Evolutionary Economic Geography: Location of production and the European Union

Nearly half a century ago, Walter Isard, the founder of regional science, suggested in his seminal book Location and Space Economy two promising approaches to further development of economic geography: the monopolistic competition approach and the evolutionary approach. As is well known, since the late 1980s, the monopolistic competition approach has been vigorously explored in the New Economic Geography. Then, the exploration of evolutionary approach started in the late 1990s, following the development of evolutionary economics and complexity theory. Although it is still in its infancy, evolutionary approach has been contributing significantly to the enrichment of economic geography, adding new perspectives to the long-term process of structural change in space and in time, In Evolutionary Economic Geography, Miroslav N. Jovanović presents in lucid manner the state of arts in economic geography with a special focus on evolutionary approach. The book covers not only theory but also regional policy and location of production in the context of global economy. I enjoyed reading this wonderfully informative book and learned a lot, and so will you. Masahisa Fujita President of the Research Institute of Economy, Trade and Industry, Tokyo

Evolutionary Economic Geography The purpose of this book is to provide a guided tour through the theoretical foundations of spatial locations of firms and industries in an evolutionary economic framework. It addresses the issues of how a location of business in geographical space is selected and where economic activity may (re)locate in the future. The analysis is in the context of technological progress, innovation, disequilibrium and endemic uncertainty. Jovanović raises pertinent questions such as how willing, motivated and able firms (and governments) are to adapt to constantly evolving new opportunities and challenges over time and to experiment and translate these perceptions into profitable actions. How is their ‘competitive position’ evolving relative to others and changing over time when faced with a stream of constantly arriving new opportunities, innovation, threats and obstacles? Considerations are always supported with a plethora of examples and cases from real life. Jovanović argues that the economy is a complex and constantly adaptive system which is almost always outside equilibrium. Building on this, he suggests that there is an important lacuna in our understanding of evolutionary spatial economics and that there is much space for further multidisciplinary research in this academic and practical area. This book offers an evolutionary and disequilibrium analysis of the subject and makes parallels, where appropriate and possible, among economics, geography, physics, biology and art. It considers key areas in theory, market and production structure, spatial location of domestic and foreign firms, as well as regional policy. In addition, there are references to policy intervention; importance of investment in local social stability, education and training; as well as to uncontrollable variables that are beyond the influence of firms, industries, regions or public authorities. The author offers various evolutionary insights and alternatives to the pure neoclassical equilibrium economic model. This book will be of great interest to students and researchers engaged with disequilibrium economics, evolutionary and complexity economics, economic geography and location of firms and industries in geographical space. Miroslav N. Jovanović is Economic Affairs Officer at the United Nations Economic Commission for Europe, Geneva and Lecturer at the European Institute of the University of Geneva.

Routledge Studies in Global Competition Edited by John Cantwell, Rutgers, the State University of New Jersey, USA and David Mowery, University of California, Berkeley, USA 1 Japanese Firms in Europe Edited by Frédérique Sachwald 2 Technological Innovation, Multinational Corporations and New International Competitiveness The case of intermediate countries Edited by José Molero 3 Global Competition and the Labour Market Nigel Driffield 4 The Source of Capital Goods Innovation The role of user firms in Japan and Korea Kong-Rae Lee 5 Climates of Global Competition Maria Bengtsson 6 Multinational Enterprises and Technological Spillovers Tommaso Perez 7 Governance of International Strategic Alliances Technology and transaction costs Joanne E. Oxley 8 Strategy in Emerging Markets Telecommunications establishments in Europe Anders Pehrsson 9 Going Multinational The Korean experience of direct investment Edited by Frédérique Sachwald 10 Multinational Firms and Impacts on Employment, Trade and Technology New perspectives for a new century Edited by Robert E. Lipsey and Jean-Louis Mucchielli

11 Multinational Firms The global–local dilemma Edited by John H. Dunning and Jean-Louis Mucchielli 12 MIT and the Rise of Entrepreneurial Science Henry Etzkowitz 13 Technological Resources and the Logic of Corporate Diversification Brian Silverman 14 The Economics of Innovation, New Technologies and Structural Change Cristiano Antonelli 15 European Union Direct Investment in China Characteristics, challenges and perspectives Daniel Van Den Bulcke, Haiyan Zhang and Maria do Céu Esteves 16 Biotechnology in Comparative Perspective Edited by Gerhard Fuchs 17 Technological Change and Economic Performance Albert L. Link and Donald S. Siegel 18 Multinational Corporations and European Regional Systems of Innovation John Cantwell and Simona Iammarino 19 Knowledge and Innovation in Regional Industry An entrepreneurial coalition Roel Rutten 20 Local Industrial Clusters Existence, emergence and evolution Thomas Brenner 21 The Emerging Industrial Structure of the Wider Europe Edited by Francis McGowen, Slavo Radosevic and Nick Von Tunzelmann 22 Entrepreneurship A new perspective Thomas Grebel 23 Evaluating Public Research Institutions The US Advanced Technology Program’s Intramural Research Initiative Albert N. Link and John T. Scott

24 Location and Competition Edited by Steven Brakman and Harry Garretsen 25 Entrepreneurship and Dynamics in the Knowledge Economy Edited by Charlie Karlsson, Börje Johansson and Roger R. Stough 26 Evolution and Design of Institutions Edited by Christian Schubert and Georg von Wangenheim 27 The Changing Economic Geography of Globalization Reinventing space Edited by Giovanna Vertova 28 Economics of the Firm Analysis, evolution and history Edited by Michael Dietrich 29 Innovation, Technology and Hypercompetition Hans Gottinger 30 Mergers and Acquisitions in Asia A global perspective Roger Y. W. Tang and Ali M. Metwalli 31 Competitiveness of New Industries Institutional Framework and Learning in Information Technology in Japan, the US and Germany Edited Cornelia Storz and Andreas Moerke 32 Entry and Post-Entry Performance of Newborn Firms Marco Vivarelli 33 Changes in Regional Firm Founding Activities A Theoretical Explanation and Empirical Evidence Dirk Fornahl 34 Risk Appraisal and Venture Capital in High Technology New Ventures Gavin C. Reid and Julia A. Smith 35 Competing for Knowledge Creating, Connecting and Growing Robert Huggins and Hiro Izushi 36 Corporate Governance, Finance and the Technological Advantage of Nations Andrew Tylecote and Francesca Visintin

37 Dynamic Capabilities between Firm Organisation and Local Systems of Production Edited by Riccardo Leoncini and Sandro Montresor 38 Localised Technological Change Towards the economics of complexity Cristiano Antonelli 39 Knowledge Economies Innovation, organization and location Wilfred Dolfsma 40 Governance and Innovation Maria Brouwer 41 Public Policy for Regional Development Edited by Jorge Martinez-Vazquez and François Vaillancourt 42 Evolutionary Economic Geography Location of production and the European Union Miroslav N. Jovanović

