position as one of the top footwear producers in the world, this progress has been marked by strong internal differences. Process and product upgrading has taken place for most of the firms in the cluster, but functional upgrading seems elusive, particularly to firms


Local enterprises in the global economy

Table 5.2 Governance in the footwear value chain and implications for local upgrading of footwear producers Quasihierarchies

Market relationships

Dimensions of governance and implications for upgrading



Latin America


Market share Buyers’ concentration Dependence on intermediaries Unequal division of essential activities in the chain

High High High High

Medium High Medium Medium to high

Low Medium to low Medium to low Low

High Low Low Low

Upgrading Product and Process upgrading


Medium to high

Functional upgrading


Medium to high Low to medium

Medium to high High

Medium to high

Source: Authors’ own research

that are mainly engaged in exporting through the quasi-hierarchies. Generally speaking, firms exporting to the US and Europe have been concentrating most of their efforts on production rather than design, branding and marketing. Firms in the domestic market have their own brands, but feel unable (with few exceptions) to enter the US and European markets with these brands. In short, the widespread functional upgrading that Gereffi (1994) observed in Asia when clusters enter a buyer-driven value chain has failed to happen. We corroborate Humphrey and Schmitz’s (2000) indication that this failure is linked to the more predominant pattern of chain governance – quasi-hierarchy. Moreover, we found that a quasi-hierarchical type of governance may be established in any value chain, not only in the US and European chains. Even in the domestic and Latin American chains, quasihierarchy may exist when producers relinquish most (or all) control over activities such as branding, design and marketing to their buyers. However, we also found increasing evidence that market-type relationships have become a way for producers to ‘escape’ from the dependence path established by quasi-hierarchy. Furthermore, given the alleged pervasiveness of this type of value chain governance, success stories of functional upgrading were significantly more numerous than one would expect. Thus, there were several signs of more hopeful developments taking place in the cluster.

The underground revolution in the Sinos Valley


Therefore, even if chain governance matters for the type of upgrading achieved by local producers, this relationship is not necessarily clear-cut. We discuss in the next section how we were able to perceive these signs towards upgrading which make this chapter perhaps a bit more positive regarding local clusters’ chances for functional upgrading even when exposed to high degrees of quasi-hierarchy.



In the previous section we focused on the dominant patterns of chain governance and the associated dominant patterns of upgrading. The strong presence of the US quasi-hierarchical style to value chain governance has usually been blamed for the lack of a generalized functional upgrading effort in the Sinos Valley. Yet, as we have seen in the previous section, the US value chain is not the only one exerting its influence on the cluster. There are (at least) four distinct value chains operating in the Sinos Valley, and more importantly, many firms work within several types of chains simultaneously. Therefore, this study is offering a new view of qualified optimism on the upgrading prospects of the Sinos Valley by taking a closer look at two main factors. First, the variety of upgrading opportunities due to the presence of different chains and second, the use of these opportunities by individual firms operating in different value chains simultaneously. Most studies on the Sinos Valley (Schmitz, 1999; Knorringa and Schmitz, 2000) have concentrated on the global value chains that the cluster belongs to and neglected national and Latin American chains. Those studies could therefore not capture the type of upgrading (functional) that was taking place in the national value chain as well as in the other value chains of Latin America. This bias has to do with a pervasive idea in academic and policy making circles, that ‘learning by exporting’8 is superior to any learning acquired within the national frontiers (particularly, but not exclusively) of a developing country. If Brazil were a small country with an undemanding market, these achievements in functional upgrading by domestic-oriented firms could be dismissed as unimportant. But that is hardly the case. The Brazilian footwear market is large (480 million pairs in 2002 according to Abicalçados, 2002), sophisticated and therefore very attractive to any international producer. However, (mostly) Brazilian companies are supplying it. These companies are not the net exporters, these are companies that devoted themselves to the national market and upgraded by working and learning with it. As a rule, buyers from the USA and Europe do not allow their Brazilian suppliers to sell the same products that are made for the Northern markets


Local enterprises in the global economy

directly into the Brazilian market, so any domestic venture has to be fully designed, marketed and manufactured by the Brazilian producers on their own. According to the ‘learning by exporting’ paradigm, Brazilian suppliers should not have any problem transferring the skills learned by exporting to the manufacturing of footwear for the domestic market. Indeed, the evidence from our research shows that they are able to transfer their skills in product and process upgrading, but little else. As mentioned before, there have been numerous attempts by the traditional exporters (belonging to the US value chain) to penetrate the domestic market, but success has been modest at best. The type of functional upgrading skills required to succeed in the domestic market (and in international markets) is not learnt when exporting through quasi-hierarchies. On the other hand, these skills seem to flourish when firms work through market relations. This important learning advantage of domestic-oriented manufacturers had not been detected in most studies on the Sinos Valley. The reasons for this are twofold. First, the market-based value chains had not been thoroughly analysed, particularly in regard to their potential to promote upgrading. Second, this advantage has become more important and visible now, when the traditional advantages of quasi-hierarchies (higher margins, regular payments in foreign currency, security and less responsibility) are being eroded by the commoditization effects of globalization, and some well-known flaws are becoming more evident (over-dependency on the buyer, lack of control over the firm’s future, fickle commitment on the side of the buyer, constant pressure to lower costs and prices, and so on). How then does one explain the advances in functional upgrading made by a few net exporting firms? This apparent contradiction can be explained by two factors: (i) the long history of knowledge diffusion in the Sinos Valley (better explained by the literature on clusters than value chains); and (ii) the fact that most producers work for several value chains simultaneously. Therefore, it is the experience of functional upgrading in the national value chain that is being transmitted to experiments in export markets and not vice versa as is usually presumed. The literature has not sufficiently recognized the significance of operating in several chains simultaneously. By working for several value chains these companies are exposed to the different styles of governance that stimulate different types of upgrading. This is very powerful when combined with a deliberate intent to apply the newly acquired capabilities in the more flexible value chains (usually the market-based ones). Sinos Valley producers that follow this strategy distinguish themselves from others by penetrating new and old markets with their own brands and engaging in product and process upgrading whenever it is convenient for them and not only when directed by the governors of the chains (in the case of the quasi-hierarchy).

The underground revolution in the Sinos Valley


This argument is not entirely new; it has been made before by Lee and Chen (2000) and by Kishimoto (2002) when they talk about ‘leveraging competences’. Lee and Chen provide a particularly useful insight when they observe that a company’s development strategy does not have to be linear in all cases. That is to say, following the path from OEM (Original Equipment Manufacturers) to OBM (Own-Brand Manufacturers) status in one single value chain is not the only way, or indeed the ideal way, for all firms. Taking the case of resource-constrained contract manufacturers, Lee and Chen explain that firms may be OBM in one market and still be OEM in another, while always pursuing a growth strategy based on the synergies between different business activities. There seems to be one short-term caveat regarding the profitability of this strategy. Lee and Chen (2000) suggest that firms following OBM strategies in some markets would receive higher incomes, which in turn would support the rest of the operations. We did not find evidence of increased incomes for firms pursuing functional upgrading activities in the Sinos Valley in comparison to others that remained solely unbranded manufacturers. Investments in marketing, design and branding are high at the beginning so perhaps another survey should be made in a few years’ time to measure whether the returns justified such expenditure. It could be argued, however, that because of the increasing commoditization of footwear production and tough competition from several clusters around the world, pursuing an OBM strategy in one or more markets could be a survival strategy rather than a means of just raising profitability. In brief, evidence from governance patterns and upgrading efforts from all the main chains (export and domestic) present in the Sinos Valley provided the basis for this study’s comprehensive analysis on the cluster’s upgrading scenario. Thus, it became apparent that along with its wellknown achievements on product and process upgrading there has been a degree of functional upgrading in the footwear cluster of the Sinos Valley. Perhaps it has not been a total re-positioning of the cluster at a global level (becoming the next ‘made in Italy’ type of cluster) but something has changed since the mid-1990s. The Sinos Valley’s self-depicting phrase ‘we don’t sell, we are bought’ is no longer valid.



For almost three decades the Sinos Valley experienced the contradictory situation of being a successful cluster (in economic terms) but highly vulnerable due to its strong dependence on US buyers. This vulnerability was seen as a small price to pay when compared to the instability of the


Local enterprises in the global economy

Brazilian domestic market that was riddled with currency crises and hyperinflation. Therefore, any factor that reduced the capability of Brazil to sell footwear to the US was regarded as a threat, and eventually, as a crisis in the Sinos Valley. In this context, it becomes clear that the Brazilian producers carefully avoided any action that had the potential to alienate US buyers. Upgrading strategies of a functional type were seen as a challenge to the lead firms in the quasi-hierarchies, since these actions would encroach on the core competences of these lead firms (marketing, branding, design). The fear of possible retaliation from US buyers, combined with steady, safe incomes in US dollars, made Brazilian producers view any attempt at functional upgrading an unnecessary risk. But from the early 1990s onwards, the globalization winds were blowing strongly and the foundations of this comfortable arrangement started to weaken. Producers in quasi-hierarchies had adopted all process and product upgrading techniques indicated by their buyers and still felt at risk of losing important shares of their main export markets to China9 (particularly in the US) and to Portugal and Spain in Europe. In the domestic market, low-cost Chinese imports were also making local Brazilian producers increasingly concerned. A timid but inexorable process of risk diversification emerged within the cluster. Suppliers of the quasi-hierarchies began coveting larger shares of the domestic and Latin American markets (sometimes by developing their own brands). Domestic-oriented firms realized they could export more of their products and complement their incomes with more foreign currency. This process has been subtle in order to reassure traditional buyers in the quasi-hierarchies that they are still important customers. Thus, we have termed this process an ‘underground revolution’. In this section we will concentrate on the individual efforts undertaken at firm level. The next section will tackle the collective and institutional dimensions of this change. The origins and effects of this underground revolution are better explained using a value chain perspective. As mentioned above, from the early 1990s falling footwear prices and growing competition from global footwear producers had created an environment of greater uncertainty and increasing volatility of markets. In order to satisfy the new global requirements, local producers and input suppliers concentrated (with the encouragement of their buyers) on improving their manufacturing abilities (product and process upgrading) and lowering costs. All these efforts pushed the cluster even further into the ‘world class’ manufacturing status, according to a survey of global buyers carried out by Knorringa and Schmitz (2000). However, if footwear producers thought that these improvements would

The underground revolution in the Sinos Valley


guarantee global buyers’ loyalty, they were disappointed. On the contrary, in the Sinos Valley it had become possible to reconsider relationships because of the ‘generalized competence’ achieved by footwear producers and input suppliers. It became possible for buyers to ‘shop around’ to reduce costs, and producers had to do likewise with input suppliers. Internal ties loosened in the cluster as they had been loosening globally between global buyers and supplying clusters. Excellent levels of manufacturing became more ubiquitous and cheaper the more buyers looked around. Many global buyers felt that it was in their best interest to diversify their sourcing operations in order to protect themselves from sudden price rises and economic or political crises that could affect deliveries. This meant that they could protect themselves from any unforeseen problems arising from their sourcing regions. For Sinos Valley’s customers, this precaution became particularly acute between 1994 and 1998, when Brazilian footwear was relatively expensive due to the strength of the new currency (the real) and the weakness of the euro. Even Brazilian export agents had no qualms in transferring part of their sourcing operations to China and/or buying from the European competitors of the Sinos Valley. The main catalyst for change, however, came in 1998/99 when a very large US apparel group (whose main business is not footwear) bought the major customer of the Sinos Valley cluster. This customer accounted for 40 per cent of total cluster exports and up to 80 per cent of the output of the five largest and most powerful firms in the cluster. The sale meant a substantial reduction in the buyer’s commitment to the cluster, to the extent that the five leading firms (formerly reluctant to change strategies) were forced to consider value chains other than the US chain. This event shook the roots of the cluster’s powerful establishment because it affected the large, US-oriented footwear producers as well as the small and medium ones in contrast to previous crises. Even the most conservative firms realized that market diversification and progression into other essential (and higher-paid) chain activities such as marketing and design had become imperative. Buyers in general loosened their demands on ‘exclusivity’ from producers, even encouraging the latter not to be ‘overly dependent’ on one buyer. Reserving certain ‘moments of production’ for preferred customers emerged as the new custom. Therefore, the need for a diversification strategy was ‘passed down’ the value chain forcing all producers to find new markets and new outlets for their products. Input suppliers for the footwear industry deserve a mention here, since they were among the first firms in the cluster to react in the face of decreasing orders, declining prices and increasingly demanding expectations from


Local enterprises in the global economy

footwear global markets. They realized that they could not rely on the footwear industry alone to keep their businesses running, so they began selling their products directly to other industries and even to global footwear competitors of the Sinos Valley. Input suppliers have been very successful in their diversification strategies. There may be a link between this success and the long experience this sector has accumulated whilst working for customers operating in different value chains and even supplying different industries.10 Diversification is the key to the underground revolution. Producers in the Sinos Valley are attempting to increase their chances of success by acquiring new skills and trying to utilize them in several value chains. Compared with earlier decades, there is more emphasis on keeping a good ‘market mix’ than relying on one main market. Ideally, each firm would like to have a share of the domestic market and export as well. This balance is being approached from different starting points. Export-oriented firms working in the quasi-hierarchies are attempting to enter the domestic market and domestic-oriented firms are trying to increase their exports. Furthermore, both types of firms are trying to protect their traditional niches from newcomers. For the domestic-oriented producers, keeping their brand is an integral part of their strategy. They are transferring the lessons learnt in the domestic market to their exporting ventures, that is export agents are avoided and when required, they are hired by the producer so that they act on behalf of the producer instead of a foreign buyer. A Brazilian export agent in Novo Hamburgo explained a risky functional upgrading strategy: A couple of years ago, a local shoe producer decided to enter the US market with his own brand and offered to hire my services as an export agent. I accepted the deal not only for the commission, but also because I was already thinking to launch my own brand in external markets and this was an excellent opportunity to learn the ropes before actually doing it for myself. The risks of such an incursion are high: a project to position a brand in the US can cost around 5 million US dollars over two years, without any guarantee of success.

Therefore, another barrier that Brazilian brands face when trying to penetrate the US market is the cost. Lack of resources may be one of the reasons why Brazilians have not been more aggressive in appropriating branding and marketing, and since their US buyers are very powerful, a failed branding exercise can leave the entrepreneur in a very vulnerable position. Additionally, if any of a firm’s US customers felt alienated by the upgrading attempt, there is a high chance that the relationship will cease; this in itself constitutes another strong deterrent for aspiring functional upgraders.

The underground revolution in the Sinos Valley


Another interesting trend is that export agents are also trying to upgrade. They have all the obvious capabilities to enter markets (both domestic and external). They are quite knowledgeable in some of the key activities in the footwear value chain, that is marketing, design, brand, quality control, chain coordination, and so on. Although they are not manufacturers, producing to specifications is a competence which can be easily ‘hired’ in the Sinos Valley. It is quite possible that the new competition for local branded producers wanting to export will come from export agents. Apart from exporting with own brands, there is another functional upgrading strategy that is available to a few domestic-oriented firms – becoming governors of their own chain within the domestic market. We refer here to the case of some firms in the Sinos Valley that are following the example of designer footwear companies from São Paulo and Minas Gerais. This entails ending production and concentrating on the chain coordination activities including marketing, design and brand name. Again, export agents and trading companies are also following this trend locally. However, it is important to note that the internal distribution network in Brazil is so complex that export agents competing with domestic producers still need to hire the services of local sales representatives to deal with retailers. As for the really large branded footwear producers in Brazil, they are independently searching for new markets, that is without assistance from export agents. A few of them have pursued this strategy for a number of years by establishing their own marketing and sales offices abroad, even in the US and Europe. Such a strategy is too costly for smaller firms so they have forged direct alliances with distributors in the US. These alliances are forged after many business trips to trade fairs and often complex negotiations in order to convince distributors to sell incipient but increasing proportions (approximately 15 per cent) of their branded output in the US, without abandoning the domestic market. Some of them are paying high sums for distributors to advertise Brazilian footwear in the central and western cities of the US. At the other end of the spectrum, unbranded Brazilian producers exporting to the US and Europe are experimenting with functional upgrading in the domestic market. As mentioned earlier, these firms are trying to enter the domestic market by creating their own brands, for example one of the largest producers in the US value chain has adopted this strategy. This path, however, appears to be as difficult as it is to search for new markets abroad. Since the early 1990s, domestic-oriented firms fortified their position by increasing investments in their market. Such investments were enhanced in the period 1994–1997. As a result, some domestic firms developed established brand names and captured large market shares in the


Local enterprises in the global economy

domestic arena. Thus, for export-oriented firms, as much as for any newcomer in general, entering the domestic market is a difficult task. In sum, firms in the Sinos Valley saw their belief in the loyal buyer disintegrate in the late 1990s thus intensifying competition within the cluster and with other producer regions in the world. This challenge to the cluster’s stability came from the top of the US value chain but local firms have been able to rearrange their inter-firm relationships and their activities due to the pre-existing ‘generalized competence’ of producers and input suppliers. The strength of the cluster, combined with the loosening of vertical ties in the quasi-hierarchies made it possible for the Sinos Valley to adapt to the changing requirements of its various markets. One of the reasons this did not happen earlier is the fear of buyer retaliation. Indeed, a combination of qualitative and quantitative evidence (Bazan and Navas-Alemán, 2001) suggests that the fear of buyer retaliation still limits many producers’ intentions (particularly in the quasihierarchies) to pursue a functional upgrading strategy. Second, regardless of such trepidation, some producers are either finding ways to avoid confrontation or are finding the determination to assert new upgrading strategies (changing buyers if necessary). Third, most firms in quasi-hierarchies do not feel prepared or confident enough to pursue a functional upgrading strategy in their main value chain, whereas firms in market-based chains seem more inclined to venture into all types of value chains (with a fair degree of success). Finally, the threat of buyer retaliation (while still a possibility) has been subdued by the loosening of ties in all value chains and partially offset by the need to diversify risks and ensure survival in the markets.

6. THE INSTITUTIONALIZATION OF THE UNDERGROUND REVOLUTION The previous section concentrated on the implications of chain governance at the firm level. This section explores the connection between chain governance and collective upgrading strategies. As mentioned in section 2, many highly specialized local support institutions are located in the Sinos Valley. They have been instrumental for product and process upgrading in the cluster by complementing the contributions of global buyers. Yet Schmitz (1998) concluded that they did not succeed in promoting functional upgrading. Our research showed that this had greatly changed by year 2000, when several collective upgrading initiatives were established. During our survey, we came across alliances and healthy competition amongst many institu-

The underground revolution in the Sinos Valley


tions in order to come up with the best sectorial programme for local functional upgrading.11 Such openness towards programmes that openly announced their aims to create a collective brand for the cluster (such as ‘BY BRASIL’), to encourage local producer participation in international trade fairs, and to support the development of local designs, was a stark contrast with the earlier situation. According to Schmitz (1998), the five most powerful firms in the cluster (tightly committed and allied to their US buyer) had blocked a very similar programme called ‘Calçado do Brasil’ (Footwear from Brazil). It was clear that this significant change at the institutional level in 2000 increased the scope for collective action at the cluster level. This turn of events raised the questions: Why did this change happen? Why did it happen at this time? Again the chain perspective helps to understand this process. As mentioned in the previous section, a change of ownership occurred at the top of the dominant value chain, that is the takeover of the key US buyer who previously bought half of the Sinos Valley’s exports and who maintained a close relationship with the top leading firms in the cluster. This change of ownership, as well as the increased sourcing from China, resulted in a declining commitment to purchasing from the Sinos Valley. As a result, local lead firms realized that some of the initiatives they had blocked in the past would now be advantageous. They relinquished their grip on Abicalçados (the local footwear trade association), which in turn gave the other associations that were influenced by it a breathing space. This withdrawal made an immense difference to the cluster’s upgrading initiatives since the leaders of Abicalçados had previously played a prominent role in controlling the cluster rather than promoting it. For the old leaders, keeping the status quo in the quasi-hierarchies was seen as the priority. This came out clearly from in-depth interviews with association officials and some business leaders. Before we describe the variety of programmes sponsored by local trade associations, it is important to note that the most relevant associations are Abicalçados, ACI (the industrial-commercial association of Novo Hamburgo, capital of the Sinos Valley), AICSUL (the tanneries association), ASSINTECAL (the association of footwear input suppliers) and ABRAMEQ (the association of machinery producers). Bazan and Schmitz (1997) have shown how these associations have played an important role in the cluster’s upgrading trajectory. In many respects this profusion of associations has been positive for the cluster, for example, by representing the interests of the different sectors, promoting and supporting them nationally and abroad, and bringing resources into the cluster. However, it has also generated intense competition and differentiation into factions that look after their particular interests and have at


Local enterprises in the global economy

times lost sight of the common interests of the cluster. Such fragmentation has proved another formidable obstacle for collective initiatives towards functional upgrading in the cluster.12 In this context, the emergence of a new programme for the whole footwear value chain has not materialized. The origins of one of the most successful recent collective upgrading strategies in the cluster are in the failed attempts at the ‘Calçado do Brasil’ programme. Two of the trade associations involved at that time, ASSINTECAL and ABRAMEQ, prepared a new project aimed at creating a collective brand called ‘BY BRASIL’ and obtained funding from the Brazilian Federal Government through APEX (Brazilian export promotion programme). The agreement was signed in December 1999. According to this agreement, the Federal Agency would invest 5 million reais (US$ 2.8 million13) and the two local associations would invest the same amount for the period up to 2002. In May 2000 the programme had reached 90 firms, most of whom were from the Sinos Valley. There is also a generational renewal of leadership in most of the cluster associations. The generation identified with the quasi-hierarchical paradigm left when it was no longer necessary to control the trade associations. Once the conservative elements retired, input suppliers and machinery producers led the way for technological innovations and launched the cluster brand (‘BY BRASIL’) abroad. Due to their extensive experience in diversifying markets and customers (that is working for several value chains simultaneously), these two sectors have been at the forefront of the ‘underground revolution’, and far ahead of the majority of footwear producers. In the words of an ASSINTECAL official: ‘We want to be strong input suppliers within a strong Sinos Valley footwear industry. There is no point in us being strong and having a local footwear sector that is weak. But in order for us to be strong, we have to diversify and sell abroad.’ Other associations are working hard to establish their own APEX project. Abicalçados renewed leadership followed the example of the input suppliers and developed a new programme to obtain APEX support. This time, the project was submitted by Abicalçados and approved by the federal agency who agreed to invest 6 million reais (US$ 3.1 million) while Abicalçados invested 12 million reais (US$ 6.2 million). The project seeks to help Brazilian shoe producers explore new export markets by giving them specific training, supporting firms to participate in international fairs and showrooms and improving firms’ capabilities in areas such as technological development, quality assurance, productivity and design (interview with an Abicalçados official, 2000). At the time of our survey the project was at an early stage. Another collective initiative comes from the group ‘Shoes from Brazil’.

The underground revolution in the Sinos Valley


It comprises a core of 11 traditionally domestic-oriented branded firms that are attempting to penetrate markets as distant as the Arab Emirates. This group of highly successful firms has increased their traditional presence in GDS (the most important trade fair for the European footwear industry, taking place twice a year in Germany), their presence at other fairs like the MICAM in Milan and also the Al Hida’a in the Arab Emirates.14 At the time of this survey, the group had joined forces with a São Paulobased trade fair organizer (COUROMODA) and begun participating as exhibitors in several trade fairs worldwide. These initiatives are largely selffinanced by firms that have decided to build their own private network. Another group of eight firms has joined a consortium to maintain a showroom in New York. Inspired by these groups, the local business association (ACI) is also developing its own APEX project aimed at helping 60 small and medium-sized domestic-oriented producers to participate in a similar showroom and, thus, enter the external market with their own brand name. The strategy in this case is to take advantage of a market niche that is concentrated in higher quality fashion shoes sold by select boutiques in the US. As for governmental efforts to foster local private upgrading initiatives, Schmitz (1998) found they had been largely absent in the 1980s and 1990s. Again, this had changed by 2000. As well as the APEX initiatives funded by the federal government, the state government has provided direct support for small-sized firms. This support was structured in a programme called ‘Programme for the leather and footwear Value Chain of Rio Grande do Sul’. The programme comprises five major areas – improving quality of products and processes, commercial promotion, support to collective action as consortiums and cooperatives, subsidized credit and fiscal incentives. The industrial extension centre is one of the main programme activities that help firms with process upgrading. There is also a Centre for Design and Innovation, consortia for collective purchasing of raw materials and other inputs, export consortia and institutional (and financial) support for participation in trade fairs. According to local producers and institutions, the most successful activity of the state government’s programme was support to participate in trade fairs. In 1999, with the joint support of the ‘BY BRASIL’ programme, 70 small and medium-sized input supplier firms participated in the Hong Kong input suppliers fair. The Brazilian collective stand in this important Asian fair was a modest first step, but it encouraged new business for Brazilian smaller input suppliers from that year onwards. In short, changes at the top of the dominant value chain in the cluster triggered a series of reactions that loosened the ties down the chain. Combined with the determination to upgrade by a new generation of association and


Local enterprises in the global economy

business leaders, it produced a series of initiatives aimed at repositioning the Sinos Valley cluster in the world market.



The recent value chain literature suggests that chain governance structures the upgrading possibilities of local firms and clusters. This chapter has examined this proposition by comparing the type of upgrading pursued and achieved by local firms supplying value chains that are governed in different ways. It also examines the interaction between global chain and local cluster governance, in order to identify the scope for local upgrading initiatives. Our results show that the US value chain embodies the concept of a quasi-hierarchy, while Europe constitutes a more flexible version. The domestic market is characterized by market-based relationships and the Latin American market is a hybrid, that is quasi-hierarchical for unbranded producers using export agents and market-based for branded producers. Footwear producers operating in market-based value chains are more likely to achieve functional upgrading than producers in the tightly coordinated US quasi-hierarchy. Softer quasi-hierarchies such as the European value chain seem to be slightly less obtrusive for functional upgrading. Product and process upgrading are actively encouraged and supported by quasihierarchical buyers. Domestic-oriented firms pursued product and process upgrading strategies as well, but these were usually linked to their investments in functional upgrading. The recent significant advances in functional upgrading in export markets have two explanations. First, prior ‘apprenticeship’ in the national market and, second, operating in several chains simultaneously. This mix of opportunities, combined with the producer’s determination to upgrade appear to be the main factors explaining why some firms were able to progress into higher value added activities (functional upgrading). A good example of the advantages granted by this strategy of ‘leveraging competencies’, appears to be the head start input suppliers took in diversifying markets and customers, which increased their profile in international markets and made them less vulnerable. At the cluster level, the emergence of a series of initiatives for collective upgrading seems to suggest that the loosening of vertical ties within global chains (due to increasingly volatile global conditions and local ‘generalized competence’) may have opened a window of opportunity for the repositioning of the cluster in world markets. Therefore, this chapter’s main contributions are fourfold. First, the

The underground revolution in the Sinos Valley


systematic and detailed comparison and analysis of the four main value chains present in the Sinos Valley, their governance patterns and the corresponding upgrading patterns. Second, highlighting the importance of learning in the national market, particularly for functional upgrading. Third, the analysis of the effect that working in several value chains simultaneously has for upgrading opportunities at the firm level. Finally, a more optimistic approach to the upgrading possibilities for local clusters working with a dominant quasi-hierarchical value chain is presented. In conclusion, the value chain perspective has proved valuable to explain how global events influence local individual and collective upgrading efforts within the Sinos Valley cluster.




4. 5. 6.

7. 8.

In the Sinos Valley 53.9 per cent are domestic-oriented, 11.34 per cent are external oriented and 34.6 per cent are both domestically and externally oriented. These proportions among sampled firms are respectively 27.5 per cent, 40.0 per cent and 32.0 per cent. In the case of input suppliers, proportions were not observed, as most firms are domestically oriented. According to the criteria used for sampling firms, an error of six percentage points was admitted. These numbers are open to dispute because there is a high turnover of firms (especially due to the footwear crises in the 1990s) and also because some sources include (or ignore) the small workshops (locally called Ateliers) that are so important for the functioning of the cluster. ACI data does not include most of the workshops, but according to other statistics from the state government there are more than 1400 footwear firms in the cluster with less than 20 employees (MTB/RAIS, 1999). This ‘convening power’ undermines the authority that local producers are supposed to have over their own supply chain. When selling to the quasi-hierarchies, input suppliers reported that they must meet global buyers’ demands instead of the local producers’. Often this means lowering prices to meet the stringent low-cost strategies of the global buyers. Input suppliers have sought alternative markets and clients as a response (Bazan and Navas-Alemán, 2001). For more detailed figures on buyer and sales concentration, see Bazan and NavasAlemán, 2001. Innovative modelling is a term that aims to capture the added value an experienced producer can provide by translating original designs into technical specifications with minimal (if any) support from the buyer. According to our findings, among the 24 firms that sell for the domestic market, 91.6 per cent use sales representatives (representante comercial). Sales representatives are also used by 14 per cent of the 15 firms that sell to Latin America. In other regions, such as Europe and the US, direct sales to retailers (less common) are made either through the client’s office in Sinos Valley (17.6 per cent of sales from 17 firms that sell to Europe and 4.5 per cent of the 22 that sell to the US). Two domestic producers that sell to Europe and two that sell to the US do so through their own sales offices abroad (authors’ survey, 2000). For domestic-oriented firms, we included as ‘new markets’ other states within Brazil where the producer had started selling his/her footwear. For different critiques of the merits of this paradigm (which escapes the scope of this chapter) please see Clerides et al., 1998; Tewari, 1999; Tybout and Westbrook, 1995 and Rodrik, 1999.

138 9.

10. 11. 12.

13. 14.

Local enterprises in the global economy It is worth mentioning that the quality of Chinese footwear has been improving quite rapidly, in some cases due to the technical support of Brazilian technicians hired by US buyers to teach the Chinese workforce how to make shoes. At the time of this survey, it was common to find several ads in the local Sinos Valley press offering good salaries in US dollars for technicians wanting to relocate for one or two years in China. For a more detailed discussion on chain governance and input suppliers’ upgrading patterns in the Sinos Valley, see Bazan and Navas-Alemán, 2001. Elements of product and process upgrading are also present in these programmes but we have discussed elsewhere that the cluster is already considered to be a ‘top of the range’ manufacturer, so this aspect is not so remarkable. Perhaps the failed programme depicted by Schmitz (1998), ‘Calçado do Brasil’ was the closest the cluster has ever been to having an inclusive programme for the whole footwear value chain; it was designed by five local trade associations (footwear manufacturers, tanners, input suppliers, machine producers and export agents). Exchange rate: Rs 1.92⫽1 US$ in December 1999. By the year 2000, Saudi Arabia alone had consumed US$ 600 million in footwear (Shoeinfonet, 2000)

REFERENCES Abicalçados (2002), Resenhas Estatísticas 2001, 2002. Associação Brasileira das Industrias de Calçados, Novo Hamburgo. www.abicalcados.com.br (12/03/03). ACI (2000), Database of footwear firms in Rio Grande do Sul, CD-ROM from the Statistics department. Commercial and Industrial Association of Novo Hamburgo. Bazan, L. (1997), ‘Etnia, cooperação e conflito: mediações da identidade nas relações industriais – um estudo de caso no setor calçadista do Vale dos Sinos’, unpublished Masters Dissertation, Department of Sociology, Federal University of Rio Grande do Sul, Porto Alegre. Bazan, L. and L. Navas-Alemán (2001), ‘Comparing chain governance and upgrading patterns in the Sinos Valley, Brazil’, unpublished paper presented at Workshop on Local upgrading in global chains’, Brighton: Institute of Development Studies, February, www.ids.ac.uk/ids/global/vw.html. Bazan, L. and H. Schmitz (1997), ‘Social capital and export growth: an industrial community in Southern Brazil’, IDS Discussion Paper 361, Brighton: Institute of Development Studies. Clerides, S., K. Lach and J. Tybout (1998), ‘Is Learning-by-Exporting Important? Micro-dynamic evidence from Colombia, Mexico and Morocco’, Quarterly Journal of Economics, vol. CXIII, August, 903–47. Gereffi, G. (1994), ‘The organisation of buyer-driven global commodity chains: how US retailers shape overseas production networks’, G. Gereffi and M. Korzeniewicz (eds), Commodity Chains and Global Capitalism, Westport, CT: Praeger. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Humphrey, J. and H. Schmitz (2000), ‘Governance and upgrading: linking industrial cluster and global value chain’, IDS Working Paper 120, Brighton: Institute of Development Studies. Kishimoto, C. (2002), ‘The Taiwanese personal computer clusters: trajectory of its

The underground revolution in the Sinos Valley


production and knowledge systems’, unpublished D.Phil. thesis, Brighton: Institute of Development Studies. Klein, E. (1991), La Cadena de Distribución y la Competitividad de las Exportaciones Latinoamericanas: Las Exportaciones de Calzado del Brasil, LC/G1669, CEPAL, Naciones Unidas, Santiago de Chile. Knorringa, P. and H. Schmitz (2000). ‘Learning from global buyers’, Journal of Development Studies, 37 (2), 177–205. Lee, J.-R. and J.-S. Chen (2000), ‘Dynamic synergy creation with multiple business activities: toward a competence-based growth model for contract manufacturers’, in R. Sanchez and A. Heene (eds), Theory Development for Competencebased Management, Advances in Applied Business Strategy, Stanford, CT: JAI Press, pp. 209–28. MTB/RAIS (1999), Bases estatísticas Relação anual de informações sociais’, CDROM, Ministério do Trabalho e Emprego, Brasil. Nadvi, K. (1997), ‘The cutting edge: collective efficiency and international competitiveness in Pakistan’, IDS Discussion Paper 360, Brighton: Institute of Development Studies. Nadvi, K. and H. Schmitz (eds) (1999), ‘Industrial clusters in developing countries’, Special Issue of World Development, 27 (9). Rabellotti, R. (1997), External Economies and Cooperation in Industrial Districts: A Comparison of Italy and Mexico, London: Macmillan. Rodrik, D. (1999), ‘The new global economy and developing countries: making openness work’, Policy Essay No. 24, Overseas Development Council, UK. Ruas, R. (1995), ‘O Conceito de Cluster e as Relações Interfirmas no Complexo Calçadista do Rio Grande do Sul’, in J. Fensterseifer (ed.), O Complexo Calçadista em Perspectiva: Tecnologia e Competitividade, Porto Alegre: Ortiz. Schmitz, H. (1995), ‘Small shoemakers and Fordist giants: tale of a supercluster’, World Development, 23, 9–28. Schmitz, H. (1998), ‘Responding to global competitive pressure: local co-operation and upgrading in Sinos Valley, Brazil’, IDS Working Paper 82, Brighton, Institute of Development Studies. Schmitz, H. (1999), ‘Global competition and local co-operation: success and failure in the Sinos Valley, Brazil’, World Development, 27 (9), 1627–50. Shoeinfonet (2000), ‘The footwear business in Brazil’, interview with Couromoda published by www.shoeinfonet.com (06/09/ 2000). Tewari, M. (1999), ‘Successful adjustment in Indian industry: the case of Ludhiana’s woolen knitwear cluster’, World Development, 27 (9), 1651–72. Tybout, J. and M.D. Westbrook (1995), ‘Trade liberalisation and dimensions of efficiency change in Mexican manufacturing industries’, Journal of International Economics, 31 (August) 53–78.

6. How globalization affects Italian industrial districts: the case of Brenta Roberta Rabellotti 1.


The turn of the century coincides with a divide for the Italian shoe industry and more generally for the fashion industry. Due to increasing globalization, major changes are taking place: international relocation of production, increasing concentration in distribution, mergers of fashion firms and acquisitions of additional brands and the formation of financial multi-product giants (Saviolo and Testa, 2000). Nothing remains the same, not even in a long established industrial district such as Riviera del Brenta, a highly specialized and strongly export oriented footwear cluster, located near Venice. This chapter is concerned with the impact of these global transformations on local competitiveness in Brenta, one of the most important Italian footwear districts. The aim is to integrate the typical industrial district approach,1 traditionally focused on analysing the local sources of competitive advantages, with the global value chain approach which stresses that activities such as design, production and distribution are often located in different regions or countries (Gereffi, 1999; Kaplinsky, 1998). In the typical industrial district most of the activities along the value chain have traditionally been undertaken locally, and competitiveness of producers has mainly come from intra-cluster vertical and horizontal relationships generating collective efficiency, namely increasing returns from incidental economies of agglomeration and active cooperation (Rabellotti, 1997; Schmitz, 1995 and 1999). Recent changes in production systems, distribution channels and financial markets, accelerated by the globalization of product markets and the spread of information technologies, suggest that more attention needs to be paid to external linkages.2 Furthermore, industrial boundaries are blurring and the shape of industries is no longer conforming to the standard industrial classification 140

How globalization affects Italian industrial districts


(Mytelka, 2000). This is the case of the footwear sector, analysed in this chapter, which is increasingly integrated in the fashion industry, dominated by a few multi-product oligopolies, and exploiting economies of scale and scope in activities such as distribution, marketing and branding across traditionally separated industrial sectors such as shoes, clothing, glasses, perfumes and leather accessories. Accordingly, to understand the effect of these changes on a cluster such as Brenta it becomes necessary to adopt an analysis which pays attention to linkages with actors external to the district. This study of the effects of globalization on Brenta tries to fill this gap in the literature on industrial clusters using some of the insights of the global value chain approach. By doing so it takes into account activities operating outside the cluster and in particular it highlights the significance of the relationships with key external actors.3 An important aspect stressed in the literature on value chains is that the various activities in the chain are subject to some degree of governance or coordination (Gereffi, 1994). This coordination may take place through arm’s-length market relations or non-market relationships. In the latter case, following Humphrey and Schmitz (2000), we distinguish between three types of governance: (i) the network implying cooperation between firms of more or less equal power which share their competencies within the chain; (ii) a quasi-hierarchy involving relationships between legally independent firms in which one is subordinated to the other and where the leader in the chain defines the rules that the rest of the actors have to comply with; and, (iii) hierarchy governance when the local firm is owned by an external firm. Among the different forms of governance, the literature on value chains, which is mainly concerned with developing countries, stresses the importance of the quasi-hierarchy type, distinguishing between those cases when coordination is undertaken by buyers (‘buyer-driven chains’) and those in which producers play the key role (‘producer-driven chains’) (Gereffi, 1994). Moreover, several authors conclude that the increasing concentration of retailing in developed countries makes buyer-driven chains a growing phenomenon (Dolan and Humphrey, 2000; Gereffi, 1999). In this chapter we aim to contribute to this debate by broadening the analysis to include developed countries, showing that quasi-hierarchy may also be a relevant form of governance in linkages existing between producers from one of the main shoe clusters in Italy and the lead fashion firms governing the chain. The following questions are addressed: Is globalization pushing Brenta towards new value chains? What types of governance characterize the relationships between local and outside actors? Do the chain leaders come from inside or outside the districts? Does the integration of industrial clusters in global value chains enhance or weaken local upgrading strategies?


Local enterprises in the global economy

The concept of upgrading is used here in the sense proposed by Humphrey and Schmitz (2000). Process upgrading means transforming inputs into outputs more efficiently by reorganizing the production system or introducing superior technology; product upgrading can be defined as moving into more sophisticated product lines; and functional upgrading is acquiring new, superior functions in the chain, such as design or marketing. Furthermore, from our empirical analysis in Brenta a different possible form of functional upgrading coming to the fore is the externalization of low value added functions combined with a focus on more advanced activities or higher value added segments of the market. This chapter is based on primary data collected in the Riviera del Brenta shoe cluster in summer 2000. Forty shoe manufacturing enterprises were interviewed, randomly selected from a list provided by ACRIB (Associazione Calzaturifici della Riviera del Brenta), the local entrepreneurial association of footwear producers.4 Qualitative information was also collected through in-depth interviews with trade organizations, fashion firms, retailers and other key informants. Furthermore, in September 2000 during the international shoe fair in Dusseldorf we interviewed six buyers belonging to some of the main European buying groups. To understand the changes confronting Brenta, the findings emerging from the information collected in 2000 were compared with the findings of a previous study we carried out in the early 1990s, which gives a detailed picture of the district at that time (Rabellotti, 1997). Here, we disaggregate the cluster by value chains and analyse the two main ones. One chain is dominated by firms which own the top brands and operate in the luxury market. The other is dominated by groups aggregating the majority of independent footwear stores in Germany. Compared with our previous work on Brenta (Rabellotti, 1997) we are shifting the focus from the internal to the external linkages of the cluster. In the process we are trying to close a gap in the literature, which traditionally privileged the study of local relationships.5 Nevertheless, we agree with Schmitz (2000) that the challenge is not to shift from one to the other but rather to understand the interaction of internal and external linkages. Therefore, the main aim of this chapter is to investigate Brenta’s problems and opportunities arising from globalization, analyse external linkages, local relationships and how they interact. After an overview on Brenta and its main markets, the chapter concentrates on two key chains. First, in the top brand chain, which has become increasingly important, we analyse the recent entry of some global fashion leaders into the footwear sector. A number of fashion companies originating in other sectors have penetrated the footwear industry looking for skilled manufacturers in order to outsource the production of shoes sold

How globalization affects Italian industrial districts


with their brand names. This is why many producers in Brenta have begun to work as subcontractors, abandoning key activities such as design and sales, which are the core competencies of fashion companies. The latter have become the new lead firms of the chain. In the second chain analysed, high fashion and top brands play less of a role but high quality is important. We call it the German chain because Germany and neighbouring countries constitute the main market. A key feature of this market is that independent retailers are organized in large, powerful buying groups. These groups are network organizations supplying credit and information to their members, helping them to reduce transaction costs and risks. The chapter then analyses how global linkages have affected local linkages, showing that the latter have weakened in recent years. In backward linkages the weakening is due to the increasing outsourcing to Romania and other nearby low-wage countries. In forward linkages it is due to the forced abandonment of activities such as design and marketing, notably in the top brand value chain. The final section draws together the main conclusions and reflects on how Brenta’s future evolutionary path is influenced by its recent redefined participation in global chains.



In Brenta the origins of the footwear industry date back to the beginning of last century. During the footwear industry boom after World War II, the sector progressively absorbed most of the rural workforce available in the area. In the 1960s, the local enterprises expanded and increased their exports, specializing in the upper segments of the market. In 2000, 88 per cent of the shoes produced in the area were medium-high and high priced women’s shoes with an average ex-factory price of 58 euros. Since the second half of the 1980s, the area has suffered from increasing competition in the international market and sales have stagnated, fluctuating between 7.9 and 8.8 million pairs (mainly due to exchange rate fluctuations). In value terms, however, sales continued to increase in most years (Figure 6.1).6 It may be useful to add that Brenta’s performance is in line with that of the rest of the Italian footwear industry, which suffered from stagnating European demand and from increasing international competition (ANCI, 2001). In terms of the structure of the district,7 there were 968 firms in 2000. These comprised 430 shoe producers, 396 firms manufacturing inputs such as heels, soles, lasts, 67 design firms and 75 trading companies (Table 6.1).


1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

500 450 400 350 300 250 200 150 100 50 0

Local enterprises in the global economy

Source: ACRIB (various years)

Figure 6.1

Total sales in Brenta (1995 prices, euro millions)

The number of firms has decreased continuously from the second half of the 1980s as has the number of employees (see note 6). Shoe factories in Brenta are characterized by their small size. According to the latest available industrial census (Istat, 1996) 78 per cent of the firms employing 40 per cent of the workers have fewer than 20 employees, and 18 per cent, corresponding to 34 per cent of total employment in the shoe industry, have between 20 and 49 employees. The average size is 15 employees. The total value of shoe production is about 762 million euros, close to 10 per cent of the total turnover of the Italian footwear industry. Over the years, Brenta’s orientation towards the external market increased from about 70 per cent of sales exported at the beginning of the 1980s to 88 per cent in 2000. This represents more than 10 per cent of total Italian exports in the sector (ACRIB, 2001).



The majority of firms in Brenta are small. The size distribution of our sample (40 firms) reflects this situation. Seven firms (17.5 per cent of the total sample) have fewer than 20 employees, 20 firms (50 per cent) have between 20 and 49 workers, a further nine employ between 50 and 99 people and only four companies (10 per cent) have more than 99 workers (Table 6.2).8 The small size of sample firms is not related to age because, although 20 companies were established before 1970, ten of them still have fewer than


How globalization affects Italian industrial districts

Table 6.1

The district in 2000 No. of firms

No. of employees

430 396 67 75 968

8310 3054 200 259 11823

Footwear companies Accessory manufacturing companies Shoe designers Trading companies Total Source: ACRIB (2000)

Table 6.2

Size distribution of the sample

No. of employees ⬍20 20–49 50–99 ⬎99 Total

No. of firms


7 20 9 4 40

17.5 50.0 22.5 10.0 100.0

Source: Author’s survey

50 employees. Furthermore, the age distribution of the sample reflects the evolution of the district, with only four companies created during the last decade. Like other Italian shoe clusters, Brenta is therefore characterized by a predominance of firms that in the past often chose to remain small, subcontracting out specialized phases of production, sometimes for fiscal reasons. Many of the sample firms are outsourcing some of their production and this means that their capability is far greater than could be expected from their size. Particularly, as seen later, in recent years Brenta’s firms have begun to outsource to Romania and other neighbouring countries, pressured by the need to reduce costs and by the increasing scarcity of local skilled labour. Firms in Brenta do not like to talk about this process of external relocation because they are worried about the possible negative impact on their image as top quality producers and, indeed, there are signs of decreasing quality. Nevertheless, data about the number of pairs per worker clearly show that in several cases part of the production is not carried out inside the firm. For example, in our sample among firms in the category of 20–49 employees, six produce less than 50000 pairs per year and another six produce


Local enterprises in the global economy

between 50000 and 100000, three between 100000 and 500000 and one firm produces more than 500 000 pairs.9 Thus, in our sample the number of pairs per capita produced yearly varies enormously, between a minimum of 750 and a maximum of 20000. Moreover, as expected there is a strong (and statistically significant) negative correlation (–0.595) between the number of pairs per worker and the price of shoes.10 Nonetheless, the small firm size has recently become an obstacle to investments in marketing, branding and distribution, which are increasingly becoming crucial activities in the global footwear market. This will be explored in more detail below. Output of the sample firms seems to have remained relatively stable. According to 45 per cent of the enterprises, the number of shoes produced and the value of total sales over the years 1995–2000 had remained more or less unchanged and the export share is stable in 70 per cent of the sample. With regard to profits, the sample distribution is more varied: 30 per cent of the firms revealed stable profits during 1995–2000, 45 per cent decreasing profits and the remaining 25 per cent reported that their profits had increased (Table 6.3). Table 6.3

Performance indicators (% of sample firms)





35.0 40.0 22.5 25.0 70.0 52.5 30.0

45.0 45.0 70.0 30.0 27.5 42.5 55.0

20.0 15.0 7.5 45.0 2.5 5.0 15.0

Number of pairs Value of sales Exports Profits Average price Quality of product Number of employees Source: Author’s survey

Finally, in order to investigate upgrading strategies, we explored process and product innovations introduced by sample firms in 1995–2000. As regards process, 52.5 per cent of the sample firms introduced many innovations and 42.5 per cent some innovations (5 per cent do not innovate at all) (Table 6.4). In most cases (90 per cent of the sample firms), process innovation consisted mainly of introducing new, more advanced machinery, while in 45 per cent of the sample there was also some reorganization of the production process and/or management system. In 37 per cent of the sample there was also the introduction of quality control systems which in most cases means employing a technician in charge of this task.11 Forty per cent of the sample introduced many product innovations and


How globalization affects Italian industrial districts

Table 6.4

Process and product innovation (number of firms)

No innovation Little innovation A lot of innovation Total

Process innovation

% of sample firms

Product innovation

% of sample firms

2 17 21 40

5.0 42.5 52.5 100.0

3 21 16 40

7.5 52.5 40.0 100.0

Source: Author’s survey

52.5 per cent only some innovations (Table 6.4), namely improvement of quality (67 per cent), introduction of new product lines (60 per cent) and export towards new geographical areas (50 per cent). Less frequently, sample firms have entered into a new segment of the market (22 per cent of the sample) or introduced a new brand (27 per cent). After this first general presentation of the sample firms, the next sections analyse how and where they sell their products, identifying the main global value chains, discussing the issue of governance in each of them and differentiating enterprises belonging to diverse chains according to upgrading and performance.



According to some estimates (Pambianco, 2000), the value of the world footwear market is around US$ 47 billion, 6 per cent of which is in the upper segment of the market, 9 per cent in the medium–high, 25 per cent in the medium and the remaining 60 per cent in the low-price segment. Therefore, the high and medium–high segments are worth about US$ 7 billion, 33 per cent of which corresponds to the market share of Italy (Table 6.5). In Riviera del Brenta the main market has traditionally been Europe, particularly Germany, and to a more limited extent France, the UK and the rest of the EU. Among sample firms, 35 per cent sell between 50 and 90 per cent and 17.5 per cent more than 90 per cent of their production to Europe. Sales to Italy are less than 10 per cent for the majority of the sample (53 per cent of firms) and less than 50 per cent for another 23 per cent of firms. In the European market, Brenta’s companies are selling to a variety of customers. In the UK they typically sell to large buyers or department


Table 6.5

Local enterprises in the global economy

The world market for footwear (US$ billion)

Market segments High price Medium-high price Medium price Low price Total

World market


Italian production


% Italian production in world market

3 4 12 28 47

6.0 9.0 25.0 60.0 100.0

1.0 1.3 2.8 1.9 7.0

14.0 19.0 40.0 27.0 100.0

33.0 32.5 23.0 7.0 15.0

Source: Pambianco estimates available at www.pambianco.com

stores, in France and Italy they supply mainly to independent retailers and in Germany to buying groups (see section 7). Regarding the rest of the world, Brenta’s market penetration is more difficult due to the small size of local enterprises, geographical distances and the large investments involved. Only nine firms (23 per cent of the sample) export between 10 and 50 per cent of their production to the USA and 12 (30 per cent) to other countries, mainly in the Far East, Russia and the Middle East. It should be added that exports to the USA have increased considerably during the last two years, boosted by the weakness of the euro.12 Nevertheless, based on our empirical findings and confirmed by key informers such as some members of the US National Shoe Retailers Association, Brenta’s penetration in the US market is still very much dependent on volatile events, such as a favourable exchange rate. The strong potential of the largest market in the world is not thoroughly exploited by Brenta’s enterprises. In our sample, only seven firms have some form of relationship with US buyers and none of them are trading a significant share of production through this channel of distribution. The independent shoe retailers represent a more promising market.13 Hoping to increase their market share principally in this segment, some Brenta enterprises have recently opened a collective showroom in New York (see section 9). However, small Brenta firms are running into difficulties in their recent attempt to build up strong and stable relationships with their US customers. This is because there is a linguistic barrier (about which US retailers complain strongly) and the opportunities of direct interaction are rare. On the other hand, US independent retailers, interviewed during an organized business trip to Brenta, expressed both their satisfaction with the high quality of Riviera’s shoes and their concerns that they would have to change suppliers if prices went up, in other words, if the dollar–euro exchange rate became less favourable.

How globalization affects Italian industrial districts


Brenta’s presence in the Far East, the Middle East and Russia is even more volatile, and sales mainly occur through export agents. A potentially important market is Japan; however, protectionism still represents an obstacle to further development.14 Furthermore, the demand from other Asian countries, such as Hong Kong, is slowly recovering after the financial crisis in 1997. Russia also remains a very unpredictable market. Overall, none of the sample firms maintain very stable relationships with customers in these distant markets; many firms see them as opportunities to diversify their sales but none of them systematically invest in marketing. Thus it appears that sample firms are able to maintain many exit options, selling in many different geographical markets and to a large variety of customers, such as independent retailers, department stores and large buyers. The only exceptions are eight firms selling more than 50 per cent of their production to large buyers in the European market (seven in Europe and one in Italy). The ability to combine many customers is also confirmed by the limited amount of production sold by each firm to its two main customers (to whom on average sample firms sell only 16 per cent of their production).15 The above analysis of the main markets of Brenta has neglected a category of increasingly important customers for the district – the high fashion companies. Many firms in Brenta have recently begun to work as subcontractors to some leading global fashion firms. The next section analyses this value chain characterized by high fashion and global reach.



The world market in luxury goods is estimated at about 46 billion euros in 2001, 60 per cent of which corresponds to clothing and 8 per cent to footwear. In this rich and rapidly increasing market, the Italian market share is 30 per cent, followed by France with 25 per cent. In recent years, the luxury fashion system went through important changes, turning it into an oligopoly dominated by a few multi-product giants. The growth strategy of many companies has been characterized by a similar pattern. First, successful firms established their brand names in specific product lines (for example three among the most important and largest companies in the industry, LVMH, Gucci and Prada began producing and selling leather goods). Second, they capitalized on their brand names and diversified to other segments (in the cases named above, they entered into clothing, footwear, glasses, perfumes, wines). Finally, they have also begun to grow through the acquisitions of other well known existing brands.16 During the last five years, mergers and acquisitions in the fashion


Local enterprises in the global economy

industry have increased from 31 in 1997 to 155 in 2001. In the shoe sector, there was a total number of 15 operations (Pambianco, 2002). The economic logic behind these growth strategies is a search for scale and scope economies in activities other than manufacturing, such as branding, marketing, advertising, opening of mono-brand shops in the most exclusive and expensive streets in the world. The increasing concentration of fashion enterprises in the intangible phases of the value chain may be explained by the growing concentration of rents in these activities. According to Kaplinsky (2000), in the past decade the barriers to entry in manufacturing have begun to fall and consequently the rent going to production activities has shrunk in favour of rents accruing to activities outside the area of production. This explains the growing concentration of investments in areas like branding, advertising, marketing and sales where capital cost barriers to entry are high. Creating and maintaining a global brand is very expensive and only those with access to international financial markets are able to afford it. Therefore, leading firms are expanding through mergers and acquisitions, capitalizing on their core competencies such as design, advertising, marketing and brand naming, which are no longer sector specific. This implies that we can consider the footwear or clothing industries as subsystems of the global fashion system. The rapid growth of the leading enterprises in the luxury industry is confirmed by the following example. The second largest Italian firm in the top brand industry, Prada, has augmented the value of sales from about 500 million euros in 1997 to 1.6 billion in 2000. This is an increase of more than 100 per cent in three years (Table 6.6). The net profits gained in the industry are also quite remarkable. In Prada they reached 27 per cent of the value of sales, 16 per cent in Gucci and 10 per cent in LMVH. The top brand value chain can be regarded as a sub-type of the ‘buyerTable 6.6 The value of sales of some top luxury-fashion companies operating in the footwear industry in 2000

LVMH (France) Gucci (Italy) Prada (Italy) Tod’s (Italy)

Value of sales (Euro million)

Share of world market %

11 581 25031 16231 1 2521

25.2* 5.4 3.5 0.5

Note: *The total value of sales of LVMH include wine, champagne and spirit sales. Source: Pambianco estimates available at www.pambianco.com

How globalization affects Italian industrial districts


driven’ chain because the lead firms are the owners of top global brands, controlling activities connected with intangible characteristics of the products such as design, brand name, marketing and distribution. Gereffi’s work is more focused on ‘buyer-driven’ chains ‘in which large retailers, branded marketers, and branded manufacturers play the pivotal roles in setting up decentralised production networks in a variety of exporting countries, typically located in the Third World’ (Gereffi, 1999, 41–2). Although he discusses the importance of the creation of brands, he is more concerned with global brands in mass markets, such as Liz Claiborne, Nike and Reebok. In the luxury market, barriers to entry are supposed to be higher and returns from branding and marketing very high. Therefore, following Kaplinsky (1998 and 2000), economic rents in this chain are assumed to be higher than in other types of chains. The next section presents one of our key empirical findings. It shows that, in the top brand chain, developed country producers (and not only developing countries as in many Gereffi studies) are affected by increasing buyer concentration. It follows the analysis of the rapid expansion of the top brand chain in Brenta, investigating the uneven relationship between local footwear producers and top global brand leaders.



Recently, the high and medium–high segments of the footwear industry have increasingly attracted the interest and the financial capital of wellknown top brands and luxury multi-product oligopolies from outside the shoe world.17 Some world-class luxury companies, looking for highly skilled manufacturing capabilities to begin footwear production, identified the Riviera del Brenta as a preferred area in which to find subcontractors. The beginning of this trend corresponded with a difficult time in Brenta because local firms were facing the end of the positive impact on exports of the 1992 devaluation of the lira. Among our sample firms, 17 enterprises (corresponding to 42.5 per cent of the sample) work as subcontractors to high fashion companies, producing shoes sold with globally known top brands. In five cases (12.5 per cent), they work exclusively as subcontractors, while four of them (10 per cent) make between 50 and 89 per cent of their total production for high fashion companies and the remaining eight (20 per cent of the sample) make less than 50 per cent (Table 6.7).18 According to our interviews, it appears that in many cases the amount of production made as subcontractors is still rather changeable, depending on seasons. Furthermore, some firms do not have direct contact with high fashion companies,


Table 6.7

Local enterprises in the global economy

Production for high fashion companies among sample firms (%)

% of total production 0 1–49 50–89 ⱖ90 Total

No. of enterprises

% of total sample

23 8 4 5 40

57.5 20.0 10.0 12.5 100.0

Source: Author’s survey

Table 6.8 Internal functions in sample firms working as subcontractors to high fashion companies No. of sample firms* Design Product development Purchase of components Sale

Not undertaken

Partially undertaken

Totally internal

11 (65%) 2 (12%) 1 (6%)0 14 (82%)

4 (23%) 2 (12%) 3 (18%) 2 (12%)

2 (12 %) 13 (76%) 13 (76%) 1 (6%)0

Note: *The total number of sample firms working as subcontractors is 17. In parentheses is the ratio on the total number of subcontracting enterprises in the sample. Source: Author’s survey

instead they are subcontractors to other local enterprises who are the direct subcontractors. In most of the cases investigated, fashion companies provide the design and Brenta manufacturers take care of all production phases, including product development19 and purchase of raw materials and components. After that, shoes are sold by fashion companies with their brand names. In our sample, design is totally controlled by fashion companies in 65 per cent of the firms working as subcontractors, while 35 per cent of them contribute to design. This often means that fashion companies give producers some ideas and sketches to be transformed into a shoe.20 Sales are undertaken by fashion companies in 82 per cent of the cases (Table 6.8), while product development and purchase of inputs are carried out in most of the cases by the subcontracting firms. Nevertheless, according to many firms interviewed, fashion companies are increasingly becoming involved in these activities by directly selecting suppliers, sometimes even through acquisi-

How globalization affects Italian industrial districts


tions of firms, and extending their control on quality and delivery conditions backwards along the chain. From what has been said so far, it appears that Brenta has been undergoing a process of functional downgrading. Traditionally the design and acquisition of inputs were controlled locally, and carried out inside the firms or inside the district. More recently, with the advent of the luxury fashion companies, local enterprises are moving out of design and sale. There are also signs of luxury fashion companies extending their control backwards along the chain. Integration in the luxury fashion value chain is thus causing a process of functional downgrading at district level in those activities that are the typical core cross-sector competencies of luxury fashion companies, namely design, branding and sales. Nevertheless, although Brenta is showing a trajectory of functional downgrading resulting from the integration in top brand chains, it has to be underlined that many of the leading companies in those chains are Italian. Therefore, by moving from a narrow district perspective to take into consideration the evolution of the Italian fashion system as a whole, our conclusions may be very different. The Italian luxury goods industry is definitely undergoing a process of functional upgrading and concentration in rent-rich activities by exploiting its cross-sector core competencies in design, branding and marketing. This is very different from what is often occurring in developing countries. Small Brenta producers are abandoning some key activities which have moved to the headquarters of the chains’ leaders in Milan. In the case of producers taking part in global chains in the developing world, these activities are never carried out within the country, instead they are fixed in New York, London or other cities in the developed world. Apart from design, the functional downgrading that is occurring in Brenta concerns activities in which local enterprises, probably including the majority of small Italian footwear firms, have traditionally been rather weak, that is branding, marketing and sales strategy. In our sample, 60 per cent of firms do not perform any marketing at all and the existing brand names are sometimes recognized at national level (mainly in Germany) but never globally. A reason for the very limited local investments in these activities is the average company size. Firms are too small to afford very expensive strategies in marketing or advertising and to impose a brand name in the global market. Local entrepreneurs are aware of their weaknesses and are beginning to accept that in the global market high production skills are no longer enough to sell their products; brand names and aggressive marketing strategies have become unavoidable competitive factors. Therefore, many of them agree that becoming subcontractors to luxury fashion companies is a


Local enterprises in the global economy

way to face the challenge of globalization despite the cost of functional downgrading. Furthermore and quite unexpectedly, our empirical analysis shows that this choice is not an impoverishing strategy. In fact, Table 6.9 shows a statistically significant positive relationship between performance21 and the share of production sold to high fashion companies. Furthermore, Table 6.10 confirms that during the last five years the share of firms mainly working as subcontractors and recording an increase in any of the observed performance indicators is higher than those in the rest of the sample. There is, of course, a problem of causality in interpreting these findings because it is impossible to know a priori whether positive performance is caused by participation in the luxury value chain or quite the reverse, high fashion companies select their subcontractors among better performing enterprises. In order to address the causality problem we tested the relationship between performance and production of subcontractors to top brand companies with OLS (ordinary least squares) regression analysis. Among the independent variables included (indicators of process and product innovation and number of employees) the amount of production as subcontractors is the only statistically significant variable explaining performance. Nevertheless, the cause–effect relationship remains unclear given that if the amount of production as subcontractors becomes the dependent variable, then performance is the only statistically significant coefficient among regressors. On the other hand, although our sample survey does not provide a sufficiently robust empirical test of the hypothesis that performance is positively influenced by the participation in the luxury value chain, we may corroborate our argument by looking at the economic theory of rent.22 In fact, the extent of subcontracting shoes is derived from the demand for luxury shoes and given that final consumers are prepared to pay a high price then top brand companies are also willing to pay a relatively high price to their high quality subcontractors, sharing with them (a small) proportion of their rent. In other words, we argue that top brand companies are exploiting final consumers’ willingness to pay very high prices for luxury goods, earning a rent or a super-normal profit above production costs. This rent, it seems, is to some extent shared within the chain in order to guarantee high and consistent quality and respect of delivery conditions.23 Moving on to upgrading, Table 6.9 shows that there is a positive and statistically significant relationship between the amount of production made for high fashion companies and the degree of product upgrading. This finding is confirmed in Table 6.11, which illustrates that subcontracting firms introduce more product innovation than the rest of the sample. The

155 Process innovation


⫺1.000 ⫺0.213 ⫺0.068 ⫺0.092

Number of employees


⫺1.000 ⫺0.126 ⫺0.161 0.080


Ex-factory % of production price as subcontractor


⫺1.000 ⫺0.447**

% of production sold to European buyers


Source: Author’s survey

Notes: ** correlation is significant at the 0.01 level; * correlation is significant at 0.05 level a Some of the variables in this study (process and product innovation) are ordinal variables which can be ordered qualitatively in a limited number of ordered classes. Therefore, to analyse correlation we calculate the Kendall coefficient which is the more appropriate for ordinal variables. However, alternative correlation measures, such as Spearman and Pearson coefficients, were also calculated and no differences in terms of statistically significant associations were found.