Evolutionary Economic Geography Location of production and the European Union

Miroslav N. Jovanović

LONDON AND NEW YORK

First published 2009 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2008. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” © 2009 Miroslav Jovanović All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Jovanovic, Miroslav N., 1957Evolutionary economic geography : location of production and the European Union / Miroslav Jovanovic. – 1st ed. p. cm. – (Routledge studies in global competition) Includes bibliographical references and index. 1. Economic geography. 2. Production (Economic theory) 3. Industrial location. 4. International economic integration. 5. European Union. I. Title. HF1025.J658 2008 338.6'042094–dc22 2008003404 ISBN 0-203-89318-2 Master e-book ISBN

ISBN: 978-0-415-42346-5 (hbk) ISBN: 978-0-203-89318-0 (Print Edition) (ebk)

Својим синовима Јовану и Николи посвећујем све на свету ове књиге , која је безвредна у односу на љубав, срећу и наду коју пружају.

Ai miei figli, Jovan e Nikola dedico ogni cosa, tranne questo libro, che non vale tutto l’ amore, la felicità e la speranza che mi danno.

To my sons, Jovan and Nikola, I dedicate everything in the world except this book which is worthless compared with the love, happiness and hope that they provide.

Contents List of illustrations

xiii

Foreword

xvi

Preface

xix

Acknowledgements

xxi

List of abbreviations

xxii

1 Introduction

1

2 Theory

6

3 Regional policy

156

4 Market structure and location of production 195 5 International firms

276

6 Conclusions

366

Annex I

377

Annex II

380

Annex III

385

Bibliography

387

Index

408

Illustrations Figures

1.1

The Gini Index.

5

2.1

Adaptive walk.

52

2.2

Random jumps.

53

2.3

Combining adaptive walk and random jumps.

53

2.4

From positive to negative lock-in regional path-dependent economic development.

60

2.5

Productivity/applications curves for a general-purpose technology.

62

2.6

Productivity/applications curves for a succession of generalp urpose technologies.

63

2.7

Productivity curves for different patterns of two successive general-purpose technologies.

64

2.8

Diffusion of mechanical power in the United States manufacturing firms.

65

2.9

Concave (a), convex (b) and spherical (c) distributions of potential industry locations.

70

2.10 Urban and rural population of the world, 1950–2030.

100

2.11 Ideological distribution of the parties in the US House, 1969–70 122 and 1999–2000. 3.1

Relation between regional manufacturing populations and real

158

wages with varying transport costs. 3.2

Transport costs and equilibria.

158

3.3

Welfare in a trade-diverting customs union.

168

4.1

Effects of innovation on the use of factors.

204

4.2

Trends in R&D intensity by area as a percentage of GDP, 1991–2005.

209

4.3

Characteristics of computers.

230

4.4

Technologies with constant, increasing and decreasing returns to scale.

238

4.5

Cost curves with constant, increasing and decreasing returns to scale

238

5.1

Comparative advantage and location of economic activities.

319

5.2

Transnational corporations and the growth process.

321

Paintings

1 School of Andrea Mantegna, Opportunity and Penitence (ca. 1503– 73 6). Tables

2.1 Two basic theoretical concepts of how an economy operates

39

2.2 Some unresolved issues associated with path dependence

68

2.3 Clusters costs and benefits

95

4.1 Non-tariff barriers

255

5.1 Similarities and differences between the two globalisation waves 308 5.2 Investors’ motives and determinants for FDI in the target location

314

5.3 Profit in the aircraft industry without intervention

340

5.4 Profit in the aircraft industry with intervention

341

5.5 Potential benefits and real concerns that bring TNCs to the target 352 country/location

Foreword To an increasing extent, economists become interested in geography. Since economists are aware that economic activities tend to concentrate in space, they feel an urge to integrate that fact in their theories and empirical work. Miroslav Jovanović has understood that key message more than anybody else, and contributes a lot to its dissemination. Last year, he already edited three volumes entitled Economic Integration and Spatial Location of Firms and Industries, each of which contained a multitude of key contributions bringing together the disciplines of economics and geography. This new book can be viewed as a continuation along that path. What makes this book interesting is that it brings together various approaches that have criticized the mainstream neo-classical framework in the last decades. It is something that cannot be repeated enough, and therefore, should be embraced accordingly. What these approaches dislike is the way traditional neoclassical equilibrium theory has neglected the role of history, path dependence and increasing returns, among other things. I fully agree that these alternative approaches come indeed much closer to reality. This is especially true for the evolutionary perspective on economic geography, which brings back life into regional economics and economic geography, as Geoff Hodgson would put it. In addition, these approaches tend to take an interdisciplinary perspective, not closing itself off and refusing to dialogue with other disciplines, which has damaged mainstream economics so much in the last decades. It is therefore a great joy to see that many researchers (many of which are heterodox economists) currently embark on this exciting new voyage, and Miroslav Jovanović gives full witness of these new developments and ideas. Implicitly, the book gives a very broad definition of what Evolutionary Economic Geography is. It ranges from New Economic Geography and Endogenous Growth Theory, which are still deeply embedded in the standard neoclassical equilibrium framework, to Complexity Theory and Evolutionary Economics, which are based on a disequilibrium framework, and which employ a range of methodologies going from mathematical modeling to descriptive case studies. The two latter approaches have been extensively discussed in a recent special issue on Evolutionary Economic Geography published in the Journal of Economic Geography in 2007. So, I am not sure I entirely agree that all these approaches can be labeled evolutionary, since not all of them are firmly grounded in evolutionary theory. Having said that, Jovanović makes interesting connections between these different approaches which are thought-provoking, and applies these insights to a range of topics such as the location of multinational enterprises, the formation of agglomerations and clusters, trade patterns between countries and convergence/divergence of regions, among others. What makes the book especially attractive is that many examples are presented to clarify key concepts. The book is full of empirical illustrations, providing explanations for the rise of big companies like Microsoft, IBM, Honda and Lloyd’s, the evolution of cities like New Orleans, Antwerp and Venice, the evolution of the ideological representation of political parties in the US house, how wars affect the location of

industries and how chance events triggered the growth of cities like Cannes, Geneva and Basle, among many others. Now and then, Jovanović makes interesting and quite unexpected connections when he refers to Buridan’s ass, Morton’s fork, the Bar Problem of Arthur, and the Medieval School of Andrea Mantegna. Doing so, his conceptual statements come to life so to speak, and this makes the book very readable. Of course, as always, many questions and unresolved issues remain, and Jovanović is one among the first to admit that. He is right when he states that one of the main scientific challenges is without doubt to cope with complexity issues caused by unpredictability, non-determinacy, contingency, non-linearity and chance events. What also deserves more attention is to demarcate better what Evolutionary Economic Geography really encompasses. To start with, there is a need to separate the New Economic Geography of Krugman from Evolutionary Economic Geography, because they rely on very different theoretical frameworks and assumptions, and they have developed and applied very different concepts of time and space. It is true that they share some general properties, such as path dependence, increasing returns and imperfect competition. Nevertheless as, among others, Ron Martin (Cambridge Journal of Economics, 1999), Ron Boschma and Koen Frenken (Journal of Economic Geography, 2006) have explained at length, the New Economic Geography is still firmly grounded in the maximizing-equilibrium analysis, while the Evolutionary Economic Geography rests on the foundations of bounded rationality-disequilibrium principles. To make matters even more complicated, there is no consensus whatsoever with respect to what evolutionary economics really stands for. In fact, various evolutionary perspectives coexist, and it is crucial to be very specific about that before applying evolutionary insights to regional economics and economic geography. While I fully share the view that Evolutionary Economic Geography has a huge potential to provide a more realistic alternative, most of its promises have yet to be fulfilled. Among many other unresolved issues, we still have to specify what path dependency really means, how it can be measured empirically, and how to test it. Many researchers are also currently embarking on network analysis. An evolutionary perspective provides a promising point of departure to analyze the spatial evolution of networks, but we have yet little understanding of how different types of networks evolve spatially, and why. The same is true with respect to the topic of the evolution of clusters. While there is an extensive literature on clusters, we still know very little of how new clusters come about, and what benefits and costs clusters do provide to local firms along their life cycle. And there is, of course, the issue of policy making. While Evolutionary Economic Geography heavily criticizes neoclassical equilibrium theory’s focus on market failures as a key underpinning of public policy, evolutionary approaches are completely silent on how to pursue welfare policy, monetary policy, et cetera.

In other words, much work still needs to be done. In that context, Jovanović has delivered a very thought-provoking book that provides a major input for further discussion and rethinking in what Evolutionary Economic Geography really means, and what it might offer to academics in disciplines like economics and economic geography, as well to provide decision makers and policy makers more in general. Ron Boschma Professor Regional Economics Utrecht University Utrecht, June 2, 2008

Preface This book provides a new framework for the discussion of an old economic problem: where to locate a firm in space. The new spatial programme is an evolutionary approach to economics and economic geography. Traditional neoclassical equilibrium theory is an elegant and quite convincing academic exercise. Its conclusions are straightforward and offer simple and clear insights regarding the location of firms in space. Its general economic policy recommendation is: always remove all market imperfections. In spite of its complex formulation, the evolutionary approach to spatial economics is much closer to reality. This book gives a theoretical and analytical framework on evolutionary economic geography and may provoke research curiosity. It attempts to contribute to a multidisciplinary exchange of views among economists, geographers and businesspeople within a new evolutionary approach. It also ambitiously includes various contributions to spatial topics that are widely scattered: microeconomics, planning, development, economic geography, regional science, urban economics, location theory, industrial organisation, international trade and integration, foreign direct investment, transport economics, business economics, innovation studies, public finance, price theory, imperfect competition, economics of scale, labour economics, environment and resource economics. The common research denominator in all these fields is the spatial dimension, which is seen as an opportunity, a medium for interactions, as well as a limitation. In the 1980s, the inclusion of imperfect competition and economies of scale rejuvenated spatial economics. It also helped to explain why firms agglomerate and form clusters. Even though the ever-increasing demand for sophisticated quantitative analysis makes results harder to explain and comprehend, while policy advice is more difficult to derive, the introduction of new analytical tools assists spatial economics in finding its proper place in mainstream economics and becoming a hot research topic. The main insight of this book concerns the way one new economic or business system replaces another, and where. Economists need to step out of their ‘old analytical box’ of clean and lean abstract theoretical, mathematical and aspatial models that bring the analysed system into equilibrium. They need to go to real life and the actual operation of the economic system. Neoclassical equilibrium models were based on assumptions that rendered analysis totally unrealistic and useless to some professions. For example, a geographer would argue that societies and real geographical space are so complex that they cannot be expressed simply as mathematical formulae. ‘Pure’ economic models refer to geography only to the extent that it supports the economic analysis of the ‘market clearing logic’ and equilibrium based on very specific and many (unrealistic) assumptions that ease mathematic modelling. This dangerously damages the practical relevance of their conclusions. Economists need to expand their research agenda out of the usual issues that refer to entry, exit and total factor productivity. Certain economists have already broadened the analysis to include issues such as institutions, history, firm heterogeneity in the use of factors, multiple equilibria, evolution, technological and time–space trajectories, spatial location theory, endemic disequilibrium and various

market imperfections. Geographers need to leave their ‘old analytical box’, in which they were interested in economics only to the extent that it assisted them to understand and explain the spatial organisation of societies. Instead of ignoring or bypassing one another, economists and geographers need to try to talk about topics of mutual interest, potential agreement and obvious difference. They may also explore the relevance of complexity theory for the solution of common problems. Regarding problems in real life, underdeveloped countries and regions worry that high value-added economic activity will be located and concentrated in the core developed regions. Currently, prosperous countries and regions worry that at least a certain part of economic activity will move to other regions where wages are lower. All countries and regions worry that a large part of manufacturing production will take place in China. The Chinese worry about their lack of primary resources, and so they locate their investment in places where they find the missing economic links. Thus, all regions are anxious because of the ‘assault’ on what used to be their established geography of production. This attack on the existing local spatial order comes not only from other locations, but also from evolution in technology and innovation, variation in competition, fickle tastes and, in certain areas, an ageing population (changes in demography). The book is organised as follows. Chapter 1 introduces the subject. Chapter 2, the core part of the book, covers the theory behind evolutionary economic geography. It presents the basic concepts, as well as relations between economics and geography, biology, evolution and physics. It surveys theories of locations of firms without consideration of their ownership. Special attention is devoted to complex systems. Spatial units such as clusters, cities and regions are also presented in turn. The chapter also covers the impact of history, expectations and war on the location of firms. Chapter 3 covers regional policy and includes considerations about its objectives, justification and instruments. It also shows the impact of international economic integration on the location of firms and industries. Chapter 4 is devoted to the effect of market structure (competition) on the spatial location of production. It tackles issues such as basic concepts, innovation, specialisation, returns to scale, standards, rules of origin and non-tariff barriers. Ownership, i.e. foreign possession and control of firms, is treated in Chapter 5. This chapter examines international firms, i.e. transnational corporations, and explores the theoretical and practical reasons for foreign direct investment and location of trans-border business activities: what motivates firms to locate abroad? Special attention is also devoted to the evaluation of the effects of globalisation as a great economic and political story of our times. Chapter 6 concludes the book. The presentation in this book has a slight European slant as this is the geographical area most familiar to the author. I believe and hope that Evolutionary Economic Geography will prove useful to scholars, students, researchers, civil servants, business executives and others in widening their knowledge and awareness of the process and interrelation between evolutionary theory and the spatial location of firms and industries in the economy of today and tomorrow. However, if it also attracts the curiosity and attention of those studying economic development, international business and policy-makers, then this is to be welcomed. The bibliography at the end of the book may serve as a departure point for a curious student or researcher into the demanding, expanding, surprising and academically rewarding world of evolutionary economic geography.