Product innovation




⫺1.000 ⫺0.461** ⫺0.028 ⫺0.142 ⫺0.256*

⫺1.000 ⫺0.104 ⫺0.042 ⫺0.051 ⫺0.319** ⫺0.229*

Kendall correlation coefficients a

Performance Product innovation Process innovation Number of employees Ex-factory price % of production as subcontractor % of production sold to European buyers

Table 6.9


2 28.6%

6 25.0%

Firms mainly selling to buyers**

Rest of the sample

8 33.3%

1 14.3%

7 77.8%

Value of sales

4 16.7%

1 14.3%

4 44.4%


4 16.7%

1 14.3%

5 55.6%


18 74.0%

2 28.6%

8 88.9%

Average price

Source: Author’s survey

13 54.2%

2 28.6%

6 66.7%

Quality of product

Notes: ** Firms producing more than 50 per cent of their production as subcontractors to high fashion companies ** Firms selling more than 50 per cent of their production to buyers

6 66.7%

Firms mainly working as subcontractors*

Number of pairs

7 29.2%

1 14.3%

4 44.4%

Number of employees

24 100%

7 100%

9 100%


Table 6.10 Performance among sample firms (number and % of firms recording an increase of performance indicators during the last 5 years)

157 0

1 14.3% 12 50.0%

4 57.1%

1 11.1%

Little innovation

12 50.0%

2 28.6%

7 77.8%

A lot of innovation

1 4.2%

1 14.3%

1 11.1%

No innovation

Source: Author’s survey

14 58.3%

5 71.4%

2 22.2%

Little innovation

Product innovation

Notes: ** Firms producing more than 50 per cent of their production as subcontractors to high fashion companies ** Firms selling more than 50 per cent of their production to buyers

Rest of the sample

Firms mainly selling to buyers**

1 11.1%

No innovation

Process innovation

Process and product innovations among sample firms

Firms mainly working as subcontractors*

Table 6.11

9 37.5%

1 14.3%

6 66.7%

A lot of innovation

24 100%

7 100%

9 100%



Local enterprises in the global economy

role played by top brand companies in product innovation is confirmed by the fact that 60 per cent of sample firms stress the importance of their assistance in this field. However, high fashion companies, which do not have specific skills in shoe production, are not assisting their subcontractors in process innovation, although their degree of upgrading is still better than the rest of the sample (Table 6.11). Sample firms identified a number of advantages coming from their activity as subcontractors. The most important advantages are: the size of orders (all firms except one); their regularity (all firms except two); and, for 70 per cent of firms the prestige of working for a globally recognised top brand. A further advantage, very much emphasized by the interviewed firms, is the reduction of costs because they no longer have to produce a sample set. Among the disadvantages, imposition of tight delivery conditions and timing, lack of direct market access and the loss of independence were stressed by 75 per cent of sample firms. These disadvantages are strongly related to the issue of governance within the chain. Coordination of the value chain is clearly in the hands of top brand companies who keep their control on rent-rich activities such as design, branding and marketing. They are also increasingly becoming more directly involved in shoe and component production. Nevertheless, only two sample firms have clearly defined their relationship with high fashion companies as quasi-hierarchical, while the remaining ones reported that there is some degree of cooperation. A clear assessment of how hierarchical these relationships are is a difficult task. There are mixed signs: in many cases firms are not fully dependent on the top brand value chain, producing less than 50 per cent of total production as subcontractors. In some cases they even contribute to design but, on the other hand, most of them suffer from functional downgrading and are losing their direct link with the market. Furthermore 75 per cent complain about dependence on high fashion companies. We may conclude that the most common type of governance within this chain is somewhere in between network and quasi-hierarchy. The clear leaders in the chain are the top brand companies. They are definitely setting the parameters that the rest of the actors have to comply with but in many cases they are also cooperating with their highly qualified partners to obtain top quality products and besides, very importantly, they are willing to share with them part of their rent in order to acquire their production skills. The local entrepreneurial association is pushing Brenta firms to increase cooperation with high fashion companies by making themselves unavoidable partners and eventually acquiring licences.24 The question is how long

How globalization affects Italian industrial districts


this strategy will be successful, because some high fashion companies are already setting up their own production facilities and others are beginning to decentralize production of more standardized shoes to Romania and other low-wage countries. They will probably need Brenta firms as subcontractors of top quality shoes for a long time but competition is tough within the district and in Italy more generally. Following this strategy, Brenta firms may seek to build up a cooperative relationship with the chain’s leaders, capitalizing on their excellence in production but they may progressively lose their design capability and the direct market access. To conclude, it is useful to summarize our main empirical findings in connection with our reference point on value chains (Gereffi, 1999). The top brand value chain confirms Gereffi’s trend toward ‘buyer-driven’ chains. However, there are also some new findings. Producers from developed countries are involved in the chain, and their participation, far from having an upgrading effect (much stressed in Gereffi’s work), has a functional downgrading effect instead. Nevertheless, although firms have abandoned some key functions, their performance is still very positive because they have the prospect of sharing with the chain’s leaders (most of which are Italian) the high rents of the luxury industry.



Almost half of our sample sells more than 50 per cent of its total production to Europe, mainly to Germany. In Germany the shoe market is dominated by independent retailers (74 per cent of total sales). However, most German retailers are organized in large buying groups, some of them associating more than 1000 shops and extending to other European countries, mainly Switzerland, Belgium and the Netherlands.25 Therefore, at first sight the German market seems to be dominated by buyer-driven chains. Yet, as we will demonstrate in the rest of this section, the buying groups’ main purpose is not to drive the chains and establish ‘quasi hierarchy’ relationships with their suppliers. On the contrary, their purpose is to: a) facilitate market-based relationships by reducing transaction costs, providing what Zucker (1986) calls ‘institution-based trust’; and, b) facilitate networks in which ties are information-rich. Buying groups offer financial assistance to their members (the retailers) by providing long- and short-term credit for the retailers and guaranteeing their payments to suppliers. Furthermore, buying groups maintain direct contact with shoe producers through their specialized buyers for different market segments, visiting them at least twice a year, corresponding with the summer and winter sample sets.


Local enterprises in the global economy

There are usually two different forms of relationships with shoe producers. There is a network of authorized suppliers from whom retailers can purchase at agreed payment conditions, obtaining credit from the group, and there are some footwear producers selling directly to the group, offering their own sample set to members. Therefore, retailers can choose to buy directly from producers or purchase from the group. In both cases they receive credit if needed. Furthermore, the groups organize several events outside official trade fairs to present suppliers’ sample sets and to arrange meetings between retailers and producers. These events usually take place in different parts of Germany or neighbouring countries and are specialized for market segments and the producers’ country of origin. For example, there are specific exhibitions to present Italian shoes. The largest buying groups have local branches in Italy (usually in Veneto or Tuscany) taking care of logistics and maintaining more frequent contacts with suppliers. Sometimes this implies quality controls during the production cycle, but more often controls take place on finished products. Moreover, local branches deal with problems of returned products and after-sales service. Among the buying groups interviewed, the largest ones operate in a variety of market segments, while two smaller groups are more specialized in the high and medium-high ones.26 In this segment, Italian producers are the main suppliers of all interviewed buying groups, although all of them said that German producers are recently regaining market shares. The increasing competition of German producers is explained by buyers in terms of improved design and better sales and after-sales services. In most cases, German shoe enterprises are very large firms, domestically manufacturing only a minor share of what they sell with their brand name and outsourcing the rest to producers outside Germany. They focus on design, branding, advertising and sales, offering a very varied sample set to satisfy different types of customers, for example, from youngster and sport to classic and formal shoes. According to the buyers interviewed, they offer very good service, which means reliability in delivery timing and quality, rapidity in reassortment and after-sales assistance. As shown in the next section, these are areas in which Brenta producers are rather weak. Moreover, German producers have become very good at interpreting domestic tastes, as well as maintaining intensive and fruitful contacts with their customers. According to Brenta producers, they also have a strong price advantage, because the distribution mark-up on German shoes is half that of Italian shoes. In conclusion, it may be useful to recall that in recent years demand for shoes in Germany and neighbouring countries has stagnated, which seems to have penalized the top market segments. In these segments, Italian pro-

How globalization affects Italian industrial districts


ducers have also encountered increasing competition from German firms. Therefore, Brenta’s producers, traditionally exporting a very large share of their production to Germany, have been particularly hit by this situation. The next section analyses their relationships with German buying groups in more detail.

8. BRENTA PRODUCERS IN THE GERMAN VALUE CHAIN In the medium–high and high segment of the German market, most imports come from Italy and particularly from Riviera del Brenta and Tuscany. The interviews showed that German buying groups are the key players operating in Brenta.27 All of them have a large number of enterprises (an average of 200 firms) included in their list of approved suppliers, but stable relationships exist only with a minority of producers. On average, each buyer interviewed maintains continuous relationships with around 20 firms located in Brenta. The connection with the large majority of approved suppliers is a market-based relationship with producers proposing their sample set and retailers buying directly from them on the basis of a standardized contract, agreed by manufacturers with the buying group. Each buying group has preferred suppliers with whom competencies are shared. This includes information exchange as producers, retailers and buyers have many opportunities for meeting at special events organized by the buying groups. Furthermore, buyers and some retailers often visit firms. In some cases there is collaboration between producers, buyers and retailers on design. Occasionally, shoes are sold with either a brand name of the buying group or of an important retailer. Nevertheless, German buyers complain because Brenta producers are not learning quickly how to take advantage of all information they could receive via new information networks. In fact, most of the buying groups have recently set up information networks linking some suppliers and shops which collect and provide instant sales information. Most of the German suppliers are connected and use this information to speed up delivery and offer reassortment. One of the buyers interviewed said that, among their 20 stable suppliers in Brenta, only one is connected to their network. Low propensity to innovate among buying groups’ suppliers in Brenta is confirmed by our empirical findings: sample firms selling more than 50 per cent to buyers appear to be less innovative than the rest of the sample (Table 6.11). With regard to performance, results are also less positive than for the rest of the sample (Table 6.10).28 Moreover, there is also a negative,


Local enterprises in the global economy

although not statically significant, correlation between price and the amount of production sold to buyers (Table 6.9). Indeed, in Brenta firms complain strongly that they are unable to pass on increases in cost. In our interviews we also asked buyers to evaluate the performance of their Brenta suppliers regarding product quality, innovative design, response time, punctual delivery and service.29 As expected, firms in Riviera excel in design and are doing rather well in product quality. However, many buyers stressed that quality has recently been decreasing. This seems to be due to the relocation of production to other countries. In effect, given their average small size, firms are running into difficulties with quality control on the production carried out abroad. Increasingly, Brenta enterprises are using their internal production capacity to make shoes as subcontractors to high fashion companies and outsourcing the rest of their production abroad. Decentralization to Romania and other neighbouring countries is a necessity for reducing costs, given that price competition is severe even in the high quality German market. In this respect, outsourcing may be interpreted as a strategy of Brenta firms’ functional upgrading, that is moving low value added activities abroad and focusing on production for the rent-rich luxury market at home. German buyers do not rate the reliability of Brenta suppliers’ delivery very highly. They complain about delays and a response time that is too long between order and delivery.30 In Brenta, the responsibility of delays is given to their suppliers of components and raw materials who satisfy their high fashion customers first. According to our survey, this privilege is due to the different timing of production, with the luxury value chain anticipating the rest of the market, as well as, inevitably, to the stronger bargaining power of the top brand oligopolists in imposing their delivery conditions. Therefore, this is an area of potential conflict within the district between production for the luxury value chain and the other chains. The other weakness, according to the German buyers, is Brenta’s poor performance in after-sales service, reassortment and substitution of defective goods. In these aspects, Brenta enterprises are unable to compete with German producers, who invest many resources in these activities. To conclude, it appears that the German value chain is characterized by market transactions, but these often include some cooperation between more or less equal partners sharing their competencies. Brenta enterprises maintain a high degree of independence because they can sell to several buying groups, and from these buying groups they can sell to many retailers. In other words, they maintain many exit options. Furthermore, there is no conflict in functional specialization. Brenta manufacturers take care of the entire cycle from design to production as well as the buying groups of

How globalization affects Italian industrial districts


distribution and marketing. Cooperation is mainly aimed at improving product quality and some of the buyers want to intensify this cooperation. However, Brenta manufacturers sometimes see this desire of increasing cooperation in product development as an intrusion on their creativity. This is very much in line with the findings by Schmitz and Knorringa (2000): ‘. . . the Italians are considered unwilling to learn from foreigners. A telling example was the reaction of an Italian company to a Brazilian consultant who had been asked by a buyer to investigate logistical and quality problems: the Italians were offended’.



While Brenta never lived up to the romantic view of the Third Italy, much advertised in the past literature on industrial districts, it has nevertheless been traditionally characterized by strong local linkages at both a vertical and a horizontal level. As documented in a previous survey carried out in Brenta in 1992 (Rabellotti, 1997), local shoe producers had intensive and cooperative linkages with their local suppliers and subcontractors and frequent informal contacts with other shoe firms in the district. Compared with Marche, another major footwear district analysed at that time, entrepreneurs in Brenta showed a more cooperative attitude, also attributing more importance to these informal relationships. In addition to cooperative vertical and horizontal links among firms, Brenta is also characterized by the existence of several collective institutions, that is ACRIB, the local entrepreneurial association established more than 30 years ago; an export consortium (Consorzio Maestri Calzaturieri del Brenta), created in 1967 and a technological and training institute, named Consorzio Centro Veneto Calzaturiero, set up in 1986.31 Among many initiatives organized by these bodies, a significant example is the collective organization at international fairs, managed by the Consorzio Maestri Calzaturieri. For example, at the September 2000 international shoe fair in Dusseldorf, 81 exhibiting producers from Brenta participated in a space collectively organized with a common look and identified by the logo of the consortium. However, a significant change can be noted by comparing local governance in the early 1990s with that at the turn of the century. Links within the cluster have become less important and links with outside actors have become more important. In 2000, among the sample firms, only 25 per cent think that relationships with other local shoe enterprises are important, while the remaining 70 per cent attach no importance to them at all. Furthermore, during recent years, cooperation among shoe enterprises has


Local enterprises in the global economy

decreased within the district according to 30 per cent of sample firms. A better, but far from good, evaluation is given to the relationship with the local entrepreneurial association ACRIB, which is considered very important by 15 per cent of the enterprises, important by 37.5 per cent and unimportant by 40 per cent. Among local links, only relationships with suppliers32 assume a real importance. They are very important for 65 per cent of sample firms and important for 15 per cent (Table 6.12). Regarding the importance attributed to the reputation of the district, most of the sample firms believe that their main buyers and high fashion companies do not attach much importance to this aspect. This is confirmed by the interviews with German buyers, who did not select ‘Made in Brenta’ or ‘Made in Italy’ among the main factors of competitiveness. In fact, the cluster has never made a real effort to promote its collective image, apart from joint participation at the fairs. Furthermore, the ‘Made in Italy’ factor is progressively losing its impact on the market because customers are looking more for individual top brands and because in many cases shoes are no longer made in Italy but in nearby low-wage countries. In contrast, links with actors external to the cluster are considered increasingly crucial. In particular, relationships with main customers are regarded as very important by 65 per cent of the sample firms and important by 15 per cent, and those with high fashion companies are very important for 70 per cent of the sample firms working as subcontractors and important for 13 per cent (Table 6.12). In summary, external linkages have recently assumed a greater importance while relationships within the cluster have become less important, except for links with suppliers which remain crucial. Among the entrepreneurs interviewed there is a diffused feeling of disillusionment regarding local cooperation and the possibility of repositioning themselves in the world market through joint initiatives.33 Nevertheless, there have been a few joint projects recently, among them the opening of a showroom for 13 firms in New York, sponsored by ACRIB with 50 per cent of its cost funded by public subsidies, which have been particularly successful. Boosted by the euro devaluation of recent years, this initiative has been a real success for at least seven partners with more than 200000 pairs sold by the show room in its first year of activity, compared to an initial objective of 50000. Following this successful experience a new project has been established to set up a trading company in joint venture with a local partner in Shanghai, to sell to the top segment of the Chinese market. This project should also receive financial support from the Veneto regional government and from the Italian Ministry of External Trade. Another project sponsored by ACRIB is the setting up of an electronic information network connecting shoe enterprises, suppliers of components,


65.0 7.5 15.0 65.0 70.0

12 7 15 6 2

30.0 17.5 37.5 15.0 13.0

2 28 16 4 3

5.0 70.0 40.0 10.0 17.0

0 2 3 4 0

Source: Author’s survey

Note: *17 sample firms are working as subcontractors for high fashion companies. The percentage is calculated on the 17 firms which answered the questionnaire.

26 3 6 26 12

Very important % of sample Important % of sample Unimportant % of sample Missing (No. of firms) firms (No. of firms) firms (No. of firms) firms (No. of firms)

Importance of linkages according to sample firms

Suppliers Shoe enterprises ACRIB Main customers High fashion companies*

Table 6.12


Local enterprises in the global economy

retailers and some local branches of banks. The e-commerce initiative, which is in a pilot phase, is aimed at supplying business-to-business services. Furthermore, there is a regional government project34 involving ACRIB and other local institutions that trains immigrants from outside Europe, providing them with the essential services (that is housing, residence permits for their families) needed to facilitate their integration into the local community. This initiative addresses a crucial issue in Brenta; given that the district is located in a region characterized by full employment35 and the footwear industry is not considered very appealing in terms of pay and working conditions, there is a structural shortage of skilled labour. Availability of highly qualified workers was traditionally one of the main advantages of firms located in industrial districts like Brenta. Consequently, the difficulty in finding skilled workers is having a negative impact on the district by increasing labour costs, decreasing flexibility and becoming an obstacle to further expansion. This is why local initiatives addressing this constraint are particularly important for the future of Brenta. To conclude, there has been a change in the relative importance of local and global linkages. Vertically, there is the weakening of local linkages, that is backward links with subcontractors due to increasing decentralization of production to Romania and other nearby countries and forward links due to increasing presence in the district of high fashion companies. Horizontally, there are a few collective initiatives promoted by ACRIB but compared with the past, Brenta enterprises are investing more in their external linkages with buyers or high fashion companies than in local relationships.



This chapter has studied the impact of global transformations occurring in the fashion system on local competitiveness in one of the most important Italian shoe districts. This was done by integrating the typical industrial district approach with the global value chain approach. Primary data were used to address the following questions: Is globalization pushing Brenta towards new value chains? What types of governance characterize the relationships between local and outside actors? Do the chains’ leaders come from inside or outside the districts? Does the integration of industrial clusters in global value chains enhance or weaken local upgrading strategies? This section presents the main conclusions.

How globalization affects Italian industrial districts


1. In Brenta most of the shoe enterprises feed into a variety of chains. If we exclude from our sample firms producing more than 50 per cent of their production as subcontractors of high fashion companies (22.5 per cent of the sample) and enterprises which sell more than 50 per cent of their production to large buyers (17.5 per cent) we are left with 60 per cent of firms combining several different chains. Therefore, our first conclusion is that Brenta shoe enterprises often face several exit options; dependence on one value chain is limited to a reduced number of cases. This is likely to be a main difference between clusters in developed and developing countries (Humphrey and Schmitz, 2000). 2. The two main value chains for Brenta are the ones leading to the high quality German market and the one dominated by global luxury brands. The former has been relevant to Brenta for a long time. The relationships within this chain are mainly market-based with some sharing of competencies between producers and their customers. Referring to Gereffi’s definition of chains, the German chain cannot be defined as buyer-driven because buying groups are not ‘setting the parameters’ for manufacturers in the chain. The peculiarity of this chain is that buying groups, gathering together large numbers of independent retailers, are helping to reduce transaction costs and risks and to improve information access. 3. The top brand value chain is of more recent origin. Its increasing importance follows a global boom of luxury, branded goods in clothing, leather accessories, shoes, perfumes, and so on. This is a crosssector global value chain, characterized by very high entry barriers, given the costs of creating and maintaining a global brand. It can be defined as a sub-type of Gereffi’s buyer-driven value chain, with some peculiarities. 4. To be part of the chain, Brenta’s shoe producers accept a functional ‘downgrading’ by abandoning design and sales, which are the key competencies of the leaders of the chain, and focusing on production. Their relationships with top brand companies can be defined as somewhere in between network and quasi-hierarchy, but it is clear that the leaders of the chain are not located in Brenta. This story draws attention to at least two new insights, that is not only upgrading but also downgrading can occur within global value chains and this may happen even to leading producers of developed countries. 5. These firms, which have given up their design function and have become subcontractors, perform better than the other local producers in terms of sales and profits as well as process and product upgrading. It appears that the luxury brand companies share some of their high rents with their skilful subcontractors.


Local enterprises in the global economy

6. Furthermore, although a process of functional downgrading is occurring at district level, by taking a more systemic perspective our conclusions are different. The Italian high fashion luxury industry is generally undergoing a process of functional upgrading and concentrating investments in rent-rich activities linked with intangible characteristics of the products. The global leaders in the chain exploit cross-industry economies of scale and scope in branding, marketing and advertising. Most of the global leaders of this value chain are Italian companies. Therefore, if the Brenta district is experiencing functional downgrading, the Italian fashion system, in contrast, is experiencing functional upgrading. 7. The integration of Brenta into the top brand value chain is causing conflicts within the district. First, some producers use their internal production capacity to satisfy the demand of shoes coming from high fashion companies and outsource (often to Romania) the production of their own sample set with a decrease in quality. Second, suppliers of components first fulfil orders coming from the top brand chain, generating delays for the remaining firms in the cluster. Nonetheless, these changes occurring in the district can also be interpreted as a form of functional upgrading, with Brenta enterprises moving the less profitable production abroad in order to use their internal capacity to satisfy the demand from the more profitable top brand chain. 8. There are also conflicts arising between both linkages within the district and outside it. Comparing our findings with a previous survey in the early 1990s, it is clear that Brenta enterprises are now attributing less importance to relationships with other local firms than before. Backward and forward local linkages are weakening while external linkages with buyers and top brand companies have recently assumed a major role. 9. The conclusion that can be drawn is that intangible activities are increasingly becoming the major assets in the top brand industry. In the past, Italian industrial districts, such as Brenta, have built their excellence on a mix of skills in design, fashion and production but the small size of firms has limited their capability to face the massive investments required to control intangibles in the global market. These intangible activities have become the core competencies of a few large cross-industry companies dominating the top brand value chain, which is assuming a leading role in Brenta. Its expansion allows local footwear firms to continue exploiting their traditional comparative advantage of highly skilled producers, maintaining a good performance. Nevertheless, the top brand oligopolies, with their huge profits and large financial capital availability, are the global leaders of this segment of the market.

How globalization affects Italian industrial districts


10. It is not clear what this will mean for the future of Brenta. Local firms are winning a place in the rent-rich global top brand market but by focusing only on their production skills they offer capabilities which can increasingly be found in other clusters in the world and this may slowly erode their competitiveness and independence. To date, a mix of factors prevents luxury companies searching for alternative subcontractors in countries such as Romania or Brazil. They include the size of rents which downplays the cost factor, and their lack of experience in the shoe sector. These factors induce them to search for very skilful subcontractors that are already able to produce for the top quality market, and, of course, the higher transaction costs involved in the relationships with more distant and less qualified subcontractors. Nevertheless, these conditions may change and the dynamic comparative advantage of Brenta’s producers may one day vanish.




3. 4. 5. 6.

I am very grateful to Hubert Schmitz who inspired this research project and then commented in great detail on successive drafts of the chapter. I also would like to thank Tania Pasin for assistance in interviewing enterprises. Furthermore, Brenta’s entrepreneurs, ACRIB (Associazione Calzaturifici della Riviera del Brenta), the Regional Government of Veneto, the US National Shoe Retailers Association, the buyers interviewed at the Dusseldorf shoe fair, Sabino Ventura, Roberto Vergelio and Heinrich Zumnorde are kindly acknowledged. A special thank you goes to John Humphrey, Raphie Kaplinsky, Jorge Katz and participants at workshops held at the Institute of Development Studies and at the University of Molise for providing valuable comments on earlier drafts. Financial support from the Volkswagen Foundation and the Department for International Development is gratefully acknowledged. The shortcomings of the chapter are my responsibility alone. By typical industrial district we refer to the Marshallian concept and particularly to its Italian variant. For a collection of classical papers see Pyke, Becattini and Sengenberger (1990). A recent collection of papers on industrial clusters in developing countries can be found in Nadvi and Schmitz (1999). Markusen (1996), broadening the definition of industrial district, discusses four types of districts. In the ‘satellite platform’ type, consisting of a congregation of branch facilities of externally based multi-plant firms, she acknowledges the importance of external linkages. The need to use the market channel approach to study industrial clusters was first advocated by Knorringa (1999) in his work on Agra, a footwear district in India. We asked ACRIB to select among their members a list of firms selling in different geographical markets and to different types of customers. The sample of firms was then randomly selected from this list. Notable exceptions are Weijland (1994) and Knorringa (1999). Unfortunately, time series data about the district are available only up to 1999. Since 2000, ACRIB has changed the definition of the district including a larger geographical area and therefore data are not comparable with the previous time series. Nevertheless, according to our calculations sales continued to increase while the number of firms and employees decreased from the second half of the 1980s.

170 7. 8.

9. 10. 11.

12. 13. 14. 15. 16.



19. 20. 21.

22. 23.

Local enterprises in the global economy For a detailed analysis of the main characteristics of the Brenta district also see Rabellotti (1997). The category of firms employing fewer than 20 employees is under-represented in respect to size distribution presented in Table 6.2. This is because smaller firms are underrepresented in ACRIB. Most of these microenterprises are subcontractors to larger local firms. Four firms belonging to this category did not declare how many pairs of shoes they produce yearly. Both the Kendall coefficient and the Spearman rank correlation coefficient were calculated. They are the same in terms of statistically significant association and therefore only the Kendall coefficient is reported here as well as in the rest of the chapter. Among Italian shoe firms, the adoption of a quality standard like ISO 9000 is extremely rare. According to the director of CIMAC, a technological institute in charge of the diffusion of quality standards in the footwear industry, in Italy there are about 50 certified firms and among them only about 20 produce leather shoes. In Brenta there are five certified enterprises. Personal communication with the director of the local entrepreneurial association (ACRIB). Total retail footwear sales in the USA in 1999 were $38.4 billion. The market share of the retail category is about 10 per cent of total sales (NSRA, 2000). In Japan the tariff on imported shoes ranges from 23.6 per cent to 41.3 per cent for leather footwear. In the EU it ranges from 4.6 per cent to a maximum of 8 per cent. 66 per cent of the sample firm’s main customers originate in Europe, 11 per cent in the USA, 6 per cent in Italy and the remaining 17 per cent constitutes the rest of the world. Prada is an example of an acquisition of brand names. Prada recently acquired Church, a well known British footwear enterprise. The French group LVMH owns more than 50 brands in different industries. For example, in the fashion industry names like Louis Vitton, Dior, Kenzo, in the wine industry Moët, Krug, Dom Pérignon and in the distribution industry Le Bon Marché and E-luxury (e-commerce). In Italy among the top ten companies in the luxury industry there is only one footwear firm, Tod’s, ranking at 9th position with a sales value of US$ 202 million (Table 6.6). It originated in Marche, the largest Italian footwear cluster, and was recently listed at the Milan Stock Exchange. According to the local entrepreneurial association, in 2000 the amount of production made by Brenta’s enterprises as subcontractors to high fashion companies had reached 50 per cent of total production in the area (personal communication with the director of ACRIB). If these estimates are correct, subcontractors are underrepresented in our sample. There is a distinction between creative and technical design; in the chapter the former is simply called design while technical design, including size developing, is named product development. In the rest of this section, the proportion of sample firms is intended as a ratio of the 17 enterprises which work as subcontractors to high fashion companies. The index of performance is generated by attaching equal weights to each performance indicator (see Table 6.3). The five possible values – from strong increase to strong decrease – were coded on a range of ⫹2 to ⫺2 respectively, with no change coded 0. The index for each firm was then constructed by adding up the actual values and dividing them by the number of variables. To test the robustness of this indicator, an alternative performance indicator was also obtained with a principal component analysis, estimated as a linear combination of the original variables. The correlation coefficient between the two indicators is 0.9 (significant at 0.01 level). A summary of rent theory in the history of economic thought, with a particular focus on land, is presented in Camagni (1992). An anonymous referee suggested we refer to the management literature in order to explain the relationship between the lead firms and their subcontractors. The concepts of clan and socialization as the principal mechanism of mediation to overcome ‘goal

How globalization affects Italian industrial districts


25. 26.

27. 28. 29. 30.

31. 32.

33. 34. 35.


incongruence’, first introduced by Ouchi (1980) and that of strategic networks (Jarillo, 1988), may indeed be useful analytical tools to adopt in a future follow up of this study, focused on exploring the relationship from the point of view of the luxury brand companies. According to some sector experts, acquiring licences could be an alternative strategy to get into the luxury value chains. This strategy was adopted in the 1970s by one of the leading Brenta firms and has been rather successful in the past. This firm owns seven licences of globally known brand names such as Yves Saint Laurent and Calvin Klein and it takes full care of their design and distribution. Nevertheless, there is a recent trend among fashion leaders to reacquire licences. For example, one of the licences of this named firm was not renewed after many years. Furthermore, in 2001, the French group LVMH bought 45 per cent of the shares of this same enterprise. For instance, among the buying groups we interviewed, one has a membership of more than 3000 shops in Europe (1500 in Germany) and another around 1900 shops. In Germany, women’s leather footwear of high and medium high price (more than US$ 80) represent less than 12 per cent of total sales (data presented by the German Shoe Retail Trade Association at a press conference in GDS, International Shoe Fair, Dusseldorf, September 2000). The buying groups interviewed were selected from a list provided by the local entrepreneurial association. Only one group was not interviewed from this list. Table 6.9 shows a negative, although not statistically significant, correlation coefficient between the performance indicator and the share of production sold to European buyers. None of the buyers interviewed attributed any importance to quality standards, such as ISO 9000, as a factor of market competitiveness. This confirms empirical findings obtained by Schmitz and Knorringa (2000) who interviewed buyers from the US, the UK, the Netherlands and Denmark. According to their survey, Italian footwear producers in comparison with those in Brazil were not performing very well in ‘reliable product quality’, ‘speed of response’, and ‘punctual delivery’. A detailed description of their main activities can be find in Rabellotti (1997). Most of the suppliers of components are located in Brenta. In contrast, raw materials’ suppliers are typically located in other clusters that specialize in leather production. Finally with regard to machinery, production is concentrated in Vigevano, an old footwear cluster transformed in a mechanical district, but in Brenta there are usually local dealers with whom shoe firms interact (Rabellotti, 1997). A similar result was found by Schmitz (1999) in the largest Brazilian shoe cluster, the Sinos Valley, where producers with the closest ties to global buyers were least interested in collective local initiatives. This project is part of a local agreement (‘patto territoriale’), an instrument of industrial and infrastructural policy at local level recently introduced in Italy and much implemented in the Southern regions. According to the last available population census (1991), in Brenta 96 per cent of the working population is employed.

REFERENCES ACRIB, various years, Rilevamenti Statistici, Strà: Associazione Calzaturifici della Riviera del Brenta. ANCI, various years, L’Industria Calzaturiera Italiana, Milan: Associazione Nazionale Calzaturifici Italiani. Camagni, R. (1992), Economia Urbana, Rome: La Nuova Italia Scientifica. Dolan, C. and J. Humphrey (2000), ‘Governance and trade in fresh vegetables: the


Local enterprises in the global economy

impact of UK supermarkets on the African horticulture industry’, Journal of Development Studies, 37 (2), 147–76. Gereffi, G. (1994), ‘The organization of buyer-driven global commodity chains: how US retailers shape overseas production networks’, in G. Gereffi and M. Korzeniewicz (eds), Commodity Chains and Global Capitalism, London: Praeger. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Humphrey, J. and H. Schmitz (2000), ‘Governance and upgrading: linking industrial cluster and global value chain research’, IDS Working Paper 120, Brighton: Institute of Development Studies. Istat (1996), Industrial Census, preliminary results available at www.istat.it. Jarillo, J.C. (1988), ‘On strategic networks’, Strategic Management Journal, 9, 31–41. Kaplinsky, R. (1998,) ‘Globalization, industrialisation and sustainable growth: the pursuit of the nth rent’, IDS Discussion Paper 365, Brighton: Institute of Development Studies. Kaplinsky, R. (2000), ‘Spreading the gains from globalisation: what can be learned from value chain analysis?’, IDS Working Paper 110, Brighton: Institute of Development Studies. Knorringa, P. (1999), ‘Agra: an old cluster facing the new Competition’, World Development, 27 (9), 1587–604. Markusen, A. (1996), ‘Sticky places in slippery space: a typology of industrial districts’, Economic Geography, 72, 293–313. Mytelka, L. (2000), ‘Local systems of innovation in a globalized world economy’, Industry and Innovation, 7 (1), 15–32. NSRA (2000), Business Performance Report, Columbia: National Shoe Retailers Association. Nadvi, K. and H. Schmitz, (eds) (1999), ‘Industrial clusters in developing countries’, Special Issue of World Development, 27 (9). Ouchi, W. (1980), ‘Markets, bureaucracies, and clans’, Administrative Science Quarterly, 25, 129–45. Pambianco (2002), ‘Osservatorio delle principali operazioni di Merger & Acquisition avvenute nei settori del Made in Italy a livello mondiale’, mimeo, available at www.pambianco.com. Pambianco, various years, ‘I bilanci del Made in Italy’, mimeo, available at www.pambianco.com. Pyke, F., G. Becattini and W. Sengenberger (eds) (1990), Industrial Districts and Interfirm Cooperation in Italy, Geneva: International Institute for Labour Studies, ILO. Rabellotti, R. (1997), External Economies and Cooperation in Industrial Districts. A Comparison of Italy and Mexico, London: Macmillan. Saviolo, S. and S. Testa (2000), Le Imprese del Sistema Moda, Milan: Etas. Schmitz, H. (1995), ‘Collective efficiency: growth path for small-scale industry’, Journal of Development Studies, 31 (9), 1627–50. Schmitz, H. (1999), ‘Increasing returns and collective efficiency’, Cambridge Journal of Economics, 23 (4), 465–83. Schmitz, H. (2000), ‘Does local co-operation matter?’, Oxford Development Studies, 28 (3), 323–36. Schmitz, H. and P. Knorringa (2000), ‘Learning from global buyers’, Journal of Development Studies, 37 (2), 177–205.

How globalization affects Italian industrial districts


Weijland, H. (1994), ‘Trade networks for flexible small rural industry’, in P.O. Pederson, A. Sverisson and M.P. van Dijck (eds), Flexible Specialization: The Dynamics of Small-Scale Industry in the South, London: Intermediate Technology. Zucker, L.G. (1986), ‘Production of trust: institutional sources of economic structure, 1840–1920’, Research in Organisational Behaviour, 8, 53–111.


Upgrading in the tile industry of Italy, Spain and Brazil: insights from cluster and value chain analysis Jörg Meyer-Stamer, Claudio Maggi and Silene Seibel1



The chapters in this book investigate the relationship between clusters and value chains, and in particular the effect that the integration of clustered firms into value chains has on collective upgrading efforts inside clusters. This focus was chosen because we observed that earlier research on clusters, including our own, did not systematically explore the external connections, and because we observed that value chain researchers did not systematically analyse options for local action within clusters. This chapter addresses these issues in the context of a case study on the wall and floor tile industry. We chose this industry for three reasons. First, it is a strongly clustered industry. A large part of world production originates from just two clusters – Sassuolo in Italy and Castellón in Spain, and these two clusters dominate the technological evolution of the industry. Second, it is a highly internationalized industry. The two clusters have a very strong export performance, both in terms of final products, and capital goods and key inputs. Third, there are emerging clusters in advanced developing countries, and the technologically dominant position of developed country clusters does not seem to translate into an ability to stem the upcoming competitors. Addressing the tile industry exclusively from either the cluster or the value chain perspective would provide relevant data, but in the end it would neglect crucial features and thus not lead to an adequate understanding of its evolution and upgrading perspectives. A cluster perspective would tend to focus on horizontal and vertical linkages and cooperation within the location, but overlook the external linkages. A value chain perspective would focus on the vertical linkages, irrespective of location and would try to determine whether, in the terminology introduced by Gereffi (1996), this 174

Upgrading in the tile industry of Italy, Spain and Brazil


industry is a supplier-driven or a buyer-driven chain. But even though value chains are very important for its understanding, the tile industry does not fit into either of these chains introduced by Gereffi. The tile value chain is not dominated or driven by global buyers or strong multinational corporations. The relationships are more even and combine network and market relationships. Gereffi and others have advocated the analysis of power, but in the tile value chain strong power positions have been at best transitory. There is little space for rents except for innovation-based rents, which tend to be competed away very quickly. We will interpret the recent restructuring of the tile value chain as the result of atomized efforts to create rents based on market power. This, we will argue, involves a process that erodes the conditions for successful local network governance without actually creating dominant players among the actors. We will show that the struggle for dominance within the value chain is reinforcing local rivalry and eroding existing institutions that execute localized collective action. In fact, it appears that the struggle for domination among the established actors is creating favourable conditions for latecomers who can, to some extent, play the incumbents against each other. Underlying both the changes in the clusters, in particular the Italian cluster, and the value chain is a process of consolidation that can be interpreted from an industrial life cycle perspective. This adds a third dimension, apart from clustering and value chains, which we did not consider at the outset of the research but which turns out to be an important factor in explaining the changing governance of the tile industry. The structure of this chapter is as follows. Section 2 gives an overview of key features of the tile value chain. Sections 3 and 4 analyse the evolution of the two leading clusters in Italy and Spain. Section 5 addresses the issue of latecomer development by examining a cluster in Santa Catarina, Brazil. The final section draws together the main conclusions. The empirical material on which this chapter is based is set out in a detailed research report (see Meyer-Stamer et al., 2001).



China is the largest world producer of tiles, and it is also the biggest consumer. However, imports and exports are extremely low. As a result, it has little influence on the global tile value chain. In order to understand this global value chain, it is important to look at the five leading tile markets (next to China), namely Brazil, Spain, the US, Germany and Italy. Three of these markets are major producers, that is Italy, Spain and Brazil, while Germany and the US are the main importing countries. Italy and Spain are by far the most important exporters of tiles.



Local enterprises in the global economy

Key Facts about the Supply Side of the Tile Value Chain

In order to understand the evolution of the tile value chain, it is useful to know a few technical facts. Tiles consist of clay and glazing materials. The production process starts with the milling of clay. Fine clay powder is shaped into the tile body by extremely powerful presses. In the conventional production process, the tile body is covered with glazing material which gives the visual appearance. Apart from conventional tiles, there are porcelain tiles which are not glazed; their visual appearance depends on the consistency of the powder (which includes additional inputs, apart from clay) and the quality of the pressing and burning process. Both glazed and porcelain tiles pass through kilns where they are burned at temperatures beyond 1200°C. There are two types of supply industries which play the main role in pushing the technological frontier of the tile business forward.2 First, there are suppliers of capital goods. They are located almost exclusively in Italy. They have in-house development departments for both machinery and tiles, and the large ones also have experimental production lines. They constantly develop incremental innovations for each step of the production process. Second, there are producers of glazing materials. Technically, this is a subsector of the chemical industry. It is a sector that has undergone profound restructuring in the past 20 or so years. Until the 1980s, glazing manufacturers came from different countries, and not always from places with strong tile industries. Leading glazing manufacturers were Colorobbia (Italy), Degussa (Germany),3 and Ferro and Johnson (US). Today the sector is dominated by Spanish firms, all of whom have their headquarters and main laboratories in the Castellón region. The four firms mentioned above have also located their main tile-related operations there. Both types of companies have production or sales affiliates in several other countries as well. Competition within and between these two supplier industries is defining a key pattern in the tile business, namely frequently changing products. Incrementally different designs are now launched on a quarterly basis, while radically different designs are launched annually. 2.2

The Demand Side

In order to get an adequate understanding of tile distribution and sales it is essential to recall that tiles are part of the construction material industry. Second, tiles are competing with other materials used to cover floors and walls, be they inside or outside buildings. This gives rise to two issues, that is, who is the customer and who attends the customer.

Upgrading in the tile industry of Italy, Spain and Brazil ●


There are basically three types of customers: tile-purchase decisionmakers, namely consumers; architects; and construction companies (in countries such as Germany this includes small specialized tiling firms). Each of these customers have different criteria for making a purchasing decision. Consumers, with limited information, make their decision based on aesthetic and price criteria, as well as demanding basic advice at the point of sale. Architects are wellinformed, they require more sophisticated information, and have a more refined set of aesthetic criteria. In this respect, they are the most sophisticated customers. Construction companies are mostly interested in low prices. This diversity creates challenges for tile producers in terms of advertising and product information. There are three, and in some countries four, types of points of final sale. First, there are independent shops specializing in tiles. They cater to the medium- and high-price segment. They often have alliances with tiling firms, some even have their own tilers. Second, there are home-centres and DIY shops, such as Home Depot, Obi, and Le Roy Merlin. They cater to the low- and medium-price segment. Third, there are construction companies. Finally, in the US there are floor covering shops which used to sell mainly carpets and vinyl but have started to sell tiles as well. All of these points of final sale may purchase tiles from wholesalers, but it is increasingly common for manufacturers to deal directly which final sellers, especially large chains of home-centres and DIY shops.

What are the main tendencies in terms of commercialization? So far, commercialization of ceramic tiles is not particularly concentrated. A large tile manufacturer may have as many as 4000 customers. However, industry insiders are anticipating, and in mature markets such as Germany they are already clearly discerning, a concentration process in commercialization, in particular a strong growth of home-centre chains and DIY shops. This process would have two consequences. First, intermediate actors, such as import agents and wholesale traders, may suffer or even disappear. Second, it is likely that there will be a polarization in tile demand, with a strong demand both for cheap tiles (sold in home-centres and DIY shops) and for fashionable, design-intensive, high quality tiles (sold in specialized shops), with the middle segment slowly disappearing. As regards brands, few firms found it worth creating a brand image with individual customers. In terms of shops, brand image seems to be more important.



Local enterprises in the global economy


Ninety per cent of Italy’s tile production capacity is located in the EmiliaRomagna region, with 80 per cent being concentrated in 80 km2 of ten municipalities around Sassuolo in the province of Modena. The industry builds upon a centuries-long tradition in craft-based ceramics manufacture. Today’s firms were mostly created in the post-war period, when the reconstruction of Italy created a strong demand for construction materials. Business founders often had a background in agriculture, and the availability of clay on a rural estate sometimes generated the idea to start tile production. Today, the Sassuolo cluster is mature. A consolidation process has been going on for many years. Consolidation leads to a change in upgrading strategies and an emphasis on value-chain related issues when it comes to creating a competitive advantage. It also leads to erosion of local network governance. 3.1

Stage of Industrial Evolution: Consolidation

As shown elsewhere (Meyer-Stamer et al., 2001), Sassuolo became a mature cluster in the 1990s. The number of mostly family-owned small–medium sized enterprises peaked at around 480 in the mid-1970s and dropped to 260 by the late 1990s. In contrast, the number of employees has been more or less stable. Since the 1980s, the industry has been going through a process of concentration, with no new tile manufacturers entering and many producers being taken over by other firms. However, although this is the dominant trend, it is not the only one. Restructuring of the cluster also displays a tendency towards polarization between large producers with a broad product portfolio and producers that are catering for niches, in particular special parts (for instance, small tiles for reliefs and decoration). There are two typical patterns of concentration. First, minority crossholdings between firms were transformed into more formal ties, creating a holding or group. The second pattern arose because most firms were family-owned, and in some cases families opted for selling the firm rather than going through the trouble of an intra-family management succession. One would expect that concentration would involve an attempt to create economies of scale. This has been the case regarding production, but much less so in terms of branding and sales. Instead, the formation of a group typically led to the better utilization of existing production capacity. As overall demand tended to grow, firms expanded. However, competitive pressure forced the groups to improve on efficiency of production. With the formation of a group, the production of a given brand of tiles could take

Upgrading in the tile industry of Italy, Spain and Brazil


place in any of the group’s factories, leading to improved utilization of production capacity. Recent changes in terms of segmentation strategies refer to the issue of branding. There are two distinct manifestations. First, some firms experiment with brand image transfer from other industries. This is part of an effort to widen the high end of the market. It is not yet clear whether this is really an important trend. Second, some large firms are producing private labels for large customers, especially home-centres; this appears to be an important trend. It is part of an effort to stimulate growth at the low end of the market. In other words, producers have a clear notion that upgrading by moving exclusively towards the upper end of the market is not a viable option. Instead, they are trying to widen both the high and the low end, to the detriment of the medium segment. The process of concentration has implications for collective action inside the cluster. As outlined earlier, until the 1980s, the predominant pattern was one of small and medium-sized family-owned companies. At that time, collective action was crucial to overcome competitive disadvantages due to small size, for instance in terms of building a brand image for Italian tiles and accessing foreign markets (Russo, 1985). Nowadays, more than half of the employees in the cluster work in companies which are part of a group (Brioschi, Brioschi and Cainelli, 2001), that is, that no longer qualify as small and medium-sized enterprises, and with increasing size they rely less on collective action. 3.2

Structure of the Value Chain: Downstream Activities

For the Italian tile producers, a key issue in upgrading has been the restructuring of downstream parts of the value chain. Starting in the 1970s, Italian firms systematically began to explore export markets. Some leading firms undertook serious efforts to create a brand identity abroad, but it was mainly the Tiles from Italy label which made the difference, because Italian tiles generally have both a different appearance and a more diversified size than those produced locally in target markets. Tile manufacturers also started to collaborate with external designers and artists to create innovative designs. At this time, this was essentially achieved by combining such a superior production technology and efficiency with superior production design and marketing that the Italian tile industry established a clear leadership position, driving traditional competitors such as German firms out of the market. More recently there has been a causal link between concentration and internationalization. Exports have doubled in the 1990s. Italian firms argue that there are minimum size requirements for a company to be able to be


Local enterprises in the global economy

present in several export markets, which are due to production capacity and size of the sales force. Internationalization so far means having sales representatives, and in many cases distribution warehouses, in target markets. Internationalization of production, however, is so far an unusual phenomenon and limited to a handful of the largest groups. Italy’s tile export market is highly concentrated, with about half of total exports going to Germany, France and the US. There has been low growth in European markets, whereas the US market displayed strong growth during the 1990s. Italy’s market share in the US is twice that of Spain, its main competitor. It is essential for Italian firms to defend their leadership in the US market if they want to increase their exports. In order to do so, they are increasingly concerned with the downstream part of the value chain. Italian tile manufacturers, especially the leading firms, play an active role in shaping international value chains, especially during the commercialization stage. This is particularly evident in the US, whereas in Europe (especially Germany), home-centres seem to be the more aggressive actors. The differences in market sophistication strongly influence the power of distributors vis-à-vis tile producers. In the German market, where tiles have a long tradition, the sales structure is much more mature than in the US, which is an emerging market in terms of tiles. In the US, Italian tile producers are systematically training sales personnel to deal with customers, who are starting from a low level of sophistication. In a mature market like Germany, this kind of approach would not be promising, but in the US this kind of service promises to create a competitive advantage over rivals like Spain. What then is the governance structure in the downstream part of the value chain? Despite the differences we emphasized above, the common feature across different national markets is the prevalence of network relationships. There are few spot markets for tiles; the predominating pattern is one of long-term supplier–customer relationships. The relative power of suppliers and customers differs between countries. Customers are stronger in the mature markets, but even there no side has a dominant role. Therefore, following the definition offered by Humphrey and Schmitz (2002, 26), the downstream part of the value chain is neither characterized by market relationships, nor is it a quasi-hierarchy. 3.3

Structure of the Value Chain: Upstream Activities

Now we examine the upstream relationships and their importance for upgrading among tile firms. The most important actors on this side are producers of glazing materials and capital goods. Today, Spanish producers

Upgrading in the tile industry of Italy, Spain and Brazil


dominate glazing materials production. Even the single remaining important Italian producer, Colorobbia, has moved central functions to Castellón. This is because Italian tile manufacturers have only outsourced routine production of glazing materials. With the strong tendency to produce porcelain tiles, there was less demand, and in particular less sophisticated demand, for glazing materials in Sassuolo. More relevant to the understanding of the evolution of the Sassuolo cluster, and very important for the overall tile value chain, is an analysis of the specialized capital goods producers. In fact, Italian firms are dominating the production of capital goods for the tile industry. More than 56 per cent of their productive capacity is located inside the Sassuolo cluster, and many of the other firms are not far away (for instance SACMI, the largest manufacturer, is located in Imola, some 40 km away). Thus, capital goods are not only a crucial part of the tile value chain but also of the cluster. Since the 1970s, capital goods producers have developed major innovations such as wet grinding, pressing with high tonnage machines, roller kilns, and increasingly sophisticated control instruments (Burzacchini, 2000). Wet grinding allowed a much better control of the mass that enters into the press, thus homogenizing product quality. Improved pressing had the same effect. The introduction of roller kilns shortened the firing process, thereby not only improving quality but also reducing production costs and improving the control of the firing process. Another important innovation (albeit pioneered by tile producers in Spain) was the single firing process that replaced the traditional two firing processes. More recently, in 2000, a machinery company in the cluster launched a revolutionary new product and production technology, called Lamina, which may pose a serious threat to conventional tiles. The development of such radical and incremental innovations occurred during a process of close interaction between tile manufacturers and capital goods firms. Although the capital goods manufacturers are the driving force when they come up with new types of equipment, they often install it at one of the tile factories and refine it through on-site experimentation. This will involve the free acquisition of a new piece of equipment, or at least at a substantial discount, for the tile manufacturer. This is the main explanation for the claim, sometimes heard in other countries, that Italian tile firms pay less for equipment. They also enjoy some privileges in terms of access to latest innovation. However, this does not involve exclusive relationships. The capital goods manufacturers sell their products to whoever is willing and able to pay. For instance, in the first half of the 1990s they sold many turnkey plants to Chinese firms, and there is no indication that this involved any outdated equipment. One might expect that there are highly conflictive relationships between


Local enterprises in the global economy

capital goods and tile manufactures, since the former are very active in creating competitors for the latter. But this conflict has not become too manifest. Even though Chinese or Brazilian firms may have the same equipment as Italian tile manufacturers, the style and elaboration of their products do not come close to that of Italian producers. The reason lies in tacit knowledge. First of all, tacit knowledge by definition is difficult to transfer. The importance of such knowledge is underrated by competitors. For example, Brazilian producers conceptualize technology transfer as the acquisition of equipment plus some training, whereas Italian suppliers are well aware of the fact that it is the other way around, that is some equipment plus a lot of training. But the suppliers are happy to leave customers abroad in the dark rather than enlightening them, since this reduces the level of conflict they have with domestic customers who are concerned about the emergence of new competitors using Italian technology. The relationship between capital goods and tile manufacturers in Sassuolo can be characterized as a mixture of networks and market-based relationships. There is clearly a very strong element of learning-by-interacting, but this normally involves a given capital goods producer and a limited set of pilot customers. These innovation networks are not eroding. Despite the fact that Italian machinery is sold to competitors all over the world, the interaction between capital goods and tile manufacturers remains strong and plays a crucial role in creating a competitive advantage for either side. 3.4

Local Governance and the Structure of Supporting Institutions

While Sassuolo came close to the textbook model of Italian industrial districts until the 1980s, this is no longer the case. Large firms and groups are emerging, which are vertically integrated. Informal horizontal collaboration between tile producers is limited and seems to be declining. There is close, informal collaboration between machinery manufacturers and tile producers, but then the machinery manufacturers sell to customers from all over the world. To what extent is there formal collaboration to create the active dimension of collective efficiency? The two main associations in the cluster are: Assopiastrelle, the association of Italian tile manufacturers; and, ACIMAC, the association of capital goods manufacturers. By seeing Assopiastrelle’s impressive headquarters and realizing that Cersaie, the annual trade fair in Bologna, is the big event in the tile industry, one may get the impression that the Sassuolo cluster is good at collective action, thus verifying what is commonly perceived as a strong point of Italian industrial districts. Yet a closer look supports this expectation only in part. With the exception of Cersaie, Assopiastrelle is pursuing a number of activities that can be subsumed under three main

Upgrading in the tile industry of Italy, Spain and Brazil


headings: (i) services for member firms; (ii) representation of the sector; and, (iii) research and documentation of trends in the industry. The most effective services relate to marketing. Cersaie is an example of collective marketing, particularly because Italian producers notoriously obtain much larger and more conveniently located stands at the fair than their foreign competitors. The organization of a strong Italian presence at trade fairs in other countries is another example. However, there are collective failures, particularly in physical infrastructure. The decay of the Sassuolo infrastructure casts serious doubt on the alleged effectiveness of Assopiastrelle as a politically influential association. ACIMAC is also an important association. Its profile is similar to that of Assopiastrelle, but it appears more effective. In terms of services, it promotes the tile machinery industry, for example, by organizing joint stands at foreign fairs. It also supports member firms by providing information on technical standards and quality issues. Moreover, it organizes short-term training courses abroad which cater to the employees of customer firms. In terms of interest representation, it represents the sector vis-à-vis Italian embassies overseas and it lobbies at international negotiations about tariff reductions. In terms of information, ACIMAC is the most important place to access information about the global tile industry. Since ACIMAC members sell to firms all over the world, it is essential to be up-to-date with sector trends all over the world. Accordingly, the trade journal, which is sponsored by ACIMAC (Ceramic World Review), is an excellent source of information on the tile industry. With the exception of these two associations, the structure of supporting institutions is remarkably underdeveloped. There is a modest technology institute related to the sector, the Centro Ceramico Bologna (CCB), which recently set up a small affiliate in Sassuolo. CCB is part of the ERVET system of business support institutions which is administered by the regional government. It carries out research on several issues related to ceramics, but its main activity is related to testing and certification, and this seems to be clearly separated from research activities. CCB is collaborating with the Universities of Bologna and Modena to offer engineering courses with a specialization in ceramics. There is also the Scuola de Arte in Modena which is an important source of design talent for the industry, although it does not specialize in tiles. Finally, there is Cerform, the vocational training centre in Sassuolo which, after intense negotiation between municipal governments and the two associations, was devoted exclusively to the training of ceramics professionals since the mid-1990s. Its annual output amounts to some 60 graduates, which falls short of industry requirements. Collective action in the Sassuolo cluster exists, but, as we will see, to a lesser degree than in Castellón. The reasons are hard to pin down.


Local enterprises in the global economy

Complacency seems to be one cause. The other explanation could lie in changes in another part of the value chain, namely in distribution and sales. As argued above, the concentration process in the tile industry is driven in part by marketing considerations. The concentration process in itself appears to contribute to the deterioration of local governance, that is firms are getting larger and thus are more confident to be able to solve problems on their own. But there is a further aspect that connects marketing and local governance. As mentioned above, the US market is the main source of growth. A fierce struggle is going on in this market as Italian firms compete against Spanish firms, and there is particularly strong competition occurring between Italian firms. Our hypothesis is that this is creating a constellation that erodes the trust accumulated in earlier periods and creates a situation whereby collaboration between these firms is unlikely.

4. THE EVOLUTION OF THE CASTELLÓN CLUSTER The Castellón cluster differs from the Sassuolo cluster in three important respects: ●

It is not yet a mature cluster, that is there are still entries, there is little if any concentration, and companies are predominantly small- or medium-sized. Collective action inside the cluster, in particular among tile producers, is strong. Glazing manufacturers play a very important role in cluster upgrading, whereas the local capital goods industry is weak.

Tile production in Spain has grown enormously in the 1980s, with total output doubling. During the course of the 1990s, production grew almost threefold. In 1998, there were 242 tile manufacturers in Spain, of which 190 were located in the Castellón cluster (accounting for 93 per cent of Spanish production). The overall number of employees in tile firms grew from 16800 in 1995 to 23200 in 1999. The Castellón region has a long tradition in tile manufacturing. The process of industry modernization started in the late 1950s with the introduction of electrical presses that were imported from Italy and which substituted traditional man-powered presses. The next important step, beginning in the mid-1960s, was the introduction of tunnel kilns that substituted the traditional circular ‘Arab kilns’. The tunnel kilns were further upgraded in the 1970s with the introduction of second firing.

Upgrading in the tile industry of Italy, Spain and Brazil


After the second oil crisis, firms started to increase their export effort, in particular towards Portugal and France. This was encouraged by a drawback scheme introduced by the national government to reduce the deficit in the trade balance. This was important for the tile manufacturers since they continued to depend on imports of capital goods from Italy. Looking back over that period, industrialists admit that the product quality was inferior but it improved with increasing exports, reflecting a process of learning by exporting. 4.1

Industrial Evolution

The turning point for the cluster was the beginning of the 1980s, when several factors converged to improve the competitiveness of the industry. Government-sponsored efforts to promote the local capital goods industry led both to failure and success. Attempts to build up a local production base for kilns failed, but this effort led to an improved bargaining position vis-à-vis the Italian suppliers, who reduced their prices substantially and started to sell the latest generation equipment in Spain. In 1981, the region was connected to a pipeline that brought natural gas from Algeria. This was essential in order to employ latest generation kilns, and it implied a massive reduction of energy costs. The earlier generation of kilns had been based on oil. In 1984, the single firing production process was launched. The story of massive upgrading in the cluster actually starts with the single firing process. This was the first major innovation to emerge from the cluster. Before that, the industry used a double firing process, first firing the biscuit alone, then glazing and firing it again. Spanish clay is red because it has a higher iron content which leads to different sintering characteristics. This was the point of departure for single firing in the Spanish industry. Italian kilns were designed to work with white clay, using double firing, therefore, the quality of the Spanish final product did not match that of Italian tiles. Between 1981 and 1983, engineers from a tile manufacturer (Zirconio), a producer of glazing materials (Torrecid), and what was to become the Institute of Ceramics Technology (ITC) (originally part of the University of Valencia and then the University of Castellón), developed a new process. It involved the development of different glazing materials and the adaptation of Italian kilns. The result was a process that was superior to previous Spanish and Italian processes both in terms of production efficiency and quality of the final product. Single firing spread quickly throughout the cluster, reaching the majority of firms by the end of the 1980s. However, it was more than a one-off radical innovation. It shaped the paradigm of the cluster, which has three main elements:

186 ●

Local enterprises in the global economy

Competitiveness is originally based on technological excellence. It is engineers who rule in the cluster. Interaction between tile manufacturers and producers of glazing materials is strong, with the latter being an important push factor in terms of constant upgrading. Interaction between firms and local supporting institutions, in particular ITC, is strong and a key element in technological upgrading.