Acknowledgements My involvement in spatial economics has its roots in an invitation from Professor Bertram Schefold (Frankfurt) to write and present a paper on international economic integration and spatial location of industries at a graduate summer school, organised by Academia Europea (Bolzano) at the picturesque alpine resort of Bressanone/Brixen, Italy, 31 August to 10 September 1998. While international economic integration was familiar to me, I had only a general knowledge about spatial economics. I accepted the invitation with delight, thinking that it would take me only a short time to draft the paper. I was wrong. It took me very much longer. I was so fascinated by the subject that my intellectual voyage continued through economic geography and through time. Several articles, a book Geography of Production and Economic Integration (London: Routledge, 2001) and a set of three edited volumes of readings entitled Economic Integration and Spatial Location of Firms and Industries (Cheltenham: Edward Elgar, 2007) followed. The present book is the result of the limits of neoclassical equilibrium economic theory to explain the location of firms and industries in real space in a way that is satisfactory to me, at least, and to a growing number of others. Some of the inspiration for this book came from recent works by several economists. I shall cite here just two sources of inspiration: R.G. Lipsey, K. Carlaw and C. Bekar, Economic Transformations: General Purpose Technologies and Long Term Economic Growth (2005) and E. Beinhocker, The Origin of Wealth (2006), to which I refer from time to time. During my work on this book I have benefited from human capital and discussions, as well as delightful and advantageous contacts with many friends and colleagues over long years. There are, however, several to whom I owe special gratitude for a variety of valuable inspiration, encouragement, comments and assistance in the preparation of this book. . Un ringraziamento va a Marina Rossi che mi ha aiutato senza saperlo. They include Lisa Borgatti, Ron Boschma, Richard G. Lipsey, Marinette Payot and Hélène-Divna Tzico Stefanesco, as well as many students at various universities who have listened to my lectures. Jovan and Nikola Jovanović assisted greatly with graphics. The UN library in Geneva, particularly Anthony Donnarumma, provided me with most of the sources. Finally, Chris Hook efficiently handled the production of this book. I am grateful to all of them. The usual disclaimer, however, applies here: it is I who a responsible for all shortcomings and mistakes. In addition, the views expressed are my own and are not necessarily those of the organisation in which I work. Miroslav N. Jovanović Geneva, June 2008

Abbreviations CD

compact disc

CEO

chief executive officer

CFIUS

Committee on Foreign Investment in the United States

DISC

Domestic International Sales Corporation

DVD

digital video disc

EFTA

European Free Trade Association

EMU

economic and monetary union

ERDF

European Regional Development Fund

EU

European Union

FDI

foreign direct investment

FIRA

Foreign Investment Review Agency

GATS

General Agreement on Trade in Services

GATT

General Agreement on Tariffs and Trade

GDP

gross domestic product

GPS

Global Positioning System

IIT

intra-industry trade

IMF

International Monetary Fund

LDC

less developed country

MAI

Multilateral Agreement on Investment

MFN

most-favoured nation

MNE

multinational enterprise

NAFTA

North American Free Trade Agreement

NGO

non-governmental organisation

NTB

non-tariff barrier

OECD

Organisation for Economic Cooperation and Development

PANEURO

Pan-European Cumulation System

PC

personal computer

R&D

research and development

SITC

Standard International Trade Classification

SME

small and medium-sized enterprise

TNC

transnational corporation

TRIM

trade-related investment measure

UN

United Nations

US

United States

VCR

videocassette recorder

VHS

video home system

WTO

World Trade Organisation

1 Introduction Does the eagle mount up at your command, And make its nest on high? (Job 39:37) The question and the pressing problem that has troubled economists and others since the seventeenth century is: why do some areas, regions or countries grow faster than others? The spatial location of economic activity in the future is one of the most important and challenging questions in economics. It is relevant not only for the future location of jobs, their quantity and quality, but also for the creation of profit, tax proceeds, exports, public expenditure, urbanisation and the like. Progress in technology, innovation, changes in demand and certain moves towards a liberal economic policy create new challenges for theorists, policy-makers and business executives. As certain new and a number of old economic activities in manufacturing and services become footloose, highly mobile, fragmentable and internationally connected, one of the most demanding and intricate questions in such a situation is where firms and industries will locate, relocate or stay. Alan Greenspan, the former chairman of the United States (US) Federal Reserve Board, and a prominent policy-maker, said to his colleagues: ‘We really do not know how [the economy] works … The old models just are not working’ (Beinhocker, 2006, p. 22). Or in the words of Beinhocker (2006, p. 6): The economy is a marvel of complexity. Yet no one designed it and no one runs it. There are, of course, CEOs, government officials, international organisations, investors and others who attempt to manage their particular patch of it, but when one steps back and looks at the entirety of the $36.5 trillion global economy, it is clear that no one is really in charge. Hence, the time is ripe for economic theory to try something different from the past and to move on. The objective of the new evolutionary turn in economics is not to downgrade or to demolish the traditional neoclassical equilibrium approach to economics. Its goal is rather to offer a new research agenda that may prove that economics can do much better than before. The problem is that economics is well known as a field that is notoriously slow to accept new ideas and theories, and hence these new ideas may have a bigger impact only in the decades to come. Another difficulty is that economic complexity and evolutionary economic geography is still only a research agenda, and a cohesive theory is still far away in the future. With this in mind, the approach taken in this book is conceptual. The modest objective is to find, clarify and explain key tendencies rather than to provide definite answers.

Evolutionary Economic Geography

2

The spatial location of a firm is an issue only in an imperfect market. In the absence of both economies of scale (no sunk costs) and transport costs (so that location in space makes no difference), the decision as to where to locate production would be easy. With no market imperfections, firms may be divided into units of any size and operate in all locations without any cost disadvantage. What orders and directs an economy and its spatial distribution of activities? How is this done? Where does it go to? Views and arguments about these forces are divisive and different. Some consider that it is the prices system that controls a complex system such as an economy. Others prefer conscious action by economic agents and institutions, while yet others are impressed by innovations of all kinds. As an ending point of economic activity, neoclassical economic theory looks at a general equilibrium. Conversely, an evolutionary model is open to an ever-changing and never-ending dynamic process of transformations in an environment of disequilibrium and strong endemic uncertainty. Economics is passing through one of its most important turning points for a century. The current evolutionary approach to economics offers the possibility of change and an alternative to traditional neoclassical equilibrium theory. This is still a work in progress and an opening for a change in how one thinks about economics. It is an opportunity to change the traditional equilibrium framework of thought. The evolutionary turn in economics is introducing a much wider openness to other disciplines than was the case just a few decades ago. In the 1970s, economists who were thinking aloud outside the framework of neoclassical equilibrium theory and its models were a rare breed. Even today, they are not too numerous. Nonetheless, certain interdisciplinary conversations take place. The purpose of this book is to contribute to this cross-disciplinary discussion about the spatial location of firms and industries within the evolutionary framework. A relentless search for profits in a market economy never lets the economic system and its participants remain in stasis. New goods, services, jobs, firms, industries, clusters and production and management technologies are being created, while old or obsolete ones are modified or vanish. This ceaseless ‘creative destruction’ of and within the system makes the economy evolve over real time. Hence, economic evolution is linked with ever-changing and never-ending innovation and disequilibrium within the economic system. Economic geography was a sleepy backwater for a long time, not because it was uninteresting, but because problems seemed intractable. Evolutionary economic geography studies the origin, changes, direction and speed in the spatial distribution and organisation of production, as well as consumption, in a time perspective. It analyses the dynamic changes in the economic landscape. This field of academic enquiry is multidisciplinary par excellence. It is, however, not (yet) based on a general theory or even a widely accepted system of methodologies and rules. Hence, it offers wide, uncertain but potentially rewarding possibilities for deeper research and analysis. Economic geography aspires to analyse and describe economic changes and tendencies and their reasons in physical space. It considers the widest array of geographical locations. All societies and market-oriented firms must decide what, for whom, how and where to produce. Our concern in this book is linked with the spatial ‘where’ part of this decision-making process. In this complex situation, the crucial

Introduction

3

questions for research, analysis, business decisions and economic policy include the following: • Where will economic activity locate (and stay) in the future? • What are the dynamics of the development and growth processes? • Are these processes deterministic or unpredictable? Are they leading towards (a higherlevel) equilibrium or is it a perpetual, uncertain and unpredictable process? • Why is production (and consumption) concentrated in urban areas or regions? Why do urban and other areas differ in size? • What are the reasons for the uneven spatial distribution of economic activity and prosperity?1 • Is it smart to keep the existing industries where they are? What is the opportunity cost of doing this? • Does market expansion through international economic integration encourage agglomeration, clustering, spatial concentration of industries, adjoint locations of linked production and ‘thick’ market effects? Or, alternatively, does agglomeration encourage economic integration? • Does integration cause convergence or divergence in the geography of production, income levels and/or growth rates in the participating countries? • How do these forces interact? Where do they lead the economy? • What is the role and effect of public policy (intervention)? • What are the current and future properties of economic adjustment (spatial and industrywide reallocation of resources) to the changing situation? • Where will ‘hot’ and ‘cold’ spots for the location of business be situated? • Who will benefit from these developments? These are all important and difficult issues to analyse and resolve. Evolution is a process of unfolding, self-transformation and self-organisation; it follows oscillations and the appearance of events in orderly succession over a long period of time. This process is irrevocable, and the historical card, song or CD is played only once. Complexity and evolutionary economics deals with the actions, reactions, competition among various expectations, strategies and whims of economic actors. It is also concerned with the phenomenon of how these fit into the aggregate out-ofequilibrium and everchanging picture. With this in mind, it is clear that relations between economics and evolutionary biology were known and recognised a long time ago. However, these interactions were put aside by many researchers in the field of economics. Evolutionary economic geography deals with the evolution of economic units (such as firms, industries, clusters, cities, regions, states and integration arrangements among states) in geographical space. It considers not only their location and relocation in physical space, but also their creation, operation in the market, growth and demise over time. For instance, where a firm moves in business and in space depends on the starting position: where was its location and how was it created? Thus, history and expectations may assist in the explanation and comprehension of the (likely or wanted) future. Evolutionary economic geography differs from the traditional model in several important dimensions. The new evolutionary model makes the case that production specialisation in a given locality is based not only on certain local comparative

Evolutionary Economic Geography

4

advantages, but also on a self-reinforcing lock-in effect, path dependence, accumulated knowledge, agglomeration, clustering and linkages (indivisible production). In addition, while the traditional models reason that a reduction in trade costs among locations may favour local specialisation, the new evolutionary economic geography claims that the effect on local specialisation is ambiguous. The final outcome is industryspecific and depends on the functional intra-industry production linkages, market structure, consumer preferences (homogeneity of tastes), factor market (availability and mobility of factors and flexibility in prices) and expectations. Evolutionary and complexity economics is compounded with constant shocks, uncertainties and dynamics, as well as evolving changes on an immense scale. Persisting disequilibria are a big game in town. One may feel uncomfortable with this situation, but the evidence against it is not convincingly presented. Evolutionary concepts consider economic changes in conditions of bounded rationality. It is a constant search for optimal solutions in conditions of scarcity, competition and endemic uncertainty. The process deals with creation, differentiation, selection (including mistakes), adaptation, innovation, retention, amplification, replication and constant search for creative– destructive optimal solutions in a pool of many and various possibilities. Compared with neoclassical equilibrium theory, complexity and evolutionary theory of economic behaviour is another general and conceptual look at how the world works. This evolutionary theory is concerned with: • uncertainty, against which one cannot get insurance; • (micro-)analysis of endogenous technological change, innovation and their spillovers; and • non-maximising behaviour of profit-oriented firms. This introduction sheds light on the continuing importance of research efforts, public debate and the search for appropriate policy to face the challenges of spatial distribution of economic activity in the light of evolutionary changes. Let us now see what theory says about evolutionary economic geography. Notes

1 The Gini index (Figure 1.1) compares the actual distribution of activities with an even (standard) distribution. It measures the degree of similarity or inequality in a distribution (for example, spatial location of firms or income). This index is a ratio between of area between the Lorenz curve of distribution and the curve of the uniform distribution. Zero measures perfect equality, while perfect inequality corresponds to 1 (one region has all the firms or one person has all the income). Applied to the spatial concentration of firms, this index does not take into account the size of firms. Some industries may be organised in a few large firms (chemicals, finance, car assembly), while others may be composed of a number of small firms (food, textiles).