The second half of the 1980s laid the foundation for the massive expansion of the cluster in the 1990s. Tile manufacturers and producers of glazing materials expanded production capacity and upgraded products. In the case of tile manufacturers, rapid growth during the 1990s was due to both a strong local economy and a strong increase in demand from abroad. Producers of glazing materials and tile manufacturers continue to enter the cluster. The typical pattern of family-owned firms also continues. The vast majority of firms are small- or medium-sized businesses. The typical firm has some 200 employees. So far there is no concentration trend in the tile industry, and multibrand firms are rare. There appears to be sufficient sales potential for everybody, and there has been no generational change that might stimulate mergers or takeovers. Moreover, in terms of production, an internationalization trend is not apparent. There are several reasons for this. First, Castellón displays remarkable dynamic locational advantages, particularly in terms of both formal and informal communication and collaboration between firms. It is not by chance that Marazzi, the largest Italian firm, set up its Spanish subsidiary in Castellón. Second, the average firm size is smaller than in Italy, and in fact most firms are small–medium sized enterprises (SMEs), with insufficient managerial capability to deal with the challenges of managing a multinational operation. Third, there is no necessity to internationalize since the sector is doing exceptionally well with its current modus operandi. Despite their smaller size, manufacturers in Castellón prefer to have a broad spectrum of products, that is both floor and wall tiles, low- to highend, and glossy as well as rustic. To explain this, manufacturers point to the fact that distributors tend to deal with a limited number of producers, and as a result gaps in the manufacturer’s product portfolio may lead to a discontinuation of business relations with a given distributor. Some manufacturers pursue some form of rough market segmentation, others a country segmentation (this kind of product is much accepted in Germany). However, the segmentation effort inside the industry seems to be very limited so far; the overall pattern is similar to the product-driven strategy that can be observed in Italy.

Upgrading in the tile industry of Italy, Spain and Brazil


The marketing competence and sales system of Italian firms appears to be superior (Generalitat et al., 1999, 87), but there seems to be only a limited effort among Castellón firms to upgrade in this respect. They have, however, copied the Italians by promoting a ‘Tiles of Spain’ label. 4.2

Structure of the Value Chain: Downstream Activities

Even though the export ratio of the Spanish tile industry is lower than that of Italy, Spain is by far the second most important tile exporter. Both production and exports are growing, but since domestic demand is also growing, the export share remains stable at around 55 per cent of value and 51 per cent of volume. Spanish exports are less concentrated than those of Italy. However, the rapid growth of the US market, and the strong presence of Spanish producers there, means that this may change in the future. In 1999, the seven largest markets accounted for 50 per cent of exports, and the 14 largest accounted for 66 per cent. The Spanish tile industry is involved in a global value chain that has a network structure, that is it involves neither arm’s length relationships nor hierarchical relationships. This applies both to inputs and sales. It seems that tile manufacturers are in a strong albeit not dominant position vis-àvis buyers. In Spain, there are more than 200 tile manufacturers, and the largest among them attend the majority of the 3000 distributors that currently exist in Spain (Bigné, 1998), as well as a number of customers abroad. It is difficult to imagine that either side could acquire a dominating position. However, it is notable that, in contrast to Italian producers, Spanish firms display little propensity for forward integration. Given the fact that there is a clear concentration process at the commercialization stage, this may prove to be a strategic error in the long term. At the same time, it must be acknowledged that even the large Spanish firms are probably too small and do not have the necessary capital to pursue a strategy of forward integration into commercialization, particularly in foreign markets. In an analysis of the cluster conducted in 1999, the consultants of Cluster Competitiveness recommended that firms should seek a clearer market position, forward integration into sales, and internationalization (Generalitat et al., 1999). For an outsider, this appears to be a plausible suggestion, not least because it would replicate the successful Italian strategy. However, the suggestion is less plausible for the cluster firms. The firms are already hugely successful with their current business model. The growth and profitability of the Spanish industry is higher than those of their Italian competitors.


Local enterprises in the global economy

At the same time, it is notable that the Cluster Competitiveness analysis is an important point of reference for industry actors, especially since it was based on an exercise which strongly involved them. In the future, however, a learning and discussion process could occur in the industry that leads to a re-evaluation of the recommendations, especially in the light of the discernible restructuring of the building materials commercialization chain. 4.3

Structure of the Value Chain: Upstream Activities

When comparing Sassuolo, differences also arise in Castellón’s upstream activities. Whereas Italian manufacturers used predominantly white clay, the Spanish prefer red clay since it is cheaper and more readily available. Atomization (milling), and in several cases the production of biscuits, has been outsourced to specialized firms that have usually been jointly established by tile manufacturers. There are clear economies of scale in atomization, and an efficient atomization plant would have been too large for most of the tile manufacturers in the 1980s. While this may appear perfectly rational, it is important to note that the manufacturers that jointly own an atomizing operation are often rivals in all other respects. There is, in other words, less vertical integration in Castellón than in Sassuolo. Spanish firms continue to depend on Italian manufacturers for most types of capital goods; the major exception is glazing equipment, where the close proximity to leading glazing producers created an opportunity for local machinery producers. However, today this is based on mutual dependency. Demand for capital goods is higher in Spain than in Italy, and Spanish producers appear to be more demanding in terms of technology than producers in Italy. Moreover, they tend to be very competent in terms of specifying their requirements, rather than just purchasing what the Italians want to sell. Accordingly, it is paramount that Italian capital goods manufacturers have a close contact with Spanish tile manufacturers in order to remain at the cutting edge. The fact that ACIMAC is considering setting up a branch in Castellón is a good indication of the relationship that is evolving. Just as the interaction between tile producers and capital goods manufacturers has shaped the evolution of the Sassuolo cluster, the interaction between tile and glazing producers establishes the paradigm of the Castellón cluster. The producers of glazing materials have gone through an impressive upgrading process since the 1980s. Not only was there an extraordinary growth in exports, which grew sixfold between 1990 and 1999, but also a massive internationalization push. Many of the 24 firms from Castellón set up factories or at least distributors in many of the main tile producing countries.

Upgrading in the tile industry of Italy, Spain and Brazil


Glazing materials producers continue to be very important for constant product upgrading, even though firms try to strengthen their internal design capacity in order to have some unique designs. Nevertheless, the design departments of glazing producers continue to be larger than those of tile manufacturers, and the top graduates of design courses join glazing rather than tile firms. It appears that the power role of design specialists is stronger in glazing than in tile firms; production engineers dominate the latter. Therefore, the core competence in design is moving from tile to glazing manufacturers. While a typical tile manufacturer in Castellón may have three designers, the leading producer of glazing materials has 40 in Castellón and even more when affiliates are included. The total number of employees is less than 1000. They are distributed across factories in eight countries with representative offices in a number of others. 4.4

Local Governance and the Structure of Supporting Institutions

One of the distinctive features of the Castellón cluster is the density and competence of the supporting institutions. First, there are several business associations. ASCER is the association of tile manufacturers and all local tile manufacturers are affiliated. It appears to be the main actor in the cluster in terms of facilitating collective action and strategic initiatives. It provides information for, and about, the industry. It also articulates the industry’s demands vis-à-vis government, from the local through to the regional, national and EU level. Furthermore, it organizes joint purchasing for gas, electricity, telephone and mobile telephones. ASCER recruits approximately 20 full- and part-time professionals. However, in contrast to its Italian counterpart in the case of Cersaie, it does not participate in the organization of Cevisama, the Spanish ceramic tile fair. In a similar way to other business associations in the cluster, ASCER does not appear demonstrative, but it fulfils its essential tasks in a very effective way, both in terms of political representation and services to affiliated firms. Regarding the latter, joint purchasing is probably the most tangible service provided. However, further evolution of joint purchasing is difficult to predict due to the emergence of e-commerce and the entry of private e-commerce operators within the purchasing business. ANFFECC is the association of 24 glazing producers. It is administered by a legal firm; a feature which indicates that it operates differently from ASCER. It primarily operates as a lobbying organization vis-à-vis government. This is because glazing production has a serious environmental impact, therefore constant negotiations with government bodies are essential. In terms of other activities, ANFFECC acts as a convenor at commissions with professionals from firms. Two years ago, ANFFECC also started


Local enterprises in the global economy

to organize short-term courses for firm employees. ASEBEC is the association of the capital goods manufacturers. Its status reflects the fact that the local capital goods industry is relatively weak. Thirty-five firms, with an average of 24 employees each, are affiliated. The main professional organization is the Associación de Tecnicos de Cerámica. There is also the Chamber of Industry and Commerce (Cámara Oficial de Comercio, Industria y Navegación de Castellón). Since 1992, its activities have included organizing the biannual technical–scientific forum Qualicer. Apart from these associations, there is a well-developed set of other institutions. The Ceramics Technology Institute (ITC) emerged from an institute for chemical technology that was founded at the University of Valencia in 1969 and increasingly focused on the tile industry in the 1970s. In 1983, part of the institute was relocated to the Castellón university campus. In 1984, the Research Association of the Ceramics Industry (Asociación de Investigación de las Industrias Cerámicas, AICE) was founded to facilitate cooperation between ITC and firms. In 1991, the first chemical engineers to have specialized in ceramics graduated from the institute. In 1992/93, the name changed to ITC, and the institute was integrated into the newly independent University of Castellón. ITC’s activities comprise training professionals for the tile and glazing industry, conducting tests for firms, and working both independently and with firms on R&D projects. ITC has its own pilot plant for experimentation with tile production issues. The Institute for the Promotion of Ceramics Design (Asociación para la promoción del diseño cerámico, ALICER) was founded in 1993. Its main activities include training and support for firms. ALICER offers a five-year course at tertiary level. Glazing firms employ the majority of the 20 students that graduate every year. In contrast, joint projects with firms are almost exclusively conducted with tile manufacturers. Before it became an independent institute, ALICER was a department within ITC. It was created because the perception was spreading that limited design competence was a severe competitive disadvantage for Spanish tile firms. As the department grew, ITC’s management decided to spin it off since they felt that design was beyond the main focus of ITC. Another important supporting institution is Bancaixa, the region’s bank, which is the main source of credit for the firms. Since the performance of the sector is essential for the bank’s performance, it plays a leading role in organizing a process of reflection inside the cluster. In 1999, it organized a series of seminars with key actors from the cluster, as well as from Italy, and it co-sponsored the analytical work of the Cluster Competitiveness consultancy firm (Generalitat et al., 1999). Taking all this into account, a picture of strong local governance struc-

Upgrading in the tile industry of Italy, Spain and Brazil


tures emerges. The upgrading of the cluster is driven by a collective effort, with the private sector taking the lead.

5. LATECOMER INDUSTRIALIZATION AND VALUE CHAINS IN THE TILE INDUSTRY: THE CASE OF SANTA CATARINA, BRAZIL In this section we will investigate upgrading in the tile cluster in Santa Catarina (SC), Brazil. What does the evolution of the SC cluster tell us about the functioning of the global value chain in the tile business, particularly in terms of the interaction between the elements that are further upstream? We have mentioned that there is an internal tension in the clusters of Sassuolo and Castellón. The producers of capital goods and glazes appreciate the existence of customers in emerging countries, whereas the tile producers are unhappy about the emergence of new competitors. We will show that in a place like Brazil the rivalry among and between capital goods and glaze producers creates a constellation that benefits the emergence of a competitive tile industry. As firms in the cluster pursue a similar competitive strategy to the one we observed in the Italian case, that is forward integration, upgrading is based on individual company efforts, not on collective action. 5.1

Value Chain Restructuring and Competitive Advantage

The state of SC is Brazil’s most significant tile cluster. Located in the city of Criciúma, there are a total of three large tile producers, approximately ten medium-sized producers and seven glaze producers. The first phase of evolution was during the 1950s to 1970s, when local firms acquired a basic production capability. The second phase, in the 1970s and 1980s, was marked by an expansion of production capacity in order to satisfy a growing market (today this is the Western world’s largest market), regardless of product quality. The Brazilian tile market was a seller’s market at this time. The third phase began when the sector fell into a deep crisis in 1989, caused by macroeconomic crisis. By 1991, sales had dropped by a third, one of the large firms filed for debtor’s protection, and a number of other firms nearly followed suit. Firms reacted by defining technological upgrading as the way out, opting for quality instead of quantity. As shown elsewhere (Meyer-Stamer, 1998), collective action played an important role in this upgrading of processes and products. Since the late 1990s, the companies’ upgrading effort has concentrated on marketing and sales. The basic idea was to create a structure that


Local enterprises in the global economy

persuaded the customer of the superior quality of tiles from SC when they were at the point of sale, thus creating a competitive advantage vis-à-vis two other clusters in the state of São Paulo. SC companies have invested in training the sales force, evidenced by their decision to select the best ceramics technicians to work in the sales department, whereas in the previous period these people would have been placed in a supervisory role on the shop floor. Most of the companies have also invested in exhibition space (showrooms) and, to some extent, in specialized upmarket shops. Two of the large firms have set up exclusive franchising networks. Inside the shops, well-trained salespeople attend customers, and architects offer design proposals free of charge that involve combinations of floor- and wall tiles. By analysing customer complaints, firms realized that there was little sense in producing high-quality tiles if the tiler was not sufficiently competent at selling them. The most radical manner of dealing with this problem is forward integration: producing not only tiles but also marketing and technical assistance. There are three stages through which firms arrive at this point: 1. 2.


Firms start to train tilers at their own expense, organizing courses both at their home location and elsewhere in the country. In some cases, this includes providing the tiler with up-to-date tools. Firms not only train but also certify tilers, and offer customers a five to ten year warranty when they employ the certified tiler to set the tiles using the firm’s own special cement. In particular, this full package is offered by one of the large firms with their own shops, so that the customer only has the shop to deal with, hence, they pay only the one bill. Firms start to train and employ their own tilers, offering the full package to construction companies. This is not only based on quality considerations but also on the observation that the cost of setting the tiles is higher than the cost of producing them, hence, the full package opens up opportunities to increase margins.

5.2 Upgrading and Local Governance in Santa Catarina: a Value Chain Perspective Brazilian tile firms are integrated into international value chains, but in a completely different way from, say, footwear or garment manufacturers in the same country. Whereas the latter rely on foreign firms in downstream activities, that is in marketing, the tile firms rely on foreign firms in upstream activities, that is in the provision of capital goods and crucial inputs. Dependence on foreign suppliers of capital goods is not a particularly

Upgrading in the tile industry of Italy, Spain and Brazil


unusual feature for an industry in a newly industrializing country. In fact, in tile production almost all firms depend on Italian capital goods manufacturers, including firms in Spain. Firms in Brazil report that access to the latest vintage equipment is not restricted, though machine producers collaborate with Italian tile producers in the development of new machines. A representative of the machine producers’ association confirmed that, to the dismay of Italian tile manufacturers, machinery is sold to whoever is able to pay. Input manufacturers, especially producers of glazing materials, are the other technology drivers in the tile industry. The second half of the 1990s saw a profound change in terms of the structure of the glazing business in Brazil. Until the early 1990s, three transnational corporation (TNC) affiliates produced glazing materials in Brazil (from Italy, Germany, and the US). With the strengthening of the Spanish tile cluster, Spanish glazing manufacturers founded affiliates in Brazil, mostly in SC. Today, the cluster comprises seven companies that produce glazes locally; two of them are local firms. There are a total of 28 glaze companies in Brazil. This means there are more glaze producers in Brazil than in Spain. As a result, total Brazilian tile production is lower; this means there is intense competition between glazing manufacturers. In order to create a competitive advantage, they changed their behaviour in a profound way. According to tile manufacturers, in the old days a glazing manufacturer would come up with a new glazing variety, drop it at the reception and leave it to the tile manufacturer to figure out what to do with it. Beginning in 1996/1997, the glazing manufacturers set up development and design teams, offering a full service to tile manufacturers, comprising the design, the glazing material needed to produce it, and technical assistance in mastering new designs and solving problems in the production process. We found no indication that Brazilian tile manufacturers face discrimination. Instead, it appears to be the other way around, that is Spanish glazing manufacturers draw on know-how available at headquarters in Castellón and at Spanish R&D centres if they cannot solve production problems locally. Drawing on all these observations, it becomes clear that Brazilian tile producers benefit from the fierce rivalry among Italian machinery and Spanish glaze producers. As firms from both groups are constantly looking for new ways to broaden their customer base and establish a competitive advantage, the idea that they might withhold up-to-date technology from Brazilian customers is unlikely. To phrase it differently, the tile sector in SC benefits from the structure of the upstream part of the international value chain in tile production. The pattern of interaction with machinery producers differs from that of glaze producers. The relationships with machinery manufacturers appear


Local enterprises in the global economy

to be mostly market-based, that is there is little if any joint development (learning-by-interacting). The situation differs with respect to glaze producers. There is relatively close cooperation between tile and glaze producers, with technicians from glaze producers often included in solving production problems at their customer’s site. The pattern of interaction between tile and glaze producers is thus predominantly of the network type (though it is important to note that this is not about innovation networks; the pattern of interaction in Brazil is different from that in Spain). Looking at the internal structure of the cluster in the course of the 1990s, there was a strong shift in the relative importance of intra-company and collective upgrading efforts. Collective action was important in the first half of the 1990s but decreased thereafter. This is closely related to the companies’ effort to move further downstream in the value chain. The competencies required for this are localized within the companies and they try to keep their specific efforts secret. There is little collective action left. The joint initiatives of the early 1990s either achieved their objective or fizzled out. Successful collective efforts, documented elsewhere (Meyer-Stamer, 1998), did not create a lasting ‘collective efficiency’ mindset inside the cluster. The reorganization and strong competition in the marketing stage of the value chain have undermined this.



Analysing the tile industry from a combined cluster and value chain perspective is a useful exercise. The cluster perspective is helpful in understanding learning-by-interaction and the creation of locational advantages. The value chain perspective is helpful in putting the evolution of the industry into perspective, as efforts to create a competitive advantage are increasingly chain-related, whereas the relative importance of locational advantages created through deliberate collective action is declining, at least in Italy and Brazil. This chapter has shown that product and process innovation in the tile industry is to a large extent driven by suppliers. This does not only apply, as one would expect, to the suppliers of new machinery that encourage process and product innovation. It also applies to glaze producers that facilitate product innovation. Two groups of suppliers play a key role in innovation, namely Italian capital goods manufacturers and Spanish glaze producers. Their goal is to create demand for their products: ●

Machinery producers do not only come up with new machines to do the same things better. They also develop new products that require

Upgrading in the tile industry of Italy, Spain and Brazil


new machinery. A machinery producer that comes up with a new product in this way creates the demand for its machinery. In some cases the machinery producer even creates a captive market for subsequent services. Glaze producers do not just produce standardized glazing materials. Rather, they have large in-house design departments which offer entire design collections to tile producers, usually for free. Accepting a glaze producer’s design proposal means purchasing their glazing materials as well.

This innovation behaviour creates a situation that involves four groups of actors (Figure 7.1). They are the tile manufacturers in Sassuolo, the tile Machinery producers (mostly Sassuolo)

Glazing producers (mostly Castellón)

Tile manufacturers in Sassuolo

Tile manufacturers in Castellón

Source: Author’s research

Figure 7.1

Key actors and interaction in the tile industry

manufacturers in Castellón, the machinery producers that are mainly located in Sassuolo and the glaze producers that are mainly located in Castellón. There are four types of interaction: ● ●

Strong localized rivalry within each of the four groups. Competition between the tile manufacturers from the two different locations which are essentially competing for the same markets. Formal and informal cooperation between groups. This does not just happen locally but also, as indicated by the diagonal arrows in Figure 7.1, between clusters. Rivalry between the machinery producers and glaze producers. The machinery producers prefer to develop innovation that reduces the relevance of glazes for a differentiated product, and ideally makes


Local enterprises in the global economy

glazes altogether unnecessary. The glaze producers concentrate on innovations that permit new, differentiated products without major investment in new equipment. In other words, one group tries to undermine the other group’s market position. Latecomers such as the cluster in Santa Catarina benefit from the rivalry within and between the leading clusters. While these latecomers are rarely partners of the machinery and glazing material suppliers in their experimentation of new products, they have access to the new products with only minor delays. It is important to recognize that the competitive strategies of suppliers induce a profound restructuring of the value chain. There was a time when innovative, differentiated design, based on an in-house effort, was the main competitive weapon of tile manufacturers. However, much of the design function has been taken over by machinery and glaze producers, which are thus expecting to develop additional sources of innovation- and differentiation-based rents. For the tile producers this means that they have to look for other ways to create competitive advantage and rents, and they find them downstream, in terms of customer services, distribution and sales. From this perspective, there is a tendency by the main groups of actors in the tile business to follow a generalized movement downstream. Tile manufactures do not only perceive that a stronger position at the distribution and sales stage improves their competitive position vis-à-vis their competitors. They also perceive that the margins at this stage tend to be higher, so that a strong position here may enhance the company’s profitability. In the end, one may also interpret forward integration as an effort to create the conditions for rent that is based on controlling access to the final customer. This pattern is clearly discernible in the cases of the Sassuolo and the Santa Catarina cluster. However, it remains doubtful whether the tile producers can sustain this strategy. Any change in the structure of the value chain is not the outcome of individual action alone. One of the characteristic features of the tile business is reflexive structural change:4 there is constant discussion, both inside the clusters and between them, organized by ACIMAC (for example ACIMAC 2000). This discussion is not just about general industry, technology and market trends but also about changes in the structure of distribution and sales. As the construction materials retail sector is something of a latecomer in terms of consolidation, actors in the tile industry closely observe trends in the distribution and sales segment of other industries where restructuring is more advanced. They also try to pre-empt a kind of structural change which would weaken their competitive position and strengthen that of other actors; the obvious adversary being large retail

Upgrading in the tile industry of Italy, Spain and Brazil


chains. At the same time, it is unlikely that the efforts of tile producers to build up a dominating position in commercialization will go unnoticed. As there is more than a handful of large players in the construction material sector, one might expect that they would pre-empt any single competitor’s effort to establish themselves as a dominant actor. While the outcome of the tile producer’s functional upgrading strategy remains uncertain, the importance of another analytical dimension becomes more visible. Rivalry within and between key stages of the value chain leads to chain restructuring. This is most clearly discernible in the Italian case, where the industry is reaching maturity. These two elements, chain restructuring and maturation of the industry, are causally linked. Maturation coincides with consolidation, that is the creation of large companies or economic groups. In contrast to small or medium-sized companies, they have the resources that are necessary for forward integration into distribution and sales. This changes the pattern of relationship between tile manufacturers and makes cooperation less likely. Forward integration creates a pattern of competition with winners and losers, quite different from the mostly-winner pattern of a past with high growth rates. Some manufacturers may succeed in their attempt to establish captive sales chains, but most of them will not. Moreover, as producers consolidate, they perceive less necessity for horizontal cooperation. Both factors weaken local network governance inside a cluster. Spanish producers have not yet opted for forward integration, and from this perspective it is no coincidence that collective action is still strong inside this cluster or that the cluster is still in a growth phase, that is an earlier stage of the life cycle. The evolution of the tile industry thus reinforces one of the key points of this book, namely that analysing clusters from only a local collective action perspective is not advisable. By including the value chain perspective, it takes us an important step further in understanding the latitude of local upgrading efforts. It is essential to introduce further concepts into cluster research, and the life cycle concept is perhaps the most relevant one. As shown by Grabher (1993) and Staber (2001), some clusters decline. This underlines the importance of understanding the life cycle of clusters. Research on Italian industrial districts has emphasized the profound change in internal structures and external relationships that occurred in the 1980s and 1990s (Belussi, 1999). The evolution along the life cycle has implications not only for the internal governance structure of clusters but also for the structure of the value chains they are part of. Investigating this kind of interrelationship may take the understanding of the dynamics and potential of clusters an important step forward.


Local enterprises in the global economy

NOTES 1. Field research for this case study was conducted in 2000. It involved interviews with firms and institutions in Italy, Spain and Brazil and visits to the construction materials fair in Brazil and the tile fair in Italy. 2. Other suppliers, such as producers of clay, are not discussed here because they neither shape the technological evolution of the tile industry nor are they particularly powerful. 3. In August 2001, Degussa sold its glaze-related business unit to Ferro. 4. This relates to the concept of reflexivity as formulated by Giddens (1984, 3): ‘. . . it is useful to speak of reflexivity as grounded in the continuous monitoring of action which human beings display and expect others to display’.

REFERENCES ACIMAC (2000), 2º Meeting Mondiale della Ceramica (2nd world meeting on ceramics), Conference proceedings, Modena, Italy, 26 May. Belussi, F. (1999), ‘Policies for the development of knowledge-intensive local production system’, Cambridge Journal of Economics, 23, 729–47. Bigné, E. (1998), ‘Obstáculos al Comercio. Remoción de Obstáculos Comerciales’. Foro de Debate entorno a la Industria Azulejera (forum on the tile industry), Bancaixa Obra Social, Segunda Sesión, 10 November. Brioschi, F., M.S. Brioschi and G. Cainelli (2001), From the Industrial District to the District Group. An Insight into the Evolution of Local Capitalism in Italy, Milan: Istituto di Ricerca sulla Dinamica dei Sistemi Economici. Burzacchini, P. (2000), ‘Porcelain Tile, its History and Development’, Ceramic World Review, No. 37, pp. 96–103. Generalitat V., Bancaja and Cluster Competitividad (1999), ‘Iniciativa de refuerzo de la competitividad’, El cluster azulejero en Castellón (the tile industry cluster in Castellón), Valencia. Gereffi, G. (1996), ‘Global commodity chains: new forms of coordination and control among nations and firms in international industries’, Competition and Change, 4, 427–39. Giddens, A. (1984), The Constitution of Society. Outline of the Theory of Structuration, Cambridge: Polity Press. Grabher, G. (1993), ‘The weakness of strong ties: the lock-in of regional development in the Ruhr area’, in G. Grabher (ed.), The Embedded Firm. On the Socioeconomics of Industrial Networks, London, New York: Routledge, pp. 255–77. Humphrey, J. and H. Schmitz (2002), ‘Developing country firms in the world economy: governance and upgrading in global value chains’, INEF Report 61, Duisburg: Institut für Entwicklung und Frieden. Meyer-Stamer, J. (1998), ‘Path dependence in regional development: persistence and change in three industrial clusters in Santa Catarina/Brazil’, World Development, 26 (8), 1495–511. Meyer-Stamer, J., C. Maggi and S. Seibel (2001), ‘Improving upon nature. Creating competitive advantage in ceramic tile clusters in Italy, Spain, and Brazil’, INEF Report 54, Duisburg: Institut für Entwicklung und Frieden. Russo, M. (1985), ‘Technical change and the industrial district: the role of inter-firm

Upgrading in the tile industry of Italy, Spain and Brazil


relations in the growth and transformation of ceramic tile production in Italy’, Research Policy, 14, 329–43. Staber, U. (2001), ‘Spatial proximity and firm survival in a declining industrial district: the case of knitwear firms in Baden-Württemberg’, Regional Studies, 35 (4), 329–41.

8. Local upgrading strategies in response to global challenges: the surgical instrument cluster of Tuttlingen, Germany Gerhard Halder1 1.


The small town of Tuttlingen in Southern Germany has occupied an unrivalled position in the international surgical instruments industry. It now faces unrelenting price competition from producers in low wage countries and more exacting demands from increasingly powerful customers. Together, these factors are forcing prices to decline and quality to increase. Consequently, Tuttlingen firms have had to re-assess their position. Pressures in the traditional surgical instrument industry coincide with opportunities to make new products, notably instruments for minimally invasive surgery and surgical implants. This chapter examines if and how the Tuttlingen cluster of surgical instrument providers has responded to these challenges. First, there is a challenge to reduce costs. Public health care systems are under pressure in most industrialized countries, since demographic and disease profiles have changed, and new and costly methods have been introduced (Knappe et al., 2000). However, the cost challenge is also supplydriven, since new locations in developing countries entered the market (Nadvi and Halder, 2002). Second, advances in several technological fields have led to challenges for new product development. New materials and surface treatment including biomedical devices, fibreoptics and laser technology, microsystems technology and the marriage of electronics, computing and information technology with surgery, have resulted in new medical equipment (Grönemeyer, 2000). Responding to these challenges requires substantial upgrading, and as Humphrey and Schmitz (2000) noted, upgrading requires substantial investment. This investment can not be restricted to financial assets, nor to machinery and materials. Bell and Albu (1999) have analytically distin200

Local upgrading strategies in response to global challenges


guished between the production system and the knowledge system to unravel the need to go behind material flows. In particular, radical upgrading often requires altering the knowledge base. The academic literature shows many examples of how knowledge floats, in particular in agglomerations of firms. The literature on industrial clusters (Porter, 1990; 1998; Enright, 1996) emphasizes the stimulating effect of competition on knowledge evolution. In contrast, the industrial district approach leads us to look at the potential gains from cooperation and the importance of knowledge spillover and socioeconomic features (Brusco, 1992; Becattini, 1991; Pyke and Sengenberger, 1992; for comparison, see Martin and Sunley, 1996). In line with this body of work, the ‘collective efficiency’ approach highlights the success of active cooperation within clusters in a problem solving way (Schmitz, 1995; Schmitz and Nadvi, 1999). However, recent research on global value chains has highlighted the importance of external linkages for cluster success (Humphrey and Schmitz, 2000; Gereffi and Kaplinsky, 2001). Other authors have stressed that joint action at the cluster level rarely changes the stock of a cluster’s knowledge in a radical way (Bell and Albu, 1999). Instead, knowledge spillover and joint action at cluster level spreads already existing knowledge, and leads to incremental rather than radical upgrading. Thus, increasing globalization and rapid technological development requires the management of external relationships. External knowledge has to be attracted and combined with the stock of the cluster’s knowledge leading to new solutions. However, in the course of these observations, local relations or local joint action becomes by no means irrelevant. In contrast, we find that small firms do not have the resources to establish external relations in a promising way. Here, collective initiatives may be a way to overcome small-scale restrictions. Conceptually, this chapter builds on recent literature which has stressed the need to combine the analysis of internal and external knowledge flows (Bell and Albu, 1999), and the coordination of global and local relationships (Humphrey and Schmitz, 2000). It analyses how local and global relationships affect the ways in which Tuttlingen confronts the challenges it currently faces. These challenges relate to production (competition from cheaper labour producers) and knowledge (innovations and new technology, changes in health services in general and surgery in particular). These pressures signify that although Tuttlingen may be a dominant player at the level of production within the boundaries of the surgical instrument value chain, within the framework of the wider medical engineering value chain it is of far less importance. The key questions driving this chapter are: What upgrading strategies by individual actors can be identified in the Tuttlingen cluster? What upgrading strategies require cluster-based joint action as collective


Local enterprises in the global economy

strategies? Are there upgrading strategies that require external relations? Which strategies strengthen the cluster as a whole? Which are leading to increasing differentiation within the cluster? And, finally, what implications arise for the cluster? The findings of the chapter are rooted in primary field research conducted in 2000 and 2001. In 2000, a total of 64 interviews were carried out, including 18 qualitative interviews with institutions and key informants. Thirty-five manufacturing firms and 11 trading firms were surveyed using a semi-standardized questionnaire. Additional information was obtained through the author’s participation in a consultancy study for the Steinbeis Foundation, as well as visits to the sector’s leading international trade-fair, ‘Medica’, at Düsseldorf in November 2000, and to two firms in the former surgical instruments cluster in Sheffield, UK. Additional fieldwork in 2001 drew on 30 qualitative interviews, ten of which were conducted with key informants and 20 carried out with firm founders and large firms. The chapter is structured as follows: section 2 introduces the Tuttlingen cluster, presenting its main actors and explaining its specific features concerning material flows and distribution systems. Section 3 examines upgrading at the cluster level, whilst section 4 poses the same question in relation to external ties and value chains. Finally, section 5 brings together the main findings and insights.

2. THE TUTTLINGEN CLUSTER OF SURGICAL INSTRUMENTS This section provides a brief overview of the Tuttlingen cluster, including structural features and its main actors. A more detailed description of the cluster is provided elsewhere (Halder, 2002). Nadvi and Halder (2002) provide an overview of the global surgical instrument sector. The Tuttlingen cluster of medical engineering consists of a core of about 300 producers. They may be individual craftsmen or different sized industrial firms that manufacture complete medical devices. Together, they achieve an annual turnover of approximately US$ 610 million. The cluster’s success during the last 20 years is linked to its ability to diversify and expand into segments beyond classical surgical instruments. Some firms have emerged as recognized success stories, becoming leading firms both in terms of size as well as technological advancement, despite starting as small craft firms. Figure 8.1 shows the product range of the Tuttlingen cluster. The size of the circles reflect an estimation of the scale of the distinct segments. In Tuttlingen, classical surgical instruments have been the traditional

Local upgrading strategies in response to global challenges


medical engineering surgical instruments implants


classical instruments

Figure 8.1

diagnostics and others

minimal invasive instruments

Products of the Tuttlingen cluster

product. However, this subsector has declined over successive years and there has been a shift to minimal invasive instruments. Other new products include endoscopes and implants. The latter has required more sophisticated materials and techniques combined with a higher degree of mechanization. In terms of minimal invasive instruments, there has been a growing local market for small firms. While many firms producing classical instruments are also engaged in the production of minimally invasive instruments, the production of endoscopes as well as diagnostics equipment and implants are limited to larger firms in the cluster. The producers in the cluster are surrounded by 200 process-specialized subcontracting firms that are typically very small. Approximately 30 firms supply manufacturers with inputs, including two specialized die-making and forging firms. In this respect, the cluster benefits from the metalworking tradition and machine tool production of the surrounding regions. Finally, some 20 firms provide a range of producer services, ranging from specialized software, to translation offices, logistics and transportation. Most firms in the cluster are very small. Over 90 per cent of firms employ less than 20 people.2 However, the total employment effect of this large group of firms is only about 20 per cent of the workforce (including firm owners). In contrast, eight large firms account for 64 per cent of the cluster’s manufacturing employees.3 This asymmetric structure is even more pronounced when one considers that the two largest firms in the cluster make up approximately 50 per cent of the total manufacturing employment in the cluster. The size structure of core manufacturing firms in medical engineering is shown in Table 8.1.


Local enterprises in the global economy

Table 8.1 Medical engineering in Tuttlingen: number of firms and workers engaged in manufacturing in 1999 (by firm size)4 Firm size (by no. of employees)


1 to 19

20 to 99

100 and more Total

No. of firms Share of total firms No. of employees Share of total employees Average firm size

208 43.1% 0 0% 0

233 48.2% 1065 17.1% 4.6

34 7.0 % 1157 18.6% 34.0

8 1.7% 4001 64.3% 500.1

483 100% 6223 100% 12.9

Source: Statistisches Landesamt Baden-Württemberg, Landesarbeitsamt BadenWürttemberg, 1999

The cluster’s largest firm accounts for 32 per cent of the cluster’s manufacturing workforce, producing instruments, implants and other surgical products. In terms of employment, surgical instruments production is the main activity of the cluster (apart from the largest firm) with 30.5 per cent of total manufacturing employment within the medical technologies sector in Tuttlingen. However, nine of the ten largest firms in the cluster specialize in the new sectors of endoscopes, implants and surgical apparatus. In contrast, small firms employing less than 20 people predominantly work in the instruments subsector. Firms whose primary specialization is in endoscope manufacture account for 20.6 per cent of total manufacturing employment, and those in implants and apparatus production engage 17 per cent of the workforce5 (Nadvi and Halder, 2002). According to official statistics, there are 76 traders with a total of 636 employees. However, many producing firms also trade instruments, while retaining their main activity in production. According to a 1993 postal survey, 27.3 per cent of producers were also engaged in trading (Mekelburg, 1994). Other producers have completely changed their main activity to trading, although they often continue to describe themselves as producers to enhance and protect their reputation. In a functional sense, therefore, trading activities in Tuttlingen are much more significant than official statistics suggest, and they are increasing every year.6 Among the wide range of private firms and public institutions, some emerge as main actors. Besides the two largest firms mentioned above, there are two organizations, Medicon and Gebrüder Martin, through which 25 mainly small producers (with about 900 employees in total) sell their products. As Tuttlingen is not the regional capital, in terms of administration or population, important institutions related to the cluster are located outside. The Chamber of Industry and Commerce (IHK) in Villingen-

Local upgrading strategies in response to global challenges


Schwenningen and the Chamber of Crafts (HWK) in Konstanz are good examples of this because of their engagement in vocational training, market information and technology transfer programmes (see section 3). Moreover, they provide a link to regional economic interests and act as an intermediary with actors at the federal state level as well as the regional level.7 At the local cluster level, key institutions are the Association of Craftsmen, which plays an important role in ensuring high quality advanced training with common rules and testing standards, and the local governments of Tuttlingen, which provide infrastructural support and contribute to the local economy. One of the main gaps in the cluster is the lack of research institutions. Nevertheless, the cluster benefits from the dense research landscape and the variety of technical transfer institutions within the federal state of BadenWürttemberg as a whole. This state has 11 Max Planck Institutes for fundamental research, 13 Fraunhofer Institutes for applied research, 20 Industrial Contact Research Institutes, 220 Steinbeis Foundation technology transfer centres for SMEs, nine universities and 39 polytechnics (Cooke and Morgan, 1994; Heidenreich and Krauss, 1998). The most important clinical research institutions for Tuttlingen are the University Hospitals in Freiburg, Tübingen and Ulm, which are all within 90 minutes’ drive of the cluster. Although the density of research and technology transfer institutions is impressive, pure numbers do not imply effective delivery of technical advice (Heidenreich and Krauss, 1998). Most research institutions are small in size and focus on a particular technological field, and can only provide comprehensive consultancy services in response to very precise demands on the part of the company. ‘The ability to define innovative questions for the transfer centres is not something that can necessarily be taken for granted, especially among smaller (. . .) companies’ (ibid: 230). As a result, technology transfer is only used by a limited group of companies, and this effectively strengthens technically stronger firms. A central feature of a deep cluster is economies of agglomeration, and Tuttlingen is no exception. The gains from a pool of specialized workers, the emergence of specialist suppliers and the spillover of knowledge (often referred to as ‘Marshallian external economies’, see for example Martin and Sunley, 1996) remain important for the cluster’s success. Surveyed firms and key informants support this. Other cluster-gains for Tuttlingen firms include the specialized training system, the presence of many traders and buyers that link small producers to markets worldwide, the achievement of economies of scale and scope, the low transaction costs, and last but not least the reputation in the global market of health-care delivery. These passive gains of clustering can therefore be observed in Tuttlingen.


Local enterprises in the global economy

The total world market for medical engineering products in 1998 was estimated to be approximately US$ 113 billion (F&O, 1999). However, the world market of classical surgical instruments is expected to account for US$ 650 million (Nadvi and Halder, 2002). Over the last century, an everincreasing number of surgical instruments have been developed. Key informants estimate that Tuttlingen produces about 30 000 different types of surgical instruments and that annual production consists of approximately 15 to 20 million pieces.8 This would mean an average of 500 to 660 pieces per batch of a single instrument every year. This accounts for the presence of small craft firms, as it offers many market niches with small batches and uncertain demand. In the cluster we find both product specialization as well as process specialization regardless of firm size. Process specialization is common in labour-intensive production steps (for example polishing) or capitalintensive stages (for example heat treatment or laser-working). A widespread practice for large firms is to use both in-house capacities as well as outsourcing to deal with unsteady demand or to complement the in-house product range. Among the firms visited that produced mainly surgical instruments, 87 per cent outsourced some production steps to other firms. These included all of the larger firms (20 employees and more9) interviewed. Usually subcontractors work for several firms but irrespective of this, production ties are generally long-standing (Halder, 2002). To fulfil large orders, producers and traders can potentially purchase from any other firm that produces a special product or holds a stock of it. In principle, there are three modes of distribution for production firms (see Figure 8.2).10 These are integrated with the main markets. Large US firms are the most important buyers for small and medium sized firms in Tuttlingen. These health-care giants11 are usually involved in most segments of medical engineering, both in production as well as in sourcing products from all over the world. In this system, Tuttlingen producers usually act as own-design manufacturers (ODMs). Tuttlingen’s leading firms are also trying to get better access to the US market through alliances with US retailers. The European market is dominated by large producers and in particular by wholesalers. The strength of the latter is explained by their huge, wellskilled staff that provide service and advice. This is an important feature for hospitals and surgeons who do not have an overview of the full range of products available. Tuttlingen based traders are finally specialized in many national markets worldwide, where the successful sale of surgical instruments is linked to the understanding of specific cultural particularities and requirements. Although the cluster is very successful in terms of its market share, major

Local upgrading strategies in response to global challenges

Tuttlingen cluster

End customer


Homeworker Subcontractor

Homeworker Subcontractor


SME producer

US market

small trader


Other markets

large producer


European market

Figure 8.2 Model of main distribution channels changes occured during 2001–2002 and they are still continuing. The aforementioned shift from classical surgical instruments to more sophisticated products within medical engineering is a result of a couple of trajectories, which may be seen as push and pull factors. First, the restructuring of public health care delivery in most markets leads to increasing cost pressures. These pressures are both supply and demand driven. On the one hand, new developing country locations have entered the market and have systematically lowered prices of mature products (Nadvi, 1996; Nadvi and Halder, 2002). On the other hand, health care providers are under greater pressure to be cost efficient, as medical budgets come under strain. Second, technological innovation and innovation in surgery are leading to new prospects, which demanding customers are well aware of. For example, microsystems technology, robotics in surgery, tissue engineering and other biotechnologies are opening up new opportunities (Grönemeyer, 2000). This innovation push is reflected in research and development efforts, which in medical engineering accounts for an average of 10 per cent of sales, ranging from 5 per cent in the precision engineering branch of medical equipment production to 20 per cent in the field of diagnostics (Anderton and Schultz, 1999). Thus, the ability to upgrade is central for future firm success. The upgrading prospects are discussed in the following section. Backward linkages to subcontractors are somewhat different. We need to distinguish between backward linkages with local suppliers, and those with foreign suppliers. In Tuttlingen, there are subcontractors with good reputations, where normal market relations can exist. However, there are also relations whereby doubts about quality lead to quasi-hierarchical governance, with regular instructions on specification and final inspection by the


Local enterprises in the global economy

producer.12 This was supported by producers and subcontractors during fieldwork (Halder, 2002). Therefore, the ability to innovate in this situation will be a key factor for the future success of the Tuttlingen cluster. During field research, firms were asked about their upgrading success, along with changes in turnover. Among the interviewed firms, 23 (out of 33) had developed new products in the last five years. Of these, 20 firms had also raised turnover at the same time. Among those who were not able to develop new products (ten firms), only three had been able to increase turnover, while three had experienced a decline. Future expectations reiterate this development: while all firms producing in the minimally-invasive or implant field expected further increases of turnover, only 11 out of 23 firms on the ‘classical’ instrumentproducing firms side expected such increases. These findings are sensitive to product segments within medical engineering. All (six) firms whose main products are implants and four (out of six) firms engaged in minimallyinvasive instruments or endoscopes experienced an increase in turnover. In the field of ‘classical’ surgical instruments, only 14 (out of 22) have been able to increase turnover. In conclusion, firms mainly producing classical surgical instruments tend to be less innovative than those engaged in other segments, for example, minimal invasive instruments or implants.13 As a result, the innovativeness of cluster firms is clear. Yet, at the same time, most product upgrading has involved minor improvements (21 out of 33), and only five of the interviewed firms claimed to have made big improvements in this field. All of these firms were either engaged in segments other than instruments, or they were larger firms. It is not easy for small producers to make the shift to radical product upgrading in higher value segments. This is particularly true for the implant sector where advanced materials are used, a higher degree of automation is required and intense clinical testing has to take place prior to obtaining permission to enter the market. Capital is therefore required in addition to know-how in order to survive and to innovate. Thus, there is a marked shift from labourintensive to capital-intensive production. Nevertheless, some of the small firms are trying to enter this field by, for example, specializing in a single product. In the field of surgical instruments, the value chains are dealing with the production of mature goods. Even small producers have a reputation for high quality instruments; this makes them relatively reliable and low risk partners. Moreover, several producers make most types of products, which offers a choice for buyers and keeps prices down. The daily business of traders and buyers, collecting products by visiting producers, or just sending an order message in the case of regular ties, implies standardized transaction. Thus, the most pervasive form of forward-linking chain

Local upgrading strategies in response to global challenges


governance by far is arm’s length market relations, regardless of whether it is in relation to buyers, traders, wholesalers or large producers.

3. LOCAL ACTORS, LOCAL COOPERATION AND UPGRADING As outlined in the introduction, the cluster faces new challenges arising from the increasing globalization of product markets. In order to escape competition from low wage countries, especially Sialkot, Pakistan, (Nadvi, 1996; 1999), the cluster needs to reposition itself in the global economy. In accordance with Humphrey and Schmitz (2000), this chapter uses the term ‘upgrading’ to describe shifts in activity, which sustain higher earnings. Three types of upgrading will be examined in the context of this chapter: ●

process upgrading, for example transforming inputs into outputs more efficiently by reorganizing the production system or introducing superior technology; product upgrading, for example moving into more sophisticated product lines, which promises increased unit values; and, functional upgrading by acquiring new functions which promise higher rents, for example increasing design or marketing competence at the expense of lower value added activities.

Moreover, incremental and radical upgrading can be distinguished. For the latter, new knowledge is usually required. This can be attracted in several ways, using cluster dynamism or their own efforts (which are discussed in this section), or value chain ties, which are discussed in the subsequent section. Regardless of the strategy, efforts in radical upgrading are different from daily production. As Humphrey and Schmitz (2000) noted, upgrading requires substantial resources, and as Nooteboom (1999, 132) argued, there is a trade-off between the search for new knowledge and the effective exploitation of existing knowledge. Therefore, the following questions can be formulated: What upgrading efforts are undertaken by large firms? What role does inter-firm cooperation play in upgrading? In what way are public actors involved in upgrading of the cluster? First, we need to strengthen the conceptual framework for discussing upgrading.



Local enterprises in the global economy

The Production System and the Knowledge System

In order to analyse the issue of upgrading, Bell and Albu (1999) emphasize the need to focus on knowledge systems, which they distinguish from the production system. The latter can be understood ‘to encompass the product design, materials, machines, labour inputs, and transaction linkages involved in production’, whereas the knowledge system ‘encompasses those flows of knowledge and organizational systems involved in generating and managing changes in the products, processes and organization of production’ (Bell and Albue, 1999, 1723). Later they distinguish between ‘knowledge-using’ and ‘knowledge-changing’ elements. Knowledge-using elements refer to the work involved in maintaining (or even expanding) given modes of production, like the training of workers in established conventions, the imitation of production techniques and so on. This usage of knowledge thus includes incremental efforts in product and process upgrading. This section is directly concerned with knowledge-changing elements and will investigate how knowledge is changed in the context of the cluster. This will be conducted for both product and process upgrading.14 Knowledge changing is a big step. Thus, it is important to assert that upgrading, particularly knowledge changing, requires substantial investment (Humphrey and Schmitz, 2000). At the local level, the resources for upgrading can come either from individual large local firms, joint private initiatives or public–private initiatives. These three sources are not mutually exclusive. On the contrary, we have seen that successful clusters have often managed to combine these sources (see Belussi, 1999). 3.2

Large Firms and Upgrading

Knowledge-changing capacities require the management of innovation processes and the search for, selection and adoption or assimilation of new product or process technology (Bell and Albu, 1999). This has been achieved by many of the larger firms in Tuttlingen. However, only 40 of the 300 firms are estimated to have such capabilities. Even within these 40 firms, innovation capacity is very uneven. It is expected that only a handful of firms, or groups of firms, are engaged in research in a knowledge-changing way. Again, the size distribution is crucial, but this time it divides firms even more sharply. The largest firm in Tuttlingen employs more than 200 employees worldwide in R&D, and annually develops 800 to 1000 new products. The share of turnover from products introduced within the last three years accounts for 25 per cent (Aesculap.de, 2002). To emphasize the stark differences in size within the cluster, we see that the largest firm’s R&D department is

Local upgrading strategies in response to global challenges


almost as large as the third largest firm’s total staff. As the second largest firm in the cluster is engaged mainly in the technologically advanced field of endoscopes, it is expected to have a similar research capacity. To make the point, the difference between the two largest firms and the rest of the cluster is even higher in the field of knowledge than it is in the field of production. It results in the integration of new external knowledge which is combined with the firm’s own R&D efforts at knowledge innovation. For example, the step to sophisticated implants has been achieved by the marriage of internally advanced metal-working and ceramic technology from outside. However, such radical upgrading needs R&D resources, time resources and the capacity to engage in forward ties with end users and backward ties with input suppliers and specialized subcontractors. If we consider firm size, this radical upgrading in terms of new products is limited to larger firms which have the capacity to be engaged in steady product development. 3.3

Inter-firm Cooperation and Upgrading

As outlined above, on their own small enterprises cannot respond in more than an incremental way to the new challenges. Therefore, this chapter pays attention to cluster theory developed in the course of the 1990s which attaches importance to joint action (see for example Brusco, 1992; Schmitz, 1995). This includes joint action amongst private actors, and between public agencies and private enterprises and organizations. This section is therefore driven by the following questions: What upgrading opportunities are opened up through collective efforts? What types of upgrading does joint action address? What efforts are made for joint action and upgrading at the local level? To what degree does joint action take place in the Tuttlingen cluster? Most producers in Tuttlingen recognize their individual strengths and weaknesses. Much of their weaknesses originate from problems common in small-scale industries. These include: a small non-production staff, a dependence on buyers, and a narrow technological knowledge base which is limited to metal-based technologies. Joint action might be a way to reduce these disadvantages. However, joint action as a form of cooperation between actors requires trust, ‘defined as the experience-based expectation of co-operative and benevolent behaviour’ (Semlinger, 1995, 274) or at least mechanisms to avoid opportunistic behaviour, when trust is absent. In the course of many interviews, both positive and negative examples emerged although the latter prevailed. When asked about partnership, cooperation and collective forms of problem solving, a typical response from most actors in Tuttlingen, both


Local enterprises in the global economy

private and public, was: ‘you won’t find that in Tuttlingen’. Thus, competitive behaviour appears overdeveloped and cooperative behaviour underdeveloped. There are several reasons for this social behaviour. Of particular importance is historical experience, which can be traced back to the beginning of the 20th century. At that time, specialization was low, and buyers had been able to barter prices down. As the product range increased, incremental innovation and the adjustment of instruments led to imitation.15 This continues today, and has resulted in people being secretive about their work. Opportunistic behaviour (of others!) are frequent stories of lament. Consequently, both oral history and social practice have influenced Tuttlingen’s workers and firm owners’ behaviour. Contrary to the examples of some Italian industrial districts, Tuttlingen seems to lack formal or informal institutions (for example norms, routines or social practices) to avoid or at least reduce opportunistic behaviour. This may be due to the heterogeneity of the cluster with regard to structural and functional differences between private firms, and the relative privacy of independent small firms that produce small products, and whose distribution channels others cannot control. Furthermore, large firms suffer from the departure of skilled workers, who leave to start an independent firm. The loss for large firms is acute in this case because it results not only in the loss of skilled workers but also new competition, often in profitable product lines. In this situation, agreements between firms tend to be in the form of lip service. As a result, most Tuttlingen firms operate in relative isolation. There are, however, some signs that younger staff foster a more open style of communication, which may play a role in futher challenges. Despite such scepticism concerning joint approaches, some efforts have been made, both in the past and present, to overcome rivalry and mistrust. In 1989, the ACIG, a private organization, was established as a forum which offered services for small producers and traders. The main activities include a showroom for products, joint organization of trade-fair exhibitions, a yearly renewed guide to firms and a ‘cracker-barrel’ to discuss news and problems. Despite the aforementioned local attitude, some 20 people meet regularly. This raises hope of changing behaviour. The most impressive examples of private joint action were established a long time ago. Two associations, Gebrüder Martin and Medicon, established in 1923 and 1941 respectively, bring together a number of independent firms that jointly market their products in such a way that they are able to compete in markets against the cluster’s largest firm. Gebrüder Martin was founded in 1923 by seven producers as a consequence of the post-war crisis. Today, Martin employs approximately 650 local staff, including producers. The size distribution of the member firms is uneven and ranges from over 200 to under 20 employees. Together the

Local upgrading strategies in response to global challenges


group produces over 12000 different types of products; approximately 70 per cent of production is exported. The Medicon e.G., founded in 1941 by six firms, currently consists of 18 craft firms with about 350 employees. Together the group produces over 20000 different products. Within the group, firm structure is unequal, as it includes one-person firms as well as larger firms which employ up to 40 employees. In both Medicon and Martin, production arrangements have been negotiated internally within the group, so that each product is produced exclusively by one member firm. As a result, the member firms have achieved the core objectives of their collaborative efforts: economies of scale, specialization and reduced competition. Another advantage has arisen over time: both Medicon and Martin are recognized brand names, a feature which gains more and more importance, given that rents within the chains are shifting under globalization (Kaplinsky, 1998). Because of their joint size, both groups are able to offer a full-size programme for hospitals,16 and enhanced innovative capacity, thanks to the steady contact their marketing departments retain with surgeons and their development departments and laboratories. Nevertheless, such cooperation is far from conflict-free. Different strategies and prospects between the member firms have resulted in differing levels of development over time. In both cases, strong members often try to bypass rules for their own benefit. Group strategies as well as the distribution of earnings have led to disagreements. In one case, this has resulted in the exit of the most powerful member firm from one group. The search for a successor firm to keep the group alive in the early 1990s indicates how important the associations are for the member firms. In contrast, although the firm that exited is still successful and increasing in size, it has not been able to fully reach the predicted superior performance on its own. The foundation of similar associations or cooperatives has been tried several times in the cluster history (Reinert, 1951). They failed because of internal disagreements. Among surveyed independent producers, two told of their own experience with close cooperations linking a few firms together in a synergy creating way. Both examples failed due to perceived opportunistic behaviour by members. Thus, the positive example of Medicon and Gebrüder Martin is not sufficient to overcome mistrust and opportunistic behaviour as long as business for most firms runs well. Some other efforts have been made to overcome the constraints experienced by individual small firms in terms of collective action. These efforts have led to several problem-centred initiatives, most of them conducted as a cooperation between public and private actors.



Local enterprises in the global economy

Public or Public–Private Initiatives

In this section, initiatives to promote the cluster are introduced and their outcomes are briefly discussed. Most initiatives are limited in size and scope, and formal institutions are essential in mediating such joint action in order to overcome competition and mistrust. As mentioned in section 2, a range of public and private institutions are concerned with the medical engineering cluster. These institutions provide regular services for their clients, but they are also engaged in special initiatives to promote the cluster in order either to solve problems or prevent them. As medical engineering is a key industry for the region, they are engaged in improving this sector in a variety of ways. One of the key constraints for many small firms is their own lack of distribution and marketing channels and, related to that, low levels of direct producer–customer contact that could help induce innovation. Since the mid-1980s, several efforts have been undertaken to reduce this weakness (Table 8.2). In the following, only the main initiatives are discussed; they are, however, expanded upon elsewhere (Halder, 2002). ●

The most important local institution is the BBT (Berufliches Bildungszentrum Tuttlingen), which provides basic and advanced training for metal workers. Since this is the only place in Germany to learn the Chirurgiemechaniker (mechanics in surgical instruments) profession, it forms the pillar on which the cluster is built.17 Although the BBT concentrates on various different sectors, special importance is given to the surgical instrument sector. A staff of 20 teachers and 80 specialized external lecturers offer a wide programme of courses. Established in 1990, the Forum Medizintechnik aims to provide professional development for firms in several fields. Information is provided on medicine, medical engineering, production technology as well as business economics, legal issues and business management issues. Thematic topics are drawn from general information rather than an attempt to provide solutions to the recent problems and challenges of the cluster. Regional governments have established the Technology Park ‘Takeoff’. The agenda is to build an innovative environment based on three pillars. The first pillar is the provision of services including testing laboratories for common use; the second is an academy to link medical research with production, through a lecture series; and the third provides a venture capital fund for start-up firms. It is expected that the project will attract firms within technologically complemen-

Local upgrading strategies in response to global challenges



tary fields, giving the historically metal-based industry a diversifying push. The ‘Competence and Technique Centre for Minimally Invasive Surgery for Tübingen-Tuttlingen MITT’ was a result of an initiative started by Germany’s Ministry of Research and Education in 1999. The aim was to stimulate and accelerate basic research prior to competition. Part of the fund’s remit is to establish a structure to facilitate joint projects (Verbundprojekte) between research institutions and firms in basic research. A special feature is that the MITT is forming a network consisting of a University Hospital, two polytechnic schools and seven materials and processing-oriented research institutes, based at the Universities of Stuttgart and Karlsruhe. This is expected to make a considerable contribution to the sector, which suffers from the absence of a specifically dedicated research unit. As mentioned earlier, only 40 firms are expected to be able to contribute substantially to the MITT initiative with their own knowledge-intensive innovation capabilities. Other firms are not expected to have the necessary resources (knowledge, capital, staff) to engage more deeply in research. Joint action between private firms might open up some possibilities, but this would require a great ‘leap of faith’. Evaluation

This section has shown how large firms that have been able to combine external and internal knowledge in new ways to compose radical innovation have achieved the most substantial upgrading. This is a result of their ability to manage external relations. As discussed above, joint action does take place in Tuttlingen despite intense local competition, although it only occurs under special conditions. First, the grounds for joint action have to affect all firms in a similar way, and the costs of individual reaction to problem-solving have to be prohibitive.18 Second, there is a need for public institutions to mediate joint action in order to overcome local rivalry in Tuttlingen. The aim of public institutions in promoting the cluster can be summed up as creating the preconditions for competitiveness and success. This includes training of workers, access to modern technologies and, more limited in scope, different kinds of information (for example markets, laws and regulations). In this respect, the various initiatives are successful and can be seen as one pillar on which the cluster is built. Joint action has been successful in terms of enabling firms to upgrade their processes (although focusing less on product improvement). This is mainly a result of the forms of learning passed on through the efforts of the BBT or the Forum

216 Link producers and customers directly


Hospital visits

Episodical initiative, highly rated by small firms

Enhances knowledge diffusion, of particular importance for SMEs

Lectures mainly in surgical and technical knowledge

IHK, HWK, others

Forum Medizintechnik

Failed, craft firms were unable to produce in industrial ways

Link new ideas and hopeful Limited result, strengthens the inventors’ products with producers strong rather than supporting the weak

Product diversification of small craft firms

Products searching HWK, local producers governments


Business chambers Provides basic and advanced A pillar of the cluster. Source (IHK, HWK, others) training. Only place in Germany of skilled workers and diffusion to learn the profession ‘mechanics of process know-how in surgical instruments’


Tuttlingen initiative

Country government, Central showroom for SMEs, Failed because of mutual business chambers lectures, linking of producers and mistrust partly customers


Marketing and innovation centre


Founders/main actors


Table 8.2 Initiatives established and mediated by public and public–private joint action






Product (process)



Source: Own investigations

Various actors, including all main actors

Competence centre

Pooling of regional competences, accelerating basic research

Regional governments Attracting new firms, support firm founding, providing services and facilities

‘Take-off’ technology park

Process (product)

Most ambitious initiative, Product (process) possible competitive edge in future, impact too early to gauge

Too early to gauge

Improve the reputation of all Too early to gauge instruments by testing the most complicated and sensitive product lines


Sterilization initiative


Local enterprises in the global economy

Medizintechnik. Second in line of importance has been problem-specific joint action initiated by other actors, which have had differing levels of success. All initiatives and measures related to actual technical improvement or those concerning technical norms, which can be used for product upgrading, have some degree of success. From the individual firm’s perspective, the combination of the typical limitations faced by the small firms, with over-arching regional mistrust, prevents deepening engagement in private joint action. Thus, the main constraint is the lack of private joint action, and particularly the unwillingness of private firms to engage in cooperation to overcome the limitations of their small scale. But is this too ambitious? Competition always both enables and limits further success. Porter (1990) regards competition as a key to innovation, and strong competition in clusters can act as a motor to push the cluster ahead. We know from various studies, that local competition can be intense (for the instrument sector: see Nadvi, 1999). To gauge the degree of competition at Tuttlingen, we have to look at an example of successful joint action overcoming local rivalry in a cluster located next to Tuttlingen. The Heuberg area around Gosheim, 20 km from Tuttlingen, houses a cluster of about 400 metal turning firms. As the products are more or less standard, fear grew at the beginning of the 1990s that this industry might be lost as a result of increasing globalization. Nevertheless, the cluster managed to establish a Steinbeis Transfer Centre for Quality Management offering quality management, material testing services, vocational training and mediating measures for further cooperation (Semlinger, 1995: 278ff). The main difference compared to the Tuttlingen case was that a private firm initiated the process, assisted by local institutions including an association of private firms founded to promote this initiative. Although this process was not conflict-free (ibid), the basic firm behaviour seems to be quite different from that found at Tuttlingen. This hypothesis has been upheld by interviews with both key informants and firm interviews in Tuttlingen and current research of the Gosheim cluster. It is worth noticing that the Tuttlingen cluster has been very successful up to now, despite the absence of trust among the firms. Moreover, at certain points in history, some firms were able to overcome mistrust and form successive associations that enabled their members to compete even against the cluster’s largest firms. The question arises: Under what conditions does such close cooperation emerge? Here it is important to note that all, both successful and failed, cooperative initiatives that aimed to form associations (like Medicon or Martin) were established during times of difficulty for the cluster as a whole. One can argue that, where extreme external pressure is lacking, local competition prevails, hindering proactive

Local upgrading strategies in response to global challenges


solutions that could offer synergy effects for all. Thus, the question arises whether joint action to promote the cluster is sufficient to withstand future challenges. These challenges are in the areas of new technological opportunities, on the one hand, and price pressure on the other. The current initiatives of the competence centre and ‘Take off’ might be a way forward, but their success would require greater participation by private firms than has been seen in the cluster up to now.

4. EXTERNAL ACTORS, EXTERNAL COOPERATION AND UPGRADING The previous section has concentrated on actors and relationships internal to the cluster. This is in keeping with much of the academic literature on clusters which Bell and Albu (1999) have criticized for the lack of explicit attention on external relationships. The importance of these relationships for upgrading has also been emphasized by Humphrey and Schmitz (2000) in a paper that brings the cluster and value chain approaches together. This section follows the lead of these authors. Their joint cluster and value chain approach implies that the understanding of external relationships and the management of these relationships is essential for the understanding of how the cluster of Tuttlingen copes with the challenges of cost pressure and innovation. However, Tuttlingen is not a homogeneous entity; as seen in the previous section, large and small firms act and react differently to these challenges. This section first discusses the reactions and positions of the large firms and then the small firms in their respective value chains. Because process upgrading is not the main challenge in Tuttlingen, and it is by and large an issue of internal relations, the focus will be on how external relationships influence product upgrading and functional upgrading. 4.1

Large Firms and External Relationships

External relationships of firms may be analysed by looking at forward and backward linkages. I start with the latter to describe the time dimension of the shift. This is related to the fast entry of low wage locations into market in the field of re-usable surgical instruments. This is a key to understanding current competition. In 1972, Tuttlingen’s largest firm became aware of Sialkot in Pakistan as a producer of surgical instruments, even though these instruments were of poor quality. However, Tuttlingen established its own plant in Malaysia and subsequently gained from the low wages and tax advantages of Penang’s free trade zone. A few years later, other large firms followed to produce


Local enterprises in the global economy

abroad, setting up joint venture plants in Sialkot (Nadvi, 1996). Initially, ties between Tuttlingen and the new locations started with job processing. Tuttlingen firms supplied the critical input, high quality stainless steel forgings. The plants’ respective joint ventures were responsible for the intermediate and relatively labour-intensive tasks, in particular grinding, filing and polishing. Semi-finished products were then returned to Tuttlingen for the final processes. In the meantime, further locations were integrated into the production system also on an OEM (original equipment manufacture) basis (Nadvi and Halder, 2002). As the organizing hub, Tuttlingen firms are able to deliver their customers at the same conditions as ever. For large Tuttlingen firms expanding globally, however, there are new opportunities of size. A further outcome of this integration is that large Tuttlingen producers can concentrate their activities on rent-rich activities by passing down standard products that are facing price competition. In the field of surgical instruments, innovation has historically required close interaction between surgeons and manufacturers. This is important when developing instruments for new applications as well as improving existing instruments. Consequently, access to customers is a prerequisite for successful product upgrading (Lundvall, 1988). In particular, radical innovation requires close cooperation with surgeons because it is often related to new operation techniques.19 As a result, product upgrading is also interlinked with marketing because surgeons promote new techniques in scientific literature and through lectures at conventions. Furthermore, encouraging scholarly practice at University hospitals has similar effects. Analytically, there is a marriage of complementary skills (of surgeons and producers) to create new knowledge and artefacts, that is surgical instruments, implants, and other equipment. As both actors are interdependent, relations between them tend to be cooperative.20 Some large firms also employ doctors and surgeons among their staff to interact with external specialists and to evaluate internal and external progress. To establish these contacts, regular observations of external innovations are necessary. This requires participation at conventions and symposiums to keep abreast of new methods and techniques. Thus, the research and development contacts of Tuttlingen’s large firms are by no means limited to the region, or even the region of Baden-Württemberg. They maintain R&D contacts on a German-wide and even worldwide basis. Historically, however, they have had the closest contacts with university hospitals in Baden-Württemberg. It is important to note that nine of the ten largest firms in the cluster specialize in the new sectors of endoscopes, implants and surgical apparatus. Although the cluster originated in the production of surgical instruments, new product lines have been established in the cluster over the years. This may be called cross-sectoral upgrading, which can be seen as the most

Local upgrading strategies in response to global challenges


advanced and challenging kind of product upgrading because it requires a substantial shift in technology. The first step to enlarge the metal working knowledge base was made during the first half of the 20th century by the integration of small electro motors for gypsum-saws, or lights for Laryngoscopes. This first entrance was refined and diversified over successive years according to technological progress and market requirements. The usual way to enlarge the knowledge base was to attract specialized workers from outside the cluster. In the meantime, the increasing complexity of technology has limited such an ‘easy’ solution. This can be illustrated further by the case of endoscopes and implants production. The idea of keyhole surgery can be traced back to centuries ago. However, the breakthrough was only achieved when the problem of lightening the dark caves of the human body was solved. More precisely, progress in electronic lamps and optical lenses were a precondition for substantial progress in this advanced field of surgery. It is easy to see that the knowledge of optical lenses and lamps did not originate from inside the cluster. Instead, there was a marriage of these different technologies through acquisition. Tuttlingen producers of endoscopes have vertically integrated external producers of lenses for their strategic role. Similarly, Olympus, as a large producer of optical goods, bought a specialized producer of endoscopes (Winter and Ibe) in 1979. In any case, big steps have only been achieved through the combination of internal and external knowledge. In the third field of advanced products, surgical implants, we saw a combination of upgrading patterns from the former examples. Cooperative ties with end users are key for the initial development as well as for the intensive clinical testing series before permission is granted to enter markets. Therefore, before entering the market interaction with material specialists in alien technological fields like ceramics, alloys and titanium or interaction with suppliers of machine tools became critical. Therefore, the cluster needs to be an open system and inward orientation becomes dangerous for innovation. Moreover, Tuttlingen firms that follow technological development have to engage as systems integrators in surgery. This can be done at product level as well as at functional level. In terms of products, engagement in robot-aided surgery, computer-aided control and assistance of surgery, and tele-surgery, requires integration. Participation requires large internal investments as outlined in the previous section, as well as the management of comprehensive knowledge and external relationships, particularly with end users, specialized suppliers and technological consultancies. This leads to a shift in function, which alters the nature of large firms who follow that path, that is the traditional field of producing handheld surgical


Local enterprises in the global economy

instruments becomes less important for the large Tuttlingen firms and new fields gain in importance. This is increasingly concerning systems integration at a functional level, where the above-mentioned shift was enlarged and deepened. As a result of the restructuring of health care systems, hospitals and surgeons are under increasing cost pressures. There are several, interrelated outcomes of these pressures concerning the supply of medical products. First, cost pressures are directly passed down the chain and this leads to a decline in prices in general. Second, hospitals specialize in a particular field of surgery and thus avoid under-usage of equipment. Third, hospitals unite as associations to gain economies of scale, or form buying groups to increase purchasing power. Fourth, hospitals outsource functions in which they are not specialized, including facility management, sterilization of equipment and logistics. Under these conditions of concentration, new rents may emerge for large firms (Kaplinsky, 1998). As a result, we have seen many joint ventures, mergers and acquisitions in the medical field on an international scale.21 In the case of Tuttlingen, the largest firm undertakes considerable functional upgrading. Because an even larger external company has bought it, the firm is able to offer a complementary product range that includes a full programme for hospitals, and includes servicing, maintenance and logistics. Logistics are of special importance in the implant field, where short delivery times are required. As well as establishing a new implants-producing firm, European logistics will be restructured and coordinated from Tuttlingen. Moreover, the largest firm has improved its services for end customers. This means that it can offer a full instrument service including the sterilization, repair and replacement of instruments. The service involves a computerized tracking system that enables close monitoring of each individual instrument (for example age, frequency of usage, state). This leads to an optimal usage of instruments and lower costs for the hospital. According to conversations with key informants, there are signs that, in the near future, hospitals may rent or lease instruments from producers instead of buying them. In this development, large US firms are prime movers. Large Tuttlingen firms feel they are being forced in this direction. However, outsourcing of the complete facility management to Tuttlingen-based service firms owned by large producers marks a considerable functional upgrading. Again, the offer of full supply to hospitals requires a substantial investment in logistics. As a result, gaining access to end customers might be more difficult for other firms. The largest firm in Tuttlingen has developed a conference facility; this special feature is called the Aesculapium. Medical conferences and congresses take place here and new operation techniques and new equipment

Local upgrading strategies in response to global challenges


are presented.22 Here, again, the difference between the market leader and the cluster of small firms becomes evident. Other larger firms are hot on the tail of the largest firm, restructuring production and making efforts in innovation, but they are unable to upgrade functionally in the same way. Well aware of this, some of Tuttlingen’s larger firms have decided to stop upgrading and stay at the position of a technological advanced OEM supplier. In this respect, the ability to constantly upgrade the product is key for future success. The question arises, what are the prospects for the many small enterprises in the cluster? 4.2

Small Firms and External Relationships

The most significant reactions from small Tuttlingen producers to price competition were efforts to maintain and expand quality leadership in order to justify higher prices. However, as low-wage competitors, with guidance from Tuttlingen firms, were able to reduce distance in product quality (Nadvi and Halder, 2002), such market niches in the field of classical surgical instruments are declining for Tuttlingen firms. As avoiding them failed, many small Tuttlingen firms tried to emulate their competitors. However, small producers lack resources to make up their own plants. Consequently, they subcontract to Sialkot or to locations in Eastern Europe which already have a functioning production system, albeit at lower quality and much lower wages. The fall of the iron curtain in 1989 led to producers from Hungary and Poland entering the market. This meant that cheaper labour was now available within a ten hour car drive of Tuttlingen. During the 1990s, many small Tuttlingen producers established active or passive relations with low-wage locations in Pakistan or Eastern Europe. However, only some firms have entered into direct relations on an OEM or a job processing basis because maintaining direct ties with firms abroad is costly. The majority of producers established a second, cheap, product line by sourcing from specialized traders that act as intermediaries between Tuttlingen and other locations. Some other product groups, however, are completely sourced out. Thus, former Tuttlingen producers have shifted to more profitable activities such as repairing used instruments and refining externally purchased instruments. From a value chain perspective, this leads to extending the production chain, passing functions with low rents further down the chain. This behaviour can be described as passive functional upgrading, as a firm shifts its position without actively attracting new functions. In the case of Tuttlingen, former producers have become traders. Most key informants and small producers in Tuttlingen claimed that this shift of activities towards trading is the most substantial cluster change in recent years. Thus,


Local enterprises in the global economy

Tuttlingen is rendered a trading hub in the global surgical instrument industry. According to small Tuttlingen producers, product upgrading might avoid this race to the bottom. However, small firms lack the resources required to shift to a new knowledge base. For example, the technological push in endoscopes-based surgery at the end of the 1980s has led to increasing demand for minimal invasive surgical instruments (MIS). Based on sophisticated metal working, MIS offered the opportunity to upgrade products. However, the production of MIS requires a much higher degree of mechanization than that used for traditional surgical instruments. Thus, substantial investments are necessary to introduce specialized machine tools, and many small producers do not have the financial resources to participate in this way. On the other hand, this shift from labour-intensive to capital- and knowledge-intensive production keeps low-wage producers at a distance. However, mechanization also leads to higher productivity and output. As a result, markets become saturated and some small producers abandon classical surgical instruments. Some of the small firms that have the technological and financial capabilities and the potential to grow also strengthen ties with large health care providers, both outside the cluster and abroad. They are establishing joint ventures, which ensure that the small producers have further access to end customers, and the external partner has reliable delivery on an own-design manufacture (ODM) basis, in some cases also on an own-brand manufacture (OBM) basis. Table 8.3 summarizes the main upgrading strategies of Tuttlingen firms. There are, however, many other firms which have not reacted in any mentioned way. Besides a number of firms which one may describe as ‘paralysed’, there are also more reasonable explanations for others. Up to one third of the cluster’s population started business in the 1960s and 1970s. As a result, this generation is now reaching retirement age, therefore, substantial investment in upgrading does not seem appropriate. Moreover, the younger generation often refuse such hard manual labour and seek ‘easier’ jobs in marketing, R&D, logistics, or work in other sectors. As a result, many of these firms will have no successors and will be forced to close.



This chapter has focused on the external challenges faced by the Tuttlingen cluster, and the strategies of firms to meet them. The main challenges for the cluster are price pressures and innovation opportunities relating to new technologies. Price pressures arise from above and below: rising competition in


• •

• • • •

• •

Small firms mechanization quality leadership introducing a cheap second product line through outsourcing or purchasing from OEM producers abroad passive functional upgrading by extending the production chain and passing labour-intensive production steps to suppliers abroad (job processing) Establishing strategic alliances with strong external partners to gain market access

Upgrading strategies of Tuttlingen firms

mechanization business re-engineering acquisition of technological leading producers in strategic fields – otherwise close cooperation technological leadership – often combined with a decision to stay as ODM producer establishing and maintaining relations with famous surgeons for innovation and marketing opportunities concentrating on rent-rich products in other medical fields (implants, endoscopes) establishing strategic alliances with complementary firms, in particular with pharmaceutical firms for comprehensive supply to end users (hospitals) enlarging services for end customers (facility management, including sterilization, repair and replacement) establishing competences in logistics, particularly for implant delivery.

Large firms

Table 8.3

● moderation and management of • production on the basis of market • information and global sourcing ● direct engagement in production • by joint ventures to prevent poor • product quality from suppliers • abroad ● establishment of own production • facilities or joint ventures in • Tuttlingen to ensure requested • product quality



Local enterprises in the global economy

the Tuttlingen cluster has resulted from increasing pressure on public health care systems in major markets on the one hand, and producers in countries with lower wages entering the market on the other. Simultaneously, crosssectoral technologies like microelectronics and microsystems as well as progress in bioscience and material sciences promise new solutions for surgery. Demanding customers, which compete on medical success and best practices, are pushing the industry forward. Until the end of the 1980s, the Tuttlingen cluster was steadily expanding via newly established firms. The last decade, however, has led to a decline, in particular of small firms. The pressures outlined above have resulted in substantial outsourcing of mature product lines. This has affected all firms that formerly produced these instruments on their own. Small firms are outsourcing many types of mature instruments to Sialkot in Pakistan in particular, while at the same time they continue to produce complex classical surgical instruments. As initial ties with low-wage locations were established on a job processing basis, relations partly shifted to an OEM basis (Nadvi and Halder, 2002). However, only a few small producers have the resources to establish direct relations. Many others are purchasing from Tuttlingen traders which act as intermediaries. As a consequence, the Tuttlingen cluster as a whole increasingly shifts from making mature instruments to trading in them. In this context, outsourcing can be seen as an extension of the production chain, that is products and functions that promise low rents are passed further down the value chain. I refer to this as ‘passive’ because new functions are not acquired and forward linkages or higher stages of the value chain are not affected. Active functional upgrading, however, requires major investments and the management of external relationships. Large and small firms differ in relation to outsourcing. It is interesting to note that small firms established ties with firms in a well developed Pakistani cluster, while some larger firms also established facilities in regions without a tradition in the production of surgical instruments. By investing over many years, workers were instructed and trained, and some plants were increasingly able to produce at OEM conditions. The main firms in Tuttlingen design and specify the product and act as an organizing hub, while concentrating their production facilities on rent rich new products. One large firm, for example, employs about 2000 people in production at Tuttlingen and a further 4000 in Poland, Malaysia and Spain. The plants abroad have grown steadily both in number and employment during the last 20 years, while in Tuttlingen employment has grown only slowly. Key informants suggest that one reason for this strategy of large firms was the fear of losing know-how through self-employment, which is more likely in a cluster which lowers entry barriers for start-ups.

Local upgrading strategies in response to global challenges


In contrast to the majority of small firms, nine of the ten largest firms in the cluster have specialized in the new products sector, that is endoscopes, implants and medical apparatus. To participate in this radical upgrading requires, in all identified cases, connecting internal and external knowledge which in turn requires substantial investment. This is where the difference between large and small firms becomes most apparent even though it is not a rigid divide. On the other hand, in terms of the development of many new products, traditional skills are not required that frequently and they decline relatively in importance as additional knowledge in fibre optics, electronics, ceramics and so on gain in importance. Consequently, small firms that have not enlarged their knowledge base and production range in recent years, now face problems keeping up with new developments. Product upgrading towards MIS instruments can be seen as an alternative only for some, because it requires a much higher degree of mechanization and thus, capital. However, as mechanization provides higher productivity, it also accelerates local competition because markets for MIS products do not grow simultaneously. Contrary to this finding, large firms achieve considerable functional upgrading, both in forward and backward linkages. Tuttlingen’s endoscope producers have either acquired, or have entered into collaborative joint ventures with, specialized firms in the field of optical lenses from outside the cluster. Some of the firms that have been acquired were previously suppliers. By bringing these firms in-house, endoscope manufacturers successfully internalized new specialist technologies. The decision to establish hierarchical relations with former suppliers was driven by the fear of the loss of, or restriction, to inputs that give them a competitive edge. Similarly, some Tuttlingen producers of surgical implants strengthened backward linkages with specialist machine-tool suppliers, thereby enhancing their capabilities in the complex manufacturing procedures required for sophisticated products (Nadvi and Halder, 2002). Historically, the cluster’s key external relationships have been collaborations with end users. Such ties with end users continue to be important in the development of new products. Design innovation in surgical instruments requires close interaction with surgeons, who help firms specify the functional requirements of instruments. Similarly, progress in surgical implants technology, endoscopes and medical apparatus require problem definition and feedback by surgeons. Both producers and customers maintain cooperative relations because they are dependent upon each other and they have complementary resources. Current legal initiatives to increase clinical testing series in several important markets will continue to raise the importance of these relations.


Local enterprises in the global economy

A relatively new challenge for functional upgrading arises from cost pressures of end customers. Thus, hospitals are searching for better control and maintenance of medical equipment inventories and more competitive sourcing of instruments. Concentration on core competence for them means shifting inventory costs and service functions further down the chain to instrument suppliers. This means that surgical instrument suppliers need to be more conscious of logistics, and of the value added of service, maintenance and financing activities. Therefore, large firms in the Tuttlingen cluster have developed a capacity in the area of services, focusing on the tracking, sterilization, maintenance and repair of surgical instruments held by individual hospitals. As key informants suggest, leasing instrument kits will grow in importance in the near future. Cooperation agreements with specialists in logistics, and joint ventures, or mergers with external medical product suppliers to offer a full spectrum of supply represent a new competitive stage. It can be expected that this development will result in a shift of economic rents from production to trade and services (Kaplinsky, 1998). In these fields of activity, barriers of entry are much higher for small scale industry because they require financial assets, a dense web of customer services, logistics and consultancy competences. The initial and main strength of the Tuttlingen cluster was in the production of high quality hand-held surgical instruments. This had its roots in the high quality training system and in the tacit knowledge of the workers. However, low-cost competitors increasingly enter this area. Consequently, surgical instrument production has been declining in importance for Tuttlingen over the years. Moreover, the shift into new product lines requires higher automation and engineering compared with the production of classical surgical instruments. Thus, the main competitive advantage of Tuttlingen is fading, as sophisticated metal-working is not tied to Tuttlingen, but can be found in a number of different areas instead. In sum, the shift from surgical instruments to medical engineering takes away Tuttlingen’s particularity and leads to competition with different locations within the medical engineering field. This development leads to increasing differentiation within the cluster. Moreover, the fact that firms may regard efforts in external ties as more efficient than joint action at the cluster level may lead to a conflict of interest, that is whether to support cluster activities or not. Then there is the question of whether ‘no support’ means passive behaviour or active opposition. The latter could arise when large firms have direct competitors locally. These firms would not wish to help their competitors by engaging in joint action. There are past examples of this in Tuttlingen. This might also be the destiny of the Competence Centre. It is obvious from the Tuttlingen example that joint action occurs when there is a common self-protective

Local upgrading strategies in response to global challenges


reaction, unifying or otherwise, to external change or pressure. This occurred when Martin and Medicon were founded and when the sterilization of instruments became an issue of joint action. Many Industrial Districts in Italy exemplify the insidious effects of increasing heterogeneity and deterioration through a lack of common interests (Harrison, 1994; Belussi, 1999; Whitford, 2001). However, one of the main features of a cluster is the fast diffusion of knowledge. It leads to a rapid spread of innovation, but usually of an incremental nature (Humphrey and Schmitz, 2002). The example of Tuttlingen shows how cluster advantages foster small firms in times of relatively stagnating technology. Where radical technological innovation is limited, the rapid assimilation of incremental innovation is the key to competitiveness for producers. In this situation, the evolutionary effects of adoption, modification and selection of small improvements, which are accelerated through the cluster dynamic, provide the decisive competitive advantage for the cluster. Organizational flexibility and the economies of time favoured small-scale industry. However, current shifts in technology challenge smallscale industry in particular. As Nooteboom (1999) stated, there is a tradeoff for firms between the effective exploitation of existing knowledge and the search for new knowledge. This trade off is particularly difficult for small firms. Knowledge-changing has exclusive elements by virtue of its complexity and interaction with other technological knowledge. Thus, knowledge will not circulate within the cluster at the same speed as it has in the past, and knowledge systems will become more closed (Bell and Albu, 1999). Nevertheless, the traditional focus on variables internal to the cluster is not sufficient. Understanding the upgrading strategies of large and small producers in the Tuttlingen surgical instruments cluster requires the explicit incorporation of external relationships.


2. 3. 4. 5. 6.

The author thanks Khalid Nadvi and Hubert Schmitz for their various discussions on the chapter and the overall study. Thanks also to Mick Dunford, Dirk Messner, John Humphrey and Luiza Bazan for their comments on an earlier draft of the chapter. Financial support from the Volkswagen Foundation, Germany and the German Research Council is gratefully acknowledged. The usual disclaimer applies. Twenty employees is the cut-off point in official German statistics for manufacturing firms. Administrative and support staff are also included in data on producing firms. Thus, this data is sensitive to firm size. The data refers to manufacturing in medical engineering WZ 3310 and includes producers of final instruments as well as subcontractors. Based on data from the Tuttlingen labour force office and firm annual reports. Most key informants and firms surveyed stated that the increase in trading activities is one of the major changes in Tuttlingen.

230 7. 8.


10. 11. 12. 13.


15. 16. 17. 18. 19. 20. 21.


Local enterprises in the global economy However, many firms, especially small firms, question the true benefits because direct benefits seem small in their view. Some key informants estimated up to 45 000 different types of products. The exact range is difficult to count because many products are extinct, or they consist of small batch products for a few surgeons or for particular national markets. Forderer (1949, 241f) stated that the largest firm itself once produced approximately 40 000 different types of instruments. Given the size distribution of firms in Tuttlingen, ‘larger’ firms are those that employ 20 people or more. This distinction is based on both functional considerations as well as the limitations of official statistics. Firms of this size often have at least a small distribution section and are involved in direct marketing, which justifies their placement in a single category in the context of this chapter. This model reflects only the main tendencies of the distribution system, whose accurate terms are vague and rapidly changing. For example Johnson and Johnson employs about 88 000 people, or over eleven times the whole Tuttlingen surgical instrument cluster, and produces and markets a wide range of pharmaceutical and medical products. Although the majority of subcontractors are producing the requested quality without guidance. This is a result of the sophisticated training system as well as long-standing ties over several years. This finding is also sensitive to firm size, particularly if one-person firms are considered. Moreover, it depends on the life cycle of small firms, for these firms are deeply dependent on the owner’s career. During the time prior to the owner’s retirement, improvements and innovation are usually at a minimum. Functional upgrading is not considered in this section because it requires information rather than knowledge. As a result of this differentiation, functional upgrading will be dealt with in further sections. Moreover, a fourth kind of upgrading is possible: shifting into other branches. The question remains as to whether moving from surgical instruments to implants can be seen as such a shift, or if this is internal to medical engineering. I would argue that, in the Tuttlingen case, this can be seen as a kind of radical product upgrading. Implants have been a part of cluster activity for many years and are related to the cluster’s core capabilities in metal working. Small firms, small batches and incremental improvements leads to an absence of patent protection and encourages ‘innovation for free’ for competitors. In Tuttlingen this means being able to tender to supply the full range of instruments required by hospitals. This leads to the fact that surgical instrument producers located outside the cluster (even from Berlin) have to send their young workers to the BBT for training. This might be an explanation for why joint action has not played a role in the field of ISO 9000 standard compliance. Prominent examples are the development of arthroscopy concerning knee surgery and the minimal invasive surgery of the gall bladder, both of which minimized traumatic operations for patients. They have all the features of what Humphrey and Schmitz (2000) call network-based relationships. See Chapter 13. For example, Howmedica–Stryker, Synthes–Stratec, Gencyme–Snowden-Pencer, Bristol-Myers–Squibb, Zimmer–Pilling–Weck, Smith & Nephew and Beiersdorf. These mergers illustrate the diversity of firms and corporations and the growing links between the medical instrument field and the pharmaceutical sector. This could raise more synergy effects in the future, using the same distribution channels to surgeons and hospitals. A few other large firms have similar facilities, although they are smaller in size.

Local upgrading strategies in response to global challenges


REFERENCES Aesculap.de: http://www.aesculap.de/d/uberuns/index.htm, from 28 November 2002. Anderton, R. and S. Schultz (1999), ‘Explaining export success in the UK and Germany: a case study of the medical equipment industry’, Anglo-German Foundation for the study of Industrial Society, Layerthorpe, York: YPS. Becattini, G. (1991), ‘The industrial district as a creative milieu’, in G. Benco and M. Dunford (eds), Industrial Change and Regional Development, London: Belhaven, pp. 102–16. Bell, M. and M. Albu (1999), ‘Knowledge systems and technological dynamism in industrial clusters in developing countries’, World Development, 27 (9), 1715–34. Belussi, F. (1999), ‘Policies for the development of knowledge-intensive local production systems’, Cambridge Journal of Economics, 23, 729–47. Brusco, S. (1992), ‘The idea of the industrial district: its genesis’, in F. Pyke, G. Becattini and W. Sengenberger (eds), Industrial Districts and Inter-firm Co-operation in Italy, Geneva: International Institute for Labour Studies, ILO, pp. 10–19. Cooke, P. and K. Morgan (1994) ‘Growth regions under duress: renewal strategies in Baden-Württemberg and Emilia-Romagna’, in A. Amin and N. Thrift (eds), Globalization, Institutions, and Regional Development in Europe, Oxford: Oxford University Press, pp. 91–116. Enright, M.J. (1996), ‘Regional clusters and economic development: a research agenda’, in U. Staber, N. Schaefer and B. Sharma (eds), Business Networks: Prospects for Regional Development, Berlin: de Gruyter. F&O (Verband der deutschen feinmechanischen und optischen Industrie e.V.) (1999), Jahreszahlen 1998 der Feinmechanik und Optik: Cologne. Forderer, J. (1949), Tuttlingen im Wandel der Zeiten, Reutlingen: Oertel & Spörer. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Gereffi, G. and R. Kaplinsky (2001), ‘The value of value chains: spreading the gains from globalization’, IDS Bulletin, 32 (3), Brighton: Institute of Development Studies. Grönemeyer, D.H.W. (2000), Med. in Deutschland, Berlin, etc.: Springer. Halder, G. (2002), ‘How does globalisation affect local production and knowledge systems? The surgical instrument cluster of Tuttlingen’, INEF Report 57, Duisburg: Institut für Entwicklung und Frieden. Harrison, B. (1994), Lean and Mean. The Changing Landscape of Corporate Power in the Age of Flexibility, New York: Basic Books. Heidenreich, M. and G. Krauss (1998), ‘The Baden-Württemberg production and innovation regime’, in H.J. Braczyk, P. Cooke and M. Heidenreich (eds), Regional Innovation Systems, London: UCL Press, pp. 214–44. Humphrey, J. and H. Schmitz (2000), ‘Governance and upgrading: linking industrial cluster and global value chain research’, IDS Working Paper 120, Brighton: Institute of Development Studies. Humphrey, J. and H. Schmitz, (2002), ‘How does insertion in global value chains affect upgrading in industrial clusters?’, Regional Studies, 36 (9), pp 1017–27. Kaplinsky, R. (1998), ‘Globalization, industrialisation and sustainable growth: the pursuit of the n-th rent’, IDS Discussion Paper 365, Brighton: Institute of Development Studies. Knappe, E., G. Neubauer, C. Lindl, T. Zimmermann, T. Seeger and K. Sullivan


Local enterprises in the global economy

(2000), ‘Study on the value of medical devices in Germany’, Trier: unpublished study. Lundvall, B.A. (1988), ‘Innovation as an interactive process: from user–producer interaction to the national system of innovation’, in G. Dosi, C. Freeman, R. Nelson, G. Silverberg and L. Soete (eds.), Technical Change and Economic Theory. London: Pinter, pp. 349–69. Martin, R. and P. Sunley (1996), ‘Paul Krugman’s geographical economics and its implications for regional development theory: a critical assessment’, Economic Geography, 72, 259–92. Mekelburg, E. (1994), ‘Der Wirtschaftszweig Medizintechnik im Raum Tuttlingen. Eine industriegeographische Untersuchung zur Qualifizierungs- und Beschäftigungsstruktur in medizin-technischen Betrieben’, Tübingen, Geographisches Institut: unpublished diploma-thesis. Nadvi, K. (1996), ‘Small firm industrial districts in Pakistan’, D.Phil. thesis, University of Sussex. Nadvi, K. (1999), ‘Collective efficiency and collective failure: the response of the Sialkot surgical instrument cluster to global quality pressures’, World Development, 27 (9), 1605–26. Nadvi, K. and G. Halder (2002), ‘Local clusters in global value chains. Exploring dynamic linkages between surgical instrument clusters of Pakistan and Germany’, IDS Working Paper 152, Brighton: Institute of Development Studies. Nooteboom, B. (1999), ‘Innovation, learning and industrial organisation’, Cambridge Journal of Economics, 23, 127–50. Porter, M.E. (1990), The Competitive Advantage of Nations, New York: Free Press. Porter, M.E. (1998), ‘Clusters and the new economics of competition’, Harvard Business Review, November/December, 77–90. Pyke, F. and W. Sengenberger (eds) (1992), ‘Industrial districts and economic regeneration’, ILO, Geneva: International Institute for Labour Studies. Reinert, E. (1951), Tuttlingen und seine Industrie, Darmstadt: Mushaksche, Franz Mathes. Schmitz, H. (1995), ‘Collective efficiency: growth path for small scale industry’, Journal of Development Studies, 31 (4), 391–411. Schmitz, H. and K. Nadvi (1999), ‘Introduction’, Special Issue of World Development, 27 (9), 1503–14. Semlinger, K. (1995), ‘Public support for small firm networking in BadenWürttemberg’, in L.E., Andreasen, B. Coriat, F.D. Hertog and R. Kaplinsky (eds), Europe’s Next Step: Organisational Innovation, Competition and Employment, Essex, UK and Portland, USA: Frank Cass. Statistisches Landesamt Baden-Württermberg (eds) (various years), Handwerkszählung, Stuttgart. Whitford, J. (2001), ‘The decline of a model? Challenge and response in the Italian industrial districts’, Economy and Society, 30 (1), 38–65.

9. Clustering and upgrading in global value chains: the Taiwanese personal computer industry Chikashi Kishimoto 1.


In the early 1980s when Taiwan embarked on the development of a personal computer (PC) industry, few people expected that the country would obtain a dominant status in the world PC hardware production market within 20 years. In 1995, the total output of the Taiwanese information technology (IT) industry reached about US$ 20 billion and Taiwan became the third largest country in the production value of IT hardware, next only to the US and Japan. However, its status as the world’s main hardware producer is not visible to the equipment users. This is because many Taiwanese products are supplied with the brand name of foreign computer companies, and Taiwanese firms undertake the role of contract manufacturers. This does not mean, however, that Taiwanese manufacturers are only charged with simple production tasks under the supervision of foreign clients. This may have been the case in the 1980s, but since then Taiwanese producers have accumulated substantial capabilities not only in production skills but also in more complicated tasks such as product design and logistics. It is important to note that this Taiwanese industry mainly consists of homegrown small and medium-sized enterprises (SMEs) that are geographically clustered. This chapter examines the upgrading trajectory of this cluster. The Taiwanese PC industry is a particularly challenging case because of two key features. First, it probably presents the most significant case of industrial upgrading outside the OECD countries. Second, this upgrading has been accompanied by the offshore manufacture of a rapidly increasing share of its output. In order to examine upgrading processes, this chapter draws on two analytical advances that have been made in the recent cluster literature. The first is the distinction between the cluster as a production and knowledge 233


Local enterprises in the global economy

system. According to Bell and Albu (1999), the previous cluster literature mainly focuses on the analysis of the production system; a concept which encompasses inter-firm transactions involving parts, materials, machines and services directly connected with the production of goods at a given level of technology. They suggest that more attention should be paid to the knowledge system in order to clarify the technological dynamism of clusters. The concept of the knowledge system is concerned with the learning process and technological capabilities accumulated within a cluster that leads to upgrading. This chapter will show that under increasingly global competition, simple production tasks can be shed to neighbouring low-cost countries, while more knowledge-intensive tasks benefit from clustering. The second analytical advance that this chapter builds upon is the fusion of the local cluster approach with the global value chain (GVC) approach, proposed by Humphrey and Schmitz (2000). The first approach prioritizes local linkages and the second concentrates on global linkages. Yet, despite increasing recognition that the two need to be brought together, the way in which local and global linkages interact remains a disputed and underresearched question. This chapter investigates how Taiwanese PC manufacturers have attained considerable upgrading in the GVC, and how this process has been reinforced by the transformation of local cluster linkages. The chapter is structured as follows: the next section offers an overview of the Taiwanese PC cluster and highlights the main transformations. Section 3 and section 4 examine the upgrading trajectory of the Taiwanese PC industry in the GVC and the evolution of its local base respectively. The final section draws together the main findings. The analysis of these sections is based on information obtained from fieldwork conducted during the late 1990s as well as existing studies and published data. During the fieldwork, I conducted a total of 55 interviews. Nineteen interviews were carried out with monitor firms, 17 interviews were with motherboard firms, 14 with notebook PC firms, and five interviews were conducted with key informants.1

2. THE TRANSFORMATION OF THE TAIWANESE PC CLUSTER The relevance of local-level resources for upgrading in the global economy continues to be a controversial topic in the industrial development debate. Theoretically, the importance of location is likely to decrease as global markets become increasingly open and transport and communication infrastructure improves. However, authors have argued that intensifying global competition increases the importance of clustering, rather than diminishing it. According to Porter: ‘The enduring competitive advantages

Clustering and upgrading in global value chains


in a global economy are often heavily local, arising from concentrations of highly specialized skills and knowledge, institutions, rivals, related businesses, and sophisticated customers’ (Porter, 1998, 90). Countering the premise that the decentralization of production stages outside a cluster may indicate the declining importance of location, he states that the cluster’s competitiveness is recovered by moving part or all of production elsewhere to offset local wage rises as long as the cluster succeeds in grasping strategic functions and in enhancing innovative capabilities. Porter states: ‘The essence of a cluster lies in the exchange of insights, knowledge, and technology, and in offering a structure that offers the incentives and flexibility to innovate’ (Porter, 2001, 145). This section provides an overview of the development of the Taiwanese PC cluster. This overview is conducted in such a way that the validity of Porter’s proposition can be examined. The distinction between the production and knowledge systems mentioned above is particularly relevant here. Largely based on quantitative data, this section shows that the two systems have developed in different directions: although the role of clustering has decreased in the production system, it has increased in the knowledge system. That is to say, in the production system, we observe the increasing relocation of manufacturing mature products to neighbouring low-wage countries. In contrast, with respect to the knowledge system, the manufacture of the less mature products and innovation activities remain concentrated in the cluster. In the early 1980s when ordinary people were not yet familiar with the computer, Taiwanese firms entered this new business as producers of monitors, terminals and low-price PCs. In 2000, the total output of the Taiwanese IT industry reached about US$ 47 billion (including offshore production) and Taiwan became one of the leading countries in the production value of IT hardware. The Taiwanese IT industry largely consists of PCs and related products, as a result it can be called the ‘PC industry’. The indispensable role of Taiwan in the global PC supply system is reflected in the fact that Taiwan manufactures a wide range of PC-related products (we call these ‘sub-products’), and that in many sub-products Taiwan accounts for a large portion of the world total output. For example, Table 9.1 illustrates that in six main sub-products, Taiwan has more than 50 per cent of the world share. In this chapter, the small area from Taipei to Hsinchu (northern part of Taiwan) is regarded as the ‘PC cluster’.2 According to an official statistical report from 1996, there were 1344 manufacturing units in the computer industry in Taiwan, of which 1185 units (88 per cent) were concentrated in this area.3 There were approximately four main groups of actors in this cluster in the late 1990s:

236 91345 27135

1580 1 089 47019

*Original equipment manufacture/own-design manufacture.

12707 62365 27660 84372 67579 98895

2) Total output (volume)

13 548 10392 7797 5 674 2605 1 879

1) Total output (value)

77 78 51

4 70 80 43 74 91 80 86

6 88 84 48 86 95

4) Offshore production (volume: %)

77 93

53 54 25 70 39 74

5) World share (volume: %)

– 60

90 – 82 36 48 –

6) OEM/ODM* rate (%)

Source: The Market Intelligence Centre (MIC) of the Institute for Information Industry 2001, 2000 Yearbook of the Information Technology Industry (in Chinese)


1. Notebook PC 2. Monitor 3. Desktop PC 4. Motherboard 5. CD/DVD/RW 6. Switching power supply (SPS) 7. Case 8. Scanner Total


3) Offshore production (value: %)

Basic data of main sub-products in the Taiwanese PC industry (2000)

Unit: value (million US$) / volume (thousand)

Table 9.1

Clustering and upgrading in global value chains






Producers of PC systems and peripherals. The PC industry consists of many subsectors, in each of which there are dozens of specialized producers. Most of them are SMEs; however, scale is becoming more critical in most subsectors. Parts suppliers and subcontractors. Critical to the Taiwanese success story is the fact that the production of PCs and peripherals can be dismantled into different production stages, with each stage being undertaken by independent subcontractors. There were approximately 100 subcontractors in 1998. In addition, PCs and peripherals consist of a huge array of electronic parts, each one of which can be obtained from specialized suppliers. Although some key parts/components and highquality inputs are partly or entirely purchased from foreign suppliers, the broad supply base of electronic parts and the geographical proximity of suppliers and subcontractors has been a major advantage for the Taiwanese PC industry. Buyers and traders who link local producers to distant markets. As shown in Table 9.1, a substantial proportion of the total products of Taiwanese PC firms is sold through OEM/ODM (a kind of contract manufacturing, to be explained later). Most major world computer companies, as OEM/ODM clients, established an international purchasing office (IPO) in Taipei. Hundreds of indigenous traders, most of them small-scale, played an important role especially in the early years. Related government agencies and public–private supporting institutions including trade associations, research and development (R&D) institutes, and so on.

The concentration, in this small area of PC producers and related actors, facilitates an extensive division of labour and cooperation (and competition) among the different units. A pool of experienced engineers and managers has been created. Geographical proximity enables them to keep in frequent contact, while business and technological information rapidly diffuses amongst them. Such conditions enable entrepreneurs to enter business with low starting capital. It is obvious that all these conditions are closely connected with the conspicuous success story of industrial growth mainly based on home-grown SMEs. Beyond the general observation that clustering contributes to the rising status of Taiwan in world PC hardware production, there is twofold evidence of upgrading. First, Taiwan has gradually shifted its business focus from mature sub-products to new ones. This is an example of product upgrading, namely, the shift to manufacturing more sophisticated products. As shown in Table 9.2, four main sub-products (that is monitors,


Monitor (29.3) Desktop PC (15.0) Motherboard (13.4) Portable PC** (10.1) ... 67.8% SPS Graphics card Terminal Scanner Network card Keyboard

1992 Monitor (37.2) Notebook PC (17.1) Desktop PC (11.8) Motherboard (11.4) ... 77.5% SPS Graphics card Desktop scanner Keyboard CD-ROM drive Network card

1995 Notebook PC (24.9) Monitor (22.3) Desktop PC (19.1) Motherboard (12.8) ... 79.1% SPS CD-ROM drive Case Scanner Graphics card Keyboard


Notebook PC (28.8) Monitor (22.1) Desktop PC (16.6) Motherboard (12.1) ... 79.6% CD/DVD/RW SPS Case Scanner Digital steel camera (DSC)


Source: Author based on MIC, Yearbook of the Information Technology Industry (each year edition)

Notes: * The figures in brackets indicate a percentage of the sub-product in total output (including both domestic and offshore production by Taiwanese producers) of IT hardware. ** Portable PC mainly means notebook PCs, although it includes Laptop PCs and so on. *** The data for 2001 could not be obtained at the time of writing.

Desktop PC (25.5) Monitor (24.0) Motherboard (14.7) Terminal (8.4) ... 72.6% Switching power supply (SPS) Keyboard Graphics card Control card Hard disk drive Mouse


Leading sub-products of the Taiwanese PC industry (every three years)

Monitor (23.5)* Desktop PC (21.4) Terminal (14.9) ... 59.8% . . .


Table 9.2

Clustering and upgrading in global value chains


desktop PCs, motherboards and notebook PCs) have accounted for a substantial portion (about 70–80 per cent) of total PC hardware output. Among them, notebook PCs, a newly emerging sub-product, have steadily increased their weight. Similarly, among other sub-products, relatively new ones such as scanners and CD/DVD–ROM and CD–RW drives have taken the place of conventional sub-products such as keyboards and mouse units. Second, Taiwan has also achieved substantial functional upgrading. Functional upgrading means acquiring more complex functions in the GVC, for example, moving from simple production to design and marketing. Functional upgrading in the PC industry took place mainly through original equipment manufacture (OEM). OEM is a specific form of subcontracting whereby supplying firms undertake a whole production process according to the design specified by the buyer. The product is then sold under the buyer’s brand name. By relying on OEM, local firms can acquire product outlets without shouldering large investment in marketing and after-sales services, and over time they are also able to accumulate product design capability as well as production skills. Once OEM reaches this stage it is called own-design manufacture (ODM). Under ODM, the local firm carries out some or all of the product design, as well as production according to specifications supplied by the buyer. Like OEM, products are sold under the buyer’s brand name. If successful, local firms may advance beyond this to acquire other functions such as logistics and eventually they can start own-brand manufacture (OBM). As some scholars point out, this logic is not easily realized (Schmitz and Knorringa, 2000). However, the Taiwanese PC industry has passed a series of successive development stages from OEM to ODM and part OBM. This will be detailed in the next section. In line with this functional upgrading process, there is an indication that clustering has declined in importance as a simple production site. That is to say, as Table 9.1 shows, the offshore production rate has reached a considerable level in a number of sub-products. This relocation has two important features. First, the offshore production rate measured by volume is larger than the rate measured by value in all sub-products, although the difference between them is not very large (see columns 3 and 4 of Table 9.1). This implies that offshore production is more concerned with price-sensitive product segments than high-end products in any given specific subsector. Second, the offshore production rate has increased at a different pace according to different maturity stages of sub-products. As shown in Table 9.3, the rate started to rise during the earlier stages in conventional, technologically simple (or standardized) sub-products such as monitors, switching power supply (SPS) units, cases and keyboards. Yet, the rate has


Local enterprises in the global economy

Table 9.3 The rate of offshore production of main sub-products (volume: %)

Notebook PC Monitor Desktop PC Motherboard CD/DVD/RW SPS Case Scanner Keyboard









0 24 – 26

0 40 – 34 0 53 – 0 69

0 49 – 44 25 77 – 0 85

0 50 – 40 32 83 57 0 93

0 55 – 40 48 86 70 10 90

0 71 89 37 59 91 75 38 91

3 73 86 41 81 94 78 58 –

6 88 84 48 86 95 80 86 –

46 – 0 68

Source: MIC, Yearbook of the Information Technology Industry (in Chinese, each year edition), and other internal reports of the MIC

remained low, or started to increase only in the later stages, in relatively new, technologically complex sub-products such as notebook PCs,4 scanners and CD/DVD-ROM and CD-RW drives. The rate has been kept moderately high in the motherboard subsector which is relatively mature but entails rapid model change and technologically complex production control. These facts indicate that offshore production has advanced more in relatively mature sub-products, while the production of less mature subproducts (and of high-end product segments of mature sub-products) tends to remain in Taiwan. In addition, as will be examined in detail later, the original Taiwan cluster has kept, or even increased, its importance in knowledge-intensive or headquarters functions such as product design and development, sales and financial management, and coordination activities for offshore production networks. This section has introduced some important features of the Taiwanese PC cluster and its upgrading, mainly based on quantitative data. All the facts shown in this section are in line with the above-mentioned suggestions by Porter, namely that clustering does not necessarily decline. Instead, clusters can upgrade by shedding activities of low value added and concentrating on activities of high value added even in the era of global economy. This process entails repositioning of the cluster in the world economy. The next section will offer a detailed, largely qualitative analysis of the upgrading trajectory.

Clustering and upgrading in global value chains


3. UPGRADING TRAJECTORY IN THE PC VALUE CHAIN The cluster approach prioritizes internal linkages and this is now considered insufficient. In order to understand the upgrading trajectory, the external linkages of the local enterprises also need to be included in the analysis. For this purpose, I draw on the GVC approach. As mentioned earlier, upgrading in the PC value chain took place mainly through OEM. The OEM strategy has been analysed by several scholars, and opinions on whether OEM upgrading is smooth or not differs in the literature. Gereffi (1996; 1999) examines how East Asian garment producers in newly industrializing economies (NIEs) have successfully enhanced their capabilities and reached OBM status through integration into GVCs. Hobday (1995) analyses the export-led learning process of NIE firms in the electronics industry. Although he does not adopt an explicit GVC framework, he confirms the analysis offered by Gereffi. Some scholars, however, have questioned this optimistic view that OEM leads progressively to ODM and OBM (Schmitz and Knorringa, 2000; Humphrey and Schmitz, 2000). The case of Taiwan appears to conform to the view held by Hobday. Yet, detailed observation reveals important limitations of Taiwan’s success. To show this I now trace the upgrading trajectory over time from the early 1980s to the late 1990s. The first and second subsections concern the 1980s and the 1990s respectively, and the third subsection examines achievements and limitations of the Taiwanese cluster. 3.1

The 1980s – from the Initial to OEM Stages

Around the early 1980s, American computer and electronics companies set up subsidiaries in Taiwan and started to produce terminals and monitors. Taiwan was selected as a production site because it had the advantage of a relatively well-developed supply base of electronic parts and an ample supply of low-cost but good quality engineers, which had coincided with the development of the consumer electronics industry since the 1960s. In the 1980s, several indigenous electronics product makers such as Tatung and TECO began to produce terminals and monitors based on their experience in manufacturing televisions. Before long these indigenous producers came to receive OEM orders of monitors, terminals and so on from foreign companies. In the first half of the 1980s, OEM and foreign direct investment (FDI) companies were the main driving force for export expansion in the Taiwanese PC industry, accounting for 40 per cent and 57 per cent of total export value in 1984 respectively (MIC, 1989, 1988 Yearbook of the Information Technology Industry). I regard the first half of the 1980s as the initial stage.


Local enterprises in the global economy

It is important to point out that the major Taiwanese producers already possessed basic production skills and some design capabilities at that time. For example, according to one respondent from a monitor firm, when Tatung, one of the most reputable makers of electronics products, started to produce monitors, ‘They set up design as well as production technology through their own efforts. In addition, they also developed some equipment and measuring instruments on their own’. As well as these major local producers, there were other indigenous forerunners. In the latter half of the 1970s, many local producers embarked on the production of pocket calculators, electronic clocks and video game machines. These producers were mostly SMEs because the production of these items did not require large funds. Many of these firms advanced into PC production in the 1980s. For example, Acer, one of the succcessful companies of the Taiwanese PC industry, developed PCs of their own architecture in 1981 based on the experience of producing pocket calculators (Mizuhashi, 1997). Furthermore, around the same time, many local small producers became engaged in the production of illegal clones of Apple II. They had produced video game machines, but they turned to the production of fake Apple II after the government banned them in 1982 due to their alleged damaging influence on young people. Although the fake Apple II boom declined before long once Apple charged several Taiwanese manufacturers with counterfeiting, manufacturing fake Apple II provided a valuable opportunity to accumulate technological capabilities including design, production engineering, and organizing subcontractors. Through these experiences, Taiwanese manufacturers established a strong initial technological foundation from the very beginning of the PC industry history (Kawakami, 1996). In the latter half of the 1980s when the Taiwanese PC industry experienced its rapid growth phase, the inflow of increasing numbers of indigenous new entrants, especially SMEs, contributed to the formation of the flexible network system of production, which was based on an extensive division of labour and specialization. But OEM and FDI companies still made up a large proportion of total export. They accounted for 43 per cent and 35 per cent of total export value in 1989 respectively (MIC, 1990, 1989 Yearbook of the Information Technology Industry). On the basis of a good technical foundation, Taiwanese local producers actively enhanced their production skills and learned product design capability, largely relying on interaction with OEM clients and assistance from them. As a result, the second half of the 1980s is seen as the OEM stage. Foreign companies including OEM clients and FDI firms provided various forms of technological and managerial assistance to local partners. The company IBM exemplifies this development.

Clustering and upgrading in global value chains


IBM’s demanding procedures for product development, product ramp-up and quality control as well as its gruelling requirements for vendor qualification forced Taiwanese firms to radically upgrade their product quality. It also forced them to develop a broad spectrum of capabilities required for manufacturing as well as product design. In the process of qualifying as an IBM supplier, countless Taiwanese firms learned how to improve their input procurement and production control methods in order to cut costs, improve quality and to speed-up product development cycles and delivery. IBM engineers regularly visited Taiwanese suppliers, screened their production facilities and logistics and assisted them to improve their overall efficiency (Ernst, 1998, 40–1).

OEM clients and FDI companies also made a significant contribution to the proliferation of indigenous new entrants in the sphere of training and nurturing human resources. According to Kawakami, ex-employees of foreign companies ‘eventually left FMs [foreign manufacturers] to play an important role in information and technology diffusion from FMs. Some set up or joined local manufacturers after their employment at the FMs’ local branches. Some exploited the experience they accumulated while serving as engineers at local OEM manufacturers’ (Kawakami, 1996: 15). Kawakami also reports that the required experts for local manufacturers were supplied from foreign firms in the early to mid-1980s. Moreover, around 1990 when many FDI companies withdrew from Taiwan because of rising production costs, a number of experienced engineers that were released from the foreign firms entered local manufacturing, and this solved the shortage of human resources for local firms. In the sphere of marketing, besides expanding exports through OEM, in the latter half of the 1980s Taiwanese local companies began an initial attempt at own-brand overseas sales. The percentage of OBM among local firms in total exports rapidly increased from 5 per cent in 1985 to 17 per cent in 1986. This was maintained at around 20 per cent until the end of the 1980s (MIC, 1989, 1988 Yearbook of the Information Technology Industry). At that time, there were hundreds of indigenous (mainly small-sized) traders. The existence of traders as well as subcontractors provided channels for the rapid expansion of new entrants, even in the absence of heavy investment in production equipment and marketing. For example, in 1989, the number of exporters of all PC-related products reached more than 3700, but of these no more than 650 (17.6 per cent of total exporters) were engaged in manufacturing products (MIC, 1990, 1989 Yearbook of the Information Technology Industry). 3.2

The 1990s – from the ODM to ODM/Global Logistics Stages

After the easy growth phase of the 1980s, the Taiwanese PC industry encountered a setback from 1989 to late 1992. The growth rate sharply


Local enterprises in the global economy

decreased from 38.7 per cent in 1988 to 3.0 per cent in 1989 (MIC, 1996, 1995 Yearbook of the Information Technology Industry). This setback was caused by a worldwide recession in the US, Europe and Japan. It was accelerated by the appreciation of the New Taiwan Dollar (NT$) and an increase in land and labour costs, which resulted in Taiwan losing its comparative advantages as a low-cost production site. At the same time, Taiwan faced serious competitive pressure from both below and above. New lowercost competitors entered the fray in South-East Asia and China, while South Korea strengthened its position as a supplier of monitors and some key parts. Furthermore, Japanese firms started to develop more aggressive global market penetration strategies (Ernst, 1998). Original brand sales of Taiwanese PC firms came under fierce attack, and major foreign firms with investments in Taiwan transferred production sites to less developed countries. In addition, many local firms, including not only small firms but larger ones as well, went out of business. For example, during the second half of 1991, between 50 and 60 Taiwanese computer companies disappeared from the sector each month (this data is cited from Ernst, 1998, 44). In June 1992, the business situation took a new turn. Compaq Computer started harsh price-cutting and other major PC companies could not avoid following this. It made further attacks on the PC industry, which had already become depressed. However, Taiwanese firms were offered an opportunity to recover, as many PC firms from advanced countries started to send larger amounts of OEM orders to Taiwan in order to reduce costs. In the 1980s, Taiwanese OEM clients mainly consisted of middle-level PC vendors from the US (Mizuhashi, 1997), but in the 1990s, these clients included most international leading computer companies. Since the early 1990s, even Japanese computer companies, which had formerly had an antipathy to ‘low quality’ parts and products supplied by other Asian countries, have begun purchasing from Taiwan in order to survive harsh price competition (Fujita and Imai, 1995). After 1992, the Taiwanese PC industry established a different kind of international competitiveness. The primary competitive edge shifted from the low-cost and flexible manufacturing system facilitated by subcontracting to strong capabilities in product design and production control including the introduction of scale-intensive production lines. The majority of OEM was upgraded to ODM. So, I regard the first half of the 1990s as the ODM stage. More and more firms had concerns with quality control (QC). One respondent from a notebook PC firm stated it to be around 1994 that applying for ISO certifications began to become popular. Taiwan has now become one of the most important IPO centres in the world. Many of the major leading computer and electronics companies located their IPOs in Taiwan. For example, AMP, AOC, Apple, AST, AT&T, Bull, Compaq,

Clustering and upgrading in global value chains


Dell, Fujitsu, GE, HP, IBM, Motorola, NEC, Philips, Seiko-Epson, Siemens, Toshiba, Unisys, Xerox, and so on.5 In the sphere of marketing, Taiwanese firms experienced a reversal in own-brand sales when the world PC industry encountered the recession from 1989 to 1992. Although they did not give up on efforts for OBM, the share of OBM in total sales has not grown smoothly since then. In terms of the role of indigenous traders, the importance of these actors seemed to decline rapidly. There were several reasons for this. First, as manufacturing firms grew, it became normal for them to establish specialized sales and marketing departments internally, which reduced their reliance on traders. Second, establishing direct contact with main clients and key parts suppliers was also needed in order to conduct the cooperative development of new products in a rapidly changing business environment. Third, profit margins had become too low to pay traders a brokerage fee.6 The second half of the 1990s can be regarded as the ODM/global logistics stage. Through OEM/ODM partnerships with major foreign clients, Taiwanese manufacturers have further enhanced their technical capabilities both in production skills, including QC and product design and development since the first half of the 1990s. In recent times, Taiwanese manufacturers have generally been forced to enhance production management capability in order to respond to decreasing profit margins and fast technical upgrading. The offshore production of low-end products in mainland China was accelerated after the mid-1990s in order to expand production capacity as well as to reduce production costs. Since the mid-1990s, partnerships with major world clients have evolved further. In order to reduce costs and enhance time-to-market efficiency, Compaq signed the global logistics production and supply model agreement with Taiwan’s MiTAC, whereby Compaq farmed out all stages of the value chain for some of its desktop PCs, with the exception of basic product planning and marketing. In global logistics, the production process of desktop PCs is divided into several stages, and each stage is conducted in a location in which the best cost efficiency for that stage can be realized. For example, the production of motherboards is to be carried out in Taiwan or lower-cost countries such as mainland China and the Philippines because the design of motherboards is increasingly standardized and scale merits have become increasingly important in its production. In addition, other parts/components such as cases and SPS units where content is large but where price is relatively stable are also assembled with motherboards into half-finished goods (called ‘barebones’) in plants located in the same region. Half-finished goods are then shipped to final assembly plants. Meanwhile, for some parts/components that require selection among several alternatives according to user demand and where prices are high


Local enterprises in the global economy

and changing fast, for example central processing units, random access memories and hard disk drives, it is better to purchase and assemble them into final products after receiving orders. Therefore, it is necessary for final assembly plants to be located near important markets such as the US and Europe. Parts/components suppliers such as monitor makers are also required to establish storehouses near important markets. In this way, PC producers can reduce the inventory pressure of final products and avoid a loss caused by a sudden drop in parts prices. At the same time, they can rapidly respond to market changes and realize customization to a certain extent according to user demand (MIC, 1999). In the global logistics model, Taiwanese PC makers need to perform many more functions in the value chain. They are charged not only with production and design, but also with other functions such as the physical delivery and inventory control of parts/components and (half-finished) products. In addition, they are required to establish final assembly plants, warehouses and repair centres near important markets. Through shifting OEM to ODM to global logistics, they are climbing up the GVC ladder (see Figure 9.1). This evolution has meant that Taiwan has declined as a production site. High value-added Creating concepts and basic product planning

Own-brand marketing Logistics (establishing final assembly plants/ warehouses near markets, inventory control of parts and [half-] finished products, delivery, repair service)

Product design and development Managing relations with parts suppliers and subcontractors Production control, QC, cost management, etc. Assembly, production

Low value-added OEM ODM

Global logistics OBM

Source: Author’s survey

Figure 9.1

Functional upgrading in the PC value chain

Clustering and upgrading in global value chains


According to statistical data, in 2000 Taiwan accounted for only 49.1 per cent of total hardware production of all IT products. Mainland China, Malaysia, Thailand, and others made up 31.3 per cent, 4.2 per cent, 1.5 per cent, and 13.9 per cent respectively (MIC, 2001, 2000 Yearbook of the Information Technology Industry). Despite declining importance in the production sphere, the original industrial base in Taiwan remains important because the headquarters function for the offshore production system is still located there. This includes supplying important inputs, equipment and technologies, handling core activities for foreign subsidiaries such as product planning, R&D, communication with suppliers and clients, logistics, and financial management. 3.3

Achievements and Limitations of Taiwan’s Experience

In the GVC literature, it is recognized that chain structures which facilitate the fast learning of production skills may place barriers on the acquisition of more complicated functions such as design and marketing (Schmitz and Knorringa, 2000). This subsection examines achievements and limitations of Taiwan’s upgrading by focusing on the sphere of product design and marketing respectively. According to data gained from our fieldwork during the late 1990s, in the sphere of design, interactions with foreign clients are one of the most important knowledge sources for new model design and development. OEM/ODM clients include not only major foreign computer companies such as IBM and Dell, but also non-major vendors and distributors. Major clients may provide various kinds of technical assistance. According to our questionnaire survey, at least in the monitor and notebook PC subsectors, a substantial number of sample firms received technological assistance for QC, improvement of production process and product and so on from clients. However, a closer examination reveals that Taiwanese producers have already accumulated substantial capabilities in product design and development in addition to production capabilities. Furthermore, Taiwanese producers undertake the entire process from new model design and development to mass production. Therefore, most buyers’ assistance is not very substantial in nature. In a number of cases Taiwanese producers only receive ‘demands’ for adjustments of products and production processes or they have minor technical interactions. For example, according to one respondent from a monitor firm, there are not many profound technology transfers now because the monitor industry has matured. And holding enough technological capability is a necessary condition for getting orders. Similarly, in the relatively new notebook PC subsector, one Taiwanese


Local enterprises in the global economy

analyst noted that ODM had become typical in the last few years and OEM (that is just undertaking production) rarely occurred. During the ODM stage, the design of a new product model is carried out by Taiwanese firms largely based on their own experience. This detailed examination, however, also indicates that strict demands and precise feedback from major clients are conducive to the improvement of production and QC technologies and product design, and many of the latest technologies are still obtained from foreign buyers. According to one respondent, for example in the monitor subsector, the degree of dependence on foreign technological sources is low. Yet, complete ODM is not very common. Many new technologies are obtained from foreign sources. Another respondent also claimed that major PC companies such as Dell provided a new concept for future business and that Taiwanese firms carried out design and development based on this. Besides this, partnerships with world leading companies may enable Taiwanese firms to learn other aspects of business management such as personnel control, company culture and management policy (Zhou, 1999). According to one Taiwanese analyst, the main competitive source in Taiwan is the speed of product design and development, as well as flexible and low-cost production. The short cycle of new model design and development is facilitated by the hard work of R&D staff, as well as cooperative interaction with parts suppliers and in-house close communication between the R&D and other departments. Although Taiwanese firms are not necessarily superior to major foreign companies in design capability, they work at a faster pace. This is due to the fact that the Taiwanese endure long working hours. The analyst stated: ‘Foreign clients have gradually transferred tasks which they do not like to do to Taiwan. In the early years, they transferred production, now it includes design too, and logistics in the future . . . Tasks which are troublesome and provide relatively low profit margins are transferred to Taiwan’. Taiwan’s achievements are less conspicuous in the field of marketing. As mentioned above, Taiwan has already reached the ODM stage but found it difficult to progress into the OBM stage. The OEM/ODM rate has not decreased despite Taiwan’s rising status in the world PC industry (see Table 9.4). Even in the motherboard subsector that has mainly developed in connection with the rise of non-brand markets, Taiwanese firms have recently begun to try to gain OEM/ODM orders from major computer companies in response to the limitation of non-brand markets. Stressing the importance of OEM/ODM does not mean that Taiwanese firms passively depend on foreign clients as their only lifeline to distant markets. Although it is still difficult to develop outlets in leading markets such as the US, many firms have already embarked on own-brand sales as


Clustering and upgrading in global value chains

Table 9.4

The rate of OEM/ODM (%) of four main sub-products

Notebook PC Monitor Desktop PC Motherboard









77 69 49 –

77 67 40 –

79 66 37 –

82 – 53 –

83 67 65 –

84 65 72 27

– 68 76 29

90 – 82 36

Note: The OEM/ODM rate expresses the share of OEM⫹ODM products in total sales. Source: MIC, Yearbook of the Information Technology Industry (in Chinese, each year edition) and other internal reports of the MIC

well as conducting OEM/ODM. As the fieldwork confirmed, many local producers including small ones have already established specialized sales and marketing departments and established foreign branches in important markets. The strategy of OEM/ODM differs between firms, but it is a delicate task to achieve a proper balance between OEM/ODM and OBM. On the one hand, OBM results in higher profits,7 but firms must invest substantially in advertisements and after-sales service networks. On the other hand, the profit margin of OEM/ODM is so small that it can yield a profit only when production exceeds a certain large amount. Furthermore, for individual firms, partnerships with foreign clients are usually unstable. However, Taiwanese firms still seek partnerships with major clients, because large OEM/ODM orders enable them to increase total sales and offset lower profit margins. 3.4


This section has examined the upgrading trajectory of the Taiwanese PC industry from the early 1980s to the late 1990s. During the initial stage, the main exporters were subsidiaries of foreign computer/electronics companies. In the OEM stage, an increasing number of local SMEs advanced into the new business. Foreign companies provided various forms of technological and managerial assistance for local firms directly and indirectly. Some local firms initially attempted to export their own-brand products. Many OEM operations advanced to the ODM stage. Although many FDI companies withdrew from Taiwan due to rising production costs, Taiwan began to receive a large amount of OEM/ODM orders from many major world companies after 1992. In the sphere of marketing, local firms experienced a reversal in own-brand sales during the severe recession from 1989 to 1992. Since then the share of OBM in total sales has not grown


Local enterprises in the global economy

steadily although they did not give up OBM. Finally, in the ODM/global logistics stage, through experiencing partnerships with major foreign clients, Taiwanese manufacturers further established solid capabilities for production management and product design and development. In addition, major foreign clients started to make Taiwanese firms take on logistics after the mid-1990s. Offshore production in mainland China increased rapidly. On the whole, forward linkages with foreign clients have enhanced interdependence entailing substantial functional upgrading, but this has not necessarily led to increasing profit margins and strengthening bargaining power against major clients. In a sense, it seems that the functional upgrading is the result of major clients withdrawing from troublesome and less profitable tasks and transferring them to Taiwan.