Introduction

5

Figure 1.1 The Gini index

2 Theory Anyone who hears and obeys these teachings of mine is like a wise person who built a house on solid rock. Rain poured down, rivers flooded, and winds beat against that house. But it did not fall, because it was built on solid rock. (Matthew 7:24–25) 1 Introduction Where economic activity will locate in the future is one of the most important and challenging questions in economics. Progress in technology, changes in demand and certain moves towards a liberal economic policy create new challenges for theorists, policy-makers and business executives. As certain new and a number of old economic activities became ‘footloose’, highly mobile, fragmentable and internationally connected, one of the most demanding and intricate questions in such a situation is where firms and industries will locate, relocate or stay. The purpose of this chapter is to provide an overview of the theoretical foundations of the field of spatial economics, especially its evolutionary side. It considers factors and conditions that influence decisions on where to locate a business or an industry in an evolutionary space. A research agenda (Section 2) is followed by considerations about relations between economics and geography in Section 3. Basic concepts in biology, economics and evolution are found in Section 4. Theories of location (Section 5) are presented without concern about the ownership and control of firms; foreign ownership issues are covered in Chapter 5. Complex systems and evolutionary economics are the focus of Section 6. Section 7 discusses various locational possibilities in space. Select spatial units, i.e. clusters, cities and regions, are analysed in Section 8. The impact of history and expectations on the location of firms and industries is described in Section 9. Section 10 deals with the effect of war on the locational issues, while Section 11 concludes the chapter. 2 Issues Spatial economics or the geography of production and consumption studies the spatial relation among economic units. This academic field has been separately and often indirectly analysed as a sub-branch of a number of subjects within economics and geography. It has, however, been treated with different intensity and depth within those research disciplines. Contributions are scattered widely across fields which include: microeconomics, planning, development, economic geography, regional science, urban

Theory

7

economics, location theory, industrial organisation, international trade,1 foreign direct investment (FDI), transport economics, business economics, innovation studies, public finance, price theory, imperfect competition, labour economics, environment and resource economics. The (small) common research denominator in all these areas of academic enquiry was the spatial dimension in terms of geometrical distance in metres. This was usually seen as an opportunity (power to be exploited), a medium for interactions and a limitation. Space and geographical distance may be considered in terms of trade costs, i.e. all costs linked with getting the final good to the consumer (tariffs, quotas, non-tariff barriers [NTBs], transport, storage, distribution, information, contract enforcement, insurance, exchange and banking), differences in language, culture, political system and risk premia. It may also be seen in terms of relations (networks) among economic agents. In a cluster or an industrial district, for instance, firms, labour and institutions interact in formal and informal ways. They influence one another and they learn from each other. As a multidisciplinary research field, new2 spatial economics has tried to integrate these previously separate research areas.3 This is in line with the new trend in science which includes merging various methods in research. 3 Geography and economics Dialogue between geographers and economists was minimal for a long time. Neoclassical equilibrium economists lived in the world of free-market forces and outcomes based on the rational choices of economic agents that brought equilibrium outcomes in space (these outcomes are exemplified and defined by an absence of further change). Limited intervention was accepted in the cases with market failure. Geographers, on the contrary, looked at the (capitalist) market economy as something that is in a perpetual state of conflict and disequilibrium, rather than harmony. The nature and composition of the economic process and decisions by players are shaped and directed by the social, cultural, administrative and political context. Hence, geographers were interested in studying and understanding the complexity of people’s livelihoods. A cultural turn with reduced economic content in part of (economic) geography became apparent after the ‘quantitative revolution’ in economics during the 1960s.4 Economists started increasingly to write professional articles for each other. To some observers, these highlight technical rigour and virtuosity, rather than relevance for life.5 Within this general set-up, there were sharp criticisms between economists and geographers. Geographers viewed economics as too narrow and naive in its analysis, as market forces were settling everything on their own for everyone’s benefit. Conversely, many neoclassical equilibrium economists disregarded geographers as the ones with a lack of quantitative rigour in analysis and criticism, as well as those who were too negative about the principal neoclassical economic equilibrium workhorse: liberal markets that put all things right (almost) on their own at the end of the day. Martin (1999) critically surveyed the new ‘geographical turn’ in economics. He argued that the ‘new turn’ is not more that a reworking of the traditional location theory and regional science. The location models in the tradition of classics such as von Thünen, Weber, Lösch, Christaller and others that feature high in the ‘new economic geography’ have long disappeared from the research agendas of geographers. Geographers, never

Evolutionary Economic Geography

8

particularly persuaded in the neoclassical prediction of regional convergence, are more interested in real economic landscapes with their specific and complex histories and particularities, than in abstract models of an imaginary economy in space. The question of how the economy fits into space cannot, therefore, be answered for geographers by abstract geometric solutions and mathematical models. Geographers see more potential in the evolutionary economics that considers the change in economic structure, history, institutions and technological change. Abstract models, according to Martin, ‘attached to the real economic landscape by the thinnest of conceptual and empirical threads do not, in my view, offer particularly convincing or reassuring material from which to weave policy prescriptions’ (p. 84). The lack of communication between the two academic camps changed slightly from the 1990s. When the debate started, however, some geographers greeted interest by economists with scepticism, while others were reserved. Neoclassical equilibrium economists, in general, continued with their work as usual. However, formal dialogue and trading ideas between the two sides started with the publication of the Oxford Handbook of Economic Geography (2000), edited by Clark, Gertler and Feldman, as well as the launching of the Journal of Economic Geography in 2001. In addition, during the period 1987–2004 Elsevier (Amsterdam) published four volumes under the title Handbook of Regional and Urban Economics. More recently, a part of this dialogue has continued in the set of readings entitled Economic Integration and Spatial Location of Firms and Industries (Jovanović, 2007a, 2007b, 2007c). The general neglect of spatial economics (geography of production) in mainstream economic theory was not because this research field was uninteresting, but rather because the issues were traditionally regarded as intractable. David Ricardo’s comparative advantage model provided insufficient answers to the problem. New research tools such as: • increasing returns to scale (which disturb the harmony of the whole), • imperfect competition (consumers’ love for variety and firms’ competition for limited production resources), • production linkages (presence of intermediate goods and services), • multiple equilibria (with centrifugal and centripetal forces, as well as potentially sudden ‘catastrophic’ relocation results when the economic landscape changes from a relative symmetry to complete agglomeration [all firms locate only in one spot in geographical space]), and • expectations were introduced into the field of general and spatial economics in the early 1990s. These methodological tools also helped to explain why firms form clusters. Such new developments did not mean the birth of a new subject: spatial economics has always been important. However, the ever-increasing demand for quantitative vigour is making considerations more difficult and results harder to analyse, explain and comprehend. The introduction of new analytical tools assisted spatial economics in finding its proper place in mainstream economics and becoming a hot research topic. Mutual understanding, learning from each other, professional respect, communication, criticism and cooperation among economists and geographers are necessary in the area of economic geography. Economists need to step out of their ‘old analytical box’ of clean