As proposed by Humphrey and Schmitz (2000), both global and local linkages need to be studied in order to explain how upgrading takes place and evolves. But there is very little understanding of how the two interact. This section attempts to shed further light on the interaction. In the previous section, the analytical focus was placed on the transformation of forward linkages with foreign clients. This section is concerned with how it is accompanied by the transformation of local cluster linkages. The local linkages are analysed by disaggregating them into three categories – backward, horizontal and institutional linkages. As in the previous section, this analysis is undertaken by dividing the history into two periods – the 1980s and 1990s. 4.1 The 1980s – the Construction of a Low-cost and Flexible Production Base External linkages with large distant markets may stimulate the development of local cluster linkages; this is observed in the early years of the Taiwanese PC industry. In line with export growth based mainly on FDI and OEM, local linkages also evolved. They included the formation of a solid base of supporting firms such as suppliers and subcontractors in backward linkages, the entrance of new local firms and dynamic processes of competition and cooperation in horizontal relations, and encouragement of the PC sector as a leading export industry by government and support institutions. Backward linkages: The existence of a well-developed system of parts suppliers and subcontractors signifies Taiwan’s attractiveness for foreign

Clustering and upgrading in global value chains


buyers. In the 1980s, Taiwan’s primary competitive edge was the low-cost and flexible manufacturing system that it ensured. Since the 1960s, Taiwan has had extensive experience in the manufacture of consumer electronic products such as radios, televisions and pocket calculators, and this was a major advantage for the PC industry. Accompanying the development of the electronics industry, the electronic parts and components sector has also developed. According to one respondent, after the late 1970s when labour costs increased and production plants for these conventional electronic products started to move to lower-cost countries, electronic parts producers shifted their principal clients to manufacturers of PCs and peripherals. From the start of the PC industry the local supply rate of necessary parts was relatively high. As well as the solid supply base of electronic parts, the subcontracting system in the PC industry developed rapidly in the second half of the 1980s. Large numbers of SMEs entered into the assembly of PCs and peripherals and replaced large firms as the main driving force behind business expansion. Many of them specialized in one or several stages of the production process, farming out other stages in the process. Unlike the 1990s when production automation was advancing, the assembly of electronic circuit boards in the 1980s depended largely on manual work; therefore, the importance of subcontracting remained high in order to utilize lower labour costs. In this way, the system of an extensive division of labour developed (Kawakami, 1998). The external linkages ensured access to a large export market without which the local highly specialized production system could not have developed. Horizontal linkages: Horizontal relations between fellow producers in the PC industry, like in many other industrial sectors, are characterized by strong rivalry, especially price competition. In the first half of the 1980s, the PC business was mainly dominated by a small number of FDI companies and large indigenous firms that received OEM orders from foreign clients. The share of the top 20 firms (including FDI companies) comprised 82.4 per cent of total exports in 1984. This business scene changed due to the inflow of a large number of new entrants, especially SMEs, in the latter half of the 1980s. Although the role of FDI and indigenous large firms was still important, SMEs became the leading contributors to the rapid expansion in exports. Subsequently, in 1986 the share of the top 20 firms in total exports declined to 57.4 per cent (MIC, 1987, 1986 Yearbook of the Information Technology Industry). The Yearbook (1987, 137) reports: This phenomenon is desirable in one aspect, but undesirable in another aspect. It is desirable because the sharp drop in the share of the top 20 firms reflects the rising vitality of SMEs and the expanding base of industrial development. It is


Local enterprises in the global economy

undesirable because most firms concentrate on the narrow range of similar products and compete against each other, which leads to low profit rates.

This does not mean that horizontal relations in Taiwan were restricted to price competition. Levy and Kuo (1991) describe the basic attitude of Taiwanese PC producers towards innovation in the late 1980s. According to them, Taiwanese PC producers took the ‘bootstrap strategy’, whereby a firm enters into trade often with limited funds and readily takes risky initiatives while seeking lucrative lines of business.8 The low cost of failure ensured by the network of subcontractors and traders made it easier. In addition, although the mastery of complex technology was not necessarily a part of the ‘bootstrap strategy’, firms could get the opportunity to achieve technological learning ‘via an incremental process of graduating from simpler to more complex tasks, and from tasks that involved relatively little research and development (R&D) expenditures, to those for which the R&D requirements were more substantial’ (Levy and Kuo, 1991, 368). All these conditions implied that a large number of new entrants tried out many alternative technical approaches simultaneously as a result of the low entry barrier, which was one of the advantages of clustering. Consequently, this led to rapid trial-and-error learning within the cluster as a whole. In such an environment, where the capacity to respond quickly and flexibly to emerging market opportunities was critical, this strategy seemed to work very well. As mentioned in section 3.1, OEM clients and FDI companies made a significant contribution to the proliferation of indigenous new entrants by nurturing and training human resources. Besides foreign companies, local first generation manufacturers also played an important role. Among them, the contribution of Acer is often mentioned. High-calibre people who came from Acer moved to occupy top management positions of many firms in various subsectors including PCs, peripherals, PC-related parts and even communication equipment (Zhou, 1999). In the monitor subsector, Tatung played a similar role. According to one respondent, in the early 1980s, Tatung owned an R&D department that employed 200–300 engineers. Many of these engineers moved and joined newly emerging monitor firms. The outflow of these human resources from foreign and major domestic firms alike was an indispensable factor in encouraging the rapid expansion in the number of local manufacturers in the latter half of the 1980s. Institutional linkages: The commitment of government to the IT industry became clear in the 1970s. In 1974, the Electronics Research & Service Organization (ERSO) was established in the Industrial Technology Research Institute (ITRI) in order to undertake the development of IT hardware. In 1980, the Hsinchu Science-based Industrial Park was estab-

Clustering and upgrading in global value chains


lished to facilitate investment in high-technology industries for both foreign and domestic companies. Around the same time, two national plans for the industrial development of the electronics and IT industries were published. The Ten-Year Plan for the Development of the IT Industry (1980–1989) is particularly relevant. This plan had two main goals: (1) developing the new IT industry as a strategic industry, and (2) promoting it for exports (Wang, 1995–96). According to Fuller (2002), Taiwan’s technology policy is less coloured by techno-nationalism than that of South Korea and Japan. They have accepted the presence of multinational companies (MNCs) in strategic industries and built up strategic suppliers for MNCs, rather than fostering vertically integrated national champions. Local firms have gained new knowledge through interactions with foreign clients. The state has facilitated this through instruments such as licensing foreign technologies and negotiating licensing on behalf of local firms. But the role of the state varied across the different sectors of the IT industry. Unlike the semiconductor industry where several integrated circuit (IC) manufacturers were established under the ERSO initiative, state support was modest in the PC sector. Although several computer projects were carried out by the ERSO and developed technologies were transferred to private companies, the ERSO did not appear to play a critical role. This is because of the ability of even SMEs to enter into the business during the early years, as well as intensive OEM relationships with foreign firms. 4.2

The 1990s – the Evolution of a Local Base towards Global Operations

In the 1990s, partnership with foreign clients has increased its interdependent nature as Taiwanese producers have enhanced their capabilities and succeeded in functional upgrading. In response to this external transformation, local linkages have also changed in nature. This can be seen through the enhancement of cooperative relationships with suppliers and of local key parts supply bases, the increasing importance of scale merits and inhouse technological efforts of individual firms in horizontal competition, and the continuous support of government and related institutions to strengthen the IT sector as a strategic industry and their attempts to upgrade the status of Taiwan as a regional operation centre. Backward linkages: In order to respond to increasingly strict demands from clients such as more efficient production control and quicker product design and development, the enhancement of cooperative relationships with partners in backward linkages is critical. As confirmed in our fieldwork in Taiwan, relationships between PC manufacturers and key parts suppliers were very close in nature during the late 1990s. They include


Local enterprises in the global economy

adopting a limited number of partners through a fair and strict screening, preferring long-term commitment, keeping frequent communication involving not only purchasing/sales staff but also R&D, QC and top management, and taking the ‘voice strategy’ in problem solving (that is to set up a communication system to work things out with an original partner). In addition, key parts suppliers were considered the most important knowledge source for new product design and development, because the upgrading of PCs and peripherals was initiated by the upgrading of key parts to a substantial extent. In terms of relationships with subcontractors, the same attitude was observed, although interaction with subcontractors had no importance as a knowledge source for product design and development. Some may wonder when and how such close relationships developed. Answering this question based on a rigid empirical proof is a difficult task because there is a lack of data. But a plausible inference is that it occurred (or was accelerated) in the early 1990s when many major foreign companies began to provide large-scale OEM/ODM orders to Taiwan so that they could deal with increased downward pressure on prices. Before then, orders came mainly from North American small and middle-level computer vendors. Major foreign clients normally make stricter demands on quality, functions and lead time; therefore, in order to respond, Taiwanese producers had to strengthen partnerships in backward linkages. With intense global competition, communication with suppliers has become increasingly close, although the significance of subcontracting has declined since the mid-1990s because of stricter QC demands from clients and the availability of low-cost offshore labour. Having good access to parts sources becomes one of the factors that determine business competition. In the late 1990s, the major Taiwanese producers started to introduce IT-based management systems such as electronic data interchange (EDI) and supply chain management (SCM), thus involving their suppliers. This is also a precondition for constructing global logistics. As well as strengthening cooperation in backward linkages, the Taiwanese cluster has expanded the local supply base for key parts. The rate of local supply of necessary parts was relatively high from the start of the PC industry. However, a lot of high quality electronic parts have been supplied by Taiwanese subsidiaries of Japanese manufacturers and many key parts have also been imported from foreign suppliers. Yet, indigenous firms have gradually begun to produce high-quality parts and key components. For example, the government has tried to establish a solid supply base of Thin Film Transistor-Liquid Crystal Display (TFT-LCD) (DigiTimes, 31 August 1999). TFT-LCD accounts for the largest proportion of total material costs of notebook PCs, which had been entirely imported from Japanese and Korean suppliers. In 1999, there were seven TFT-LCD pro-

Clustering and upgrading in global value chains


ducers in Taiwan, most of which relied on technology transfers from Japanese producers. This appears to be one example of Japanese firms advancing into more profitable and technologically higher products, while transferring old-generation technology to their Taiwanese partners (DigiTimes, 8 October 1999). Yet, expanding the local supply base of key parts contributes to shifting cluster activities to more profitable and knowledge-intensive spheres in the value chain. Horizontal linkages: From the late 1980s when the hyper-growth phase terminated, there were signs that the business environment was changing and scale merits became a crucial factor in business competition. OEM/ODM orders from major foreign clients tend to concentrate on several top manufacturers from each subsector in Taiwan, which results in an increasing share in the total output of Taiwan for these top manufacturers. In the mid-1990s, one Taiwanese senior industrial analyst reported: The Taiwanese PC industry has already reached the matured stage and new entry is becoming more and more difficult. It is a very hard task for new entrants to attain large-scale production capacity, to develop new products ceaselessly, to purchase parts at reasonable prices, and to sell products to the world market through establishing partnerships with respectable buyers (Hwang, 1996, 77).

This tendency has accelerated, especially since the low-price PC started to become popular in 1997. Harsh price competition was still present in many subsectors in the late 1990s. It has forced Taiwanese producers to strengthen comprehensive cost management capabilities. In the latter half of the 1980s, Taiwan established an exceptional low-cost production system facilitated by an extensive division of labour and subcontracting, and more recently, by constructing scale-intensive production lines and increasing offshore production. But competing through expanding production capacity and scrambling for large-scale OEM/ODM orders from major foreign computer companies lead to further depressed profit margins. Foreign clients take advantage of this rivalry to push down prices, although the establishment of this lowcost/low-profit system raises the entry barrier for producers from other countries. Like in the 1980s, however, it does not mean that price competition is the only feature of horizontal relationships. Innovation and flexibility are other key factors for out-competing rivals. As stated above, it had already been observed in the late 1980s that risk-taking initiatives by a number of firms and their incremental technological mastery under the ‘bootstrap strategy’ had contributed to technological upgrading of the cluster as a whole. As confirmed in our fieldwork, in the late 1990s the same mechanism played


Local enterprises in the global economy

an important role. Furthermore, mutual learning through indirect and informal routes, including job-hopping by R&D staff, was an important knowledge source for product design and development and other types of innovation although formal inter-firm technological cooperation was rarely observed among local competitors. Recently in-house learning efforts by individual firms seem to have increased in relative importance. Manufacturers can no longer survive by depending on imitation and follow-up innovation alone without a substantial in-house investment in R&D, because the speed of product design and development is critical in business competition. At the same time, the recent design and performance of products is becoming similar among local producers because the industry has become mature and they all use common key parts provided by a limited number of suppliers. Therefore, the outcome of business competition is increasingly decided by scale economies and financial power. This is the case even in relatively new subsectors such as notebook PCs. In many subsectors, conducting business has recently become more difficult for small-scale firms because they increasingly require additional resources for large-scale production, ceaseless model change and global footholds. Institutional linkages: Following on from the first Ten-Year Plan during the 1980s, the Ten-Year Plan for the Development of the IT Industry (1990–2000) was set out. This second Ten-Year Plan aimed at facilitating the application of IT technology for the improvement of citizens’ quality of life, upgrading technology and product level, and expanding the supply of IT-related human resources. The plan also viewed the software industry as the future leading force and put forward a plan to construct a Software Industrial Park (Wang, 1995–96). In addition, several PC-related R&D consortia, such as the ‘Notebook PC Strategic Alliance’ (1990–1991) and the ‘NewPC consortium’ (1993–1997), were established by the Computer & Communications Research Laboratories (CCL) which separated from the ERSO in 1990 (Mathews, 2001). Despite these polices, however, state promotion in the PC sector can be seen as modest. As noted by Wang (1995–96), government efforts were largely limited to offering a good institutional and legal infrastructure and providing fiscal and financial incentives. R&D consortia created solid technological fruits and offered local producers valuable opportunities to learn new knowledge, but could not bring substantial business success (Mathews, 2001). As confirmed by my own fieldwork during the late 1990s, business people did not usually have a high opinion of the role of state and supporting organizations including trade associations, although it was recognized that the business environment of Taiwan was fairly good in comparison to other Asian countries such as Malaysia and mainland China. Private firms

Clustering and upgrading in global value chains


also rated the importance of public research institutes and trade associations poorly as a knowledge source for product design and development (Kishimoto, 2002; 2003). Recently, however, faced with the fear of industrial hollowing which is a result of rising competitiveness in mainland China, the government has accelerated policies to strengthen Taiwan’s status as a regional innovation and operation centre over various industrial sectors. They include measures to encourage establishing business headquarters in Taiwan so that high value-added activities such as R&D and marketing can take place there, as well as to enhance the high-technology industries of which the IT industry is the core, promote investment in R&D and innovation, improve the training of technical personnel, and construct the national information infrastructure.9 4.3


Sections 3 and 4 have examined the upgrading trajectory of the Taiwanese PC cluster in the GVC from the early 1980s to the late 1990s as well as the transformation of local cluster linkages that accompanied it. The main findings are summarized in Table 9.5. The table is not conclusive due to lack of data. However, it does reveal that global and local linkages have reinforced each other and enhanced the quality of linkages in all the categories. In the 1980s, based on a relatively established initial industrial foundation that was a heritage of the conventional electronics industry since the 1960s, local cluster linkages started to evolve through export growth mainly based on FDI and OEM. The primary competitive edge of the Taiwanese PC industry was the low-cost and flexible production system based on the extensive division of labour and subcontracting among local SMEs. This early industrial dynamism was fuelled by the risk-taking initiatives of increasing numbers of new local entrants and (rather modest) encouragement by government and supporting institutions. In the 1990s, after the initial rapid growth phase terminated, Taiwanese producers were forced to establish a different kind of international competitiveness. The primary competitive edge shifted to strong capabilities in product design and production control including operational skills in the automated and scale-intensive production lines. Low cost and flexibility benefits were increasingly realized based on offshore production and inhouse organizational innovation rather than local subcontracting. As relationships with major foreign clients have become increasingly interdependent (as reflected in substantial functional upgrading), conducting business now requires far higher additional resources for scale-intensive


The first half of the 1990s

The OEM stage

The second half of the 1980s

Some local firms had already possessed basic production skills and some product design capabilities (the heritage of the electronics industry since the 1960s).

The initial stage

Backward linkages: The solid supply base of electronics parts and the geographical proximity of suppliers were a major advantage from the beginning. The subcontracting system developed in the latter half of the 1980s.

Local cluster linkages

Around 1990, many FDI firms withdrew due to rising production costs. But many company world leaders started to give large amounts of OEM/ODM orders after 1992.

Backward linkages: The enhancement of cooperative relationships with partners became critical in order to respond to increasingly stricter demands from major clients. In the late 1990s, cooperation with parts suppliers

Horizontal linkages: The inflow of many SMEs led to harsh price competition in the latter half of the 1980s. FDI firms/OEM clients provided various forms of It was also observed during this time that risk-taking technical and managerial assistance to local producers, initiatives by many firms and incremental technological as well as taking on marketing. mastery under the ‘bootstrap strategy’ were conducive to the technical upgrading of the cluster as a whole. The low-cost and flexible manufacturing system facilitated by the subcontracting network was Institutional linkages: The government promoted the IT established. sector as a strategic industry and a leading exporter since the early 1980s, however, the PC industry received Local firms began initial efforts at own-brand offshore only modest state support. Several computer projects sales (indigenous traders played an important role). were carried out by the ERSO/ITRI.

FDI firms/OEM clients were the main driving force of export expansion.

Forward linkages with external buyers

Upgrading of the Taiwanese PC cluster (the 1980s–the 1990s)

The first half of the 1980s


Table 9.5


Offshore production was accelerated largely in mainland China, with the original Taiwanese cluster shouldering the role of regional headquarters.

Local firms experienced substantial upgrading including design and logistics as well as production, but it entailed important limitations such as squeezed profit margins and difficulty in expanding OBM.

Major clients came to focus on core, profitable functions such as basic product planning and brand marketing, leaving others to Taiwanese producers.

Local firms experienced a reversal in OBM due to the world recession during the late 1980s and early 1990s (the importance of indigenous traders declined rapidly).

The primary competitive edge shifted to strong capabilities in product design and production control including operational skills for the automated equipment and scale-intensive manufacturing system. The QC procedure also began to be established.

Source: Author’s survey

The ODM/ Global Logistics stage

The second half of the 1990s

The ODM stage

Institutional linkages: The IT industry continued to be a strategic industry in the 1990s, and some additional assistance programmes were carried out. Recently, the government has accelerated policies to support the upgrading of Taiwan as a regional high-technology and operation centre.

Horizontal linkages: Scale merits became critical in competition. OEM/ODM orders from major foreign clients tended to concentrate on several top firms within each subsector. The scramble for orders from major clients led to further depressed profit margins. In the technological sphere, mutual learning through indirect routes was an important knowledge source in the late 1990s, but substantial in-house innovation by individual firms became indispensable. The recent business environment has become increasingly difficult for small firms.

became increasingly important, while the significance of subcontracting declined. The cluster expanded the local supply base of key parts.


Local enterprises in the global economy

production, rapid model change and global operations. In line with this, there has been local-level evolution such as tighter cooperation with partners in backward linkages, increased influence of large producers in local competition, and a government attempt to upgrade the status of Taiwan as a regional operation centre.



The main purpose of this chapter is to examine the upgrading trajectory of the Taiwanese PC cluster. In order to carry out this task, two analytical devices were used: i) the distinction between the production and knowledge systems, suggested by Bell and Albu (1999), and ii) the fusion of the cluster and GVC approaches, proposed by Humphrey and Schmitz (2000). Section 2 gave an overview of development patterns in the Taiwanese PC industry. It confirmed that Taiwan has experienced substantial product and functional upgrading, and at the same time the offshore production rate has reached a considerable level in a number of sub-products. A detailed examination reveals that products requiring sophisticated skills continue to be made in Taiwan and those requiring lower skills are increasingly made offshore. In other words, this means that the importance of clustering has diminished in the production system and has remained high in the knowledge system. Sections 3 and 4 show that upgrading prompted by global linkages necessitates substantial transformation in the quality of local cluster linkages. Section 3, drawing on the GVC approach, explained Taiwan’s upgrading trajectory through OEM from the early 1980s to the late 1990s. Over these two decades, Taiwan has accumulated production skills and design capabilities and has undertaken more complicated functions such as logistics. This seems to be in line with the optimistic scenario offered by Gereffi (1996; 1999) and Hobday (1995). However, as pointed out by other authors (for example Schmitz and Knorringa, 2000), upgrading through OEM is not straightforward. A detailed examination revealed that it entailed important qualifications such as squeezed profit margins and difficulty in expanding OBM. In a sense, the functional upgrading that was accomplished can be seen as the result of major clients withdrawing from troublesome and less profitable tasks and transferring them to Taiwan. Section 4 analysed the transformation of local cluster linkages that made all this upgrading possible. Put simply, a local base for low-cost and flexible production facilitated by an extensive division of labour was constructed in the 1980s, and local restructuring for more knowledge-intensive activities and global operations occurred in the 1990s. In Section 2, I intro-

Clustering and upgrading in global value chains


duced the thesis by Porter that clustering does not necessarily decline; instead, the cluster’s competitiveness is recovered by shedding simple production tasks elsewhere and concentrating on more strategic functions even in the era of the global economy. Taiwan is a good example of this and it confirms the applicability of this scenario to latecomer economies as well as difficulties that this process entails. I now turn to several important issues for future examination. This chapter analysed upgrading in the GVC from the viewpoint of local latecomer firms; however, we also need to place more emphasis on examining the structure and transformation of the GVC in terms of the lead firms. This includes questions such as: How has competition between lead firms and their competitive strategies influenced the chain structure? How has the distribution of value added changed between the chain activities? Has the distribution of economic gains along the chains changed as a result? Why is Taiwan attractive to lead firms in comparison to rivals such as South Korea, South-East Asian countries, and North American contract manufacturers? What technical elements account for differences in upgrading between industrial sectors? Finally, future studies need to pay more attention to different upgrading strategies within individual firms. In this respect, Lee and Chen (2000) provide an insightful analysis. They specify several different types of ODM and OEM, and point out different possibilities for learning and raising profits in various chains. In their model, the development trajectory is not linear (that is from OEM to ODM to OBM). Largely based on theoretical consideration, they emphasize the importance of adopting different strategies with different market channels and maximizing the gains through combining and leveraging. The analysis of Taiwan’s case, however, implies that multiple linkage strategy is not easy to implement in reality. The competitiveness of SMEs consists in focusing on a specific range of products and a niche market. Although the concurrent implementation of OEM/ODM and OBM was accomplished by some Taiwanese firms, to serve two different linkages requires a complicated production organization and management structure, which most firms find difficult to establish. One of the advantages of clustering is the possibility for the collective implementation of a multiple linkage strategy. A cluster consists of various types of firms that focus on different product segments and niche markets, and knowledge and capability are commonly utilized through external economies and joint action. In Taiwan, such collective practice has contributed to the emergence of a ‘onestop supply centre of PCs’ in terms of the cluster as a whole. This collective capability requires continuing individual efforts in order to succeed.


Local enterprises in the global economy

Thus, under increasing competitive pressure, Taiwanese firms are trying to diversify into other IT products, and are trying to reduce costs by constructing large-scale factories in mainland China. Simultaneously they are trying to develop the Chinese market as a last opportunity to establish ownbrands. It remains to be seen whether these strategies succeed. If so, it will be hard to attribute this success to any particular factor. The Taiwanese computer industry underlines above all interdependence: of OEM/ODM and OBM, of intra-firm and inter-firm learning and of fostering global and local linkages.

NOTES 1. This fieldwork was originally conducted as part of my D.Phil. thesis. For more detail, see Kishimoto (2002; 2003). 2. It includes three prefectures (Taipei, Taoyuan and Hsinchu) and three cities (Taipei, Keelung and Hsinchu). 3. Directorate-general of Budget, Accounting and Statistics, Executive Yuan 1998, The Report on 1996 Industry, Commerce and Service Census Taiwan-Fukien Area, the R.O.C., Vol. 3 Manufacturing. Strictly speaking, the computer industry appears in the statistical report under the designation of ‘data storage media & processing equipments manufacturing’. 4. The low offshore production rate of notebook PCs can be linked to the government’s prohibition against moving strategic industries to mainland China. But many business people that I interviewed stated that it was basically because of technological reasons rather than political consideration during the late 1990s. 5. For detailed information on IPOs, see ‘International purchasing centre: Taiwan’ (Components Times 1996, July, pp. 81–96) (in Chinese). 6. This explanation is based on interviews I conducted with two Taiwanese business people and Lin (1996). 7. According to an article in DigiTimes, the average gross profit rate of Taiwanese producers in the monitor subsector is less than 6 per cent, but OBM producers record their gross profit rate as 10–12 per cent (DigiTimes, 27 July 1998). 8. In the work of Levy and Kuo who study PC and keyboard sectors by comparing Taiwanese and Korean manufacturers, the ‘bootstrap strategy’ followed by Taiwanese firms is contrasted with the ‘assembly strategy’ adopted by large Korean firms. In the assembly strategy, a firm enters into business with a large starting capital and attempts to move rapidly down the learning curve, increasing productivity and thereby reducing unit costs as experience accumulates. 9. For more details, see Yearbook of Industrial Development (in Chinese) published by the Industrial Development Bureau (IDB) of the Ministry of Economic Affairs (MOEA) and Yearbook of the Information Technology Industry (in Chinese) published by the Market Intelligence Center (MIC) of the Institute for Information Industry.

REFERENCES Bell, M. and M. Albu (1999), ‘Knowledge systems and technological dynamism in industrial clusters in developing countries’, World Development, 27 (9), 1715–34.

Clustering and upgrading in global value chains


Ernst, D. (1998), ‘What permits small firms to compete in high-tech industries? Inter-organizational knowledge creation in the Taiwanese computer industry’, DRUID Working Paper, No. 98–3, Danish Research Unit for Industrial Dynamics, Copenhagen Business School. Fujita, K. and T. Imai (1995), ‘The surging Asian-made parts: requiring the reconsideration of PC design’ (in Japanese), Nikkei Electronics, No. 646: 83–106. Fuller, D. (2002), ‘Globalization for nation building: industrial policy for hightechnology products in Taiwan’, Working Paper 02.02, Cambridge: The MIT Japan Program Center for International Studies. Gereffi, G. (1996), ‘Commodity chains and regional divisions of labor in East Asia’, Journal of Asian Business, 12 (1), 75–112. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Hobday, M. (1995), Innovation in East Asia: The Challenge to Japan, Aldershot, UK and Brookfield, US: Edward Elgar. Humphrey, J. and H. Schmitz (2000), ‘Governance and upgrading: linking industrial cluster and global value chain research’, IDS Working Paper, No. 120, Brighton: Institute of Development Studies. Hwang, C-Y. (1996), Taiwan: The Republic of Computers (Japanese translation, originally in Chinese), Tokyo: ASCII Corporation. Kawakami, M. (1996), ‘Development of the small- and medium-sized manufacturers in Taiwan’s PC industry’, Discussion Paper Series No. 9606, Taipei: ChungHua Institution for Economic Research. Kawakami, M. (1998), ‘Division of labour between firms and development of firms and industry: a case study of the Taiwanese PC industry’ (in Japanese), Asia Keizai, 39 (12), 2–28. Kishimoto, C. (2002), ‘The Taiwanese personal computer cluster: trajectory of its production and knowledge systems’, D.Phil. dissertation, Brighton: Institute of Development Studies. Kishimoto, C. (2003), ‘Upgrading in the Taiwanese computer cluster: transformation of its production and knowledge systems’, IDS Working Paper No. 186, Brighton: Institute of Development Studies. Lee, J-R. and J-S. Chen (2000), ‘Dynamic synergy creation with multiple business activities: toward a competence-based growth model for contract manufacturers’, in R. Sanchez and A. Heene (eds), Theory Development for CompetenceBased Management, Stanford: JAI Press, pp. 209–28. Levy, B. and W-J. Kuo (1991), ‘The strategic orientations of firms and the performance of Korea and Taiwan in frontier industries: lessons from comparative case studies of keyboard and personal computer assembly’, World Development, 19 (4), 363–74. Lin, M-J. (1996), ‘Creating the miracle of hyper-growth: giga-byte’, in C-Y. Li (ed.), Taiwan Dragon: Case Studies of Successful Upgrading of SMEs (in Chinese), Taipei: Small and Medium Enterprise Administration of the Ministry of Economic Affairs. Mathews, J. (2001), ‘The origins and dynamics of Taiwan’s R&D consortia’, Research Policy, 1315, 1–20. MIC (Market Intelligence Center of Institute for Information Industry) (1999), Analysis of the Development of Production and Sales in the Taiwanese PC Industry (in Chinese), Taipei: Institute for Information Industry. MIC (Market Intelligence Center of Institute for Information Industry), Yearbook


Local enterprises in the global economy

of the Information Technology Industry (various editions) (in Chinese), Taipei: Institute for Information Industry. Mizuhashi, Y. (1997), ‘40-years history of the Taiwanese electronics industry I, II, III’ (in Japanese), Kouryuu, No. 558 (31 August), No. 559 (15 September) and No. 560 (30 September). Porter, M. (1998), ‘Clusters and the new economics of competition’, Harvard Business Review, November–December: pp. 77–90. Porter, M. (2001), ‘Regions and the new economics of competition’, in A.J. Scott (ed.), Global City-Regions: Trends, Theory, Policy, New York: Oxford University Press, pp. 139–57. Schmitz, H. and P. Knorringa (2000), ‘Learning from global buyers’, Journal of Development Studies, 37 (2), 177–205. Wang, W-C. (1995–96), ‘Developing the information industry in Taiwan: entrepreneurial state, guerrilla capitalists, and accommodative technologists’, Pacific Affairs, 68 (4), 551–76. Zhou, F-Y. (1999), The Story of ASUS Computer (in Chinese), Taipei: Shang-Xun Culture.

OTHER SOURCES DigiTimes, a daily trade paper specializing in the IT (including the PC) industry, published in Taipei, Taiwan (in Chinese).

10. Global quality standards and technological upgrading in the Brazilian auto-components industry Ruy Quadros* 1.


This chapter examines the implications of global quality standards for the upgrading prospects of local producers in developing countries. It draws on an investigation into the diffusion of global quality standards in the Brazilian automobile industry. The focus is on the local suppliers of auto parts but their experience can only be understood in the context of the value chain which is coordinated by their customers, the transnational component manufacturers and assemblers. The governance of the auto chain, therefore, commands attention in this chapter. The requirement for suppliers to meet certain standards and to have the standards verified by means of certification has become widespread in the auto industry in recent years. The process started with the International Organization for Standardization 9000 (ISO 9000) global quality standard, but more recently sectoral and nationally based standards have also become more prevalent, as will be discussed below. How might these standards affect linkages within the value chain, and what would this mean for technical cooperation within the chain and the upgrading possibilities for Brazilian suppliers in the automotive sector? Globalization brings increasing competition to developing country producers, both in export markets and also in domestic markets as tariffs are reduced. In the face of such competition, the most sustainable response for firms in developing countries is to upgrade. Upgrading means to produce more efficiently (process upgrading), to add value to manufactures and exports (product upgrading) and to move into more skilled activities such as product design, marketing and brand development (functional upgrading) (Humphrey and Schmitz, 2000). Functional upgrading can also be related to the acquisition of process-related technological competencies, 265


Local enterprises in the global economy

which are needed in activities (process engineering) aiming at sustaining process upgrading. This chapter asks how quality standards certification might affect inter-firm linkages in value chains and the prospects for upgrading for developing country producers. While there is little dispute in the literature on the fact that developing country firms benefit from supplying global chains in terms of process improvement (Schmitz and Knorringa, 2000) and in terms of product upgrading (Gereffi, 1999; Dolan and Humphrey, 2003), functional upgrading is altogether more problematic. What possibilities are there for such upgrading for medium-sized, locally-owned suppliers in the Brazilian automotive industry; and what contribution might certification make to this process, both directly and in terms of the relationships it fosters within the automotive value chain? In order to investigate these questions, this chapter looks into the type of chain governance that is predominant in the auto chain, and the type of technical ties that global buyers establish with Brazilian suppliers. More specifically, the study focuses on the following issues: 1.



There is convergence in the literature towards the view that process standards – of which quality standards are one example – contribute to lowering transaction costs (David, 1995; and see Nadvi and Wältring, Chapter 3, this volume). The findings presented in this chapter confirm this point. Compliance with quality standards provides an entry ticket to supplying the automobile chain and, as a screening tool, it reduces governance costs. However, such economies seem to be quite limited in terms of total transaction costs in the chain. It is shown that quality standards have not eliminated the need for costly individual auditing of suppliers (manufacturing process auditing, for instance). Moreover, there has been a multiplication of quality standards in the chain, which have been introduced by assemblers of different nationalities and have increased the burden of certification on Brazilian suppliers. The effectiveness of quality standards is dependent on their credibility (Bido, 1999; and see Nadvi and Wältring, Chapter 3, this volume). This study provides strong evidence that problems with credibility in quality standards certification undermine the achievement of its primary purpose. Lack of trust is one of the reasons for the adoption of multiple standards in the chain, imposed by different buyers. The prevailing type of relations between buyers and suppliers is important for explaining the decline in trust in quality standards certification. There are two possible implications of global quality standards for chain governance. Certification might be viewed by buyers as an indication of supplier capability to assume further responsibilities, in which case the former might be inclined to help the latter enhance com-

Global quality standards and technological upgrading


petence. Buyers might be prepared to support the technological capability upgrading of suppliers in the expectation that improvements in supplier performance would translate into benefit for the buyer. Conversely, it can be argued that certification might lead to more arm’s-length relationships, as it substitutes for direct monitoring of supplier performance. The results of this research indicate that neither has happened. Local suppliers’ compliance with quality standards has not affected the (weak) technical ties between customers and suppliers at all. Not only have assemblers and first tier suppliers not made any substantial effort to help local suppliers meet quality standards requirements (apart from insisting that suppliers meet these standards), but such adoption has not implied the development of technical collaboration in areas such as product or process engineering. The empirical evidence presented in this chapter comes from an investigation carried out in the components industry in the São Paulo metropolitan area in the year 2000. It comprised the collection of data and documents at Sindicato Nacional da Indústria de Componentes para Veículos Automotores (Sindipeças),1 the use of an exploratory postal questionnaire sent to suppliers and visits to ten selected local suppliers and interviews with their quality managers. The chapter is organized in eight sections, including this introduction. The following section outlines the characteristics of the Brazilian automotive value chain. Section 3 is an appraisal of the general diffusion of quality standards in the Brazilian auto components sector, emphasizing the problems of multiple standards and the credibility of standards. The 4th section introduces the issue of chain governance, focusing on the business insertion of local suppliers in the chain. Section 5 analyses the features of the supply contract, from customer approaching supplier, to the nature of contracts and to the forms of control exerted by customers. Suppliers’ trajectories to obtain quality standards certification are discussed in section 6 as a means of opening up a discussion on technical ties between buyers and suppliers. It reveals how suppliers evaluate and classify the importance of the various sources of information and technical assistance for adoption, that is customers, consultants, employers associations, and so on. Section 7 addresses suppliers’ evaluation of technical collaboration from customers and the possible implications of standards for such collaboration. The main conclusions are drawn together in section 8.


Local enterprises in the global economy

2. BACKGROUND: THE BRAZILIAN NATIONAL COMPONENTS INDUSTRY Brazilian-owned component producers are one part of a value chain which is dominated by transnational assemblers and increasingly transnational first tier suppliers (Humphrey 2000). The suppliers are organized in three tiers. The first tier comprises suppliers of high value-added components whose sales are almost exclusively to assemblers. These are large, global firms. Some of the largest first tier suppliers have moved to the supply of complete, assembled modules, instead of single components (modular assembly). The second tier comprises producers of light2 components, most of whom are Brazilian-owned and of medium or small size. These are the firms studied in this chapter. They supply both to first tier suppliers and directly to assemblers. Finally, suppliers to the second tier constitute the third layer. Most of them supply either commodities like fasteners or specialized metal processing services. These have not been considered in this study. The Brazilian automotive industry experienced substantial changes in the 1990s. The most significant for this chapter are summarized as follows: ●

De-nationalization. A massive wave of acquisitions transferred ownership of most Brazilian producers of major components (first tier) to foreign businesses. Liberalization and import penetration. As a result of trade liberalization and regional integration, imports increased substantially. From a share of 9.7 per cent in the local market for car components in 1990, imported parts reached the share of 27.4 per cent in 1998, introducing strong competitive pressures on local producers. New assemblers from new countries and new regions. The Mercosur area3 and Brazil in particular experienced a period of increasing investment in new plants and in the renewal of old plants. The face of the industry changed in terms of competition due to the entry of French and Japanese assemblers. The geography of the car industry was transformed as well, with the emergence of three new regional automobile poles.4 Firm restructuring and reduction in supplier numbers. Investment growth accelerated intra-firm and inter-firm restructuring in the Brazilian component industry. The pace of diffusion of new manufacturing technologies and of new organizational forms increased. While further transferring assembly operations and services to suppliers, assemblers reduced the number of total suppliers they dealt with. Firms moved closer to global product standards and production practices (Quadros et al., 2000).

Global quality standards and technological upgrading


A remarkable aspect of firm restructuring was the large and rapid diffusion of quality management practices. In this context, assemblers and first tier suppliers introduced the requirement of certification in quality assurance standards for the second tier suppliers whose experience is examined in this chapter.

3. THE DIFFUSION OF QUALITY STANDARDS IN THE BRAZILIAN AUTO COMPONENTS INDUSTRY ISO quality assurance norms were introduced in Brazil by the ABNT (Associação Brasileira de Normas Técnicas) in the early 1990s, and registered as NBR (Norma Técnica Brasileira) ISO 9000 at INMETRO (Instituto Nacional de Metrologia, Normalização e Qualidade Industrial), which is the Brazilian national standards body and accounts for the accreditation of ISO certification institutions. The diffusion of quality standards in Brazil was rapid. The annual growth rate of accredited ISO 9000 certificates was above 100 per cent between 1990 and 1998 (Melo, 1999, 44). According to ABNT, the number of valid ISO certificates rose to 6500 in December 2000 from only 18 in 1990. Such a large increase in quality certification is a good indicator of the broader and widespread diffusion of quality awareness in the business sector in Brazil and, to a certain extent, also in the public sector. The dissemination of quality awareness was reflected in the high interest displayed by the media. Diffusion started amongst large firms and moved towards their suppliers. Large firms soon realized that their efforts to attain systemic quality improvement would only be successful if their suppliers also adopted quality management practices (Melo, 1999). The adoption of ISO quality assurance standards was perceived by buyers as a simple, less costly means of disseminating a standard concept of quality and quality practices amongst firms in the Brazilian market. According to Souza (1995), as ISO 9000 certification increasingly became one of the criteria adopted by large firms in supplier selection, the number of certified local, medium sized firms boomed. As quality standards certification gradually became an ‘entry ticket’ for supplying to motor vehicle assemblers and to their major suppliers, in both domestic and export markets, Brazilian auto component producers sought to obtain accredited quality certification. Information collected at Sindipeças in August 2000 confirms that the rate of certification among members was high. A total of 340 firms (70 per cent of members) held an ISO quality certificate (either 9001 or 9002).5 In addition to this, member


Local enterprises in the global economy

Table 10.1 Sindipeças members: quality standards certification by type of certificate (August 2000) Type of certification ISO 9001 ISO 9002 QS 9000 VDA EAQF AVSQ

Number of certified firms 139 201 220 39 23 13

Source: Sindipeças

firms have also been certified for a range of other quality standards, which will be discussed below. Almost half (46 per cent) of all members held a Quality Standard 9000 (QS 9000) certificate (Table 10.1). Still, a considerable number of non-certified firms remained. Interviews revealed that most non-certified firms were small and medium-sized enterprises (SMEs) catering for the low end of the replacement market, in which competition is driven by price rather than quality. The Sindipeças data bank does not indicate the degree of overlap of these certificates, that is, the percentage of firms which accumulate two or more certificates of different types. However, we found further evidence that firms tend to accumulate two or even three certificates, in a pattern of certification which goes from the initial widespread adoption of ISO 9000 certification to an overlapping diffusion of QS 9000 certification (originating from the USA) and, more recently, to an increasing number of other nationally originated standards certification (Verband der Automobilindustrie (VDA) – Germany, Evaluation d’Aptitude Qualité Fournisseurs (EAQF) – France and ANFIA Valutazione Sistemi Qualitá (AVSQ) – Italy). The evidence is provided by responses to an e-mail questionnaire sent to approximately 180 Sindipeças members. Responses, which were received between June and August 2000 (31 responses, or 17 per cent), revealed that the dominant tendency was for firms to accumulate two or three different certificates (Table 10.2).6 The multiplication of certification raises issues relating to standards proliferation and also the credibility of standards. The evidence above reveals that component firms are facing a proliferation of quality standards and different certification requirements from assemblers, which increases the cost of qualifying for supply. This tendency is partly related to decreasing credibility of the most disseminated certificates issued by accredited insti-

Global quality standards and technological upgrading


Table 10.2 Status of firms in terms of quality standards certification (August 2000) Firms holding both the ISO 9000 certificate and the QS 9000 certificate – of which holding or preparing for a third certificate (mostly the VDA certificate) Firms holding only the ISO certificate and preparing for a second certificate (mostly QS 9000) Firms holding only one certificate (either ISO 9000 or QS 9000) and not preparing for a second Firms holding no certificate but preparing for the ISO 9000 or the QS 9000 certificate

20 9



6 3 2

Source: e-mail questionnaire replies

tutions: this undermines the economic benefits (the lowering of transaction costs) of quality assurance certification and raises the issue of its regulation and control. 3.1

Multiple Standards

Until 1996/1997, the possession of an ISO 9001/9002 certificate was a truly competitive differential in the Brazilian auto components supply market, not only because the number of certified suppliers was small, but also because the ISO 9000 certificates were a respected novelty. Since then, and as competition has brought tougher quality requirements, the global standard (ISO) has tended to be superseded (in terms of credibility) in the automobile industry by nationally originated, sector-based standards, like QS 9000 and VDA 6.1. Initially, the earlier diffusion of QS 9000 in Brazil (as compared to VDA and the others) and its acceptance by European assemblers operating in the country seemed to place it as the new ‘global’ standard. Later, the introduction of new car platforms, often associated with new plants and, sometimes, with new entrants (like Renault, VW-Audi and the Daimler Chrysler car division) entailed demand for other ‘national’ quality standards certificates related to the auto industry. The emergence of specific quality standards for the automobile industry is an international phenomenon, which has developed simultaneously within the context of the leading national automobile industries (QS in the US, VDA, EAQF and AVSQ in Europe), and its growing importance in Brazil reflects this more general tendency. These standards are based on the ISO 9001 standard, but have incorporated a number of new requirements,


Local enterprises in the global economy

such as more restrictive norms for the planning of product quality in design and for the development of suppliers. In spite of overlap, the standards differ substantially from each other. The unification of quality standards for the auto industry is a goal which has been pursued by the International Automotive Task Force, which comprises the major North American, German and French assemblers, with support from the International Organization for Standardization. By 1999, the task force managed to approve its proposal for a unified standard – the ISO/TR 16949 – Automotive Quality System, which is currently in the process of defining norms for accreditation. The clash between the global/generic standard (ISO 9000) and national/sector-based standards (QS, VDA, and so on) and the simultaneous imposition of the latter on local suppliers is most clearly perceived in a very internationalized context such as in the Brazilian car industry, where all problems and costs derived from conflicting quality requirements are borne by local suppliers. The need of suppliers to get certified twice or three times in order to attend to distinct customer requirements substantially increases the cost of certification and narrows the range of customers served by each certificate. In an increasingly global business, such as the motor vehicle industry, the proliferation of national and competing quality standards undermines the very concept of a global, generic standard (as the ISO 9000 series was meant to be) and its potential benefit of reducing transaction costs. 3.2

The Credibility of Quality Standards Undermined

In the Brazilian case, an additional and very significant reason for assemblers’ increasing demand for suppliers to obtain nationally originated certificates relates to the decreasing reliability of ISO 9000 and QS 9000 certificates issued by Brazilian Accredited Certification Institutions (ACIs). ISO and QS certificates can be issued by ACIs which have been accredited by national accreditation bodies (in the Brazilian case, INMETRO). Doubts have been raised about the reliability of ACIs accredited locally. VDA 6.1 and EAQF certificates can only be issued by international ACIs. The lack of reliability in the process of certification was brought to light in a way that illustrates governance problems in relation to quality standards. The ANFAVEA Committee for Quality Issues7 recorded mounting complaints by its members regarding the occurrence of faulty components within modules supplied to the assembly line. This particularly affected new entrants, but the concern was also shared by the long established assemblers, as product recalls had increased. When assemblers traced the flawed parts, they were led to suppliers which were QS certified. Being quality

Global quality standards and technological upgrading


assurance certified, these suppliers were supposed to keep the level of quality consistency which allowed them to supply directly to the assembly line. The ANFAVEA Committee concluded that Brazilian ACIs had not been doing their job properly. A common perception raised by assemblers and suppliers’ representatives in their respective association quality groups was that, in the search for expanding their share in a very disputed market, some ACIs reduced costs to the point at which audit quality suffered. They argued that some ACIs had been hiring only young auditors with little experience in the automotive industry (which was in conflict with QS 9000 requirements). Moreover, auditors faced excessive and demanding workloads for inadequate pay. As a result, some certification processes were carried out too hastily, with the consequent risk to the quality of audits (particularly regarding processes concerning the maintenance of quality consistency and continuous improvement). Quality professionals from the sample firms we visited agreed that some ACIs were not doing an adequate auditing job, even though they also stressed that the squeezing of margins imposed by assemblers led to reducing investment which, in turn, harmed quality (see section 4). Interviewees were particularly critical of the procedures of Brazilian ACIs. In some firms, quality managers reported that they sought a first class certificate, implying that certificates awarded by some local ACIs were second class, a situation similar to that found by Nadvi in Pakistan (Nadvi, Chapter 11, this volume). The ANFAVEA Committee decided to intervene in the process of auditing and certification of quality standards. At the time of fieldwork, this intervention was in the phase of planning and negotiation with Sindipeças. The core of ANFAVEA’s action was to deliver a document, approved with minor changes by Sindipeças, containing ‘Guidelines for the Qualification of Auditors and Preparation of External Auditing for Quality Standards Certification’. The document stated that auditors must fulfil certain requirements in order to obtain an auditor’s credential, namely:8 ● ●

● ●

five years’ work experience in the automotive industry; specific experience of the industrial process to be audited (machining, stamping, welding, and so on); good knowledge of continuous improvement tools; and, other requirements related to experience of quality management.

The problems regarding the trustworthiness of quality standards auditing and certification in the Brazilian automobile industry and the resulting intervention by manufacturers associations (and consequent costs) raise questions about the limits of global quality standards. As proposed by


Local enterprises in the global economy

Wältring (2000), international quality standards like the ISO 9000 series are coordinative norms which are decided at a public/private level. Private industry associations from the Organisation for Economic Co-operation and Development (OECD) countries and government institutions (mainly from non-OECD countries) participate in the decision-making process and in the process of accreditation of ACIs. Firms, private associations and government institutions promote the definition and adoption of quality standards (and respective certification) considering that this will eventually reduce transaction costs. The sine qua non for this is that the entire process of auditing and certification is transparent and reliable. However, the question arises of ‘who certifies the certifier’, that is, the problems of reliability related to the control of ACIs. Certifiers are accredited by national accreditation bodies, such as INMETRO in Brazil and the British Standards Institution in the UK. On the one hand, in the absence of real sanction power, it does not seem possible that the ISO can exert an effective control over national accreditation bodies. On the other hand, it can be hypothesized that state controlled accreditation bodies (as in the Brazilian case) may find it difficult to enforce accreditation norms in the absence of legal regulation. The result can be the unreliability of quality standards certification, exemplified by ISO 9000 certification in the Brazilian automotive industry. The consequence is an economic loss for the industry, as the investment in quality certification will have little return, and transaction costs are not reduced by the certification. This situation becomes more complex in the case of nationally originated, sector-specific standards – like QS 9000, AVSQ, EAQF and VDA 6.1. Most of these are standards issued by national associations of automobile manufacturers in the leading industrialized countries, which also take care of the accreditation of certification institutions. In this case, government/public accreditation bodies apparently have little control over third party auditing and certification. As the recent developments in Brazil suggest, in given institutional contexts only the direct intervention of the lead firms seems to guarantee that ACIs will respect the requirements established for accreditation. Still, the effectiveness and economic benefit of such an intervention remains to be seen, as the industry apparently cannot rely on (and benefit from) the self-control of the existing network involving local ACIs and the national accreditation body. In the Brazilian case, an additional problem, which will be further explored below, is the cost penalty imposed on suppliers by the multiplicity of global standards (which are in fact regional, if not national) and their respective certificates. This suggests that, when faced with problems of credibility, the expected lowering of transaction costs yielded by quality standards may vanish. A final and broader conclusion could be drawn from this episode, in con-

Global quality standards and technological upgrading


nection with the feasibility of truly global and general (cross-sector) quality standards. The movement from ISO norms towards standards which are specific to the industry in itself suggests that the idea of global quality standards which should apply across any type of industry and production process has suffered a setback. The emphasis put by ANFAVEA on the need for Brazilian ACIs hiring auditors with industry- and process-specific experience reinforces this point.

4. THE EXPERIENCE OF SAMPLE FIRMS IN THE AUTOMOTIVE CHAIN Having examined the problems faced in the adoption of quality standards in the Brazilian automobile chain, the following sections of the chapter address its implications for chain governance. Interviews and visits carried out in a sample of ten Brazilian auto component firms are the empirical source for this undertaking.9 The purpose of this section is to introduce the sample firms, with an emphasis on their position in the chain and on their business ties with buyers. Knowledge of the latter allows an assessment of the degree of transactional dependence between buyers and suppliers. If a supplier has a high level of transactional dependence on one or a few buyers, this may encourage the buyer(s) to establish closer technical ties with the supplier because transactional dependence lowers the risk of supplier opportunistic behaviour. The main features of the suppliers investigated and the reasons for their inclusion in the sample are as follows. First, as the focus of research is on the ties between global buyers and local suppliers, the sample comprises only components manufacturers which are 100 per cent owned by Brazilians. Second, the majority of sample firms are medium sized, in order to keep in line with the size characteristic of the second tier in the auto chain. Third, it is important that sample firms should vary in terms of products and processes as much as possible so that the technical diversity of the industry is taken into account. Finally, the sample was restricted to firms located in the Metropolitan São Paulo area, the main area for automotive component production. A description of each of the sample firms is presented in Appendix 10.1. The firms in the sample make a variety of components, from very small parts like washers and fasteners, to larger parts such as clutch disks and dashboards. These products are referred to in the chapter as light components, meaning that they correspond to low value-added components. All but two of the firms (Plastics A and B) produce metallic parts, but a variety of technologies and processes are used: stamping, welding, forging,


Local enterprises in the global economy

machining and sinterizing. The firms in the sample have been given names that reflect the parts they produce. The sample firms present some similarities regarding the characterization of their major customers (Table 10.3). First, most of them combine, among their five main customers, a majority of auto assemblers producing in Brazil and a minority of components producers.10 The bulk of the latter are in fact suppliers of complete modules, such as Dana, M Marelli, Delphi, Bosch and so on. Second, eight out of ten sample firms are VW suppliers, whereas six of them are also GM suppliers. The strong presence of these two assemblers amongst the sample firms’ customers reflects their importance in the auto industry in the state of São Paulo. The limited importance of Fiat, the second-largest car producer in the country, among the sample firms’ customers is possibly because it has successfully encouraged suppliers to locate near its plants in the state of Minas Gerais. It is important to look at the distribution of sample suppliers’ sales between their major customers. A large proportion of sales going to one or two customers implies great business dependence of the supplier on this (these) customer(s). This is one critical aspect for understanding governance in the chain. High dependence on few customers diminishes the bargaining power of a supplier in contract negotiation and its prospects for customer diversification. On the other hand, customers may perceive such supplier dependence on them as lessening the likelihood of opportunistic behaviour on the part of suppliers. Thus, in theory, they would be more inclined to incur the costs of closer technical collaboration11 with suppliers with a higher transactional dependence on them. Conversely, customers would be less inclined to invest in technical collaboration with suppliers which display a low degree of transactional dependency. As can be seen in Table 10.3, few firms have a dominant customer. Only three firms sell 30 per cent or more of their output to a single client. What are the implications of this pattern for chain governance? Taking into account sales dependence on the major customer, only Stamped A could be classified as being dependent. The majority of sample firms benefit from considerable room for manoeuvre in the market, which suggests that, in terms of transactional relations with buyers, their insertion in the automobile value chain is not of the type which would encourage buyers to invest in closer technical collaboration with them. However, transactional dependence per se does not necessarily lead to close technical collaboration. Even in the case of Stamped A, technical ties between the supplier and Fiat, the major customer, were weak, in spite of high dependence. In fact, it seems that the most important factor explaining close technical collaboration is the need for asset-specific investment by the supplier which is needed for the production of customized products. This point will be discussed in the following sections.


15% VW 47% Fiat 25% Scania 17% M. Marelli 30% Honda 8% Ford 16% VW 8% TRW JAMM 36% Fiat

1 9% Ford 18% GM 10% VW 9% VW 18% VW 7% GM 10% Scania 6% WEG Gradinotec 12% GM

2 6% Cummins 17% VW 10% GM 8% TI 10% M. Benz 7% Johnson Con 6% Behr Tenneco Jorliba Diesel 12% VW


Main customers 4 3% Scania 5% ATH Dana 7% Cofap 5% ZF Eaton VW 5% GM Bosch Viação Bristol 11% Ford

Sample firms, five largest customers and respective share of sales (1999/00)

Source: Interviews

Hose Clamps Stamped A Stamped B Washers Forge Plastics A Plastics B Sinterized Clutch Disks Fasteners


Table 10.3

2% GM 5% Hoesch N/A 5% M. Benz Bosch Toyota 3% Delphi DHB ViaçãoB.Branca 7% M. Marelli



Local enterprises in the global economy

Low transactional dependence in the Brazilian automotive chain is in part rooted in the historical development of the industry. The long period of relative stability and protection, from the middle 1950s until the early 1990s, led to the development of a considerable base of local suppliers, with which assemblers established rather arm’s-length, market relationships. There were a number of reasons for this type of relationship. First, changes in vehicle models were rare, and designs were stable and well known. Auto parts producers could compete openly in the market for supply orders, and assemblers seemed to be happy with this type of chain organization. Second, the simplicity of light components enables the assemblers to provide detailed component drawings to suppliers without suffering a high design cost penalty. Third, such simplicity allows suppliers to produce light components on generic equipment (and shift to new product drawings) with no need to invest in costly specific assets.12 This can explain why it can be economically feasible for suppliers to live with buyers’ continued practice of multiple sourcing even in a market situation characterized by more frequent product change. This finding is similar to the situation described by Helper (1993) in regard to the US auto industry before the 1980s. The low value of investment in specific assets for the supply of light components helps understand why both buyers and suppliers conduct their relations with one eye on the market.



Do standards and certification allow reduced audit and pre-qualification costs? This chapter addresses this issue by examining how the contract between supplier and buyer develops, from the initial contact. How do customers approach suppliers in the automobile chain? What is the basis for the assessment of a new supplier? How does QS certification influence this choice? When asked to rank factors which have accounted for the development of customers’ trust in suppliers, the sample firms considered two situations – the initial approach and the continuation of the supply relationship. Reputation was ranked by most suppliers as the most important factor in attracting new customers and contracts. Reputation is developed through long-term dependability, yielding both new customers and new contracts with current customers. For instance, Stamped A claimed that its reputation with GM was crucial for French entrants approaching the firm. In other words, reputation is more important than QS 9000 or any other certification for securing new businesses. Certification was ranked second, on the grounds that it is a necessary condition for inclusion in the supply chain. ‘No certification, no business’. However, the effectiveness of QS 9000 in reducing transaction costs is undermined in three ways:

Global quality standards and technological upgrading ●


It is not universal. For the new French and Japanese assemblers entering the Brazilian market, certification was said to be less relevant. According to the General Manager of Stamped A, ‘Peugeot, Renault and Toyota do not pay attention to QS 9000. They have their own questionnaire and methods for assessing potential suppliers. The Japanese said that for them QS 9000 means nothing. It seems that QS 9000 is more significant for the North Americans’. It does not replace audit and pre-qualification. Suppliers in general revealed that new customers only make a decision on pre-qualifying the supplier after they have carried out an initial audit. Suppliers call this a ‘process audit’, whereby the main focus is on process design capabilities, production equipment and the organization of production. It does not cover all of the issues of concern to customers. Many suppliers stressed the fact that QS certification only reduces monitoring of quality procedures. Monitoring of process capabilities is carried out at the beginning of the relationship and some customers do it periodically with all suppliers. Certification is not a substitute for monitoring, audit and pre-qualification of suppliers.