Theory

9

and lean abstract theoretical, mathematical and aspatial models that bring the analysed system into equilibrium. These models were based on assumptions that rendered analysis totally unrealistic and useless to a geographer. A geographer would argue that societies and real geographical space are so complex that they cannot be expressed only in mathematical formulae. ‘Pure’ economic models refer to geography only to the extent that it supports economic analysis of the ‘market clearing logic’ and equilibrium based on very specific (unrealistic) assumptions that ease mathematical modelling. Such reasoning dangerously damages practical relevance. Economists need to expand their research enquiry out of the usual issues that refer to entry, exit and total factor productivity. Certain economists already broaden analysis with issues such as organised interests (institutions),6 history, firm heterogeneity in size and in the use of factors, multiple equilibria, evolution, technological and time–space trajectories, spatial location theory and various imperfections. Geographers need to come out of their ‘old analytical box’ in which they were interested in economics only to the extent that it was assisting them to understand and explain spatial organisation of societies. Instead of ignoring or bypassing one another, economists and geographers need to try to talk about topics of mutual interest, potential agreement and obvious difference. They may also explore the relevance of complexity theory7 for the solution of common problems. Economists ought to learn more from the long history of (economic) geography, which dates back to the times of Johann Heinrich von Thünen (1780–1850). Geographers need to learn more about issues that include oligopolistic competition and economies of scale. However, if the two subjects are so specialised and too disciplinary to the extent that common goals cannot be found and that synthesis between the two fields is impossible and, perhaps, unnecessary, the two sides still need to talk to each other. They need to talk to others, too, for instance to sociologists about issues such as isolation and inclusion. Fortunately, discussion and cooperation between the two ‘major camps’ has already started. In spite of many methodological and political differences between economists and geographers, a critical exchange of views, insights and experiences may at least ease communication and mutual understanding in the future. All these general exchanges of arguments, views and concerns are now being compounded with the evolutionary or structuralist–evolutionary theory applied as much as possible to economics and geography of production. 4 Basic concepts Introduction The spatial location of a firm is an issue only in the situation with market imperfections. Without market imperfections such as transport cost (so that location matters),8 the production location decision becomes irrelevant. This is to say that without market failures, firms can split into units of any size and operate in all locations without any cost disadvantage. Recent empirical research found ‘that the degree of production indeterminacy is greatest when trade barriers and trade costs are relatively low’ (Bernstein and Weinstein, 2002, p. 73).

Evolutionary Economic Geography

10

While considering feasible locations for its business in the situation with market imperfections a firm may, according to neoclassical theory, either: • consider cost minimisation (prefer the spatial location with the lowest-cost production function and ignore the demand side); or • emphasise profit maximisation (demand/revenue and neglect everything else). In practice, neither of these choices on its own necessarily offers the optimal location in terms of profit. Therefore, when considering possible locations for its business, a firm studies both options simultaneously. The firm attempts to maximise profit (in the long term), taking into account the operating ‘cost penalty’ linked with alternative geographical locations. Apart from these basic, static, ‘mechanical’, ‘clean’ and direct neoclassical elements that are strongly entrenched in theory, it is essential to employ at least two other approaches in the consideration of spatial location of firms: • The behavioural approach This is employed because of the internal dynamics of a firm. The real situation is charged with imperfections, risks, ambiguity, diversity of beliefs and strong endemic uncertainties. Therefore, profit optimisation or maximisation may not be the ultimate choice and goal for the firm. In the situation with bounded rationality and imperfect information, firms may choose to have a profit ‘satisfactory’ behaviour as their business goal. • The institution–environment-related approach Firms make their decisions actively in a dynamic and risky environment in which there are multiple choices and limitations. Firms have to obey laws; fulfil obligations (taxes) and negotiate subsidies with the government; negotiate contracts with suppliers, clients and unions; organise and control production and distribution; consider new products, technologies and markets; decide about marketing and prices; and the like. In these circumstances, small and medium-sized enterprises (SMEs) often have relatively little leverage, while large firms fare much better. In the distant past, the resource endowment of a specific location, together with the available technology, was the principal determinant of location for the manufacturing industry. In the new situation, in which most determinants for the spatial location of firms and industries are mobile and manmade, the location determinants for (footloose, fragmentable, mobile and internationally connected) industries and firms include the following five sets of considerations: • Costs and prices (C+P): availability, substitutability, quality and prices of inputs (raw materials, energy, labour); cost of market access (trade costs);9 economies of scale; utility costs; infrastructure (physical, social and cultural); transport cost of inputs and output; earlier sunk costs in other locations (not yet depreciated); availability of investment funds; and cost of the project. • Demand (D): its real and potential size, its growth and consumer preferences. • Organisation and technology (O+T): input–output production links; externalities; competition and market structure; location decisions of other competing or supplying firms; technology and the speed of its change; and local research and development (R&D) resources and capabilities.

Theory

11

• Policy-related factors (P): incentives; taxes; subsidies; public procurement; rates of exchange,10 permissions; national defence; environmental standards and mandatory or voluntary unionisation of labour. • Social factors (S): spread of literacy and bureaucratisation of the society;11 system of education of management and training of labour; brain drain; income distribution;12 general quality of life; and retirement patterns.13 One should also include here the whims of investors. L=f(C+P; D; O+T; P; S) (Equation 2.1) Hence, the decision about location (L) of a firm or an industry in space is in the function of these five composite variables (Equation 2.1). Making a decision about the location of a firm or an industry is a complex, uncertain and risky task. It comes as no surprise that the national or regional production geography does not always change very quickly. Biology, economics and evolution In the preface to the eighth edition of his Principles of Economics Marshall stated: The Mecca for the economist lies in economic biology rather than in economic dynamics. But biological conceptions are more complex than those of mechanics; a volume of Foundations must therefore give a relatively large place to mechanical analogies; and frequent use is made of the term ‘equilibrium’, which suggests something of statical analogy. (Marshall, 1961, p. xii) Economics has always had an association with other sciences. Sometimes this association was too strong and long, as is/was the case with physics and mathematics. This liaison prevented the economic discipline emerging from its strict research equilibrium box. The association of economics with biology is not new: its legacy is more than two centuries. One has only to recall Thomas Malthus’ controversial study An Essay on the Principles of Population of 1798, its modern application to climate change, or the messages from the Club of Rome during the 1970s. What is new is the search for answers for which the neoclassical equilibrium economics is no longer a suitable theoretical model. Evolution is a gradual process in which something alters over time into something else with a different and usually – but not necessarily – more sophisticated or better form. Charles Darwin’s theory of evolution (in fact evolutionary outcomes), as presented in On the Origin of Species (1859), claims that all living creatures are related. They all descended from a common ancestor. Successful species breed in time and space, and consume the vital resources that are available in the ecosystem for their survival. Once the resources become scarcer (less free), the species mutate. Favourable genetic mutations are preserved (rewarded), accumulated and passed on as they assist in survival. Natural (biological) selection produces the survival of the fittest. Over time the original organism is changed and transformed into a completely new one. Darwin understood inheritance in a relatively loose way. Modern advances in molecular biology, biochemistry and genetics, especially after the 1950s, put this theory