Similarly, certification does not obviate the need for re-assessment of suppliers. According to many of the sample firms, the continuity of the supply relationship depends on the maintenance of dependability. This is periodically assessed through performance monitoring. Customers monitor performance by giving scores to suppliers’ deliveries, which refer to indicators such as PPM (faulty parts per million), punctuality of delivery, flexibility in meeting change in supply orders, and so on. Some customers (VW and Mercedes Benz) also carry out annual process audits. The continuity of customers’ auditing and monitoring of suppliers’ manufacturing processes and product/delivery performance is a critical finding for the purpose of this chapter. It shows that, although it is true that quality standards contribute to reducing transaction costs, this benefit is limited. The current practices in the automotive chain involve other (costly) forms of supplier assessment/selection which are not affected by quality standards. The actual contracting procedures follow similar stages in most supply relationships reported by sample firms: 1. 2.

first, component manufacturers are qualified to become potential suppliers of the customer (when the customer is new to the supplier); this is achieved through audits, questionnaires and so on, as shown above; once qualified, suppliers receive an invitation to bid for the supply of a




Local enterprises in the global economy

given component (providing a quote); in doing so, the customer sends the component specification and drawings to the supplier; the nature of the supplier’s proposal depends on the complexity of the component. The common point in all proposals is a price quote. In the case of very light parts, like washers and some fasteners, the supplier is also expected to produce samples of the component to be assessed by the customer. In the case of more complex parts, this is not necessary, but the customer may request a technical specification of the process; if the proposal is approved, a formal contract may follow. The contract comprises a technical specification of the component, the frequency of delivery, specifications regarding transportation of the component and the property of the tools which will be developed by the supplier. Plastics A, Stamped A and Forge reported that the contract also indicates that supply orders will be placed weekly, following a rough estimate for three months and a more precise one for one month, which are revised monthly. However, there is no warranty in the contract against losses suffered by the supplier if a major reduction in the volume of orders occurs.

Given the nature of the components supplied by the sample firms – light components – the manufacturing processes involved do not require special purpose machinery. Thus, suppliers do not have to face large investment risks. Yet, they do incur expenses related to designing and manufacturing the tools required for the component to be supplied, which are specific for each supply contract. Certain risk-sharing mechanisms exist in relation to the property of tools and the financing of tool development. Most customers only pay for the tools when the actual supplying of the component starts. Thus, the supplier has to face the initial cost related to tool development. In the case of more important components (in terms of value added), the customer pays for and owns the tools developed for the process, which reflects the greater level of investment and risk for the supplier. General Motors has a distinct policy: once the supplier proposal is approved, it pays in advance a share of the estimated tool development cost. This cost is increased in the case of components for platforms which have been designed in Brazil (Fiat Palio and GM Celta), because the building and testing of component prototypes is required by the customer in these cases. To conclude, the central finding of this section is emphasized. The continuity of supplier assessment by customers indicates that certification does not tell buyers about things such as dependability, punctuality, flexibility in meeting large orders, and so on. Only experience and a close relationship can provide the buyer with this type of information. Therefore, standards and certification are not a substitute for experience and a close relationship.


Global quality standards and technological upgrading

6. QUALITY STANDARDS CERTIFICATION AND CHAIN GOVERNANCE Given strong pressure from customers to certify, how did the component manufacturers acquire the know-how and systems required to obtain certification? To what extent did their customers assist in this process? This section goes further in examining governance in the Brazilian automobile value chain, particularly the issue of whether the adoption of quality standards contributes to the strengthening of ties between buyers and suppliers. The section addresses the sample firms’ trajectories in the process of attaining quality standards certification, focusing on the types of technical support they have obtained and used in order to prepare for certification audits. It might be expected that lead firms in the chain (assemblers or module suppliers) would be inclined to invest in the short term to assist suppliers in adopting quality standards requirements, on the grounds that the resultant reduction in monitoring/inspection of suppliers would reduce transaction costs. Results from the research show that this is not the case in Brazil. The situation of the sample firms in terms of certification in quality standards follows closely the pattern for multiple standards in the Brazilian auto components industry described in section 3. Eight out of ten firms in the sample had achieved both ISO 9000 and QS 9000 certification (Table 10.4). Moreover, with the exception of Hose Clamps, QS 9000 certification occurred from two to three years after ISO 9000 certification. Fasteners and Table 10.4

Sample firms, status of quality standards certificationa ISO 9001

Hose Clamps Stamped A Stamped B Washers Forge Plastics A Plastics B Sinterized Clutch Disks Fasteners

ISO 9002

1999 1997 1995 1997 1995 1997 1995 1996

QS 9000

VDA 6.1


1999 1998 1997 2000 1998 1998 1999


2000 1995



Note: ª The occurrence of each type of certification is marked by the inclusion of the year of certification in the respective box. Source: Interviews


Local enterprises in the global economy

Plastics B had already moved towards a third certification; the former had obtained the Italian certification (AVSQ), whereas the latter was preparing the German, VDA 6.1 certification. Although in the phase of ISO 9000 certification half of the firms opted for the simpler certification (9002), the shift to QS 9000 has meant that firms with such a certificate had to adopt quality assurance procedures for process design (APQP). Knowledge about how to implement standards could come from three different sources: (i) from important customers (that is from within the value chain); (ii) from business associations and the public sector; and, (iii) from market mechanisms (consultants and ACIs). The interviews revealed important aspects about knowledge sources that help to explain the issue of technical ties between customers and suppliers. When asked about the importance of sources of technical support and information for preparing for the auditing process necessary for quality standards certification, the sample firms pointed to the following hierarchy, which is also illustrated in Table 10.5: 1. 2. 3. 4. 5. 6. 7.

Consultants and consulting firms Accredited certification institutions Customers Sindipeças Other institutions Public research institutions (universities and research institutes) and government institutions Other employers’ associations

The classification of sources in Table 10.5 clearly reflects the relative importance of the actors involved in the process of diffusion and certification of quality assurance standards in the Brazilian auto components industry. For most firms in the sample (except for Plastics B and Clutch Disks), consultants had been by far the main source of information and technical support (Table 10.5), followed in second place by ACIs. These two actors were often related. In firms such as Forge, Hose Clamps, Clutch Disks and Washers, the consultants were themselves either auditors of ACIs or had been indicated by ACIs, while in Stamped B, training had been provided by an ACI. The fact that ACIs had been involved in the preparation for certification may have contributed to the problems of legitimacy and reliability of certificates (see section 4) resulting from conflicting interests. Nadvi (Chapter 11, this volume) also highlights this problem in the Pakistani surgical instrument industry. The role of customers, and particularly of assemblers, is ranked third by the firms in the sample (Table 10.5). However, the importance given to

283 18

3 4 4 3 4




4 3

4 2



5 4

1 3 3









Public research



Government institutions



Other associations

The rank is based on a scoring by order of importance, in which the highest score is 5. Thus, the higher the score, the more important the

Source: Interviews



Total Score

Notes: source.

5 4 5

5 5 5 5 5 5

Hose Clamps Stamped A Stamped B Washers Forge Plastics A Plastics B Sinterized Clutch Disks Fasteners

Consultants and consulting firms

Table 10.5 Rank of sources of technical support and information in preparation for certificationa


Local enterprises in the global economy

assemblers was justified by two reasons other than the direct technical collaboration with suppliers in order to prepare them for certification. The most frequent mention of assemblers, in relation to this question, concerned their pressure on suppliers to achieve certification. According to Washers’ Quality Manager, certification in quality assurance today (and this means a specific certification for the automobile industry, such as QS 9000) is an obligation for any supplier wishing to continue in the chain, rather than a competitive differential, as it used to be five years ago. ‘It is either have it and keep it, or give up supplying to the assemblers’. The second point, raised by two firms (Stamped A and B), relates to collaboration schemes that assemblers put into practice at the beginning of the quality wave in the late 1980s, but gave up later. The firms referred specifically to programmes developed by GM (the Ótimo programme) and by Fiat (the Sol Programme), which promoted technical collaboration with and training of suppliers in order to disseminate concepts and tools of quality management, such as Statistical Process Control and problem-solving techniques. The assemblers’ collaborative approach in the past seemed important to help small firms’ restructuring, as suggested by Addis for the case of the Ford–VW joint venture at the end of the 1980s: A novel programme by the Brazilian Support Service for Small Firms, SEBRAE, shows that with mentoring, small firms dramatically improve their performance. Curiously, although one of the first experiments with this program was in the motor vehicle industry itself, it has not gone very far in this sector. In response to competition from imports, the now defunct Autolatina [a short-lived fusion of the assets of Ford and VW in Brazil], in conjunction with a Brazilian subsidiary of a Big Eight consulting firm, Andersen Consulting, the state-level SEBRAE in São Paulo (SEBRAE/SP), and a group of small suppliers devised a programme that cut consulting costs while teaching small firms how to restructure. Most of these small family firms have become ISO 9000 certified. Since much of the consulting was done in groups, costs were lower. Simultaneously, the group dynamic encouraged firms to undertake painful restructuring and also created an often informal, but constant process of benchmarking among the small firms where each encouraged and helped the other. . . . Regardless of the exact format, the SEBRAE experiences show that when a large firm accompanies the progress of its suppliers and when they learn collectively, restructuring, productivity improvements, and the like are quite successful (Addis, 1999: 223).

This is in sharp contrast to the situation found in this study. This sort of cooperative scheme was discontinued by the middle of the 1990s. As was emphasized by the General Manager of Stamped A, at a time of increasing liberalization of auto components imports and the arrival of new foreign suppliers, the assemblers had gained the power to threaten local suppliers with a shift to imported products or to new suppliers, where they

Global quality standards and technological upgrading


were not able to meet quality, price and delivery requirements. From this moment on, the assemblers transferred the entire technical and financial burden of quality assurance to the suppliers. This was emphasized by many of the respondents. A clear message from the interviews with suppliers was that the pressure to obtain certification was the assemblers’ only contribution to their effort to obtain quality standards certification. This finding is in line with the fact that most suppliers stated that the technical support provided by the assemblers was quite limited – an issue to be discussed further in the next section. The only exception to this picture is the case of Forge, which indicated that assemblers had collaborated in the process of certification by allowing this firm to learn from the assemblers’ own experience in training and preparation for the adoption of quality standards. Sindipeças follows fourth in importance as a source of technical support for quality assurance in the view of the sample firms (Table 10.5). Its main contribution, according to the interviewees, related to dissemination of information, particularly through specific training courses (such as Statistical Process Control – SPC, Failure Mode and Effect Analysis – FMEA) and through the exchange of information between firms participating in its Quality and Productivity Group and firms participating in the Sindipeças’ benchmarking initiative (Benchmark Magazine). It is also important to mention that the support of Sindipeças was particularly praised by the firms which had decided to drop their consulting firm and become self-reliant at some point in the process of preparing for QS 9000 or ISO 9000 certification. This happened in the cases of Stamped A, Sinterised and Washers. For Plastics B, Sindipeças was the crucial source (Table 10.5), because this firm had decided not to rely on consultants from the outset. If consulting firms were the main source of support, how effective were they in enabling the component firms to meet certification requirements? How did the sample firms assess their relationship with consultants and the quality of the latter’s work? The simple ranking of sources alone cannot say much about this and it is necessary to put forward other findings related to the trajectories of the sample firms – from the initial decision to become involved with quality assurance standards through to the final achievement of certification and (in some cases) re-certification. In order to simplify this point, the trajectories towards ISO and QS certification will be bundled together. The sample firms had generally undertaken some other form of information search in response to the assemblers’ demand for supplier certification. At this stage, Sindipeças’ (low cost) courses and meetings were important sources for initiating the process. However, in most cases (eight out of ten),


Local enterprises in the global economy

contracting a consulting firm was the first step that involved substantial spending during the process of preparing the firm for the certification audit. In six of these eight cases, the firms reported major problems with the first, and sometimes a second, consulting firm. This problem was particularly acute when the sample firms sought the services of leading consulting firms with established reputations. For instance, Stamped A started its preparation by hiring a well known and expensive consulting firm, only to learn at some point in the process that the consulting package was such that it would create long-term dependence on the consultant for the maintenance of certification. Another example was technical consulting for quality assurance at Washers. This firm complained that, in spite of the high cost of the consulting work, the servicing firm used to send students rather than experienced professionals to implement the required procedures and documentation. In both cases, the firms decided to drop the consultants and to continue the process alone. Sinterized followed a similar pattern. In some cases, it was only at the second (Hose Clamps and Clutch Disks) or even the third attempt (Plastics A) at hiring a consulting firm that this solution worked. In other words, these firms spent a substantial amount of money before they found an acceptable and effective way to prepare for the certification audit. Such a waste of money and time is an indicator of poor coordination in the chain, a point which will be better explored below. In fact, the best assessed consulting experiences, which occurred in the cases of Forge and the third attempt by Plastics A, were provided by individual consultants instead of firms. These are the only cases in which the consulting service had continued up until the date of interviews, albeit on the basis of fortnightly or monthly visits. Concerns with the quality of support from consultants also led to decisions to internalize certification capabilities. In the cases of Washers and Stamped A, the decision to abandon the consulting scheme was combined with a decision to hire an experienced professional to take responsibility for the whole process of preparation. This also happened at Hose Clamps, following the second and successful consulting experience which led to the firm’s first certification. As has been stated before, Plastics B provides a completely different example, as this firm decided from the beginning to rely only on its internal capability and quickly hired the necessary staff for this. It seems that most firms have increased (in quantity and quality) their own professional staff involved with quality procedures and quality assurance. This is the most important and evident effect of quality standards in terms of functional upgrading. The more general conclusion to be drawn in this section is that leaving the provision of information and technical assistance for quality certifica-

Global quality standards and technological upgrading


tion to the market may be problematic. As the role of customers, the public sector and the relevant business associations (ANFAVEA and Sindipeças) in helping local suppliers to adopt quality standards has been secondary, suppliers turned to services provided by the private sector. Yet, the evidence in this study suggests the market for these services has serious information asymmetries (firms cannot distinguish between good and bad consultants) and the certification market is subject to perverse incentives by which poor certification has cost advantages and is not easily weeded out.

7. STANDARDS AND THE (WEAK) TECHNICAL COLLABORATION BETWEEN BUYERS AND SUPPLIERS This section goes further into the issue of technical collaboration between Brazilian auto components suppliers and their main customers. Would certification contribute to tightening or loosening the technical ties between suppliers and buyers? It was hypothesized at the beginning of this chapter that standards might have contrasting impacts on governance in value chains. On the one hand, large customers in the auto industry might take certification as a sign of supplier capability, and as a result be inclined to tighten technical linkages and help them to learn how to upgrade their manufacturing processes, which would eventually benefit the entire chain performance. On the other hand, to the extent that certification acts as a substitute for direct monitoring of supplier performance, it might lead to more arm’slength relationships. The results of the research substantiate neither of these hypotheses. With the exception of one firm (Forge), suppliers do not acknowledge their major customers as technical or financial partners, even less so in regard to the development of process engineering capabilities. This is mainly because the nature of the manufacturing process does not require it. At the same time, the previous sections have shown that quality assurance certification does not lead to more arm’s-length relationships because certification alone is not enough to ensure customer confidence in supplier capabilities, and direct monitoring of supplier capabilities and performance has been maintained. Before presenting findings on how assemblers assess technical collaboration from suppliers, it is useful to discuss the potential scope for such collaboration. The major technological capability requirements on the suppliers considered in this study are related to process engineering rather than product design. This is because assemblers, module suppliers and suppliers of high value-added components control product design in the Brazilian automotive industry. Co-design between assemblers and suppliers, a current


Local enterprises in the global economy

tendency in the automobile chain, does not encompass the participation of local suppliers of light parts. The reason for this has not been investigated in this research, but it could be hypothesized that de-centralization of the design of light components is not cost effective. Considering the various interfaces between small components which go into a larger module/component, the cost of coordination of co-design could easily surpass the cost of centralized design.13 The technological activities of the local firms of the sample were restricted to process design, which occurs at the beginning of the supply relationship for a new component. Process design here means preponderantly the design of tools for the manufacturing of the component, like moulds for castings, dies, and so on. In order to meet these requirements, all sample firms maintained a technical or engineering unit with a few process engineers. The average number of process engineers in the sample firms was 3.2. The average share of design engineers in total employment in the sample firms was 1.2 per cent. The comparable figures for two large suppliers of higher value-added components, one locally-owned and one sold to a transnational company, was above 3 per cent. Therefore, the share of process engineers in the sample firms seemed to reflect their very limited technological activity. This helps explain why suppliers in the sample presented quite limited capabilities for planning/designing changes in their manufacturing processes, in terms of either planning for the incorporation of new equipment and new process techniques, or adopting more comprehensive organizational changes in the plant (like TQC or Kaizen). In this respect, the medium firms in this study merely reflect the general situation of limited technological capability of the average Brazilian SMEs, which is a well documented characteristic (Souza, 1995). The lack of internalization of process planning competencies usually leads SMEs to rely only on equipment suppliers’ advice for purchasing of equipment, and consultant firms for the implementation of new managerial practices. The eventual result is often the adoption of solutions that are expensive and do not meet their needs (Quadros, 1988). Thus, in theory, in an industry which is presented as a leader in terms of the diffusion of new manufacturing technology, there would be a considerable scope for collaboration between large buyers and small and medium suppliers, in terms of learning about planning for changes in production processes. As seen before, suppliers of the sample view customers primarily as a source of pressure and tension rather than of technical assistance for quality certification. When asked to indicate the areas in which they had received collaboration from major customers, the firms in the sample indicated that, in general, technical collaboration from customers was limited


Global quality standards and technological upgrading

Table 10.6

Sample firms: types of collaboration received from customers Quality management

Hose Clamps Stamped A Stamped B Washers Forge Plastics A Plastics B Sinterized Clutch Disks Fasteners

Product/process design


Customer lends skilled staff









Source: Interviews

and, more importantly, it was rarely systematic. In the cases of Washers, Plastics A, Sinterized and Clutch Disks, the interviewees declared that they had no technical collaborations with their customers (Table 10.6). Two of these firms mentioned programmes previously run by GM and Fiat to help suppliers with quality issues, but these had been discontinued some years previously. Hose Clamps claimed clear technical collaboration in product design, but it was referring to the fact that customers provided detailed drawings to the supplier. Stamped B indicated that one of the major customers provided samples of a component sent by its parent company in Sweden, as this customer was going to localize the supply of components. Stamped B, Plastics B and Fasteners mentioned collaboration from customers in quality management, but they were alluding to the possibility of accessing detailed information on QS 9000 standards. Stamped A referred to having had an experience with Fiat’s programme of technical assistance in cost reduction (the ‘RC-5’ programme). Through this programme, Fiat used to send professionals to look for cost reduction opportunities at the supplier’s plant. However, this supplier eventually refused to continue in the programme, since it regarded it as having an audit and control purpose and because the customer appropriated all the benefits.14 Forge was again the only exception to this picture. It praised the assemblers’ collaborative programmes, mentioning as the most significant example the technical assistance provided by Honda for the implementation of Kaizen groups. Moreover, the Forge interviewee suggested that, based on the manufacturing experience of a given component, the supplier


Local enterprises in the global economy

provided important inputs to the customer in order to improve the component design. There are some reasons for the distinct situation of this firm. First, the example mentioned may imply that Honda has a distinct policy for supplier development, as compared to other assemblers. In another work, focusing on the policies of Japanese firms for supplier development in Brazil, the distinctiveness of Honda (in this case its motorcycle factory) was also detected (Quadros et al., 1996). Second, in comparison with the other sample firms, Forge seemed to have accumulated greater technological capabilities. This is due to the fact that Forge had had a contract for technology transfer with a leading German forging company for quite some time. At this point, it is useful to sum up our research findings, in the light of the initial hypotheses and questions. The evidence is strong that local suppliers’ compliance with quality standards has not affected at all the (weak) technical ties between customers and suppliers in the Brazilian automobile value chain. Not only have assemblers and module suppliers not made any substantial effort to help local suppliers adopting quality standards requirements (apart from pressuring them to do so), but such adoption has not led to the development of technical collaboration in areas such as process design and planning for technological change. At the same time, medium-sized, locally-owned suppliers have not benefited from the decreasing transactions costs. While coordination of activities in the value chain is predominantly based on arms’s length market relations, customers continue to monitor and supervise suppliers directly, in spite of widespread certification requirements. This imposes costs on both sides. The investigation has shown that transactional dependence is very limited and both buyers and suppliers have considerable room for seeking new partners in the market. Moreover, the occurrence of asset-specific investment is also limited. The simplicity of the components supplied by the sample firms do not require special purpose machinery. Thus, suppliers do not have to face large investment risks. The fact that production of light components is mostly based on general purpose machinery means that buyers can, in theory, easily shift to new suppliers (either local or foreign). This helps explain why there is no motivation for buyers to engage in technical collaboration with suppliers in order to promote the development of process-related capabilities in the latter.



What conclusions could be drawn of the findings of this research in the face of the main issues raised at the beginning of this chapter? Does the diffu-

Global quality standards and technological upgrading


sion of quality standards strengthen or weaken technical ties between buyers and suppliers in the automotive chain? The portion of the global automobile value chain that is located in Brazil is made up of two groups of firms which are not only very distinct animals, but animals living in worlds apart. By distinct animals I refer to the considerable asymmetries between assemblers and large components suppliers, on the one hand, and local suppliers on the other. This is not a surprising result, as the literature in Brazil and elsewhere has frequently pointed to asymmetries between large companies and SMEs. Perhaps the most relevant finding is that relations between large, powerful customers and their local suppliers are restricted in scope, unstable and distant. Private coordination within the value chain is not strong and effective enough to allow the sustainable progress of the entire chain, including the local suppliers. How has the diffusion of global quality standards affected this situation? It has certainly contributed to local suppliers’ upgrading of quality processes. But it has done little either to promote the suppliers’ acquisition of new functions (process engineering capabilities) or to develop closer ties between the two groups. Apart from putting pressure on them, customers have not really invested to help local suppliers meet quality standards requirements. Technical support for quality upgrading has come predominantly from private sector consultants. Moreover, the adoption of QS has not prompted technical collaboration in process design and planning. Governance in the Brazilian automobile value chain is based predominantly on a market, arm’s length type of relationship, characterized by significant power asymmetries. While there has been tighter monitoring of supplier quality systems in recent years and increasing emphasis of supplier performance, the absence of technical collaboration and the customer’s recurrent threat of supplier replacement indicate the continuity of arm’s length market relationships. The prospect for closer, long-term relations is not firmly established yet. Further, continuing de-nationalization of the components industry only adds to the undermining of the SMEs. Thus, coordination exerted by lead firms in the value chain does exist, but it is weak because they still rely on arm’s length market relations. Weak coordination has an economic cost which is a burden for the entire chain, therefore undermining competitiveness. Market failures in the market for consultancy services and certification itself, together with standards proliferation, impose costs on firms. They have to work through many different consultants in the search for certification, to the cost of meeting simultaneously the different (and sometimes conflicting) requirements of different systems of quality standards. As long as well established assemblers and new entrants continue to introduce their own (national) quality standards, the cost of achieving and keeping certification is increasing for


Local enterprises in the global economy

Brazilian local suppliers. As the coordinator of the Sindipeças Quality Group has summarized, ‘one has to invest in a second, virtual plant just to produce and keep all the documents required by the various auditing processes’. To put it bluntly, the lead firms in the value chain have little understanding of and sensitivity to the upgrading concerns of local firms. The implications for the inefficiency in the chain are only perceived when the quality of the final product is affected. Then customers blame the low quality of certification auditing. However, local suppliers emphasized that quality is suffering primarily because the extreme pressure to reduce costs in the short term is leading local suppliers to degrade their skills and accumulated capabilities. Although the entire chain would profit from the strengthening of long-term, technical ties between global lead firms and local suppliers, it is not happening. It is not a priority for the lead firms in the chain.



2. 3. 4. 5.




I would like to express my gratitude to Sindipeças, and particularly to Miguel de Almeida, for their support in accessing firms and understanding their certification problems. I am very grateful to Rubia Quintão for her dedicated assistance in fieldwork and in data tabulation. I also appreciate the precious suggestions from Afonso Fleury and Márcia Leite. John Humphrey provided many useful suggestions for revision. Sindipeças/Abipeças is the Brazilian national association of auto component producers. The majority of associates are producers of metal, electro-electronic and plastic components, in order of importance. However, it also includes producers of components based on other materials, such as glass and fabrics. An (unknown) share of its associates also belongs to other sectoral associations. Associates comprise both firms controlled by local capital and firms controlled by foreign capital. In this chapter light components refer to simple, low value-added components. The Mercosur free trade area comprises Argentina, Brazil, Paraguay and Uruguay. In the states of Paraná, Rio Grande do Sul and Rio de Janeiro. The ISO quality standards series comprise three types of certifiable standards. ISO 9001 and 9002 apply to manufacturing industries. The most complete standard, ISO 9001, comprises quality procedures related to both product design and manufacturing, whereas ISO 9002 is meant for firms which carry out only manufacturing activities (design supplied by third party). The ISO 9003 standard applies to firms which exclusively carry out quality tests and inspection (Bido, 1999). ISO 9000 certification in this chapter refers to both 9001 and 9002. Our sample of responses was certainly biased, as firms which had not acted at all to obtain any type of quality assurance certificate were less motivated and less likely to respond to a questionnaire about certification. However, another source (Bido, 1999) confirms the rate of certified firms and the pattern of accumulation of certificates. ANFAVEA is the national association of motor vehicle manufacturers. Major associates are the large assemblers of cars and trucks in Brazil: VW, FIAT, General Motors, Ford, Renault, Mercedes Benz, Scania and Volvo. Given the small scale of their operations in Brazil, Japanese assemblers are members but exert little influence in the association. Its committee for quality issues is composed of representatives from the assemblers who are specialists in the area of quality of supply. According to an interviewee, some of these requirements are in fact already considered in the QS 9000 series.

Global quality standards and technological upgrading 9. 10. 11. 12. 13. 14.


The interviews were completed in the second-half of 2000 and were based on a questionnaire focusing on the main issues of the IDS/INEF research programme. Clutch Disks is the exception, as it caters exclusively for the domestic replacement market. By closer technical collaboration between buyers and suppliers I mean the former’s support to enhance the latter’s technological capability upgrading. This point will be resumed in the following sections, in a more detailed account of the typical supply contract found amongst sample firms. I am grateful to John Humphrey for calling my attention to this point. This is a common complaint of automotive suppliers in Brazil.

REFERENCES Addis, C. (1999), Taking the Wheel, Pennsylvania: The Pennsylvania State University Press. Bido, D. S. (1999), ‘Implementação de Sistemas da Qualidade para a Busca de Certificação em Pequenas e Médias Empresas do ramo Automotivo’, doctoral thesis, Department of Economics, Administration and Accountancy, University of São Paulo, São Paulo. David, P. (1995), ‘Standardisation policies for network technologies: the flux between freedom and order revisited’, in R. Hawkins, R. Mansell and J. Skea (eds), Standards, Innovation and Competitiveness: the politics and economics of standards in natural and technical environments, Aldershot, UK and Brookfield, US: Edward Elgar. Dolan, C. and J. Humphrey (2003), ‘Changing governance patterns in the trade in fresh vegetables between Africa and the United Kingdom’, Environment and Planning A. forthcoming. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Helper, S. (1993), ‘An exit-voice analysis of supplier relations: the case of the US automobile industry’, in G. Grabher (ed.), The Embedded Firm, London: Routledge. Humphrey, J. (2000), ‘Assembler–supplier relations in the auto industry: globalisation and national development’, Competition & Change, 4, 245–71. Humphrey, J. and H. Schmitz (2000), ‘Governance and upgrading: linking industrial cluster and global value chain research’, IDS Working Paper 120, Brighton: Institute of Development Studies. Melo, A. A. (1999), ‘O processo de disusão dos sistemas da qualidade ISO 9000: estudos de caso em pequenas e médias empresas de campinas – SP’, Masters’ dissertation, Department of Science and Technology Policy, IG/UNICAMP, University of Campinas, Campinas. Nadvi, K. and F. Wältring (2002), ‘Making sense of global standards’, INEF Report 58, Duisburg: Institut für Entwicklung und Frieden, Gerhard-Mercator University. Quadros, R. (1988), ‘The impact of flexible automation on scale and scope in the Brazilian engineering industry’, in L. Alcorta (ed.), Flexible Automation in Developing Countries: The Impact on Scale and Scope and the Implications for Location of Production, London: Routledge/United Nations University. Quadros, R., A. Fleury and M. T. Fleury (1996), ‘O papel das empresas subsidiárias


Local enterprises in the global economy

Japonesas no Processo de Desenvolvimento Tecnológico da Indústria Brasileira’, Revista de Administração da Universidade de São Paulo, 31 (3): 19–27, July/Sept, São Paulo: University of São Paulo. Quadros, R., S. Queiroz, J. Humphrey, F. Consoni, I. Costa and R. Fonseca (2000), ‘Globalização e Capacitação Tecnológica na Cadeia Automotiva do Mercosul’, Research Report for IPEA, GEMPI – Grupo de Estudos de Empresas e Inovação, DPCT/IG, Campinas: University of Campinas. Schmitz, H. and P. Knorringa (2000), ‘Learning from global buyers’, Journal of Development Studies, 37 (2), 177–205. Souza, M.C.A.F. (1995), Pequenas e Médias Empresas na Reestruturação Industrial, Brasília: Editora SEBRAE. Wältring, F. (2000), ‘Tendencies of international standardisation processes and their governance patterns: first comments and questions’, note for the IDS-INEF workshop on ‘Interaction of global and local governance: implications for industrial upgrading’, 12–15 January, Duisburg.

Global quality standards and technological upgrading


APPENDIX 10.1 CHARACTERISTICS OF SAMPLE FIRMS Hose Clamps: Founded in 1967 and 100 per cent controlled by Brazilian owners, like all firms in this appendix. Produces many different sizes and forms of clamps. Domestic replacement market accounts for 70 per cent of sales; the balance refers to OEM sales to local assemblers. Five main customers of OEM are VW (15 per cent), Ford, Cummins, Scania and GM. Medium size: 170 employees and US$ 4 million sales in 1999. QAS certificates: ISO 9001 and QS 9000. Stamped A: Founded in 1968. Manufactures stamped parts and welded assemblies. 100 per cent sales are OEM to local assemblers. Five main buyers are Fiat (47 per cent), GM, VW, ATH/Dana and Hoesch 5 per cent. A leading firm in medium and light stamped parts. Medium size: 226 employees and US$ 7.5 million in 1999. Certificates: ISO 9002 and QS 9000. Stamped B: Founded in 1979. Main products are light and medium stamped parts/sets. Sales of OEM to local assemblers are 95 per cent; 5 per cent are exports of OEM to foreign assemblers. Main customers are: Scania (25 per cent), Xerox (25 per cent), VW, GM and Cofap/M. Marelli. Medium size: 188 employees and US$ 6.4 million in 1999. Certificates: ISO 9002 and QS 9000. Washers: Founded in 1959. Produces washers, bushes and other light stamped components. Five major customers are M. Marelli (17 per cent), VW, TI, ZF and Mercedes-Benz. OEM sales to local customers are 65 per cent; replacement market sales are 30 per cent and OEM exports, 5 per cent. Small size: 70 employees and US$ 2.7 million in 1999. Certificates: ISO 9002 and QS 9000. Forge: Founded in 1932. Forges and machines light parts for engines, gear boxes, suspension and vehicle bodies. Five main clients: Honda Motorcycles (30 per cent), VW, Mercedes-Benz, Eaton and Bosch. Leading firm in forged light parts. Sales are 100 per cent OEM, 90 per cent to local customers and the balance are exports. Medium to large size: 482 employees and US$ 10 million sales in 1999. Certificate: ISO 9002. Plastics A: Founded in 1967. Manufactures plastic air hoses, tanks and other dashboard components. Major customers are Ford (8 per cent), Johnson Controls and VW. Sales are 100 per cent OEM to local manufacturers. Large size: 550 employees and US$ 15 million in 1999. Certificates ISO 9001 and QS 9000. Plastics B: Founded in 1937. Main products are injected plastic components (dashboards and bumpers) and flexible cables. Principal client is


Local enterprises in the global economy

not in the automobile business. Five major customers in auto industry: VW (16 per cent), Scania, Behr, GM and Delphi. Sales of auto components are 100 per cent OEM for local buyers. Large size: 652 employees and US$14 million in 1999. Certificates: ISO 9001 and QS 9000. Sinterized: Founded in 1967. Manufactures various types of metal components based in a technology called sinterization (fusion of compressed metallic powder). Two major clients are from home appliances industry (white goods). Main clients are: TRW (8 per cent), Monroe/Tenneco and Bosch. Sales are 95 per cent OEM to local producers and 5 per cent in the replacement market. Medium size: 148 employees and US$ 6.4 million sales in 1999. Certificates: ISO 9001 and QS 9000. Clutch Disks: Founded in 1974. Produces plates and clutch disks for trucks and buses. In contrast to the other companies of the sample, this firm only caters for the domestic replacement market. Main customers are auto components wholesale dealers and large urban bus services. Small to medium size: 118 employees and US$ 5.4 million in 1999. Certificate: ISO 9002. Fasteners: Founded in 1965. Manufactures fasteners of different forms and sizes for various applications. Lead firm in fasteners. Sales are 100 per cent OEM to local producers (93 per cent to assemblers). Main clients include Fiat (36 per cent), VW, GM and Ford. Medium size: 288 employees and US$ 8.4 million sales in 1999. Certificates: ISO 9001, QS 9000 and AVSQ. Source: Interviews

11. The effect of global standards on local producers: a Pakistani case study Khalid Nadvi* 1.


Since the late 1980s, there has been a rapid rise in international standards addressing a wide range of concerns, from quality assurance, health and safety, labour conditions, to environmental, social and ethical norms.1 These standards reflect a growing interest with how production and delivery takes place as much as with what product or service is provided. The importance of such standards is enhanced by their incorporation in the new ‘rules’ governing international trade. Compliance is often necessary for market entry, and critical to international competitiveness for both developed and developing country producers. Global standards also point to an emerging and critical area of global governance, where various international public and private actors have come together to shape standards and promote their implementation (see Chapter 3, this volume). Yet, little is known of how global standards effect local producers, especially in developing countries, and in particular how they influence inter-firm relations, both within clusters and between developing country producers and global buyers that are linked through global value chains.2 This chapter addresses this lacuna by asking whether global quality assurance (QA) standards, and in particular the ISO 9000 standard, enhance or harm the position of local developing country enterprises in the global economy. There are two diametrically opposed scenarios: ●

Compliance with global standards can potentially offer developing country producers two advantages: (i) a means of raising their productivity and (ii) a clear signal to global buyers about their competence. As a result, they become more successful in global markets. Alternatively, global quality standards may serve to exclude developing country producers from global markets. Implementation can be costly, time-consuming and particularly difficult for developing 297


Local enterprises in the global economy

country firms, especially small and medium enterprises (SMEs) (Schuurman, 1998). Such firms often lack the managerial and financial resources to implement the standard and the technical skills to carry out the required testing and adopt the procedures that ensure quality assurance at each stage of production. This chapter reports on the processes and consequences of ISO 9000 implementation in Sialkot in Pakistan, where firms in an export-oriented industrial cluster have faced regulatory pressures to comply with global QA standards. In reviewing the impact on the local cluster, it asks the following questions: ● Do global standards exclude local producers from world markets? ● Was it difficult for firms in Sialkot to adopt ISO 9000 standards? ● Did implementation of ISO 9000 standards improve competitiveness? ● What impact did the process of implementing ISO 9000 standards have on linkages between firms in the cluster and on linkages with global buyers? ● How was the knowledge necessary to implement the standards mobilized within the cluster? ● What contributions did inter-firm relationships and cluster institutions make to the process of adoption? ● What impact does the process of implementing ISO 9000 standards have on cluster institutions and collective efficiency? The Sialkot surgical instrument cluster provides an ideal case for addressing these questions. Although a successful export cluster, its major challenge in the past decade has been to adopt international QA standards. Failure to comply implies a loss of markets, falling employment and a concomitant decline of the cluster. Compliance appears to offer the possibility for enhancing international competitiveness and encouraging upgrading. The following section describes the ISO 9000 standards and the Sialkot cluster. Section 3 reports on the implementation of quality assurance standards in Sialkot while section 4 reviews the role of compliance in enhancing international competitiveness. Section 5 analyses the impact of implementation of the standard on global value chain ties that link local producers to global markets. Section 6 considers the standard’s impact on ties within the cluster. Section 7 discusses the relationship between the standard and local institutions. Section 8 contrasts the findings that emerge on quality assurance standards with those from other types of standards, notably standards on child labour. Finally, section 9 concludes this chapter.

The effect of global standards on local producers




Historically, standards have played an important role in trade by reducing transaction costs, providing certainty and generating economic efficiency (David, 1995). Traditionally, standards focused on product characteristics. This included physical attributes, size, durability, technical specifications and conformance with health and safety issues. Thanks to product standards, firms traded across distant markets by working to common specifications. Since the 1980s, there has been a shift in emphasis from product to process standards. Process standards focus not on product characteristics but the processes by which goods are produced and delivered. These can address a range of issues from environmental management and quality assurance to social and ethical concerns. Process standards relating to quality are designed to transmit information to lead firms in value chains about their suppliers, their capabilities and their production processes, thus reducing supplier search costs. Independently certified standards can also reduce monitoring costs for lead firms. Compliance with such standards may, for example, result in lead firms no longer having to undertake their own (often costly) quality inspections of supplies. Furthermore, compliance can ensure harmonization between distinct and distant global suppliers, thereby reducing asset-specific transaction costs. ISO 9000 are international, generic (as opposed to firm- or sectorspecific) and independently monitored quality assurance standards. With over 400 000 certificates issued worldwide by 2000, it has become the most widely recognized global process standard to date (http://www.iso.ch).3 Its popularity is based on three factors. First, the widespread perception that implementation can improve efficiency, raise productivity and enhance profitability. This comes through the adoption of better quality management practices that result in lower reject levels, improved traceability that helps identify weaknesses in production, better process planning and more quality-focused training. Second, the global recognition of the ISO 9000 certificate provides a potential marketing advantage, facilitating entry to quality-conscious markets and giving certified firms a competitive edge. Third, while the standard is voluntary, compliance is often necessary in certain markets, either for competitive reasons or as a consequence of regulatory demands. Therefore, irrespective of the benefits to either buyer or supplier, it becomes a market requirement. Introduced by the International Standards Organization (ISO) in 1987, ISO 9000 is a standard driven by industry and its need for improved reliability and assurance within supply chains. This is in line with flexible, justin-time and total quality control management practices. The standard has also evolved with time. The original version, based on an earlier British


Local enterprises in the global economy

standard of 1979, was revised in 1994, and a new version (ISO 9000-2000) was released in 2000.4 For the purposes of this study, I focus on the 1994 version.5 This addresses quality assurance procedures at each stage of production and service delivery. Implementation requires an appropriate and well-documented quality control system supported by effective management. Compliance is gauged through independent certification. Certification is usually valid for three years, with periodic audits (usually twice yearly) undertaken to ensure that certified enterprises continue to comply. Numerous firms now service the rapidly growing ISO 9000 certification industry. These include national standards institutions (such as the British Standards Institution) as well as leading international firms such as SGS (Switzerland), Bureau Veritas (BVQI-UK), AOQC-Moody (US), AFAQ (France), and TUV (Germany). There is also a growing market providing advisory services to assist firms in implementing the standard. Sanctions on non-compliance are largely market-based, with suppliers risking the loss of orders from buyers. In some sectors sanctions are imposed by market regulators. In addition, auditors who issue the certificate also, in effect, put their reputation on the line. In the context of global value chains, the demand to meet ISO 9000 standards can be driven by actors either internal or external to the chain. Assistance on compliance, and sanctions for failure to meet the standard, can also come from within the chain or outside of it. What are the consequences of the implementation of such standards for the nature of ties within the chain? These questions are addressed by reference to an analysis of industry in Sialkot in Pakistan, known for its labour-intensive, export-oriented, small firm clusters. These produce surgical instruments, sports goods and leather garments (see Weiss, 1991; Nadvi, 1999a; 1999b). Some clusters have a significant share of the global market in their particular niches. They are integrated into global value chains, with local firms trading directly with leading global buyers. In recent years, Sialkot’s clusters have faced pressures to conform to international standards on quality assurance and on child labour. Standards are an emotive subject in Sialkot. Sialkot provides an ideal setting to study how local clusters linked with global value chains face global standards, and how standards impact on the organization of the cluster and chain ties. In particular, this study focuses on the surgical instrument (SI) cluster in Sialkot. It consists of 300 producers, over 1000 specialized ancillary units, subcontractors and suppliers. The cluster employs an estimated 30000 people and has annual exports of over US$ 125 million (Government of Pakistan 1998–99).6 Producers are differentiated by size with 30 large firms employing over 100 people and a further 80–100 medium-sized units

The effect of global standards on local producers


employing between 20 to 100 workers. There are also various local institutions. These include the local chamber of commerce and industry (SCCI); a trade body, the Surgical Instruments Manufacturers Association (SIMA); the local government-run technology institute, the Metal Industries Development Centre (MIDC); and the federal government’s Export Promotion Bureau. Thus, the cluster generates significant agglomeration economies as well as forms of local joint action (Nadvi, 1999a). Production is almost wholly for export, with the US accounting for 60 per cent and the EU around 33 per cent of total export volume. In summary, Sialkot is an example of a small firm cluster where collective efficiency gains have enhanced international competitiveness. Its continued success hinges, however, on the ability to comply with international QA standards.

3. MEETING GLOBAL QUALITY ASSURANCE STANDARDS During the 1990s, compliance with international quality standards became a core requirement in the international medical instruments sector. In the early 1990s, the US Food and Drug Administration (FDA) stipulated that all medical appliances must meet the US’ Good Manufacturing Practices (GMP) standards of quality assurance. Similarly, the European Commission, through Directive 93/42/EEC, stated that all medical devices sold in the EU had to comply with international quality assurance norms by 1998. This meant either ISO 9000 or EN46000 certification (a specialized European standard for quality assurance in the manufacture of medical instruments), and the CE mark (the European safety marking). These externally imposed regulatory demands posed significant challenges to Sialkot’s producers. In 1994, the FDA embargoed Pakistani surgical instruments for not meeting US standards. Faced with the loss of its largest market, the cluster responded rapidly.7 The manufacturer’s association lobbied both US regulators to ease the sanctions, and the Pakistani government to obtain financial support to implement the standard. The association recruited a US consultant to assist local firms in improving practices and in adopting quality management procedures in line with US requirements. By late 1997, over one third of the cluster was certified by the FDA as meeting its standards, and exports to the US exceeded the pre-crisis levels. As I have argued elsewhere (Nadvi, 1999b), the key to the response was joint action through the trade body, underlining the importance of local governance in confronting global challenges. In the late 1990s, similar pressures around ISO 9000 emerged for the EU market. After a slow start, ISO 9000 certification levels accelerated. Of the


Local enterprises in the global economy

300 surgical instrument producers in the cluster, 170 firms were ISO 9000 certified by September 2000. So significant has been the response that Sialkot’s surgical instrument sector is second only to the textiles industry in terms of total numbers of ISO 9000 certified firms in the country.8 The ability to meet international QA standards is also borne out by field evidence. Of the 42 surgical instrument firms interviewed in 2000, 29 had ISO 9000 certificates, 34 complied with the US GMP standards, and 25 had the European CE-mark. Furthermore, of the 13 firms that were not ISO 9000 certified, 11 were in the process of obtaining certification, with five awaiting the final certification audit. Moreover, certification was not restricted to large firms as Table 11.1 illustrates. Although a higher proportion of large firms were certified, half the sample of small firms also had the coveted ISO 9000 certificate. In addition, 80 per cent of the subset of SMEs were GMP certified and 61 per cent had the CE-mark. Finally, all certificates were issued by recognized international auditors. Table 11.1 Quality assurance certification by sampled surgical instruments firms. Percentage of sampled firms certified according to standards and firm size Small



All firms

50% 63% 44% 16%

73%.0 100.0% 80%.0 15.0%

91% 82% 46% 11%

69% 81% 59% 42%

ISO 9000* GMP CE-Mark N

Note: *Kendall’s ␶ significant at 10% only for size distribution for ISO 9000 certification. Source: Author survey

There were, nevertheless, differences by firm size. Large firms were pioneers in compliance, with most (70 per cent of the large firms interviewed) being certified in 1997 or 1998. In contrast, only 26 per cent of the subset of certified SMEs sampled were compliant by 1998. Compliance costs were positively related to firm size. Large firms, with more internal production processes and larger production volumes, needed more detailed quality documentation and more extensive auditing than SMEs. Nevertheless, as shown in Table 11.2, costs declined over time as competition amongst quality management consultants and certification agents rose within the cluster. Field evidence suggests that in late 1997, when the first large firms were certified, payments to consultants and auditors were as high as US$ 20000 per firm. By late 1999, average costs of certification for a large firm

The effect of global standards on local producers


Table 11.2 Average costs, and time taken, for ISO 9000 certification. By firm size and period of certification Costs in Rs. 000; time taken in months

Average costs Standard Deviation (N) Average costs for firms certified in 1997–98 (N) Average costs for firms certified in 1999–2000 (N) Average time taken for certification Standard Deviation (N)




All firms

240.0 109.2 (8) 300.0 (2) 220.0 (6) 7.50 3.21 (8)

412.0 242.9 (8) 633.3 (3) 317.1 (7) 11.36 5.24 (11)

815.3 381.5 (9) 955.4 (7) 325.0 (2) 8.50 4.30 (10)

495.5 355.8 (27) 765.7 (12) 279.3 (15) 9.31 4.60 (29)

Source: Author survey

had fallen to one third of this. Compliance had also come about rapidly. The average length of time taken to implement the standard was just over nine months, although the majority of small (63 per cent) and large (60 per cent) certified firms were able to do so within six months. These findings question conventional wisdom that ISO 9000 certification is a lengthy, costly and prohibitive process for small producers. Although large firms took the lead in adopting ISO 9000, significant numbers of medium and small-sized firms are now certified. Such levels of compliance are impressive and raise questions as to how it was achieved. The state played some part in promoting compliance, through technical support and a subsidy (of approximately US$ 4000 per certified firm).9 Field data, however, raise doubts about the importance of this incentive. Of the 29 certified firms in the sample, only seven had obtained the subsidy, and none considered it key to their decision to seek certification. The role of local institutions was also limited. All certified firms stated that the local technology centre gave no assistance or technical advice on QA standards. Only two of the 29 certified firms considered that the sector’s trade association played an important part in ISO 9000 compliance. Similarly, only three respondents had such views on the local chamber of commerce. These findings contrast sharply with the crisis of the mid-1990s when the trade association took on a catalytic role in bringing technical know-how on US quality assurance standards to the local cluster. External buyers also played a limited part in the dissemination of knowledge and the provision of technical assistance on ISO 9000. Of the 29


Local enterprises in the global economy

Table 11.3 Sources of know-how on ISO 9000 quality assurance certification. Percentage of ISO 9000 certified (or in final audit) SI firms by firm size Knowledge source




All firms

SIMA SCCI Buyers Consultants GMP experience (N)

27% 36% 0% 100% 55% 11%

46% 31% 0% 85% 62% 13%

20% 10% 10% 70% 70% 10%

32% 27% 3% 85% 62% 34%

Source: Author survey

certified firms, 25 stated that their buyers gave no significant assistance on ISO 9000 compliance. Instead, as Table 11.3 shows, irrespective of firm size, consultants were the key source of information and technical know-how on the standard.10 Previous experience in meeting US (GMP) quality assurance standards, which mirror ISO 9000 requirements on documentation and traceability, was also key. To conclude, the evidence suggests that the challenge to meet global standards has been met. Contrary to conventional wisdom, ISO 9000 standards have been widely and rapidly adopted by firms of all sizes. Costs have not served as a prohibiting factor. In fact, compliance costs have declined, reflecting the competitive provisioning of consultancy and certification services. The know-how for implementing quality management systems appears to have come largely from local specialist service providers and not from within the value chain (global buyers). There is a link between ISO 9000 and the earlier standards imposed by US regulators. In effect, knowledge that flowed to the cluster on the GMP standards has extended to the similar ISO 9000 standard. Given that certification is widespread, two questions arise. First, does compliance enhance performance and quality? Second, what impact does certification have on governance within value chain ties? I turn to these questions in the subsequent sections.



Given that compliance with ISO 9000 is extensive in Sialkot, has this improved the cluster’s performance? The premise is that meeting global quality assurance standards would result in better access to leading world

The effect of global standards on local producers


Table 11.4 Changes in internal practices and firm efficiency. By firm size for sampled ISO 9000 certified firms Efficiency measures




All firms

Improved stock control





Improved traceability





Improved process planning (N)

13% 8

27% 11

60% 10

35% 29

Average reject rates now S.D N

2.67% 0.52% 6

2.70% 0.82% 10

3.71% 0.95% 7

3.00% 0.90% 23

Average reject rates 3 years ago S.D N

7.50% 3.56% 6

8.50% 3.27% 10

6.57% 2.15% 7

7.65% 3.04% 23

Source: Author survey

markets, enhanced firm performance through greater efficiency in process controls, and a better reputation for the cluster as a whole. Table 11.4 illustrates that all 29 certified firms attest that implementing the quality management procedures associated with ISO 9000 improved their internal production systems. This was apparent through better control of stocks, better documentation of product orders, and better management of production batches. Some, mainly large, firms also stated that this improved their ability to plan production. Compliance was also seen as reducing reject rates, with greater emphasis on stage-by-stage inspection and documentation that helped identify systemic stage-wise problems. Thus, of the 34 firms that were certified or in the final stages of certification, 74 per cent reported that their reject rates had declined and 52 per cent that their rework rates had also fallen. On average, reject rates for certified firms fell by more than half the level found before ISO 9000 was adopted. This was consistent for firms of all sizes. The decline in reject rates provides a useful proxy on how performance has improved with adoption of the standards. With no accurate firm-level financial performance data, it is difficult to assess the impact of compliance on overall efficiency. Nevertheless, an evaluation of qualitative data on competitiveness and market access, alongside perceptions of changes in overall sales, gives some insights on changes in performance. The evidence, however, is inconclusive. ISO 9000 certification did, according to the majority of medium and large sized firms, lead to an improvement in competitiveness (see Table 11.5), but only 17 per cent of


Local enterprises in the global economy

Table 11.5 Changes in competitiveness and market access. By firm size for sampled ISO 9000 certified firms ISO 9000 certification leads to:




All firms

Increasing competitiveness Opening up of new markets N

38% 0% 8

64% 9% 11

70% 40% 10

59% 17% 29

Source: Author survey

firms thought that compliance had helped open new markets. Small firms, in particular, reported very few competitive gains from certification. This underlines the view that in this specific industry, ISO 9000 certification is not a marketing advantage but a prerequisite for market entry. This was further borne out in interviews with buyers in Europe and at the international medical instruments trade fair in Dusseldorf in November 2000. No buyer saw ISO 9000 compliance as the sole, or even leading, basis for selecting suppliers. Nevertheless, certification does matter. If we review changes in performance, in terms of sales and employment levels, it is possible to assess the diverse experience of certified and non-certified producers. This evidence, however, must be treated cautiously. First, our samples are small. Second, it is difficult to suggest causality between certification and improved performance given that firm size is likely to be a key factor in determining overall performance improvements. Nevertheless, these figures provide an indication of the possible impact of compliance. What is apparent is that certified firms are more likely to experience improving sales, employment and product quality than non-certified producers. Yet, focusing on small producers, who were equally distributed between certified and non-certified units, it is clear that while a higher proportion of certified units reported rising sales, employment and product quality levels, the difference is neither large nor statistically significant. These results add to the ambiguity reported earlier in terms of improved market access. Performance, at least in terms of rising sales, appears to have improved (see Table 11.6). But, there is little to suggest that compliance with ISO 9000 is crucial in bringing this about. Moreover, there is no indication that certification opens new markets, even though the majority, especially large firms, indicate that overall competitiveness has improved. Furthermore, with the exception of a few large certified producers, there is no indication that profit margins have risen, for either certified or noncertified firms. Finally, it is also clear that certification has not led to an

The effect of global standards on local producers


Table 11.6 Changes in performance over the past three years. By firm size for certified and non-certified firms Increase in: Annual sales Employment Product quality Net profit margins Number of buyers

Certified firms Non-certified firms Certified firms Non-certified firms Certified firms Non-certified firms Certified firms Non-certified firms Certified firms Non-certified firms

N: certified firms N: non-certified




All firms

63% 50% 25% 13% 88% 75% 0% 0% 50% 13% 8% 8%

73% 75% 55% 50% 100% 75% 18% 25% 46% 50% 11% 4%

80% 0% 50% 0% 90% 100% 50% 0% 50% 0% 10% 1%

72% 54% 45% 23% 93% 77% 24% 8% 48% 23% 29% 13%

Source: Authors’ survey

improvement in unit prices of the cluster’s finished products. In fact, only 7 per cent of the total sample of 42 firms stated that they had enjoyed an increase in average prices of their products over the past three years. In contrast, 85 per cent of non-certified and 69 per cent of certified firms felt that over this period average product prices declined. To conclude, the evidence from Sialkot suggests that compliance with ISO 9000 led to some improvements in internal practices and a reduction in average reject rates. However, it did not translate into systematic improvements in terms of sales, employment or improved market access. Furthermore, certification did not lead to an increase in the unit prices of the cluster’s exports.11 That compliance on its own has not enhanced market access, increased sales or led to rising product prices, reflects the fact that for the cluster’s leading markets, compliance is a prerequisite for market entry. Being certified in itself is insufficient to compete in such markets. If certification has not led to a significant improvement in cluster performance, what changes has it brought about in the governance of value chain ties that link local producers to global buyers? I turn to this in the next section.


Local enterprises in the global economy

5. ISO 9000 AND GLOBAL VALUE CHAIN LINKAGES In what ways does compliance with ISO 9000 influence how local firms are inserted into the global value chain? In addressing chain governance, I am keen to gauge whether the implementation of the standard has promoted arm’s-length or hierarchical forms of governance in the ties between local firms and their global buyers. It is worth re-stating that the drivers for ISO 9000 in this sector lie outside the value chain that link local producers to global buyers. Nonetheless, the demands of market regulators require all actors in the chain to upgrade practices to meet the standard. This calls for upgrading within firms and in value chain ties that link them to their suppliers and their buyers. I begin by analysing backward linkages between firms and their local suppliers and subcontractors, and then turn to forward linkages with external buyers. The reduction in transaction costs associated with the compliance of ISO 9000 standards can have diametrically opposite consequences for inter-firm ties in the value chain. One view is that it would promote arm’slength contracts. The need for quality monitoring by lead firms would decline, and the associated costs would be pushed further down the chain. The risks of quality failure could also be pushed down the chain. Furthermore, if there are a large number of standards certified suppliers, lead firms could choose amongst them, thus forcing suppliers to reduce prices. Compliant suppliers, on their part, should obtain efficiency gains from implementing the standard to pay for such cost reductions. Consequently, compliance, by reducing transaction costs and providing key information on suppliers’ quality management practices, reduces risks for lead firms and stimulates arm’s-length contracts. The alternative view is that compliance results in closer value chain ties. Adoption of QA standards by suppliers may call for technical support from lead firms, especially where lead firms set the standards. Lead firms would need to assist suppliers in accessing the necessary know-how to implement the standard, develop the required management systems for quality assurance, and bring suppliers’ practices in line with those of lead firms. Furthermore, once lead firms are assured that suppliers conform with QA standards, they can enter into collaboration in other areas such as design and development. This can be particularly important if the standard is seen as indicating a supplier’s management and technical capabilities. Thus, certification may signal to lead firms that suppliers meet minimum ‘capability’ requirements. Once this ‘capability threshold’ is achieved, it is possible to enter into constructive and collaborative technical dialogue in areas of the

The effect of global standards on local producers


Table 11.7 Firm perceptions of buyer’s key priorities. By firm size for ISO 9000 certified firms Buyer’s key priority




All firms

Quality Delivery reliability Price ISO 9000 compliance N

88% 88% 63% 50% 8%

91% 82% 46% 55% 11%

100% 90% 80% 40% 10%

93% 86% 62% 48% 29%

Source: Author’s survey

value chain where rents are higher and where upgrading by the supplier is possible. This includes product development and process upgrading. From this perspective, ISO 9000 can play an enabling function and chain governance can become more network-oriented as compliance makes it easier for lead firms and local suppliers to ‘trade’ and ‘talk’. Sialkot’s leading export markets are the United States (US) and the European Union (EU), and pressures to meet QA standards are felt in both markets. Dealing with external buyers is central to the success of Sialkoti firms. Thus, while buyers were not the drivers behind the standard, they were recognized as an important force in the implementation of ISO 9000. Of certified firms, 59 per cent stated that their buyers explicitly required the standard. When asked what factors drove the push for ISO 9000 certification, ‘buyer’s requirements’ was the most frequently cited response (given by 45 per cent of respondents). Yet, ISO 9000 certification was not viewed by local firms as the buyer’s primary concern. Instead, as Table 11.7 shows, product quality, delivery reliability and price were considered to be of greater importance. How then did compliance affect ties with buyers? To address this, it is important to first explore forward linkages. In almost all cases, firms dealt directly with buyers and not end users in healthcare delivery. Moreover, firms usually traded with a number of distinct buyers in each market. Buyer concentration was, thus, not a concern for the cluster. Furthermore, ties with main buyers tended to be long standing, averaging over a decade for the sample as whole. Firm size did, however, influence the relationship. Large firms not only dealt with more buyers than SMEs, they also had longer-term ties with their main buyers. Despite these links, buyers were not an important source of support in attaining certification. All but two of the subset of certified firms reported that buyers provided no support in implementing ISO 9000. Nor did buyers


Local enterprises in the global economy

share costs associated with certification or give advice on changes in production organization in line with the standard’s requirements. Finally, there was no evidence of increased cooperation between buyers and firms in any activity, as a consequence of standard certification. The only arena where some respondents (25 per cent) stated that cooperation in forward linkages had improved was in information exchange. These findings need to be probed more carefully. Pressures to meet ISO 9000 standards for the Sialkot cluster came primarily from the European market. The European market is also dominated by higher quality theatre instruments. The biggest players in this market are German firms located predominantly in Tuttlingen, southern Germany.12 Sialkot, however, has emerged as an important competitor to Tuttlingen (Nadvi and Halder, 2002). As a consequence, many German manufacturers directly source from Sialkot. From our sample of 42 firms, 15 reported that their main buyer was from Tuttlingen. Of these 15 firms, ten stated that their main Tuttlingen buyer was also engaged in instrument manufacturing. Seven of these 15 were large enterprises, and 11 (or 73 per cent) were ISO 9000 certified. Large firms were much more likely to be supplying buyers in Tuttlingen. In contrast with 70 per cent of large firms, only 19 per cent of small firms had a Tuttlingen buyer. Case study evidence of ties with some European, primarily Tuttlingen, buyers indicated exceptions to the general view of limited support from buyers in adopting ISO 9000. For example, one large firm, a joint venture with a leading Tuttlingen-based manufacturing group, reported that its German partner not only helped it adopt the ISO system, but also undertook internal quality assurance audits. The firm had been ISO 9000 certified by a leading German auditor, the same organization that had certified its German joint venture partner. Another medium-sized enterprise stated that its leading German buyer, a producer of dental instruments in Tuttlingen, had been helpful in adopting the QA management system. The German buyer, who was also ISO certified, provided annual reports on product quality, customer feedback and reject rates of items produced by, or subcontracted to, the Sialkot affiliate. Thus, the firm saw the relationship becoming closer as a consequence of adopting ISO 9000. Another large firm reported that its German buyer had been influential in convincing it of the merits of ISO 9000 and provided technical advice. In spite of these exceptions, there was a broad consensus that the key technical support on the implementation of ISO 9000 came from local consultants. Moreover, despite the acknowledgement of some firms that buyers gave technical support on compliance, all firms stated that buyers did not extend a price premium for certification. Instead, buyers were seen as aggressively forcing prices down. Some respondents felt that with growing

The effect of global standards on local producers


numbers of certified firms in the cluster, buyers had an incentive to ‘shop around’. Certainly, there was strong evidence of increasing local price competition. Over 80 per cent of respondents report rising local competition in the past three years, and three quarters of all respondents attributed the decline in unit product prices to local competition. What do we conclude regarding the impact of ISO 9000 standards on the governance of ties with global buyers? With the exception of a few large firms, especially those dealing with German buyers, buyers were not seen as important sources of support in adopting the standard. In fact, despite a history of relatively longstanding ties with external buyers, many firms reported a weakening of ties as buyers became more fickle in response to growing price competition amongst local suppliers. What then are the implications for the governance of value chain ties? For most firms in the Sialkot cluster, insertion into global value chains comes about either through quasi-hierarchical chain governance, or through price-driven arm’s-length market transactions. Local firms are not engaged in design or product development functions. Instead, they closely follow the dictates of buyers in manufacturing relatively standardized products. Although the range of instruments produced in Sialkot is wide, the bulk of production is focused on a narrow range of instruments. In sourcing these products, buyers are keen to lower the governance costs of managing the chain. They do not demand ISO 9000 compliance for its own sake, but for regulatory reasons. Compliance, however, lowers some of their governance costs, including supplier search costs. Moreover, as the knowledge needed for compliance becomes widely available within Sialkot, buyers can take a more arm’s-length approach with the cluster. ISO 9000, therefore, appears to have led to a loosening of value chain ties.

6. ISO 9000 AND VALUE CHAIN LINKAGES WITHIN THE CLUSTER What is the impact of global standards on inter-firm relationships within the local cluster? In particular, does compliance change the way that lead firms within the cluster deal with other firms, particularly their suppliers? Under ISO 9000, the quality management practices of subcontracting units are not independently audited. Instead, certified manufacturers must maintain a list of their subcontractors and suppliers, and document all processes undertaken by individual subcontractors for each production batch. Thus, while a subcontractor is not certified, it is the responsibility of the certified firm to ensure that a subcontractor meets accepted quality assurance norms. This can raise costs. Subcontractors need to be made aware of, and


Local enterprises in the global economy

be assisted in adopting, the quality management system of the certified firm. Failing to do so is a cost borne by the certified firm. In response, manufacturers could internalize quality-critical processes in order to lower the costs of monitoring subcontractors and to minimize risks of non-compliance arising from failings by the subcontractor. The evidence for this, however, was limited in relation to the US GMP standards (Nadvi, 1999b). While some processes were internalized, especially by large firms, this was infrequent. The findings for ISO 9000 standards are similar. Of the 29 certified firms, 26 responded to the question of changes in the extent of subcontracting during the past three years (that is, since obtaining certification). Amongst these, nine reported an increase in subcontracting, eight a decrease, and the remaining nine stated no change in levels of subcontracting out. Firm size did, however, appear to influence the response. The majority of large firms (64 per cent) reported a decline in subcontracting. Similarly, most large firms described a decline in numbers of subcontractors, and in a few cases spoke of developing dedicated ‘first-tier’ subcontractors. In contrast, the bulk of SMEs cited an increase in their number of subcontractors. Given that subcontracting units are often small, have limited resources, weak managerial structures, and are run by individuals with low levels of formal schooling, they face difficulties in adopting quality assurance documentation. This calls for assistance on the part of certified firms. Yet, quality assurance training provided to subcontractors by certified firms was minimal. Only nine of the 29 certified firms reported that they had provided any training to their main subcontractors. Of these nine, four were large producers. On further exploration, it was apparent that training did not extend beyond informal advice on keeping product batches separate and workplaces tidy. Only one large firm reported that it had conducted seminars and training sessions with a few of its subcontractors to inculcate some of what it had learned through the QA system. These observations were borne out through qualitative interviews with subcontractors. Visits to subcontractors’ workshops confirmed that some QA requirements had filtered down the chain. Workshops were cleaner and instruments were kept in batch trays with attached display tags specifying batch lots and client codes. However, none of the subcontractors interviewed reported receiving any training from local producers. All subcontractors reported that the formal documentation for each job, what was locally referred to as a ‘travel process card’, was maintained not by them but by the certified producers. In conclusion, it appears that the governance of ties with local suppliers and subcontractors has been little influenced by the implementation of the standard. Firms were required to maintain approved lists of subcontrac-

The effect of global standards on local producers


tors, and subcontracting units had to maintain more managed control of work-in-progress. Concerns about the prospects of compliance failure arising from the subcontractors’ practices did not result in a widespread reduction in subcontracting, which remains a key aspect of local production organization especially for SMEs. In some cases, larger firms did internalize certain processes as well as build closer ties with key subcontractors. This often meant bringing subcontractors in-house, or employing managerial staff to conduct periodic monitoring of subcontractors. A key finding from this, in the context of standards compliance and the global value chain, is that for global buyers compliance with ISO 9000 quality assurance standards rarely extends beyond first-tier suppliers. Yet, the fact that certification does not extend to subcontractors, or require the auditing of subcontractors’ practices, raises doubts on the value of compliance in providing proper assurance on quality management within the value chain. If the standard had required the certification and auditing of second and third-tier suppliers, it is unlikely that the cluster would have attained such high levels of compliance, or to have done so rapidly.



In exploring the impact of ISO 9000 implementation on local governance, I address two issues. First, how is the ability of local enterprises to respond to global standards influenced by local governance, that is private joint action and public support? Second, how is this local governance itself changed by the imposition of global standards? Thus, the first question probes whether cluster effects facilitate the uptake of ISO 9000. The second question considers whether compliance strengthens or weakens local governance. An important externality gain of clustering lies in the way knowledge flows within the cluster. This can facilitate innovation and the adoption of new ideas, and promote localized technical upgrading (Nadvi, 1999a; Audretsch and Feldman, 1996). In this context, clusters can potentially accelerate compliance as changes in practices of compliant firms are rapidly adopted by other firms in the cluster. Standard certification could also result in reputation gains for the cluster as a whole, potentially generating a premium for the cluster’s products. Product standards and labels often take on an explicit regional dimension. The adoption of process standards can similarly generate externality benefits for the whole cluster. This, however, requires local regulation of the standards to ensure compliance and control free-riders. In as much as certification raises the stature of the whole cluster, non-compliance by one firm can lower the standing of the whole cluster.


Local enterprises in the global economy

The other aspect of the collective efficiency gains of clusters lies in local joint action. This includes cooperation between firms as well as local private and public institutions such as trade associations, chambers of commerce, technology service institutes, research centres and relevant local government agencies. Such bodies often provide an array of collective producer services within the cluster. Such organizations can play an important function in promoting the adoption of international standards. They can disseminate information and know-how related to the adoption of quality assurance systems, especially through training and advisory services. They can provide managerial and technical services to smaller local producers who lack such resources for adopting standards. They can accredit the certification process, ensuring that the standards are fully conformed with, and the reputation of the certification safeguarded. They can also assist the local cluster in upgrading and developing new standards and norms. The collective delivery of such services through local institutions can both be cost-effective and can assist smaller producers in adopting such standards. Meeting these functions, providing various collective services, undertaking monitoring of standards and regulating local firms to avoid noncompliance, requires strong local governance. This includes regional and local government as well as local institutions, both public and private. Local institutions need to be relatively strong, capable of collectively organizing, and have a clear sense of collective gain. However, where such forms of local governance are relatively poor one can posit that the ability of local producers to adopt global process standards would be weaker. This takes us to the second theme. Do global standards weaken or strengthen local institutions? This requires empirical evaluation. What can be said is that where standards are endogenous to the local cluster they are likely to strengthen local governance. Thus, where local institutions play a key part in formulating standards or in implementing them, their interests are wedded with ensuring that the standard is effectively adopted. However, where standards are imposed from outside, either exogenous to the global value chain or by actors within the value chain but external to the cluster, it is possible that local governance could weaken. I turn now to a brief overview of the case study. As shown earlier, local institutions in the form of the trade association (SIMA), the chamber of commerce (SCCI), and the technology centre (MIDC) had not been the leading sources providing assistance and knowhow to firms keen to adopt the standard. The state had provided partial financial assistance, yet this was not seen as an important factor in promoting the implementation of the standard. This is in contrast to the role of local governance in implementing the US GMP standards (Nadvi, 1999b). Implementing GMP standards required significant local joint action. The

The effect of global standards on local producers


reason for this was simple. Until the point when the standard was imposed there was little or no knowledge within the local cluster on what the standard entailed, or more specifically how quality assurance and quality management practices could be incorporated into local production practices. This knowledge was absent for both large and small firms. Thus, a collective need, namely technical know-how on quality assurance, motivated a collective response. Small and large firms came together through the trade association to address this, bringing a technical consultant from the US to the cluster to assist firms in upgrading their practices. In effect, local joint action addressed a market failure within the cluster, that is the lack of market-based service providers extending technical assistance to local firms on QA standards. The GMP standards were applied by the US regulators in mid-1994, and within months the US FDA embargoed Pakistani-manufactured surgical instruments for failing to comply. The imposition of the standard was an exogenous shock, and it took SIMA a year to organize a collective response. This included obtaining financial support from the state, identifying a suitable foreign technical consultant and implementing a contract for cluster-wide technical support. Within two years of this agreement, the consultant had assisted two thirds of clustered firms to develop the necessary quality documentation system and become compliant with the US GMP standards. Local policy networks were essential in facilitating the flow of technical know-how on quality assurance to the Sialkot cluster. Once joint action had set this process in motion, a wide array of service providers were attracted to the cluster. Knowledge on quality management practices dispersed widely and awareness grew of regulatory demands around QA standards beyond the US, in particular ISO 9000 in the EU market. Thus, by the time pressures from buyers to implement ISO 9000 began to increase, market provisioning of technical and audit services had developed both within the cluster, and more widely in Pakistan. There are now a large number of local QA consultants in Sialkot and in nearby Lahore. Most of the leading international auditors also maintain offices within Pakistan. With extensive and competitive market provisioning of technical services associated with quality assurance standards within the Sialkot cluster, the need for cluster-wide joint action diminished. Implementation of GMP standards strengthened the position of the trade association within the cluster. However, the association’s secretariat failed to upgrade its own technical resources. Similarly, the local technology centre fell behind the cluster in terms of acquiring expertise on the implementation of process standards and on upgrading production practices. Longstanding plans to develop a metal testing facility, managed


Local enterprises in the global economy

jointly by the trade association and the technology centre, had not materialized at the time of the survey. Instead, Sialkot’s firms looked to market providers of such services rather than local policy institutions. The growing importance of market provisioning of technical know-how has, in effect, weakened cluster institutions. Nevertheless, a dynamic perspective underlines the fact that local governance was critical in setting this process in motion. Thus while local governance mattered at the initial stages of implementing QA standards, its significance declined over time. If local governance no longer plays an important part in the implementation of ISO 9000 within Sialkot, does the implementation of the global standard weaken or strengthen local governance? Local governance is likely to be strengthened if local policy networks are critical to formulating the standard, to monitoring compliance, to providing assistance on implementing the standard or imposing sanctions on non-compliance with the standard. In the case of ISO 9000 in Sialkot, none of these apply. The standard is formulated outside the cluster and is driven by exogenous actors (in this case external regulators). Certification is undertaken by independent, international, auditors and assistance on compliance is provided by an array of local and international service providers. Sanctions on noncompliance are imposed by market regulators and external buyers who exclude non-compliant firms from entering leading markets. Thus, there is no evidence to indicate that implementation of ISO 9000 strengthened local governance. But there is also no evidence to suggest that local governance is weakened by the standard. It is clear, however, that local policy networks, while initially important in disseminating assistance on compliance with QA standards, have not taken on a wider role in promoting upgrading within the cluster around these standards. Local governance can play an important role. In fact, it may need to in order to uphold the standard. As shown above, there is limited evidence that implementation of the QA standards has improved production practices within firms beyond improved documentation and process control. Adoption of quality management practices still leaves a great deal of scope for firms to upgrade further. In addition, concerns have emerged in the cluster, and amongst leading external buyers, on the ‘quality’ of standard certification in Sialkot. These include unease on the link that often exists between technical consultants on ISO 9000 and those who audit compliance. The potential conflict of interest between these sets of actors can undermine the legitimacy of the audit process. Furthermore, no auditor operating in Pakistan is locally accredited since, until recently, Pakistan did not have its own accreditation bodies. National accreditation bodies in effect regulate the auditors, and ensure that auditors meet the standards required. All leading international auditors in Pakistan are regulated by

The effect of global standards on local producers


their home country accreditation bodies. In a competitive and rapidly growing market environment for standards certification, weak or distant regulation of practices by local auditors can also weaken the legitimacy of the standard.13 Thus, an important area for local governance is the local regulation of global standards. This requires, however, significant upgrading of the capabilities of local policy networks, equipping them with the technical expertise to monitor the audit process, and providing them with the regulatory means to ensure that standards are legitimately implemented. Weakness in the local auditing of the global standards would, in time, undermine the very effectiveness of the standard. In summary, imposition of ISO 9000 has not strengthened local or global governance. On the contrary, the evidence from the field suggests that global governance in value chain ties has weakened ties with global buyers especially. There are, however, important exceptions to this, as seen for some large case study firms who built stronger ties with leading producers in Germany. Yet, there are limits to the extent of upgrading within such firms, and in the nature of technical knowledge flows between their German partners and themselves (see Nadvi and Halder, 2002). Similarly, the role of local institutions appears to have weakened as market provisioning of technical services associated with the standard, both in terms of technical advice and monitoring, spread within Sialkot. Despite this, the need for local policy networks remains, especially in regulating the implementation of the standard.

8. QUALITY AND CHILD LABOUR STANDARDS COMPARED The imposition of ISO 9000 standards has not excluded SMEs in Sialkoti from global markets. On the contrary, compliance has been rapid and widespread. However, limited upgrading, declining prices, no competitive advantages and weakening ties with global buyers suggest that compliance does not necessarily strengthen the position of developing country producers in global markets. Are these results inevitable and to be expected with all standards? No, a lot depends on where the standard comes from and how it is implemented and monitored. This point can be illustrated by contrasting the evidence on ISO 9000 with the implementation of standards concerned with child labour in Sialkot. Reviewing the evidence between the two standards highlights the importance of understanding how specific global standards are themselves governed. This includes who defines the standard, how it is implemented, how assistance on compliance is provided, and how sanctions


Local enterprises in the global economy

against non-compliance are enforced. In all these respects, there is a difference between standards that are imposed from outside the global value chain or the local cluster, and standards that are driven from within the chain or the cluster. During the late 1990s, reports in the international media highlighted the presence of child labour in Sialkot’s export manufacturing clusters, especially in sports goods manufacture. These revelations led to demands from Western consumer groups, trade unions and NGOs for a boycott of Sialkot-manufactured footballs. International buyers challenged local suppliers to meet international norms on the elimination of child labour from production. Consequently, in 1997, the Sialkot Chamber of Commerce and Industry (SCCI), with the ILO, UNICEF and leading international sports goods associations, signed the ‘Atlanta Agreement’. Its aim is to remove child labour from the soccer ball cluster through a phased programme (SCF, 1997). The Atlanta Agreement established the presence in Sialkot of the ILO’s International Programme for the Elimination of Child Labour (IPEC). Funded by aid donors, with contributions from local firms through the SCCI, it introduced ILO monitoring of soccer ball production in Sialkot. By 2000, 68 of the total of some 90 soccer ball exporting firms in the cluster, accounting for over 75 per cent of the cluster’s total production, were voluntarily monitored through the IPEC programme. In addition, the agreement supported initiatives by UNICEF and Save the Children Fund (SCF) to provide social protection to former child workers and their families. This sought to address the root causes of child labour through interventions on poverty alleviation and education. This included support for schools, training and income-generating activities. The social protection agenda now has the active involvement of local NGOs and local government and promotes a district-wide strategy for universal primary school enrolment; this is a first for any district in Pakistan. The experience with labour standards on child workers points to important contrasts with the ISO 9000. Both sets of standards were externally imposed on the local cluster. ISO 9000 has market-based, independent, monitoring of compliance. In contrast, a more complex network arrangement exists for monitoring child labour that involves local (the SCCI) and global actors (the ILO, global NGOs and the world sports goods federation). A key factor for this is the absence of well defined, credible, marketbased procedures for monitoring labour standards. Most global buyers insist that their local suppliers join the ILO–IPEC programme for monitoring purposes. In addition, some leading brand-name buyers have also opted for direct monitoring. Thus, the distinct nature of monitoring led to different impacts for the two sets of standards. Furthermore, with the wider

The effect of global standards on local producers


agenda of social provisioning around child workers, a complex network of local community bodies, local NGOs, global NGOs and international organizations emerged to deliver improved education and welfare delivery within the area. These developments strengthened local policy networks, both at public and private levels. The two different types of standards also had different impacts on value chain ties. In the soccer ball cluster, two developments emerged. First, changes in patterns of subcontracting, and second, changes in ties with brand-name buyers. The cluster historically relied heavily on rural homebased workers to carry out the more labour-intensive tasks in producing soccer balls. This was where child labour was most common. Subcontractors acted as intermediaries between exporting firms and household stitching units. With the need for effective and cost-efficient monitoring, home-based work has given way to production in designated stitching centres. Under the ILO programme, each manufacturer must register the location of each of its stitching centres, and all production must take place within these centres. These are regularly and randomly inspected by ILO–IPEC monitors. This has changed the nature of subcontracting. Dispersed home-working has given way to stitching centres specific to particular manufacturers and particular subcontractors. Subcontractors and workers are more closely linked with manufacturers. Thus, while subcontracting has generally not declined, producers’ ties with subcontractors have become more dedicated than before. Brand name buyers are most vulnerable to child labour scandals. Not only do they risk the loss of supplies of soccer balls from Sialkot, but their brand image is also under threat. Consequently, it is these buyers who have been most forceful in implementing changes in practices on the part of their suppliers, the cluster’s larger producers. In most cases, this involved an internalization of activities formerly subcontracted to home-based workers. Some large firms set up centrally located factories to which workers are transported from distant villages. This allows better monitoring of quality standards and regulating labour norms. As suppliers to the leading brand names, such firms have also had to adhere to the buyer’s code of conduct. Moreover, some of the leading brand buyers have invested in ties with local NGOs. Such NGOs have been engaged to monitor a supplier’s labour practices. It has also allowed brand buyers to undertake local philanthropy, channelling resources through local NGOs to promote education and schooling for former child workers. In some cases, brand buyers have also encouraged their main suppliers to adopt the principles and values of corporate social responsibility. This includes on-site medical facilities to workers, access to cheap credit and to fair-price shops. Thus, the imposition of labour standards resulted in a marked shift in the governance of value chain ties, leading to more hierarchical and quasi-hierarchical arrangements.


Local enterprises in the global economy

In terms of cluster governance and institutions, the wider agenda of social provisioning around child workers promoted the emergence of a complex network of local community bodies, local NGOs, global NGOs and international organizations to deliver improved education and welfare delivery within the area. These developments strengthened local policy networks, both at public and private levels. Furthermore, the arrangements for monitoring compliance strengthened local governance with an emphasis on complex network arrangements that bring together local and global as well as public and private actors. This experience, at the level of local governance, is similar to the initial experience with quality assurance standards for the surgical instruments cluster in the mid-1990s. At the time, network arrangements involving public and private actors were necessary to ensure compliance. Once the network succeeded in bringing the necessary knowledge and process changes to the cluster, a market for such services developed rapidly. It remains to be seen with the child labour issue whether current network arrangements on monitoring and social provisioning will give way to private market-based monitoring. There is a possibility for such a development, especially as leading international certification agencies formulate procedures to monitor social and ethical standards (such as SA 8000) that incorporate labour standards.



This chapter is concerned with local responses to global standards. Received wisdom suggests that compliance with ISO 9000 is costly, slow and restrictive for SMEs. The findings from the Pakistan case study challenge this. Local firms, including small enterprises, were able to comply. Whereas only two firms were ISO 9000 certified in 1997, by mid-2000 over 170 ISO 9000 certifications had been issued in the cluster. Compliance was fast (most firms being certified within six months) and cost-effective. In fact, costs declined with rising competition amongst QA consultants and auditors and the rapid flow of knowledge on QA standards in the cluster. These results are impressive, but how did they come about? Neither the state, local institutions nor external buyers emerge as critical sources of support to local firms on achieving compliance. Instead, local consultants and international auditors provided key services. Nevertheless, if one takes a historical view on the implementation of QA standards in Sialkot since the mid-1990s, a more nuanced picture emerges. This suggests that cluster effects, such as local joint action and knowledge spillovers, did matter. Compliance with US GMP standards was predicated on local joint action through the association and required cluster governance (Nadvi, 1999b).

The effect of global standards on local producers


Once the knowledge necessary to comply became widely available within the cluster, and the costs associated with it declined, market provisioning of such technical services expanded. But cluster effects continued to matter. The cluster provided a market for such service providers and reduced the costs of provisioning of such services. Moreover, dynamic externalities emerged through local knowledge spillovers. The know-how required to implement standards flowed from firm to firm within the cluster. Compliance appears not to have improved the competitive position of local producers or led to new markets. Despite high levels of certification, Sialkot’s reputation as a manufacturer of high quality products did not improve. Furthermore, while sales levels have risen, unit prices have declined since QA standards were imposed. This could be an outcome of improved process efficiency, buttressed by falling values of the Pakistani rupee during this period. Nevertheless, it is apparent that despite certification, international buyers are more aggressive in seeking lower prices from local suppliers. Compliance with QA standards is thus, at best, a certificate for market entry. Did compliance bring about upgrading? It required changes to process organization that could lead to process upgrading. There was better control of stocks and work-in-progress, and a decline in product reject rates. This provided firms with important cost savings, identified quality control bottlenecks and led to better production planning. But it is also clear that such upgrading did not extend to subcontractors. This is an important finding that raises wider questions on the QA standards and their impact on value chain ties. If QA standards promote process upgrading, they appear to do so only for first- and not second-tier suppliers. Compliance appears to have loosened value chain ties. There are exceptions, especially amongst large firms supplying high quality German producers. Generally, governance in chain ties appears to have become more arm’s-length. At the higher quality end, especially for firms supplying leading Western traders and producers, chain governance remains largely quasi-hierarchical. Buyers provide product designs, often supply raw materials, and continue to conduct quality inspections. Amongst such buyers, ISO 9000 certification is not considered a sufficient basis for determining supplier competence. Direct monitoring of suppliers alongside technical assistance is often required. Nevertheless, buyers have more choice of local suppliers, resulting in weakening links. For low quality products, especially where expatriate Pakistani traders drive the chain, price competitive, arm’slength governance is the norm. Compliance with international QA standards failed to strengthen local cluster governance. Local policy networks did play an important function in facilitating market provisioning of technical services associated with


Local enterprises in the global economy

such standards. Once this had developed, the role of local institutions diminished. The local trade association and technology centre failed to develop their own capacities on implementing process standards. Local metal testing facilities and local quality standards agencies did not emerge. Concerns on the credibility of local certifications underline the absence of effective local governance in monitoring compliance. Thus, the evidence from the Sialkot cluster and ISO 9000 suggests that chain and cluster governance have at worst weakened, and at best remained unchanged. Yet this remains an area where collective action can enhance the competitive position of local producers. Local collective action is needed to monitor the ‘certification industry’ and ensure that compliance really represents upgrading of processes. From the evidence on the surgical instrument cluster, it appears that value chain ties, both within the cluster and with global buyers, became looser. The reasons for this are rooted in the drivers behind the standard, the way in which compliance is monitored, and the manner in which assistance on achieving compliance is delivered. The demand for implementation of ISO 9000 came from market regulators outside the chain. Clearly, lead firms require compliance by their suppliers. But, once that is achieved, an even greater consideration for lead firms are questions of price, product quality and delivery reliability. With prices falling, and local competition increasing, there are signs that buyers are moving to price-driven choices across a growing pool of certified firms. A rather different set of findings emerged from the analysis of child labour standards in the soccer ball cluster. There are signs that leading brand-name buyers have been directly involved in helping, or forcing, their local suppliers to change practices. This has meant a shift to more centralized production with closer monitoring of labour standards and quality. Thus, this evidence appears to support the proposition that the impact of global standards on value chains (global governance) and clusters (local governance) is conditioned by the nature of the chain and the governance of the standard. With ISO 9000 and the surgical instruments cluster, independent monitoring, and limited innovation and product upgrading within the Sialkot end of the chain, led to global standards weakening chain governance. In contrast, with complex monitoring procedures and greater risks associated with non-compliance, child labour in the soccer ball cluster has led to more hierarchical chain governance. There are analogies between the child labour case and evidence that Dolan and Humphrey (2000), Reardon et al. (2001) and Barrientos et al. (2001) cite on standards in the fresh food and agri-products sector, with branded retailers seeking to minimize risks of supplier non-compliance through closer monitoring and inspection.

The effect of global standards on local producers


Finally, there is a role for local governance. It was important when quality assurance standards were initially required, and it is central to the cluster’s response to current pressures on child labour. Where market provisioning of knowledge and monitoring around process standards is absent, local governance is essential. Moreover, there are indications that local governance can be important in ensuring the credibility of local compliance. This is seen today in the case of ISO 9000 in Sialkot. There are suggestions that consultants and auditors, who should clearly be independent, are often closely linked. Some audit firms are said to have consultants working directly for them. The latter provide the training while the former provide the certification. Given that firms, certification agencies and consultants are all interested parties, such arrangements can result in significant conflicts of interest. In such cases, the quality of the audit, and of the certificate, is questionable. Linked to this, some local observers distinguish between first and second quality ISO certificates. According to this view, a small handful of firms seriously adopt QA practices. However, many others only go through the process nominally. While adopting some changes, through better documentation and improved batch control, they have yet to see the process as a step towards a total quality management strategy. Certainly, the fact that some firms were able to start the QA process and obtain a certificate within the space of six months raises concerns on the extent to which the process was truly understood and internalized. But are these concerns unique to the Sialkot case? Serious doubts about the ISO 9000 standards, and the process by which it is implemented have been raised elsewhere (see Seddon, 2000). Much of the critique focuses on the ‘control and procedural’ aspect of the standard, and its failure to adopt an approach that focuses on a systemic process of quality upgrading. It also raises questions about the rapidly growing certification industry. There are concerns that international auditors are inadequately regulated by their respective national accreditation agencies. As it is, such commercial firms can have vested interests in ensuring that client firms obtain certificates, and thus may be considered to be far from independent. The ISO, and for that matter the ILO, as international institutions have neither the clout nor the administrative or legal capacity to ensure that standards are effectively complied with. Thus, for global process standards to remain credible in achieving their goals, more thought needs to be applied to the governance, at both local and global levels, of global standards.


Local enterprises in the global economy


1. 2. 3. 4. 5.

6. 7. 8.

9. 10. 11. 12. 13.

Financial support from the Department for International Development, UK and the Volkswagen Stiftung, Germany is gratefully acknowledged. My thanks to the Sialkot Chamber of Commerce and Industry and the Sialkot ILO–IPEC office for their support. My sincere gratitude to the respondents interviewed for their time and cooperation. Mr Sajid Kazmi of the Sustainable Development Policy Institute, Islamabad provided excellent research assistance and co-authored an earlier version of the chapter. My thanks to him, Dirk Messner, Gerhard Halder, Judith Tendler, and, in particular, John Humphrey and Hubert Schmitz for comments on earlier drafts. The usual rejoinders apply. See Nadvi and Wältring 2002 for an overview on leading global standards. An exception to this is the study by Quadros (2002) in this volume. ISO 9000’s popularity has led to numerous national and sector-specialized variants, including QS 9000 in the automobile industry, AS 9000 in aeronautics and PM 9000 for public sector bodies. For details on the 2000 version of the standard see ISO (2000) and Seddon (2000) for a critique. The 1994 version differentiates in scope of coverage. ISO 9001 certifies quality assurance in design, development, production, installation and servicing. ISO 9002 includes all of the above with the exception of design and development. ISO 9003 only covers quality assurance in final inspection and testing. For a description of the cluster see Nadvi (1999a; 1999b). For a discussion on the impact of this crisis on horizontal and vertical production ties, see Nadvi (1999b). Personal communication with the Quality Management Cell, Federation of Pakistani Chambers of Commerce and Industry, Karachi, September 2000. Given the relative scale of the surgical instruments sector, accounting for just 4.1 per cent of textile sector exports in 1997–8, this achievement is even more remarkable. The subsidy is financed by an export development tax raised by the Export Promotion Bureau. It is paid by instalment, with final payment on successfully passing the first postcertification surveillance audit. Table 11.3 provides evidence from the subset of certified firms and the five non-certified firms that were awaiting the final audit. Unit prices in US dollars declined over the past three years, but were offset by changes in the relative exchange rate with the decline in value of the Pakistani rupee during the 1990s. See Halder (2002) on Tuttlingen’s position in the global surgical instrument industry Quadros (2002) makes a similar point in the context of Brazil’s auto-component industry.

REFERENCES Audretsch, D. and M. Feldman (1996), ‘R & D spillovers and the geography of innovation and production’, American Economic Review, June, pp. 630–40. Barrientos, S., C. Dolan and A. Tallontire (2001), ‘Gender and ethical trade: a mapping of issues in African horticulture’, Working Paper, NRET, National Resources Institute, University of Greenwich. David, P. (1995), ‘Standardisation policies for network technologies: the flux between freedom and order revisited’, in R. Hawkins, R. Mansell and J. Skea (eds), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Aldershot, UK and Brookfield, US: Edward Elgar.

The effect of global standards on local producers


Dolan, C. and J. Humphrey (2000), ‘Governance and trade in fresh vegetables: the impact of UK supermarkets on the African horticulture industry’, Journal of Development Studies, 37 (2), 147–77. Halder, G. (2002), ‘How does globalisation affect local production and knowledge systems? The surgical instrument cluster of Tuttlingen, Germany’, INEF Report 57, Institute of Development and Peace, University of Duisburg, Germany. Government of Pakistan, Economic Survey, 1998–99, Islamabad, Economic Adviser’s Wing, Finance Division, Government of Pakistan. ISO (1994), ISO 9000, International Standards for Quality Management, Geneva: International Organization for Standardization. ISO (2000), Transition Planning Guidance for ISO/DIS 9001:2000, Document: ISO/TC 176/SC 2/N 474 (2000), in: http://www.iso.ch/9000e/revisionstoc.htm Nadvi, K. (1999a), ‘The cutting edge: collective efficiency and international competitiveness in Pakistan’, Oxford Development Studies, 27 (1), 81–107. Nadvi, K. (1999b), ‘Collective efficiency and collective failure: the response of the Sialkot surgical instrument cluster to global quality pressures’, World Development, 27 (9), 1605–26. Nadvi, K. and F. Wältring (2002), ‘Making sense of global standards’, INEF Report 58, Institute of Development and Peace, University of Duisburg, Germany. Nadvi, K. and G. Halder (2002), ‘Local clusters in global value chains: exploring dynamic linkages between Germany and Pakistan’, IDS Working Paper 152, Institute of Development Studies, Brighton. Quadros, R. (2002), ‘Global quality standards, chain governance and technological upgrading of Brazilian auto-components producers’, IDS Working Paper 156, Institute of Development Studies, Brighton. Reardon, T., J-M. Codron, L. Busch, J. Bingen and C. Harris (2001), ‘Global change in agrifood grades and standards: agribusiness strategic responses in developing countries’, International Food and Agribusiness Management Review, 2 (3), 421–35. Save the Children Fund (1997), Stitching Footballs: Voices of Children, London: Save the Children Fund. Schuurman, H. (1998), ‘Quality management, ISO 9000 and government programmes’, Desarrollo Productivo 24, Santiago: Economic Commission for Latin America and the Caribbean, United Nations. Seddon, J. (2000), The Case Against ISO 9000, Dublin: Oak Tree Press. Weiss, A.M. (1991), Culture, Class and Development in Pakistan: The Emergence of the Industrial Bourgeoisie, Lahore: Vanguard Press.

12. Paradoxes and ironies of locational policy in the new global economy Jörg Meyer-Stamer 1. INTRODUCTION: WHAT IS LOCATIONAL POLICY? This chapter addresses the options for, and limitations to, localized efforts to stimulate and shape economic development, both in industrialized and in advanced developing countries.1 I define locational policy as the effort of local stakeholders, both from government and non-government, to create a favourable environment for business activities. It aims at improving the locational quality of a region, so that existing companies may become more competitive, entrepreneurs will find it an attractive location to establish a business, and external investors will prioritize it when making locational decisions. Moreover, it may include proactive measures to stimulate and support the competitiveness of companies, and shape structural change. Traditional locational policy addressed three issues: (i) making real estate available and improving the infrastructure; (ii) attracting external investors; and, (iii) facilitating communication between the business community and the public sector (Hollbach-Grömig, 1996). As locational competition became more intense and unemployment levels rose, local actors in industrialized as well as developing and transformation countries started to pursue a more ambitious approach2 which included at least one of the following elements: ●

the creation of roundtables, partnerships (including public–private partnerships) or alliances for local economic development (LED) in order to formulate and implement a strategy to improve the locational advantage or revitalize old locations (OECD, 2001; Wallis, 1996; Küpper, 2000); the implementation of cluster initiatives (Raines, 2000; Enright, 2000); the creation of dedicated local economic development agencies to coordinate and organize local level efforts (ILO et al., undated). 326

Paradoxes and ironies of locational policy


Why do local players support LED initiatives? The phenomenon of globalization is the main cause for competition between locations and the necessity for locational upgrading. First, in many if not most industries, firms now have numerous options when it comes to choosing an attractive location. As some locations succeed in quality improvement, others have to follow suit or face being left behind, with local firms moving elsewhere and external investors not shortlisting them. What is emerging is a pattern of competition between localities, that is cities, city-regions or other subnational regions. This has been coined ‘competitive regionalism’ (Jonas and Ward, 2001) or ‘territorial competition’ (Cheshire, 2001). Second, local players may consider an LED initiative even if local companies cannot use relocation as a threat. Local companies have to compete against companies from different home-bases, and if their competitors enjoy the benefits of a more sophisticated location then they pose a serious competitive threat to local companies. They could even squeeze local companies out of the market. In order to avoid this, and the consequent employment and tax losses, local players will try to enhance the competitiveness of local companies. Third, there are still many locations which are not actively integrated into the globalized economy in terms of supplying goods or services. For players in such locations, the primary motivation is to find access to larger markets so that local companies can grow and generate more jobs and tax income. The problem is that the implementation of LED initiatives does not usually run smoothly, although this is not necessarily due to the reasons elaborated in the academic literature. For instance, the LED and the cluster discussion assume that the respective locality is integrated into the economy at large and this is the starting point for LED initiatives. It is rare to find an analysis of the restrictions that the connection between a local economy and the global economy creates for LED efforts. In the international political economy literature, for example, it is more common to find the argument that the global economy determines what can be done at the local level. Thus, the usual conclusion is that very little can be done (for example Menzel, 1995). This argument can easily be falsified, as a number of successful LED initiatives have documented (for example Aghón et al., 2001). My own recent research and advisory work has been centrally concerned with LED, and I consistently found various opportunities for LED initiatives, some of which rendered surprising successes (Meyer-Stamer, 2000). Nevertheless, in this chapter I will argue that our current understanding of promoting local economic development is inadequate, and I will attempt to improve our understanding using an actor-oriented approach. I will concentrate on two features that have not been raised in the literature so far,


Local enterprises in the global economy

namely the upgrading paradox and the location paradox. The upgrading paradox addresses the goal of locational policy, namely upgrading companies to make them more competitive and upgrading the location to create employment and increase the tax base. I will show how difficult it is to get a clear understanding of what we mean by upgrading (section 2), and that upgrading is a vague goal (section 3). Moreover, the options for upgrading change with the industrial life cycle (section 4). I will then focus on the irony of upgrading, that is as locational upgrading takes place, the conditions for locational policy deteriorate (section 5). Finally, the locational paradox addresses the governance pattern of locational policy, and in particular the involvement of the private sector (section 6). Despite these problems, there are ample options for locational policy, since apart from strategic locational policy it is also possible to pursue generic or reflexive locational policy (section 7).



At first glance, the issue of upgrading appears to be straightforward enough. For a firm, upgrading means getting better – producing better products and producing them in a more efficient way. However, things are more difficult than that. Take the example of a garments manufacturer. Developing a new line of products every three months does not really imply upgrading. It is rather a routine in this business. However, if the firm has produced T-shirts, sweatshirts and polo-shirts for a number of years and then decides to also include jeans and shirts in its product portfolio so that it can offer a more complete product range and thus establish a competitive advantage, this is clearly a case of upgrading; in this case product-related. Also, if it establishes its own unit with sophisticated finishing machines, so that the products shrink less and stay in shape for longer, it is upgrading; in this case process-related. However, there are also other options for upgrading which go far beyond these upgrading activities, such as forward integration by setting up a franchise network or even opening its own shops. Recently, there have been efforts to formulate typologies for this kind of upgrading, particularly with respect to developing countries which suffer from ‘immiserising growth’ (Kaplinsky, 2000) due to insufficient upgrading. Humphrey and Schmitz (2001) argue that there are four types of upgrading: ●

Process upgrading: firms can upgrade processes – transforming inputs into outputs more efficiently by re-organizing the production system or introducing superior technology.

Paradoxes and ironies of locational policy ●


Product upgrading: firms can upgrade by moving into more sophisticated product lines (which can be defined in terms of increased unit values). Functional upgrading: firms acquire new functions (or abandon existing functions) so that they increase the overall skill content of their activities. For example, they might complement production with design or marketing, or move out of low-value production activities altogether. Intersectoral upgrading: firms apply the competence acquired in a particular function of a chain to move into a new sector. For example, competence in producing TVs is used to make monitors and thus move into the computer sector. Such horizontal moves into new sectors seem to have been central to Taiwan’s ability to gain a foothold in skill-intensive sectors (Humphrey and Schmitz, 2001). This typology is related to the example outlined above. However, it is not entirely satisfactory. To understand its weaknesses, it is useful to refer to the argument developed by Porter (1996), who conceptualizes upgrading in an unconventional way:

The quest for productivity, quality, and speed has spawned a remarkable number of management tools and techniques: total quality management, benchmarking, time-based competition, outsourcing, partnering, re-engineering, change management. Although the resulting operational improvements have often been dramatic, many companies have been frustrated by their inability to translate those gains into sustainable profitability (Porter, 1996, 61).

Underlying this, Porter argues, is the fact that firms confuse operational effectiveness with strategy. ‘Operational effectiveness (OE) means performing similar activities better than rivals perform them’ (ibid, 62). ‘OE competition shifts the productivity frontier outward, effectively raising the bar for everyone. But although such competition produces absolute improvement in operational effectiveness, it leads to relative improvement for no one. (. . .) major productivity gains are captured by customers and equipment suppliers, not retained in superior profitability’ (ibid, 63). The more benchmarking companies do, the more they look alike. The more that rivals outsource activities to efficient third parties, often the same ones, the more generic those activities become. As rivals imitate one another’s improvements in quality, cycle times, or supplier partnerships, strategies converge and competition becomes a series of races down identical paths that no one can win (ibid, 64).

The way out of this trap, according to Porter, is strategy: ‘Choosing to perform activities differently or to perform different activities than rivals (ibid.).


Local enterprises in the global economy

With most companies striving for operational effectiveness, the reality of competition in globalized markets is often running to stand still. A company which is standing still altogether, that is, that does not pursue some kind of upgrading all the time, has little chance of survival. Whereas a company which puts a lot of effort into upgrading does not necessarily improve its competitive position.3 This is particularly true for clusters, where competition is localized and particularly strong and visible. For this reason the first part of Humphrey and Schmitz’s argument (quoted above) is not entirely adequate. Increasing the skills content is something firms have to do all the time to maintain their competitive position. This is, in the words of Porter, about operational effectiveness. It is about performing activities in a different way from how they performed them in the past, but not necessarily about performing activities in a different way to their rivals. Porter would be more likely to recommend moving into market niches which have entry barriers in terms of strategy, although this does not create a strategic difference vis-à-vis competitors if it just implies copying the activities of more sophisticated rivals. It is not unusual to observe several competitors trying to move into the same market niches. By focusing on the four types of upgrading suggested by Humphrey and Schmitz, the same argument applies to process upgrading and functional upgrading. Although both are typical manifestations of operational effectiveness, they do not improve a company’s competitive position. In the future we need to rethink the concept of upgrading, and acknowledge that it must be a relational category that does not compare a company’s, cluster’s or location’s previous practice with current practice, but looks at their position vis-à-vis main competitors instead.

3. UPGRADING, DOWNGRADING AND SIDESTEPPING The ‘upgrading’ discussion has a further shortcoming: it implies that the term ‘upgrading’ means to move up. However, companies may choose different options. For instance, in her case study of the Brenta footwear cluster, Rabellotti (2001) tells a story of voluntary downgrading. In the past, shoe producers worked with buyers in Germany, for example, offering their own design and taking an active role in marketing. However, as competition in the German market became more intense, some producers opted to become subcontractors of Italian luxury brands, thereby upgrading their products, but downgrading their companies’ competences by dropping the design and the marketing function, activities which one would usually recommend as key activities for upgrading and increased earnings.

Paradoxes and ironies of locational policy


This strategy by Brenta firms was based on the fact that the luxury end of the shoe market rapidly increased in the 1990s. However, product upgrading does not necessarily enhance competitiveness and create economic success. There is the risk of upgrading products so much that they lose their market base. The German capital goods industry provides a cautionary tale. This industry eventually created such sophisticated machine tools that they became far beyond the requirements of most customers (Widmaier, 1998). In other words, not only is there the ‘sandwich situation’ frequently encountered by producers in newly industrializing countries, which are caught in between technologically superior competitors in industrialized countries and lower-cost producers in less-advanced countries; there is also a ‘thin-air situation’ encountered by high-end producers which try to upgrade further, thus creating increasingly elaborate products which are beyond the requirements of their customers. In our investigation of the ceramic tile industry, we found yet another pattern, namely simultaneous product upgrading and downgrading (Meyer-Stamer et al., 2001). Italian producers in particular are famous for their high-end tiles, with unique designs in terms of colour and surface structure. However, these producers put a lot of effort into entering into arrangements with large retail chains in locations such as Germany even though this is not the typical place to sell high-end tiles. The tile producers manufacture private brands for these chains, which are sold at low prices. This behaviour is rational for two reasons. First, the tile market seems to display, at least in countries with a mature retail structure, a tendency towards polarization, that is the high end is growing, the low end is also growing, whereas the medium segment loses importance. For companies, it is attractive to have a presence in both the high and the low end as long as the low end products do not cannibalize the high-end products. With differentiated design and the existence of private brands, this risk is low. Second, even though the margins, and perhaps even the absolute earnings, at the low end are smaller, the scale of production is larger, thus permitting a quicker amortization of new equipment. As new equipment plays an important role in product upgrading, this may create the economic base for continuous investment in new machinery and thus, ironically, for constant innovation in the high-end product segment. The pattern observed in the tile industry is depicted by the right-hand figure in Figure 12.1. How can the centre figure, that is lateral differentiation, be explained? The Swedish apparel retailer Hennes⫹Mauritz and its Spanish competitor Inditex/Zara are examples of this pattern. They have created a new segment of the market, that is low-price, highly differentiated garments (Capell, 2002; Bonnin, 2002). By renewing the stock in their shops on a fortnightly basis, they cater in particular to the demands of


Local enterprises in the global economy

Conventional view: Moving towards the high end in markets

Alternative view: Polarization

Alternative view: Lateral differentiation

Source: Author’s survey

Figure 12.1

Types of product upgrading

young customers who want to look different from their peers. In some other US stores, the renewal rhythm has been reduced to one week. From a retail perspective this means moving sideways rather than upwards. In terms of price, they compete with traditional retailers such as C&A or The Gap, whereas in terms of product differentiation they create a segment of their own. From a supplier perspective, it implies process upgrading. Lee-Young and Barnett (2001) describe the example of Li & Fung, a supply-chain coordination firm located in Hong Kong, ‘one of the speediest suppliers on Earth’ (see also Magretta, 1998). What does all this mean for companies’ efforts to upgrade? Essentially, it comes back to the Porter argument: upgrading is not a priori about a direction, such as upwards. Instead it is more about productive rent-seeking (Kaplinsky, 1998). Upgrading means doing things differently, and/or doing different things – not different compared to previous practices in the same company, but compared to competitors. For locational policy makers, this means that it is difficult to formulate activities to support upgrading activities. Upgrading is much more contradictory and confusing than the Humphrey/Schmitz-typology would suggest – the direction is a priori unclear, and therefore it is difficult, for instance, to assess the necessities in terms of factor conditions resulting from companies’ upgrading efforts.

Paradoxes and ironies of locational policy


4. TERRITORIAL UPGRADING AND THE LIFE CYCLE PARADOX My reflections on upgrading so far have been related to the Humphrey/ Schmitz typology. Their consideration of upgrading addresses companies, not locations per se. A different approach to defining types of upgrading, seemingly with an explicit focus on the territorial dimension, has been suggested by Gereffi (1999): Industrial upgrading is a process of improving the ability of a firm or an economy to move to more profitable and/or technologically sophisticated capital and skill-intensive economic niches. Industrial upgrading operates at several different levels of analysis: (1) within factories – upgrading involves moving from cheap to expensive items, from simple to complex products, and from small to large orders; (2) within inter-firm enterprise networks – upgrading involves moving from mass production of standardized goods to the flexible production of differentiated merchandise; (3) within local or national economies – upgrading involves moving from simple assembly of imported inputs to more integrated forms of OEM [original equipment manufacture] and OBM [own-brand manufacture] production, involving a greater use of forward and backward linkages at the local or national level; and (4) within regions – upgrading involves shifting from bilateral, asymmetrical, inter-regional trade flows to a more fully developed intra-regional division of labor incorporating all phases of the commodity chain from raw material supply, through production, distribution, and consumption (Gereffi, 1999, 51 f).

This typology has two shortcomings. First, Gereffi suggests that regional upgrading is the sum of the upgrading of firms in the region. To an extent this is the case, as the companies’ success is the result of certain policy initiatives, that is locational policy, technology policy, competitiveness policy, and so on. But in another sense it is not a useful typology since it is merely descriptive, that is it does not elaborate in terms of analytical foundation. Crucial analytical questions are: Did the firms upgrade because locational factors improved, or despite the fact that they remained minimal? What was more relevant: intra-firm effort, inter-firm collaboration, locational policy efforts, or overall macroeconomic conditions? Second, it is no coincidence that the typology looks like a generalized description of what has happened in East Asia during the last few decades. In fact, much of Gereffi’s research has focused on this region. This line of thinking on upgrading is very much based on the Japanese, Korean and Taiwanese experience during the 1960s to 1980s. The upgrading of the Japanese and Korean economies (and to a lesser extent that of Taiwan) occurred during a period when industries were predominantly organized along Fordist principles. This leads us to question how the phase of a long


Local enterprises in the global economy

cycle affects upgrading. It is certainly true that today we are in a different phase of the cycle than in the 1970s. For example, this is what all the discussion on flexible specialization since the 1980s has been about. Perhaps this has grave consequences for upgrading, and perhaps the East Asian experience has only limited relevance today (but will become relevant again in 20 years or so). But so far we do not really know, since the issue of the industrial life cycle has been neglected in the recent upgrading debate. However, the implications of the industrial life cycle on location have been addressed both from a practitioner’s and from a researcher’s point of view. Table 12.1 gives a practitioner’s view. Its basic message is simple and straightforward: in the early phase of the life cycle, companies rely on a sophisticated environment. In the later phases, they move to locations where production factors, in particular real estate and labour, are cheap. Interestingly, the practitioner’s view, which is mostly based on experience and inductive reasoning, is confirmed by more systematic research that addresses the issue of locational quality from an innovation economics perspective. This is summarized in Table 12.2. What does this mean in terms of upgrading, both for companies and for locations? In terms of locations, the answer seems to be simple and, to some extent, discouraging. There is a great deal that can, and indeed should, be done to support the emergence of new industries. Yet, there is little that can be done, in particular on the part of ‘old’ locations, with respect to mature and declining industries. But also ‘new’ locations, such as greenfield sites4 in developing countries that cater to relocated plants in mature industries, do not have many options in terms of locational policy. The practitioner’s viewpoint is straightforward: minimize costs of infrastructure, real estate, labour and skills development. This is exactly what has been happening in locations that were successful in attracting greenfield investments in mature industries (Kanter, 1995). What about the experience of companies in mature industries that become involved in elaborate locational policy efforts? An example would be the involvement of Volkswagen in locational development and upgrading in the region around its main facility, Wolfsburg. This, however, is a somewhat special case. The company started as a state enterprise, and state government is still an important minority shareholder. This creates a form of shareholder pressure that is different from the usual pressure that primarily addresses financial returns. One would expect, therefore, that at least locations with emerging or growing industries are favourable places for locational policy. However, this expectation is based on an analysis which looks at economic factors; the scenario changes if we introduce political factors. This leads us to question appropriate governance patterns for locational policy. Basically, there are


Source: Pieper (1994), p. 32

Highly skilled workers Knowledge infrastructure Proximity to customers

Proximity to market (up- and downstream) Specialized workers Highly skilled workers Real estate

Growth phase

Requirements on locations across the industry life cycle

Start-up phase

Table 12.1

Cheap workers Low location cost Proximity to market

Maturity phase

Cheap workers Low location cost Little regulatory cost

Decline phase



High growth rates. Employees from established firms form spinoff firms Product innovations Products are standardized. Product differentiation have primacy. In many Process development aimed (fashion) and process cases production equip- at economies of scale development dominates ment is modified by the user


Source: Gelsing (1992), p. 128

Technological development

Agglomeration economies are high. Attraction point: innovative centres

Firms are attracted to least If early: relocation is cost sites (labour, land, retarded because market taxes, etc.) strategies are better implemented from old centres. If late: reorganization of industry to less unionized labour Proximity to competitors/ Firms operate on larger colleagues less important. in- and output markets. Proximity to producers of Internal division of labour equipment of some and level of information importance rises High. Minimum optimal Low. Markets are increasscale increases and spin-off ingly organized and becomes rare negotiated


Importance of proximity

Localization pattern Close to existing pools of high-skilled labour/ founders residence


A neo-Schumpeterian model of industrial development

Stage of industry parameter

Table 12.2

Product development suppressed by short term profit dispositions. Process developments are rare



Close down operations in old industrial regions. Modernized plants in new regions


Paradoxes and ironies of locational policy


two options: hierarchy and networks. Hierarchy is the traditional pattern of public governance. Government formulates and implements a policy after a certain amount of fact-finding and interaction with special interest groups. This may be an adequate pattern for areas such as environmental policy, where government should take care of the common good and protect its citizens. However, it is not an adequate pattern when it comes to activities such as industrial policy (at the national level) and locational policy (at the local level). The argument put forward by neoliberal economists, namely that there is no reason to assume that government coordination is superior to market coordination when it comes to business promotion, is convincing. However, this does not mean that government has to limit itself to facilitating markets, since there are cases where market failure is persistent (Meyer-Stamer, 2001). Experience in industrialized countries shows that government actors are involved in policy networks which also include various non-governmental actors, and which may be quite effective in formulating and implementing sectoral policy (Messner, 1997). Policy networks are rarely designed and created intentionally. Instead, they emerge as a spontaneous response to governance requirements, for example market failures which block rapid adjustment processes in old industrial regions. So why is the involvement of policy networks problematic in locational policy? The problem is that functioning policy networks involve collective actors, rather than a large number of individuals or companies. Effective policy networks for locational policy require effective business organizations. This is where the difficulty arises and it is related to the industrial life cycle. Entrepreneurs in emerging industries feel little pressure to organize themselves and look for political support. Similarly, industries and firms that are growing rapidly do not feel the need to fight for their interests. They are so busy managing rapid growth that they do not have time for such activities. As a result, there is no immediate logic for collective action under such circumstances. Although policy makers may strive to support such industries in order to defend common interests they are still faced with the difficulty of establishing adequate communication links with new firms because they have not (yet) organized themselves. At the same time, old industries tend to be well organized for the simple reason that there is a logic of collective action, namely to lobby for defensive measures to slow down the adjustment process. Therefore, for government policy makers it is easy to tap into policy networks with mature and declining industries. However, as I have argued above, this kind of industry is not very interested in locational policy. This is the life cycle paradox of locational policy: industries which might be interested in locational policy are unlikely to be well organized, therefore,


Local enterprises in the global economy

it is difficult to establish the policy networks required for policy formulation. Old industries are well organized, but they are not interested in locational policy.

5. VALUE CHAINS AND THE IRONY OF UPGRADING So far, I have looked at the intrinsic problems of locational policy, without looking at the specific repercussions of globalization on locations. Global competitive pressure is the main reason why local stakeholders start locational policy initiatives. However, as we will see, it is also a major complicating factor for locational policy. This section addresses some implications concerning globalization of product markets for locations. More specifically, it addresses the implications of global value chains for locational policy making. Before focusing on the interaction between locations and value chains, it is important to note that there are basically two different constellations: locations may or may not be an important issue for those coordinating a given value chain. Increasingly, the coordinators of value chains are global buyers that systematically scan the globe for potential suppliers. If the location is not yet a priority for the buyer then the conditions for locational policy are fairly reasonable. This is a typical scenario in many emerging locations in developing countries (the argument developed in this section is less relevant for industrialized countries). A great deal of SME promotion is based on this scenario. The objective here is to increase the competence of local firms in terms of production, quality, technology, human resources and financial management, so that they can manufacture products of acceptable quality at competitive prices in the hope that they may attract recognition from global buyers. ISO 9000 seems to play an important role in this respect as it indicates to global buyers that a local firm has the potential to become a supplier (Quadros, 2002; Nadvi and Wältring, 2002). Prior to detection from global buyers, upgrading means learning within local markets or elsewhere to improve competitiveness in order to be noticed by value chain scouts. Government may take an important role, for instance pursuing a carrot-and-stick approach, that is both pushing and pressuring firms whilst supporting them, including dedicated efforts to raise their profile (missions abroad, presence at fairs, joint marketing and so on). What are the consequences of raising their profile and attracting orders? The most likely and immediate consequence is rapid growth. Managing rapid growth is extremely time consuming for firms. As a result, there is

Paradoxes and ironies of locational policy


little time for interaction with government or other players which are not directly related to day-to-day business. If orders keep coming in, there is also little urgency for collective action. Constellations like this have been observed in the early growth phases of the footwear cluster in Sinos Valley, Brazil (Bazan and Schmitz, 1997), and the furniture cluster in São Bento do Sul, Brazil (Meyer-Stamer, 1998). Another important aspect is that once they have raised their profile, the chain governor (that is usually a global buyer) is unlikely to expect local government to play an active role in day-to-day management. Instead, they expect government to remove obstacles that stand in the way of day-to-day business (red tape, deficient infrastructure). For the chain governor, shaping the chain is a crucial element of their effort to create a competitive advantage, and it is unlikely that they would want to share their concepts and strategies with other players, particularly not with local governments in the places where suppliers are located. The chain governor becomes the main source of information, training, advice and so on. Local suppliers prioritize communication with their new big customer. Government officials find themselves increasingly isolated from the communication loop, relying on second-hand information on the evolution of the chain. But what about private governance, and local collective action within the business community in particular? For local companies, becoming part of a global value chain may imply four different scenarios: a.



Product and process upgrading. Often this mainly concerns running to stand still. It implies joint upgrading with other participants in the value chain. This is a challenging task that involves only a limited risk. It is in everybody’s interest including the global buyer, who is also interested in fundamental activities, to improve locational quality, such as infrastructure and vocational training institutions. Strategic functional upgrading. This entails taking over functions previously handled by other companies, usually from other locations within the same value chain. This is a more risky option, as the to-bereplaced competitors will probably fight back. Global buyers may be expected to tolerate this (as long it does not threaten their own core competence), as fierce rivalry between locations strengthens their bargaining position vis-à-vis each of them. Improve their competitiveness in order to move to a different value chain. In a given sector, there are various value chains that cater for different segments of the consumer market. As long as margins are higher in more sophisticated or in differentiated markets, it may be tempting to switch from one value chain to another that serves highermargin markets. This involves the risk of falling between a rock and a


Local enterprises in the global economy

hard place; the old buyer may anticipate this and move to a different source, whereas the prospective new buyer might fail to close the deal. d. Attempting to take over the value chain or trying to take the main power position in the value chain. This is clearly the most challenging option. It may be viable in cases where the buyers’ power position is limited; the ceramic tile industry is case in point (Meyer-Stamer et al., 2001). What is the role of private sector collective action in these different scenarios? In all four cases there are strong incentives against collective action. In the case of scenario (a), one might argue for a positive-sum game5 which might persuade firms to go for collective efforts to upgrade, particularly in a situation where all the firms in the location are suffering from superior competition from another location. For example, the case of the Sinos Valley footwear cluster vs. producers in China (Schmitz, 1995). However, it is more likely that firms will think in terms of a zero-sum game, that is a firm perceives the loss of local competitors as its own gain. This is particularly likely in places where collective action has suffered from early export growth. In the case of scenarios (b) to (d), collective action is even less likely. It is highly unlikely that in a given location all company decision makers will display the same level of risk-friendliness; probably the most important risk is to be abandoned by current buyers. If the degree of risk-friendliness diverges, one might expect that some decision makers would find all of these scenarios plausible, whereas many others would not. One would expect that particularly risk-friendly, strategically oriented firms would go for one of these options, thus creating a split among business executives within the location. What is the role of government in these different scenarios? Basically, it would try not to stand in the way (that is reduce transaction costs) and to excel in the provision of basic and advanced factors. The case study of Halder (2002) on the surgical instruments cluster in Tuttlingen, Germany, illustrates this point. It seems improbable that government can play a major role, in particular with respect to scenarios (b) to (d). It is unlikely that government has the in-depth, up-to-date knowledge that is necessary to assess the viability of these scenarios. The most likely contribution of government may be to contract a specialized consultancy firm to support local businesses and associations in their decision-making process. So this is the irony of upgrading and entry into value chains: government can play a very important role in locational policy by helping local firms become so competitive that they are subcontracted by global buyers. However, as the firms get involved in the value chain, the options in terms

Paradoxes and ironies of locational policy


of government’s role in locational policy declines substantially, and it can be expected that collective action in the private sector will suffer as well.6



This section addresses the implications of globalizing companies for locational policy. The globalization of companies may occur by a local company establishing branch plants, taking over companies in other countries, or the takeover of local companies by foreign investors. I argue that locational policy makers are confronted with a paradox: globalizing companies are increasingly demanding when it comes to locational quality, but they show a declining propensity to get actively involved in locational policy. Increasing demands in terms of locational quality apply to various locational factors: high-quality and low-cost infrastructure, swift execution of licensing and permit processes, low tax burden, substantial effort in workers’ training, and so on. Companies discussed in this section sell a large part of their output elsewhere. This section does not address local companies such as developers or utilities which will often take a very active role in locational policy. For these companies, locational upgrading is a key element of their business strategy which aims at keeping and attracting customers. This discussion on the relationship between location and competitiveness focuses on industrial manufacturers and service firms that are supplying global markets. The declining propensity of companies, in particular large, multilocation companies, to get involved in locational policy has been documented in a number of case studies (Heying, 1997; Dörre, 1999). Yet, why would one expect that such companies become involved in these activities in the first place? This suggestion is based on inductive reasoning: despite globalization, companies are not usually footloose, and they do not pick locations randomly. Space and location continue to be relevant for globalized manufacturing and service companies (Porter, 2000). Companies seek specific locational qualities. This implies that companies have an interest in the creation and improvement of locational qualities, therefore, they may be willing to take an active role in this respect. Let us now take a closer look at the connection between company and location. Companies are located in a given place for four possible reasons: a. b.

historical accidents (that is because they were founded there or because they acquired a firm which happened to be located there); they are seeking proximity to other firms. A typical example would be an IT company that sets up an affiliate in Silicon Valley;


Local enterprises in the global economy

c. they strive to build up a presence in proximity to dynamic markets; or, d. they are seeking other locational factors, such as natural resources or cheap labour. For example, Renschler (1995) gives a detailed account of the criteria Daimler-Benz applied when it scanned possible locations for its SUV factory in the US. These motives do not necessarily mean that a company deliberately contributes to the improvement of locational quality. In particular, cases (b) to (d) are more about receiving benefits but they do not contribute towards them. Instead, companies will often contribute inadvertently to locational quality improvements, that is while enhancing their own competitiveness they create positive externalities. Conversely, one of the main obstacles to getting companies involved in a locational strategy is the problem of freeriding, that is companies assume that collective action renders too little outcome which they can appropriate for themselves and too much externality which benefits local competitors. There are two types of location where one would expect this problem to be less relevant: ●

hub-and-spoke-clusters (Markusen, 1996), which are essentially dominated by one company (for example Toyota City or Wolfsburg), where the ‘hub’ company can control the external effects; or, very cohesive clusters, where free-riding is minimized through social control. However, this phenomenon is becoming rare as local firms in cohesive clusters get involved in international value chains, and external firms enter into local clusters to benefit from specific locational qualities (Grabher, 1993). Strong cluster cohesion is probably more closely related to the life cycle of companies and their industry rather than to location.7

This is not to say that companies do not do anything to the benefit of their location. What they usually opt for is sponsoring – of museums, theatres, other cultural events, sports, and so on. For a large corporation, sponsoring has an unbeatable cost–benefit ratio, that is the cost is usually relatively low, whereas the visibility is high, and, moreover, the company can point at such sponsoring activities whenever somebody criticizes it for lack of local involvement. Furthermore, the cost–benefit ratio is much more predictable than in cases where companies become involved in locational policy. Understanding the structure of local policy networks, and participating in them, will involve substantial input in terms of time (that is high transaction and opportunity costs), whereas the visibility of the outcome is unpredictable. How is a company that is driven by the rationale of shareholder

Paradoxes and ironies of locational policy


value maximization supposed to justify this kind of involvement? This logic becomes even more convincing if one considers that companies tend to run operations in many different locations, and that they are likely to have an exit option that may be particularly attractive in the case of simple screwdriver operations.8



Over the past few years there has been an increasing degree of excitement about the potential benefits of local economic development efforts, both in industrialized countries and developing countries.9 The purpose of this chapter is not to dampen this excitement. However, it is important to come to a realistic assessment of the latitude of locational policy, and the typical obstacles that every locational policy initiative will have to deal with. If such an initiative does not make the swift progress that local actors had hoped, it is quite likely that the issues addressed in this chapter are to some extent responsible. Local governments may be tempted or obliged to formulate a local upgrading strategy, that is a locational policy. They frequently choose strategic locational policy, that is an effort to formulate an overall, long-term strategy for upgrading local companies and the location. This is an enormous challenge. Aligning all the relevant local players, agreeing on a problem definition, negotiating an action plan and distributing and coordinating tasks and responsibilities among various players is even difficult for those locations where local stakeholders have decades-long experience in local economic development. It is not surprising, therefore, that strategic plans for local development often do not get beyond the stage of printing a plan that will never be implemented. However, there are less challenging options. Local government may try to implement a generic locational policy that targets obstacles to business created by the various layers of government. In fact, government is often one of the biggest problems faced by companies, since it frequently introduces questionable regulations and constrains businesses with a multitude of permits and registration requirements. Streamlining local governments’ interaction with businesses in order to minimize transaction costs can make an important contribution to companies’ competitiveness, and may substantially enhance the locational quality and competitiveness. A third option is reflexive locational policy. It is an approach that we have observed in the ceramic tile cluster in Castellón, Spain (Meyer-Stamer et al., 2001). Local players did not attempt to go for the strategic approach. Instead, they participated in an exercise that involved a series of studies and


Local enterprises in the global economy

workshops, organized by the leading regional bank that took place over the duration of several months. The purpose of the exercise was to improve the information base of all the players in the cluster, so that each actor could develop their own strategy. It is a good example of a semi-public actor rectifying a market failure, because individual companies would not have had access to such high-quality information, let alone the opportunities for reflection which emerged through discussion both inside and outside the workshops. The concept of reflexive locational policy takes us back to the discussion on upgrading. If players in a given location succeed in setting up institutions (perhaps in the shape of an organization, but more likely by establishing routines) that foster constant reflection then this will contribute to the companies’ upgrading efforts. It will also establish an element of locational upgrading. It addresses issues such as the innovativeness of the local milieu, the local innovation system, or the knowledge-intensity of local development from an unusual angle. The standard approach is concerned with product and process innovation. Reflexive locational policy is about innovation in local governance, and upgrading the information base of all economic players in the location. In fact, for public-sector players involved in locational policy, it solves the dilemma of upgrading to some extent, as it is not always up to them to decide which route companies are to follow in their effort to establish a competitive advantage. Local government may be ignorant about businesses’ strategy decisions and yet still pursue a useful generic locational policy. This may even include the provision of sophisticated locational factors, such as specialized training, R&D or finance, since there is no a priori reason why such factors have to be supplied by public agencies, especially in a place where reflexive locational policy works. As a result, the uncertainty for private providers of such services is substantially reduced. It is important to point out that the three types of locational policy, that is generic, reflexive and strategic, cannot easily be ranked in terms of relevance or effectiveness. Furthermore, there is not necessarily a ladder, that is a sequence whereby local stakeholders go through a learning process that starts with a generic locational policy and leads, via a reflexive locational policy, to a strategic locational policy. The type of locational policy that is most adequate for a given territory is highly location- and life-cycle-dependent. A location that is thriving due to dynamic firms and market processes is probably well served with a generic locational policy. Actors in a location with a highly diversified economic structure will probably find it difficult to start a reflexive locational policy. This kind of approach appears more promising in locations with a somewhat narrower specialization profile, in particular industrial clusters. Strategic locational policy is a very difficult

Paradoxes and ironies of locational policy


approach, which is most likely to happen in places suffering from profound crises. Every location ought to pursue a generic locational policy. Some may find it useful and viable to pursue a reflexive locational policy, while very few will succeed in formulating and implementing a strategic locational policy.

NOTES 1. This chapter draws on research from the joint IDS/INEF project ‘The interaction of global and local governance: implications for industrial upgrading’, as well as research conducted between 1998–2001 in the project ‘North Rhine–Westphalia: structural economic change and regional and locational policy’. The chapter also draws on my experience as a consultant on local economic development in some municipalities in Santa Catarina and in the Greater ABC region, both in Brazil, in Mpumalanga province, South Africa, and in several municipalities in Argentina and Thailand. 2. For the US see Bradshaw and Blakely (1999), for Germany see Meyer-Stamer (1999), and see Helmsing (2001) for developing countries. 3. This explains the otherwise somewhat discouraging story of Nadvi and Kazmi (2001), who find that although surgical instruments firms in Sialkot, Pakistan, have gone through tremendous upgrading during recent years, their position in the global value chain has not changed. Likewise, Gibbon’s account of failed efforts by garment manufacturers in Mauritius to integrate into downstream activities in the value chain shows that these firms are only just competent enough for the ‘running to stand still’ scenario, but not sufficiently competent for functional upgrading (Gibbon, 2000, 42). 4. This term refers to new sites, built on metaphorical ‘green fields’. 5. Games where both parties can gain, as opposed to zero-sum games, where one party’s gain is another’s loss. 6. This would be an alternative interpretation of the case presented by Leite (2002) about different records of locational policy in the ABC region of Brazil with respect to the car industry and the plastics industry. According to her presentation of the plastics industry, it appears to be an industry that has not yet been detected by external buyers. 7. This is an important point made by Bazan and Schmitz (1997). 8. This refers to basic assembly plants where parts/components are attached to each other (with a screwdriver). 9. For industrialized countries see European Commission (1998), OECD (2000). Regarding developing countries see Helmsing (2001). An overview of organizations involved in LED work is available at www.meyer-stamer.de/led-links.html.

REFERENCES Aghón, G., F. Alburquerque and P. Cortés (2001), Desarrollo Económico Local y Descentralización en América Latina: Un Análises Comparativo, Santiago: CEPAL/GTZ. Bazan, L. and H. Schmitz (1997), ‘Social capital and export growth: an industrial community in Southern Brazil’, IDS Discussion Paper 361, Brighton: Institute of Development Studies. Bonnin, A.R. (2002), ‘The fashion industry in Galicia: understanding the “Zara” phenomenon’, European Planning Studies, 10 (4), 519–27.


Local enterprises in the global economy

Bradshaw, T.K. and E.J. Blakely (1999), ‘What are “third-wave” state economic development efforts? From incentives to industrial policy’, Economic Development Quarterly, 13 (3), 229–44. Capell, K. (2002), ‘Hip H&M’. Business Week, 11 November. Cheshire, P.C. (2001), ‘Territorial competition: lessons for (innovation) policy’, paper prepared for International Workshop on ‘Innovation clusters and interregional competition’, Kiel, 12–13 November. Dörre, K. (1999), ‘Local heroes. Globalisierung, Participation und Mikrosoziale Regulation’, in Gerhard Fuchs, Gerhard Krauss and Hans-Georg Wolf (eds), Die Bindungen der Globalisierung. Interorganizationsbeziehungen im regionalen und globalen Wirtschaftsraum, Marburg: Metropolis, pp. 92–135. Enright, M. J. (2000), ‘Survey of the characterization of regional clusters: initial results’, Hong Kong: University of Hong Kong. European Commission (1998), ‘Second report on local development and employment initiatives. The era of tailor-made jobs’, Brussels. Gelsing, L. (1992), ‘Innovation and the development of industrial networks’, in B.A. Lundvall (ed.), National Systems of Innovation. Towards a Theory of Innovation and Interactive Learning, London: Pinter Publishers, pp. 116–28. Gereffi, G. (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48, 37–70. Gibbon, P. (2000), ‘Back to the basics through delocalisation: the Mauritian garment industry at the end of the twentieth century’, CDR Working Paper 00.7, Copenhagen: Centre for Development Research. Grabher, G. (1993), ‘The weakness of strong ties: the lock-in of regional development in the Ruhr area’, in G. Grabher (ed.), The embedded firm. On the socioeconomics of industrial networks, London, New York: Routledge, pp. 255–77. Halder, G. (2002), ‘How does globalisation affect local production and knowledge systems? The surgical instrument cluster of Tuttlingen’, INEF Report 57, Duisburg: Institut für Entwicklung und Frieden. Helmsing, B. (2001), ‘Local economic development. New generations of actors, policies and instruments’, A summary report prepared for the UNCDF symposium on ‘Decentralization and local governance in Africa’. The Hague: Institute of Social Studies. Heying, C.H. (1997), ‘Civic elites and corporate delocalization. An alternative explanation for declining civic engagement’, American Behavioral Scientist, 40 (5), 657–68. Hollbach-Grömig, B. (1996), Kommunale Wirtschaftsförderung in den 90er Jahren: Ergebnisse einer Umfrage, Berlin: Deutsches Institut für Urbanistik. Humphrey, J. and H. Schmitz (2001), ‘Governance in global value chains’, IDS Bulletin, 32 (3), 19–29. ILO, UNOPS, EURADA and Cooperazione Italiana (undated), ‘Local economic development agencies. International co-operation for human development, democratic economies and poverty reduction’. Jonas, A.E.G. and K.G. Ward (2001), ‘City-regionalisms: some critical reflections on transatlantic urban policy convergence’, Economic Geography Research Group, Working Paper Series, 01/01. Kanter, Moss, R. (1995), ‘Thriving locally in the global economy’, Harvard Business Review, 73 (5), 151–160. Kaplinsky, R. (1998), ‘Globalisation, industrialisation and sustainable growth: the

Paradoxes and ironies of locational policy


pursuit of the nth rent’, IDS Discussion Paper 365, Brighton: Institute of Development Studies. Kaplinsky, R. (2000), ‘Spreading the gains from globalisation: what can be learned from value chain analysis?’, IDS Working Paper 110, Brighton: Institute of Development Studies. Küpper, Ingo U. (2000), Standortpolitik für die Städte. Kommunale Wirtschafts- und Beschäftigungsförderung in Deutschland, Cologne: Deutscher Städtetag. Lee-Young, J. and M. Barnett (2001), ‘Furiously fast fashions’, The Industry Standard Magazine, 11 June. Leite, M. (2002), ‘The struggle to develop regional industry policy: the role of the plastics and auto sectors in the regional chamber of ABC, São Paulo’, Brighton: Institute for Development Studies. Magretta, J. (1998), ‘Fast, global and entrepreneurial: supply chain management, Hong Kong style. An interview with Victor Fung’, Harvard Business Review, 76 (5), 103–14. Markusen, A. (1996), ‘Sticky places in slippery space: a typology of industrial districts’, Economic Geography, pp. 293–313. Menzel, U. (1995), ‘Die postindustrielle Revolution. Tertiarisierung und Entstofflichung der postmodernen Ökonomie’, Entwicklung und Zusammenarbeit, 36 (4), 100–104. Messner, D. (1997), The Network Society. Economic Development and International Competitiveness as Problems of Social Governance, London, Portland: Frank Cass. Meyer-Stamer, J. (1998), ‘Avoiding collective efficiency. Growth and crisis in the furniture cluster in São Bento do Sul (Santa Catarina, Brazil)’, Duisburg (mimeo, available at www.meyer-stamer.de). Meyer-Stamer, J. (1999), ‘Lokale und regionale Standortpolitik – Konzepte und Instrumente jenseits von Industriepolitik und traditioneller Wirtschaftsförderung’, INEF Report 39, Duisburg: Institut für Entwicklung und Frieden. Meyer-Stamer, J. (2000), ‘Estrategias de desarrollo territorial basadas en el concepto de competitividad sistémica’, El Mercado de Valores, 60 (9), 48–60. Meyer-Stamer, J. (2001), ‘Was ist Meso? Systemische Wettbewerbsfähigkeit: Analyseraster, Benchmarking-Tool und Handlungsrahmen’, INEF Report 55, Duisburg: Institut für Entwicklung und Frieden. Meyer-Stamer, Jörg, Claudio Maggi and Silene Seibel (2001), ‘Improving upon nature. Creating competitive advantage in ceramic tile clusters in Italy, Spain, and Brazil’, INEF Report 54, Duisburg: Institut für Entwicklung und Frieden. Nadvi, K. and S. Kazmi (2001), ‘Global standards and local responses’, paper for Workshop on ‘The impact of global and local governance on industrial upgrading’, Brighton, 13–17 February. Nadvi, K., and F. Wältring (2002), ‘Making sense of global standards’, INEF Report 58, Duisburg: Institut für Entwicklung und Frieden. OECD, (2000), ‘Best practices in local development’, Paris. OECD, (2001), ‘Local partnerships for better governance’, Paris. Pieper, M. (1994), Das interregionale Standortwahlverhalten der Industrie in Deutschland – Konsquenzen für das kommunale Standortmarketing’, Göttingen: Verlag Otto Schwartz. Porter, M.E. (1996), ‘What is Strategy?’, Harvard Business Review, 74 (6), 61–78. Porter, M.E. (2000), ‘Location, competition, and economic development: local clusters in a global economy’, Economic Development Quarterly, 14 (1), 15–34.


Local enterprises in the global economy

Quadros, R. (2002), ‘Global quality standards, chain governance and the technological upgrading of Brazilian auto-components producers’, IDS Working Paper 156, Brighton: Institute for Development Studies. Rabellotti, R. (2001), ‘The effect of globalisation on industrial districts in Italy: the case of Brenta’, IDS Working Paper 144, Brighton: Institute of Development Studies. Raines, P. (2000), ‘Euro-cluster: final report’, Glasgow: University of Strathclyde. Renschler, A. (1995), ‘Standortplanung für Mercedes-Benz in den USA’, in H. Gassert and Péter Horváth (eds), Den Standort richtig wählen. Erfolgsbeispiele für internationale Standortentscheidungen, Stuttgart: Schäffer-Poeschel, pp. 37–54. Schmitz, H. (1995), ‘Small shoemakers and Fordist giants: tale of a supercluster’, World Development, 23 (1), 9–28. Wallis, A.D. (1996), ‘Regions in action: crafting regional governance under the challenge of global competitiveness’, National Civic Review, 85 (2), 15–24. Widmaier, U. (1998), ‘Der deutsche Maschinenbau im Umbruch?’, WSIMitteilungen, 51 (2), 92–101.

13. Chain governance and upgrading: taking stock John Humphrey and Hubert Schmitz 1.


The upgrading prospects of clusters of local enterprises differ according to the type of global value chain they feed into. This is a central theme of the book running through most of its chapters. This final chapter pulls together the evidence and draws out the strengths and limitations of the chain approach to upgrading.1 Our focus on chain governance does not imply that other factors are not important for upgrading. Our proposition is that the upgrading opportunities of local enterprises are often structured by the relationships in global value chains. If this is so, then the chain perspective has a profound effect on the order of questions asked in the debate on upgrading in clusters, switching the focus away from the internal organization of clusters towards the linkages with the global economy, in particular with global buyers. The chapter is structured as follows: section 2 identifies the main positions found in the literature on the connection between chain governance and upgrading. It then provides typologies of chain governance and upgrading. These are necessary for moving the debate forward and formulating hypotheses for new research. Section 3 sets out the propositions with which we started the research project on which this book is based. Sections 4 to 8 then pull together the evidence, drawing mainly on the chapters in this book but complementing them with other recent studies. Section 9 discusses the limitation of the chain approach and the final section brings together the main findings and new questions that arise.



In order to explore the connection between chain governance and upgrading it is useful to start with the view of Gereffi who has been the leading author in this field. For Gereffi (1999), the key feature of the current phase 349


Local enterprises in the global economy

of globalization is ‘the functional integration and coordination of internationally dispersed activities’ (Gereffi, 1999, 41). With regard to developing countries, he put forward two central propositions. First, their trade with developed countries is increasingly coordinated by global buyers. Second, developing country producers that integrate in such buyer-driven chains can expect rapid upgrading. These conclusions were informed by his research on the garment sector. Our earlier research confirmed the first proposition. Research on the footwear sector concluded that ‘an increasing number of countries engage in contract manufacturing for a decreasing number of global buyers’ (Schmitz and Knorringa, 2000, 200). Research on the horticultural sector that focused on the fresh vegetable exports from Africa to the UK, established the high degree of control exercised by the major buyers, the UK supermarkets (Dolan and Humphrey, 2000). However, it would be wrong to generalize from these cases as shown by, for example, Tewari’s (1999) study of an Indian garment cluster or Mitsuhashi’s (2001) study of Thai furniture clusters. Exports from these clusters are not coordinated by powerful global buyers. Similarly, the work of Gibbon (2001) and Ponte (2002) on agricultural exports shows that global traders do not intervene in operations further back along the chain in the way seen in horticulture. A more differentiated view on chain governance is therefore necessary. In order to develop our approach, we took two steps. First, we introduced typologies of chain governance and upgrading. The idea was to have tools of analysis with which one can explore whether certain types of chains are associated with particular types of upgrading. Second, we sought to identify the circumstances in which global buyers would seek to govern (or explicitly coordinate) value chains, expecting that this would help us to understand why buyers foster the upgrading of suppliers in some cases and not others. The distinctions and arguments were developed in Humphrey and Schmitz (2000; 2002).2 Here we limit ourselves to the key points. 2.1


Our concern is with inter-firm relationships in the context of repeat transactions in international trade. We distinguish between four types of relationships in global value chains: ●

Arm’s length market relations. Buyer and supplier do not develop close relationships because the product is standard or easily customized. This implies that buyers’ requirements can be met by a range of firms and that the switching costs for both parties are low. Networks. Firms cooperate in a more information-intensive relation-

Chain governance and upgrading


ship, frequently dividing essential value chain competences between them.3 The interaction is coordinated and the relationship is characterized by reciprocal dependence. The buyer may specify certain product performance standards or process standards to be attained, but would be confident that the supplier can meet them. Quasi-hierarchy. One firm exercises a high degree of control over other firms in the chain, frequently specifying the characteristics of the product to be produced, and sometimes specifying the processes to be followed and the control mechanisms to be enforced. This level of control can arise not only from the lead firm’s role in defining the product, but also from the buyer’s perceived risk of losses from the suppliers’ performance failures. In other words, there are some doubts about the competence of suppliers in the chain. The lead firm may exercise control not only over its direct suppliers but also further along the chain.4 Hierarchy. The lead firm takes direct ownership of some operations in the chain.5 The case of the intra-firm trade between transnational companies and their subsidiaries falls into this category.

The chapters in this book deal primarily with cases that fall between market and hierarchy, and so does this final chapter. Relationships between parent companies and their subsidiaries are not discussed here – the literature on this relationship is substantial but largely inconclusive with regard to the upgrading prospects of subsidiaries (see, for example, Ariffin and Bell, 1999). Market-based relationships are discussed both because they remain important in some cases and because the contrast provides insights for understanding the intermediate cases.6 We started the research using the two cases lying between market and hierarchy set out above, that is network and quasi-hierarchy, but on reviewing the case material we found that we needed a further category: modular production networks. This concept was developed by Sturgeon (2002). It captures the following configuration. Buyers purchase customized products that they design, and therefore must exchange information (on product specification, scheduling, and so on) with the suppliers. However, the interaction between buyer and supplier is made less complex because the information on products and processes can be codified in technical norms, and the suppliers use generic machinery that can be used for various customers. The suppliers also have the necessary skills for the tasks they undertake, which reduces the buyers’ need for monitoring. As a result, even though products are specific to each customer, the level of transactional dependence is low on both sides. This is why Sturgeon refers to this case as modular networks. Different suppliers can be inserted or removed from the


Local enterprises in the global economy

value chain.7 Buyers have various suppliers across their product range, and suppliers work for various customers. The reason for adding this category will become apparent in the course of the chapter. The above distinctions help us to ask whether some types of chains offer local producers better upgrading prospects than others. However, in order to address this question it is necessary to distinguish between different types of upgrading. We distinguish four upgrading types: ●

Process upgrading: transforming inputs into outputs more efficiently by reorganizing the production process or introducing superior technology. Product upgrading: moving into more sophisticated product lines (which can be defined in terms of increased unit values). Functional upgrading: acquiring new functions in the chain (or abandoning existing functions) to increase the overall skill content of activities. The functional upgrading route frequently discussed in the literature is the transition from assembly to OEM (original equipment manufacture) to ODM (own-design manufacture) to OBM (own-brand manufacture). Inter-sectoral upgrading: using the knowledge acquired in particular chain functions to move into different sectors.8

This fourfold distinction is finding rapid acceptance in the international debate (as shown for example in UNIDO, 2002). These categories are, however, not without problems, as stressed by Meyer-Stamer in Chapter 12 of this volume. Therefore, we will re-examine our chain and upgrading categories at the end of this chapter. 2.2

Reasons for Chain Governance

The tool kit presented above could be used for analysing many situations. The one we are particularly interested in concerns the relationships between buyers in developed countries and suppliers in developing countries. In such chains, global buyers often set and/or enforce the parameters under which others in the chain operate. In other words, relationships are characterized by quasi-hierarchy. Our question is what upgrading implications arise from such chain governance. Does it help or hinder the upgrading of local producers? In what circumstances can we expect one outcome or the other? Answering this question requires posing a prior question. Why would global buyers go to the trouble and expense of monitoring and supervising international supply chains? No firm will incur the expense of developing

Chain governance and upgrading


arrangements with specific suppliers in order to purchase products that the market freely provides. We suggest that quasi-hierarchical relations are likely to emerge when the buyer seeks to define the product and/or the buyer is exposed to considerable risk if the supplier fails to perform: ●

Product definition.9 Buyer specification of the product is most likely to arise when the buyer has a better understanding of the demands of the market than the supplier. It often occurs when the buyer uses product differentiation as part of its competitive strategy. The more the buyer focuses on product differentiation, for example through design and branding, the greater the need to provide the supplier with precise product specification and to monitor that these specifications are met. Losses from supplier failure. The increasing importance of non-price competition based on such factors as quality, response time and reliability of delivery, together with increasing concerns about safety and standards, means that buyers (both retailers and manufacturers) in developed countries have become more vulnerable to the shortcomings in their suppliers’ performance.

This does not mean that price has ceased to be important. On the contrary, pressure on prices has been relentless especially for products which can be sourced from developing countries (Kaplinsky, 1998), and this leads global buyers10 to look constantly for lower-cost production sites. Buyers integrate new producers into value chains, exposing these producers to the demands of more sophisticated markets. This creates the problem of the ‘latecomer firm’, defined by Hobday (1995, 34) as facing two disadvantages. It is ‘dislocated from the mainstream international markets it wishes to supply’ and ‘dislocated from the main international sources of technology and R&D. These dislocations create the need for quasi-hierarchy. This point deserves elaboration. Keesing and Lall (1992) argue that when producers in developing countries are expected to meet requirements that frequently do not (yet) apply to their domestic markets, a gap arises between the capabilities required for the domestic market and those required for the export market. The need for quasi-hierarchical governance follows for three reasons. First, the new supplier’s lack of knowledge of overseas markets means that the buyer has to be involved in product design. Second, close monitoring and control may be required to ensure that products and processes meet the required standards. Third, if the gap has to be closed quickly, buyers will need to invest in a few selected suppliers and help them to upgrade. In order to reap the benefits of this investment the buyers have an interest in making the suppliers transactionally dependent on


Local enterprises in the global economy

them. The low cost of the new supply base provides a return on the investment. However, as the capabilities of local suppliers improve and diffuse, the need for monitoring and control declines. In summary, quasi-hierarchical coordination is costly and develops when the buyer perceives a high risk of supplier failure. Risks are high when the competence differential is high. By implication, where this differential does not exist, or where it disappears over time, buyer and supplier develop other relationships. Therefore, the direction of causation is not solely from chain governance to upgrading. As suppliers upgrade, governance patterns can change.



Equipped with the analytical distinctions set out above and with an understanding of why chain governance arises, we derive a number of propositions on chain governance and upgrading, and on differences between developed and developing countries. We regard such comparison as essential for identifying the specific implications that arise for developing country firms from participating in the global economy. This section presents the propositions formulated at the start of the research on which this book is based. Later sections of this chapter will then show to what extent these propositions are confirmed by empirical analysis. Our propositions are organized according to the type of chain.11 ●

Proposition 1: In quasi-hierarchical chains, developing country producers experience fast product and process upgrading but make little progress in functional upgrading.

Rationale: the risk of supplier failure leads global buyers to assist local producers in upgrading production capabilities but upgrading beyond production might clash with the core competence of the buyers. ●

Proposition 2: Local producers in developed country clusters do not operate in quasi-hierarchical chains.

Rationale: this follows from proposition 1. Local producers in developed countries are close to the market and generally possess high levels of competence. The risk of supplier failure is therefore low. ●

Proposition 3: Network-based chains support an open-ended upgrading path but local producers in developing countries rarely find themselves in such chains.

Chain governance and upgrading


Rationale: network-based chains where they do arise are characterized by intense knowledge-based interaction aimed at developing new products and processes. This is the rationale for the sharing and complementing of competences between firms. The required competence levels are rarely found in developing countries. ●

Proposition 4: In market-based chains, producers experience neither support for, nor blockages to, upgrading (from within the chain).

Rationale: since products can be obtained freely on the market, there is no need for buyers to invest in relationships with suppliers. Conversely producers are not tied to buyers, and obstacles to upgrading do not arise from within the chain. This set of propositions defined our starting point for investigating the connection between chain governance and upgrading. The remainder of the chapter will show to what extent these propositions are supported by empirical evidence. We will show that some propositions were supported, while others were not, and we will explain why not. The material drawn upon comes primarily from the previous chapters of this book which present case material from the footwear, computer, ceramic tile, surgical instrument and auto component sectors. Where useful, we draw on additional evidence from other recent research.

4. UPGRADING OF DEVELOPING COUNTRY PRODUCERS IN QUASI-HIERARCHICAL CHAINS The research conducted to examine the above propositions and presented in the previous chapters confirms the importance of quasi-hierarchy. This is clearest in the early export phase of the Brazilian shoe cluster (Chapter 5) and of the Taiwanese computer cluster (Chapter 9). Other cases are not so clear cut, and later sections will highlight the unexpected surprises. This section focuses on the upgrading experiences of developing country firms where they were integrated in quasi-hierarchical chains, and examines proposition 1. ●

In quasi-hierarchical chains, developing country producers experience fast product and process upgrading but make little progress in functional upgrading.

The conclusions on this proposition are important not just for the value chain debate, but also for the wider debate concerning the distribution of


Local enterprises in the global economy

gains from globalization. It therefore occupies a central place in this chapter. As noted above, our starting point is Gereffi’s optimistic view that producers entering quasi-hierarchical (in his terminology ‘buyer-driven’) chains have good prospects for upgrading within production and subsequently into design, marketing and branding. On the basis of his research on the garment chains he suggests that East Asian suppliers working for large US buyers were on an upgrading trajectory from OEM all the way to ODM and even OBM.12 Gereffi attributes this to ‘organizational succession’, a process by which manufacturers start producing for buyers catering for the low end of the market and then move up to buyers targeting more sophisticated market segments: ‘This succession of foreign buyers thus permitted manufacturers to upgrade their facilities as they met buyer demands for more sophisticated products’ (Gereffi, 1999: 53). In our terminology, not just product and process upgrading, but also functional upgrading seems to have resulted from integrating into quasi-hierarchical chains in East Asia. 4.1

Consensus on Process and Product Upgrading

How generalizable is this finding? Most authors would agree with Gereffi that local producers experience significant product and process upgrading.13 Local producers learn a great deal from global buyers about how to improve their production processes, attain consistency and high quality, and increase their speed of response to customer orders. This upgrading effect is particularly significant for local producers new to the global market. In this book, Bazan and Navas-Alemán (Chapter 5) confirm rapid process and product upgrading for Brazilian shoe producers exporting to the US and Europe; and Kishimoto (Chapter 9) underlines the contribution of foreign buyers in the early export phase of the Taiwanese computer producers. This upgrading effect is not automatic. First, it requires continuous investment by the local firms themselves in people, organization and equipment. This is underlined by all serious accounts of why the OEM strategy helped East Asian producers to break into advanced country markets.14 But the local producers’ own efforts are rarely enough. The foreign buyers are critical for accessing distant knowledge and markets. Second, buyers do not always provide support for this upgrading. Some accounts suggest that buyers present challenges to suppliers but do not provide support to help meet these challenges. For example Gibbon (2000) gives an account of the relentless pressure that foreign buyers put on the clothing manufacturers of Mauritius to improve their processes and prod-

Chain governance and upgrading


ucts. Similarly, Quadros (Chapter 10) stresses that while the customers of the Brazilian auto-parts producer continuously challenged their producers to lower prices and improve quality, they provided no assistance for this process. There could be two explanations for this lack of buyer support. One explanation relates to chain governance. As shown in Chapter 10, there is a low level of transactional dependence in this automotive chain, and this implies great uncertainty as to whether the customer can reap the benefits of investing in supplier capabilities. The second explanation is that these Brazilian producers are not newcomers. In the case of the Brazilian auto industry, a combination of easier access to imported auto-parts (following trade liberalization) and the rapid spread of quality assurance systems and competences meant that buyers could find competent suppliers. Instead of upgrading under-performing suppliers, they would abandon them and seek other suppliers elsewhere (Chapter 10). In other words, in discussing the buyers’ tutoring role we need to understand both the overall availability of competent suppliers and the power of buyers. In the case of the locallyowned Brazilian auto-part manufacturers analysed by Quadros, the customers were large transnational assemblers or first-tier suppliers. The parts manufacturers were largely small and medium-sized enterprises that the buyers could replace easily. This point seems critical and deserves elaborating with other sources. Hobday and Schmitz stress the importance of buyer support for new export manufacturers in the East Asian and Brazilian industry respectively. Under the early OEM deals, the foreign corporations frequently supplied training, technical specifications and advice on engineering and capital goods. The OEM system proved an enduring technological training school for latecomers in the NIE [newly industrializing economies], enabling hundreds of small firms to overcome barriers to entry (Hobday, 1995, 192). [The foreign buyers] studied the market which necessitated visiting shoe shops in the United States and Europe as well as international trade fairs. They . . . [set up] model shops in the Sinos Valley to produce samples. They inspected product quality and production schedules on site; they provided technical assistance; they organized the transport and payment arrangements (Schmitz, 1995, 14).

These efforts are justified by the cost advantages of opening up new supply locations. However, when competences in these new supply locations become more widespread, buyers have no need to invest in the capabilities of specific firms. Therefore, their support diminishes over time, as shown by Bazan and Navas-Alemán (2001) and Kishimoto (2002). There are other factors explaining differences in buyer contribution to process and product upgrading. For example, ex-manufacturers are more


Local enterprises in the global economy

capable tutors than buyers without production expertise; and suppliers of high quality products receive more tutoring than suppliers of low quality products. While these differences need to be recognized they do not challenge the emerging consensus that insertion into quasi-hierarchical chains provides a fast track to process and product upgrading. 4.2

Controversy on Functional Upgrading

There is no agreement, however, on whether such insertion provides a route to functional upgrading, that is occupying the design, branding and marketing functions in the chain. The most thorough analysis of the OEM → ODM → OBM transition is provided by Hobday. The conclusions are optimistic, particularly for the OEM → ODM transition: . . . by the late 1980s foreign buyers and TNCs had begun purchasing goods under so-called ODM, allowing local companies to exploit their design talents and thereby gain more of the value added. Sometimes the latecomers designed goods independently, using their own knowledge of the international market. In other cases, they worked closely with foreign buyers and TNCs. The emergence of ODM signified a new phase of latecomer technological progress, indicating that local firms had internalized much of the ability to understand market needs, then to design, develop and make electronic products for overseas markets. As with OEM, the ODM system allows the foreign buyer or TNC to brand and distribute the goods . . . enabling the latecomer to circumvent the need for heavy marketing investments (Hobday, 1995, 193).

Hobday’s model of latecomer export-led learning includes the transition to developing own brands and own overseas marketing but there is more evidence of firms reaching the ODM stage than attaining the OBM stage. This cautiously optimistic view is confirmed by Kishimoto’s research on the Taiwanese computer industry (Chapter 9). The common feature of these studies is that they consider producing for global buyers as a promising starting point for moving up the value chain. This optimistic view contrasts with another set of studies which has emphasized the obstacles which firms face in functional upgrading. These obstacles are of two types: buyer resistance and resource requirements. Research on global footwear chains suggests that local producers (in China, India and Brazil) encounter barriers to developing their design and marketing competence. They face obstacles because such upgrading encroaches on their buyers’ core competence (Schmitz and Knorringa, 2000). Bazan and Navas-Alemán (Chapter 5), in their more recent study of the Brazilian footwear manufacturers, confirm that even leading export manufacturers refrained for many years from making substantial invest-

Chain governance and upgrading


ment in design and marketing content to remain subordinated to their US buyer. They feared that advancing into design and marketing would upset their main buyer who accounted for 80 per cent of their output and close to 40 per cent of the cluster exports. Only after a long delay (of almost two decades), did they make inroads into own design, branding and marketing, prompted by rapidly declining profit margins in contract manufacturing for their main buyer. Gibbon (2000) found no functional upgrading in the tightly controlled garment chains leading from Mauritius via the Far East to the US. In the chains leading to the EU, a number of local producers did try to move up the value chain and develop their own brands but they had little success. In their study on the buyer-driven chain, which connects the Mexican cluster of Torreon to the United States, Bair and Gereffi (2001, 1895) conclude that local manufacturers of blue jeans were ‘generally confined to translating the buyer’s specifications into practical knowledge that is necessary for production. No manufacturer in Torreon markets its own apparel brands in the United States . . . and no Torreon producer of US brand is able to sell its branded output directly in Mexico’. Similarly, truncated upgrading can be observed in the East Asian electronics industry. A study by Chiu and Wong of the Hong Kong electronics industry concludes that ‘most OEM suppliers remained locked in low-end production’ (2002, 12), and that: The weakness of local suppliers in marketing and the tight control of overseas buyers in distribution are just two sides of the same coin. Underlying this business arrangement is such power asymmetry that a buyer’s approval is always prior to anything done on the part of a supplier, leaving most suppliers with few choices but to take buyers’ orders and sales forecast as the primary source of market information. . . . Information asymmetry of this kind helps explain why a local supplier is prone to get locked into the subcontractor role (Chiu and Wong, 2002, 11).

How can we explain this failure to upgrade beyond the sphere of production? One of the reasons seems to be buyer power. As stressed by Palpacuer (2000), the source of power in global value chains lies increasingly in nonproduction activities, notably in branding, marketing, product development and the coordination of inter-firm relations. The lead firms of such chains focus on and invest in these activities as they regard them as their core competence. One would therefore not expect these lead firms to share this core competence with their suppliers. Bazan and Navas-Alemán (2001) confirm that for a long time in the Brazilian footwear industry, local suppliers were discouraged from functional upgrading by their main US buyer. While the conflict between the ambitions of manufacturers and the


Local enterprises in the global economy

interests of buyers is obvious in marketing, it is not so clear cut in design. Some buyers consider design as part of their core competence, others do not.15 We cannot establish here whether such differences have systematic sectoral determinants or whether they are due to enterprise strategy. We need to record, however, that recent studies, notably Hobday (2001) and Kishimoto (2002), provide more examples of transitions to ODM than other forms of functional upgrading.16 Buyer power is not the only obstacle to functional upgrading. The investment required is substantial and entails risks. This is apparent from Bair and Gereffi’s (2001, 1895) study of Torreon, which notes that, ‘One company that we interviewed planned in the future to launch its own line of apparel in the US market, but the amount of capital necessary to promote and market a new brand make such endeavours risky’. While upgrading in the sphere of production is often possible in small steps, in particular where clustering facilitates specialization and the coordination of upgrading efforts at different points in the cluster itself,17 bigger steps are required for functional upgrading. The Mauritius clothing producers exporting to Europe found this out at their cost. In most cases they retreated from their initiatives because they had underestimated the expertise and financial resources required in launching their own brand. ‘They also underestimated the extent to which distinct knowledge and skills were required to operate even a small retail chain’ (Gibbon, 2000, 33). This is not a new recognition. Both (Lall, 1991) and Roberts and Tybout (1995) have stressed the marketing barriers facing export manufacturers in developing countries. Working for foreign buyers and accepting quasi-hierarchy is a tempting solution, even if it means low profit margins. 4.3


There is agreement that insertion into quasi-hierarchical chains provides latecomer firms with a fast track to process and product upgrading. There is no agreement, however, on the prospects for functional upgrading. Some authors regard the upgrading process as open-ended, others have identified blockages: the perceived power of the buyers and the discontinuous leap required to move from production to design, branding and marketing. The latter two have been particularly difficult whereas advances into design have been more common. Thus, two questions arise that need answering. First, why have firms in some clusters been able to overcome these more difficult barriers even though they started off in quasi-hierarchical chains? In other words, how have they been able to escape the lock-in that can arise from operating in such quasi-hierarchical chains? This question will be answered in later sections. Second, if barriers arise due to the quasi-hierarchical

Chain governance and upgrading


governance of the chain, do enterprises in different types of chains make more progress in terms of functional upgrading? An answer to this question will emerge in the course of later sections which deal with upgrading in network-based and market-based chains.

5. QUASI-HIERARCHICAL CHAINS AND FUNCTIONAL DOWNGRADING OF DEVELOPED COUNTRY FIRMS In trying to understand what difference global chain governance makes to economic development, we asked why buyers would seek to exercise chain governance, given that such governance is costly. Our reasoning was: first, chain governance is due to the risk of supplier failure; second, the risk of failure is higher in developing countries; and it is lower in developed countries because of the ‘latecomer firm’ issue; third, buyers govern chains when their suppliers cannot design suitable products themselves, and this is more likely when suppliers are remote from the markets they are serving. This, in a nutshell, is the origin of proposition 2 put forward in section 3 and to be examined more closely in this section: ●

Local producers in developed country clusters do not operate in quasi-hierarchical chains.

This proposition was falsified by the analysis offered by Rabellotti in Chapter 6. She first examined Brenta in the early 1990s, when it was a cluster of predominantly small enterprises exporting mainly to the European market, in particular Germany and neighbouring countries. In spite of high wage costs, the enterprises were able to compete by offering their own designs and brands and producing small batches of shoes of very high quality aimed at the high end of the market (Rabellotti, 1997). These characteristics had not changed by the year 2000, when Rabellotti started a follow up study. Two features had changed. First, the wage differential between Italy and other competitors had increased and some of the most labour-intensive operations were farmed out to Eastern Europe. Second, some of the best local enterprises had become contract manufacturers for top brand companies operating in the luxury leather good markets. This was a significant break from past practice. Instead of developing their own designs and using their own brands, they were producing to the specifications of their top brand customers. Rabellotti concludes that this amounts to functional downgrading. This producer–customer relationship comes close to quasi-hierarchy. The question that then arises is: Why did buyers and producers seek this


Local enterprises in the global economy

sort of relationship? A lack of production and design competence cannot have been the reason. Brenta has some of the most capable shoe makers in the world. The same applies to Marche and other Italian footwear clusters. The reason presumably lies in product differentiation as well as the very tight specifications and high quality requirements stipulated by the customers. Global brands such as Prada and Gucci have specific requirements when sourcing shoes. They need reliable supply to the specification at precise times. They also need shoes which coordinate with their accessories range, and so design has to be coordinated with the overall fashion ‘look’ being offered by the company. Therefore, sourcing through market-based relations is not an option. Rabellotti stresses, however, that the existing relationships do not fall neatly into our chain categories. On the one hand, the top brand companies set the parameters for other enterprises in the chain. This includes not only the shoe makers but sometimes also the suppliers of inputs. On the other hand, the producers do not just obey orders but participate sometimes with their own suggestions for design and production. So there are network features in this chain. The other question is: Why are some of Brenta’s best producers willing to become contract manufacturers and embark on a path that leads to functional downgrading? Rabellotti offers a convincing argument: it pays. Apparently, operating in these chains is more profitable than in other chains where producers continue to produce to their own designs. This gives rise to interesting new questions, one empirical, one conceptual. First, are the suppliers of the top brand companies short-sighted in accepting good profits and relinquishing design and marketing? How can they be sure of keeping these customers? There are other highly competent suppliers in Spain, Brazil and elsewhere who could be brought on stream in a very short time. Time will tell. The conceptual question concerns the notion of functional downgrading. Given that the rationale for upgrading is avoiding the race to the bottom and raising the returns on economic activity (Kaplinsky, 1998), how useful is it to use the term ‘downgrading’ for a strategy which provides returns that are superior to the returns in the most immediately relevant alternatives? We come back to this issue in the concluding section, which discusses the terms and distinctions which have been used in this chapter and in others in this book.



Network governance involves a stronger mutual commitment between firms than in a market-based relationship, but the relationship is less asym-

Chain governance and upgrading


metrical than in quasi-hierarchy. Such relationships are more likely to arise when producer and user have complementary competences such that both parties can contribute to innovation. Operating in a chain characterized by such network-based relationships is clearly desirable but how likely is it for developing country firms? Our initial proposition, stated in section 2, was: ●

Network-based chains support an open-ended upgrading path but local producers in developing countries rarely find themselves in such chains.

The empirical investigations only partially confirmed this, leading to a more differentiated conclusion. Operating in chains characterized by network-based relationships is not a likely prospect in the early stage of the product cycle when new products are being developed through close interaction between producers and users. Such innovation networks are more likely to develop in advanced countries. However, developing country firms can develop network-based relationships with their customers. This is most likely to take the form of modular production networks. The chapters in this book provide evidence for both statements, and this section seeks to systematize the evidence. We start with innovation networks. There is a body of literature which emphasizes that innovation results from an interactive process between producers and users, and particularly between equipment suppliers and manufacturers. Lundvall (1998), Cassiolato (1992) and others have shown that collaboration between users and producers is important in the design and debugging of new products and that the producer’s competitiveness becomes structurally linked to the user’s competitiveness. This technical change literature has been largely ignored in the recent value chain research, mainly because the latter has tended to focus on forward rather than backward linkages. This has been rectified in Chapter 7 of this book, which explicitly investigates the backward linkages and confirms that such innovation networks typically arise in developed countries. Meyer-Stamer, Maggi and Seibel (2001) show that the competitiveness of the ceramic tile producers of Sassuolo in Italy has depended critically on innovations carried out in close collaboration with the equipment producers. Similarly, the tile makers of Castellón in Spain have competed by innovating with new glazing materials, a process carried out in conjunction with the producers of these materials. This process is facilitated by their ability to market new products, creating a demand for the results of innovation. In contrast, the Brazilian tile cluster does not experience such close interaction with their suppliers of equipment and glazing materials. The same applies to other tile clusters in developing countries. The inputs are sourced through market-based relationships with suppliers – usually from developed


Local enterprises in the global economy

countries. This does not preclude the developing country producers from upgrading their processes and products. Indeed, in the Brazilian tile industry the manufacturers do design and market their own products. However, it does mean that internationally they are followers rather than leaders. This is a critical difference since it is the relative position in the global industry that counts for competitiveness in export markets. Ceramic tiles are an example of a sector in which being a leader requires the complementary competences of the equipment and/or material suppliers. Other well-known examples are the interaction between Danish makers of dairy equipment and dairy products or the close connection between Italian tanneries and producers of leather goods. These examples come mainly from backward linkages, illustrating the importance of learning-byinteraction between manufacturer and equipment supplier or component supplier. Similar mutual benefits can apply in forward linkages with end users, even though they are less extensively researched. A fascinating example is offered by Halder in Chapter 8, which investigates the surgical instrument cluster of Tuttlingen in Germany. The competitive advantage of Tuttlingen firms lies in their close collaboration with pioneering surgeons and hospitals. Halder stresses that establishing and maintaining these relationships requires considerable investment and that only the larger firms within the Tuttlingen cluster have the resources to do so. The main competitor of the Tuttlingen cluster is the surgical instrument cluster of Sialkot in Pakistan. However, the Pakistani producers have no direct contact with hospitals and surgeons and do not engage in collaborative development of new instruments (Nadvi and Halder, 2002). This comparison of the German and Pakistani clusters confirms that developing country enterprises are less likely to participate in such innovation-conducive, network-based relationships. We are not, however, suggesting that networks based on symmetrical relationships and complementary competences are absent from developing countries. The chapters in this book show that they exist but they are networks of a different kind: they are modular production networks, as defined by Sturgeon (2002). He suggests that the greater attention given to the core competences of lead firms has resulted in a relative neglect of the outsourced non-core operations. Precisely because outsourcing has become more generalized, new developments have occurred in the supply base. ‘To meet the growing demand of full-service outsourcing solutions, suppliers have in many cases had to add entirely new competence areas, increasing their scope of activities while improving quality, delivery and cost performance’ (Sturgeon, 2002, 455). In the case of contract manufacturing in the electronics industry (the main focus of Sturgeon’s analysis) there remains, however, a clear division of labour between contract manufacturers and their customers. The latter

Chain governance and upgrading


are global household names (Nokia, Apple, IBM, Toshiba, and so on), involved in design and branding, while the latter are largely invisible providers of products and services, such as Solectron and Flextronics. The latter have acquired important competences in the areas of supplier development, process technology, logistics and flexible production. They sell these competences to a variety of customers, but they generally avoid encroaching on the customers’ own competences. Many of the assembly plants owned by the contract manufacturers are in developing and transition economies, but the companies themselves have global reach and are frequently based in the advanced economies, particularly North America. Insofar as global production is fragmented due to trade barriers or the need for proximity to markets (for example production for North America in Mexico and production for Western Europe in Central and Eastern Europe), these manufacturers need to have the capacity to produce in multiple locations around the world in order to meet their customers’ requirements. Furthermore, there are considerable advantages in being located close to the design centres of their customers and to global innovation centres, such as Silicon Valley, where new technical standards and standards for inter-firm communication are developed.18 Nevertheless, the analysis of the Taiwanese computer cluster in Chapter 9 does indicate that some developing countries can enter this type of network. Kishimoto suggests that the Taiwanese computer cluster now constitutes a modular production network in the sense described by Sturgeon. Firms have the entire range of required production skills; all the specialists are available locally; they can produce infinite product variety, any combination of attributes stipulated by the customer can be dealt with; and where required they provide incremental R&D for adjustments. However, major R&D continues to be carried out in Japan or the US. According to Saxenian and Hsu, the division of labour between the US and Taiwan is unlikely to disappear: New product definition and leading edge-innovation will remain in Silicon Valley. However, Taiwanese companies continue to enhance their ability to design, modify and adapt as well as rapidly commercialize technologies developed elsewhere. As local design and product development capabilities improve, Taiwanese companies are increasingly well positioned to take new product ideas and technologies from Silicon Valley and quickly integrate and produce them in high volume at a relatively low cost (Saxenian and Hsu, 2001, 915).

To what extent can we see parallel processes developing in other sectors? In particular, to what extent have leading firms in the Sinos Valley shoe cluster become producers in a modular footwear value chain? Insertion in a quasi-hierarchical chain provided them with a fast track to process and


Local enterprises in the global economy

product upgrading. Compared with the Taiwanese computer cluster, transactional dependence on buyers is frequently higher in the Sinos Valley, while product R&D is lower, and organizational innovation is probably on a par. Certainly the cluster today can produce any type of women’s leather shoe to the specification of the buyer by drawing on the competences available in the cluster. In terms of quality, speed and flexibility, the Brazilian producers can match the Italian competition. This is a major difference compared with 20 years, or even ten years ago, when competition was more price based, buyers were dominant and provided essential inputs into production and logistics. We would therefore suggest that the Brazilian footwear cluster is undergoing a transition from captive to modular production network. The discussion of how this transition takes place is taken up in sections 8 and 9.

7. UPGRADING IN MARKET-BASED VALUE CHAINS It was argued in section 2 that there were two main reasons for the development of explicit (or non-market) coordination of global value chains: the buyer’s role in product definition and the risks to the buyer of supplier non-compliance with product or process requirements. It follows that market-based value chains tend to develop when these two factors do not apply. In particular, when: ●

Products are standard (produced to well-known designs available to many companies) and can be evaluated cost-effectively with regard to critical characteristics at the point of sale. The buyers are ‘design takers’, who buy products from suppliers who take responsibility for design and production. In this case, the supplier may be considered to have greater competences in these areas than the buyer. This is most likely to occur when the buyers are relatively small. They may lack the competences to define product and process parameters, or the volume requirements may be too small to justify the costs of explicit coordination. In value chain coordination, there are clear economies of scale.

In these situations it can also be assumed that the suppliers are competent. Various suppliers exist, and buyers will choose those that can perform the required tasks. In spite of the rising level of trade in differentiated (non-standard) products,19 many products are traded through arm’s-length market relation-

Chain governance and upgrading


ships. In these cases, the supplier has to be able to meet the requirements of customers without inputs with regard to product design or process. If the supplier is to take responsibility for product design, then it has to know what the buyers want before a specific order is made, which requires familiarity with the user market. Clearly, market-based value chains can be found. How do they emerge, and does the evidence from them support the hypothesis put forward in section 3? ●

In market-based chains, producers experience neither support for, nor blockages, to upgrading (from within the chain).

The importance of both market knowledge and customer size for market-based export value chains is underlined by Bazan and NavasAlemán (Chapter 5), who explicitly compare the quasi-hierarchical chain linking producers to Europe and North America with market-based chains linking producers to both the domestic market and regional markets within Latin America. In the case of these latter markets, small buyers purchase ready-designed shoes and either sell them under their own labels, or under the supplier’s own brand. Market-based relationships in the (large) domestic market enabled substantial capabilities in design and marketing to be built up, and Brazilian producers were then able to export to the Latin American market products which they had designed and branded. Brazilian brands and designs are very visible at trade fairs in Latin America, but much less so at trade fairs in Europe or the US. Tewari’s (1999) analysis of exports of the Ludhiana (India) knitwear cluster to advanced country markets reinforces this point. Some of the leading knitwear exporters in the Ludhiana knitwear cluster prefer to work with small foreign traders, rather than large retailers, because these relationships are more symmetrical and give them the space for learning gradually in the context of small orders. However, in order to do this the producers need to develop their own products (rather than produce to somebody else’s specification), hiring designers who travel to Europe and have first hand knowledge of the final market (Tewari, 1999). In addition to this, Tewari emphasizes that the firms which were first able to export to advanced country markets were those which had previously developed products for the high end of the domestic market. It was the experience of this market which gave them the design and quality capabilities needed to break into the European market. In this case, therefore, the characteristics of the domestic and export markets were similar enough for knowledge gained in one market to be applicable in the other. This will vary considerably according to product and country. In a similar vein, Hsing (1999) found that some Taiwanese fashion shoe


Local enterprises in the global economy

companies distinguished themselves by shifting from producing to order, to manufacturing products which they had designed themselves. She argues that this functional upgrading was achieved by working with small trading companies. There is, however, counter-evidence. Nadvi and Halder (2002) found that surgical instrument makers (in Germany and Pakistan) who sell through small traders (local or foreign) are not amongst the most innovative firms. Their ability to export through market relationships may depend on producing highly standardized, low-value products. In the Sinos Valley, too, a good number of small footwear manufacturers connect to the European or North American markets through small traders but they constitute the fringe of the cluster and are not particularly innovative. Some of these export manufacturers were longing to be adopted by a large buyer and produce to their specification.20 What does this mean for the validity of our hypothesis? The evidence confirms that market-based relationships do not entail blockages to functional upgrading. Such upgrading seems to be facilitated by dealing with small rather than large customers. However, local producers do not necessarily make the required investment for functional upgrading where such circumstances exist. The chain approach cannot explain why they do in some cases and not others. The connection between chain organization and functional upgrading does, however, emerge clearly in an inter-chain comparison. This can be observed at trade fairs. Producers operating in quasi-hierarchical chains are unlikely to exhibit, because they do not have their own design, and marketing is taken care of by their buyer. Producers who take it upon themselves to find new customers and orders, need to develop a design (even if copied) and invest in marketing (have a stall at the fair). Take the example of the twice yearly shoe fair in Düsseldorf, the biggest shoe fair in the world and the main fair for the European market. In spite of their large and increasing market share, developing country producers are poorly represented at the fair. In contrast, a large contingent of exhibitors comes from Italy and Spain, whose producers tend to have market-based relationships with their customers (the retailers).21 This observation from trade fairs underlines our basic point that in market-based relationships local firms are unlikely to be locked into the sphere of production. This ‘freedom’, however, comes at a price. The producers themselves need to invest in design, branding and marketing, and the sums involved are often bigger than for process or product upgrading. Large firms can make the leap on their own, small firms find this much more difficult, and often rely on collective initiatives. Significantly, the strong presence of small Italian and Spanish producers at the Düsseldorf trade fair is prepared and organized by their business associations.

Chain governance and upgrading


8. UPGRADING BY OPERATING IN DIFFERENT TYPES OF CHAINS SIMULTANEOUSLY This final chapter explores how far the chain approach takes us in explaining different types of upgrading. To this end, each section has focused on particular hypotheses. A conclusion which can be drawn from examining these hypotheses is that the explanatory power of chain variables increases with the extent of explicit coordination of the chain. In particular, section 4, concerned with quasi-hierarchical chains, shows a clear pattern: producers in developing countries experience fast process and product upgrading while functional upgrading conflicts with the chain governor’s core competence and is therefore often blocked. In some cases, however, local producers are able to overcome these difficulties. Explaining these cases is the purpose of this section. In order to do so we need to widen our focus. The analysis so far has assumed that enterprises operate in only one type of chain. In many cases, this is true not just for particular firms but for entire clusters. Concentrating on the implications of this dominant type of chain governance is a sensible way to proceed. However, the chapters in this book suggest that we risk not recognizing some of the most interesting upgrading experiences unless we consider those cases in which firms operate in several types of chains simultaneously. In the Taiwanese computer industry (Chapter 9) local producers have progressed from producing to the specifications of their buyers to owndesign manufacturing (ODM). Customers were willing to relinquish design and producers were willing to invest in design capabilities. The more difficult step was from ODM to OBM. While only achieved by some, it seems that those who succeeded pursued a double strategy – maintaining their OEM/ODM production while starting to experiment with and build up their OBM operations in a different market. This and other Taiwanese experiences have given rise to Lee and Chen’s thesis on the leveraging of competences across chains. They argue that firms were able to acquire new capabilities by applying lessons from one chain to another. Firms could, for example, take a design supplied by one customer and then make adaptations and use the modified design to supply other customers in other markets (Lee and Chen, 2000). Bazan and Navas-Alemán also show in Chapter 5 that upgrading by operating in several chains simultaneously has begun in the Brazilian shoe industry. However, the leading export manufacturers greatly delayed in pursuing this strategy because they were locked into relationships with existing buyers from the US. Interestingly, it was above all second-tier and/or second-generation exporters that began exploring different markets,


Local enterprises in the global economy

in particular the Latin American market. As shown in Chapter 5, some Brazilian firms that were world class suppliers of very big US buyers found it difficult to succeed in the smaller markets of Latin America or establish themselves in the Brazilian market. Manufacturing to tight specifications for the main customer requires an internal organization geared to this purpose and builds up capabilities which are highly developed but narrowly-focused (limited to the sphere of production). Entering new markets requires different sets of capabilities. Where producers sell to powerful customers, they cannot compete directly with them and must find other markets when diversifying and upgrading. A question that arises from Chapter 5 is whether learning and market diversification occurs at the enterprise or cluster level. Interestingly, the firms which were most successful in functional upgrading and exporting to new markets were companies which had acquired their design and marketing experience in the national market. Some of them had never exported to the US market, but by operating in the Sinos Valley, they were able to absorb many of the process and product innovations which those manufacturers targeting the US market had brought to the cluster under the tutelage of their buyers. In this sense, they benefited from the export orientation of the cluster, but the capabilities for functional upgrading were acquired in the national market. This means that the widely held view of the superiority of learning-by-exporting needs to be revised. The chain and upgrading distinctions employed in this chapter help in making such a revision.

9. ADVANTAGES AND LIMITATIONS OF THE CHAIN APPROACH This chapter has tried to distil from the previous chapters those insights that help to understand how chain governance influences upgrading. However, chain governance is not the whole story, and this section maps out briefly what other factors would need to be considered to explain differences in upgrading experiences of exporting producers. This will necessarily be schematic but it is essential to recall the wider upgrading agenda to which this chapter seeks to contribute. Most contributors to this book rely on a combination of the cluster and chain approach. In fact, the rationale for the book was that earlier cluster studies were insufficient by themselves to explain the upgrading experiences of local producers. There is now ample literature on the strengths and limitations of the cluster approach, so a mere listing of main factors will suffice. The cluster effects which help upgrading can be grouped into inci-

Chain governance and upgrading


dental and deliberate effects. The incidental effects were laid out by Alfred Marshall (1920) who stressed the easy access to input suppliers and traders, the benefits of a pool of specialized workers and the spillovers of knowledge. More recent literature would add the ease of parallel experimentation (Malmberg and Maskell, 2002) and the importance of local rivalry (Porter, 1990). The deliberate effects concern horizontal and vertical cooperation between firms (Nadvi and Schmitz, 1999) and supportive local government (dei Ottati, 2002). The chapters in this book (or the research on which these chapters build) provide ample evidence that these factors matter. However, all the contributors agree that the analysis of local agglomeration and cooperation effects has to be complemented by the study of relationships with actors outside the cluster, especially in export-oriented clusters. This is what led to the analysis of clusters in the context of global value chains. A conclusion that can be derived from these chapters is that the combination of cluster and chain effects is powerful. The success of the Taiwanese computer industry is a good example. Like export manufacturers elsewhere in the developing world, they started in hierarchical (subsidiaries of transnational companies) or quasi-hierarchical chains. Like export manufacturers elsewhere they experienced fast process and product upgrading, but they seem to have been much better in using the newly acquired capabilities as a springboard to move into ‘higher’ chain functions, notably design, and to explore other chains where they could make their own marketing experiments. This requires strategic intent. The question is why such strategic intent is more common in East Asia than in other parts of the developing world. While systematic comparisons remain rare, it is clear that East Asia has produced many more success stories of firms progressing from OEM to ODM and sometimes even to OBM (UNIDO, 2002). The chain approach – even if combined with the cluster approach – is ultimately not sufficient to explain this difference. Other approaches have been put forward to explain the East Asian success, all with their distinctive contributions, including Hobday (1995), Lall (2001), Kim and Von Tunzelmann (1998), Saxenian and Hsu (2001) and Westphal (2002). There are overlaps with the approach used in this book but also differences, partly because these works are reactions to different agendas. Specifying which are the critical factors explaining the East Asian upgrading success remains controversial. Single factor explanations will not do, but all authors would probably agree that the policy environment occupies a high place in the hierarchy of issues to be considered. Here we do not so much mean explicit innovation policy (for example, financial support for training and technology institutes) as implicit innovation policy through the incentive structure (trade, tax and monetary policy). In a cross-country


Local enterprises in the global economy

comparative project, Sagasti (1978) concluded long ago that the implicit policies are more important. We believe this continues to be the case. The East Asian countries did not suffer from the kind of disabling policy environment that characterized many developing countries in the 1980s and 1990s.22 Frequent and drastic changes in the incentive structure make it difficult to develop or maintain the strategic intent required for major upgrading initiatives. Take the case of Brazil: Between 1980 and 1990, Brazil . . . had eight different inflation stabilisation plans, fifteen different policies towards wages, eighteen changes in the rules regulating foreign exchange, fifty-four changes in the rules controlling prices, twenty-one different proposals on the negotiation of foreign debt, four different currencies, . . . eleven separate indexes for calculating the devaluation of local currency, and . . . five wage and price freezes (Ferraz et al., 1992, 57–9).

Management time is scarce and operating in this kind of environment makes it very difficult to focus on major upgrading initiatives. In many developing countries, the policy environment has driven industrial managers into reluctant speculation rather than strategic upgrading. This is a fundamental factor impacting on upgrading and needs to be acknowledged when upgrading experiences from different countries are considered.



The world map of manufacturing has changed fundamentally over the last two decades. Developing countries, particularly in Asia and Latin America, have become major exporters of manufactured products, forcing developed countries to abandon some industries and restructure others. Winners and losers can be found in both sets of countries, but it is not easy to identify common traits. It would certainly be wrong to declare large firms as winners and small firms as losers. The cluster literature, in particular, has provided a lot of case material showing that small and medium sized enterprises can participate successfully in the global economy, but this cluster literature needs a two-fold change in gear. First it needs to recognize that some of the former small or medium sized enterprises grow large and that this may well be a good thing, given that non-incremental upgrading requires substantial resources. Second, it needs to incorporate in the analysis the relationships with actors from outside the cluster. This has been a central concern of most chapters in this book. This final chapter has sought to pull together the evidence on how these external relationships influence the upgrading prospects of local producers. To this end we have drawn on the global value chain approach. As set out

Chain gover