Evolutionary Economic Geography

12

into crisis. Molecules, even in the simplest living organisms, are so numerous, rich and complex that they cannot find any parallel whatsoever either in the non-living world or in any machine created by men. Evolution and change may produce a tendency of organisms to endure hardship, rather than benefit from mutation. Some may even argue that monkeys evolved from the degradation of men. Even in philosophical terms, relations among species, available resources, changes in technology and the balance of the earth’s ecosystem are almost impossible to explain under evolution’s ever-changing view. Hence, evolutionary theory is only one approach or opinion about the origin of life and the way evolution takes place in nature. Certain facts can be readily supported and accepted, but not others, as there are missing links in the ‘evolutionary chain’ of events. Lacunae in our knowledge are in the shadow of ‘the benefit of the doubt’. Theology fills these gaps with faith. There is a new connection of economics with psychology and neuroeconomics, in addition to biology. The Journal of Evolutionary Economics and the Journal of Bioeconomics have been associated with this type of economic considerations from 1991 and 2000, respectively. This new relationship is much more demanding (and perhaps problematic) than was the case with its association with technical sciences that started a century ago. It is quite difficult to quantify the new models, which is in contrast with the traditional equilibrium approach. It is even more difficult to establish correlation or substantial relation (not to mention causation) between forces that operate in the natural world and the ones that order social relations. Although analogies and intellectual exchanges between biologists and economists can be quite risky, confusing and perhaps misleading, an economist may be inspired by the purpose and direction in the evolution and development in the natural world. In any case, one may say at the outset that there are at least two clear resemblances between economics and biology: • there is a resistance to change (often quite strong); and • an external shock is often necessary for a fundamental and lasting change. As for profound differences between biology and economics, some are also obvious from the start: • Knowledge: its creation, accumulation, expansion and transfer. Nothing like this exists in nature. • Purpose: apart from mere biological survival and multiplication, members of a society normally have higher purposeful and objective-seeking behaviour. They intend to reach higher social utility functions. • Exchanges and ‘contracts’: in the natural world, these are not strictly enforceable. This is not the case in many contractual situations during exchanges among people or other social subjects. Evolutionary concepts consider economic changes in conditions of bounded rationality. There is a constant search for optimal solutions in conditions of scarcity, competition and endemic uncertainty. The process deals with creation, differentiation, selection (including mistakes), adaptation, innovation, retention, amplification, replication and a constant search for creative– destructive optimal solutions in a pool of many and various possibilities. Human inventiveness and rationality have some impact on the speed, direction and nature of the evolutionary process, but they do not substitute it. The

Theory

13

evolutionary concepts try to grasp the (purposeful) behaviour of players in the economic arena (locations, firms, industries, people, institutions) over time. Of particular interest is the shape and direction of their changes: actions, reactions, adaptation and the general and specific course of interaction and change in time and space (the natural and organisational environment). What are the gifts of locations, firms, industries, people and institutions to sense and identify opportunities? How willing, motivated and able are they in adapting to new challenges over time, in experimenting and translating these perceptions into profitable actions? Why do locations differ? How is their ‘competitive position’ evolving relative to others, and how does it change over time when faced with new opportunities and obstacles? Outcomes of changes in the economic environment are usually unpredictable in the medium term and totally unpredictable in longer term.14 Nonetheless, these concepts are quite useful and persuasive regarding the explanation of the location of certain new and footloose industries. These concepts, even without a commonly accepted meaning or priorities, include a number of notions and metaphors that mark the evolutionary process and its outcomes. Even though certain of these concepts on their own may seem too simplistic to some, they are useful in making this complex system at least partly tractable. These concepts include such diverse matters as: • multiple and rapidly shifting equilibria; • non-linear economic processes; • human agency15 or creativity (rather than deterministic mechanism); • learning and knowledge; • innovation and novelty; • endogenously and exogenously generated changes; • self-organisation (spontaneous appearance of order after oscillations);16 • coordination failure; • differentiation; • selection (including mistakes); • adaptive and purposeful behaviour; • amplification; • retention; • replication; • fitness; • cumulative causation; • acceleration; • increasing returns; • positive and negative feedback; • neighbouring effect;17 • spillovers (unpaid side-effects of economic activity); • untraded interdependency; • chance (or luck); • surprises; • randomness; • strong endemic uncertainty; • incomplete information; • ambiguity;

Evolutionary Economic Geography

14

• indeterminancy; • path: creation, dependence and destruction; • lock-in effect;18 • organisational and business routines (organisational memory) and competences; • differences in opinion among experts and in expectations among businessmen. Martin and Sunley (2007, pp. 577–8) point to an important difference between complicated and complex systems. A complicated system has a large number of component parts, like a mechanical watch or a car. These parts may be taken apart and put together again. This complicated system may be explained in terms of its parts and their relation in operation. However, a complex system is different. It has non-linear relations among its parts. This system may not be explained in terms of its component parts alone. Effective complex systems find and keep a balance between inertia and innovation. Too much flexibility may bring chaos, while an excess of rigidity may bring motionless equilibrium. A complicated system need not necessarily also be a complex one, while a complex system need not necessarily be a complicated one. Nonetheless, a complex system may easily be complicated too. A complex, evolutionary and adaptive system (such as geographical space) has properties that include the following: • Borders between a complex system and its environment may not be easily discerned. They are not fixed. The system is open for an exchange of energy, matter and information with its environment. The borders of the system often depend on the purpose of analysis and availability of data. • There is a limited scope to decompose a complex system into stable and self-sustained components. • A spontaneous change in the behaviour on the micro level may have a strong impact on the macro structure. • Complex feedbacks among components are non-linear and various. • The system has adaptive potential, both for changes that emerge from within and for those that arrive externally. • The system is non-deterministic. Even if one is fully familiar with the operation of all individual components, the behaviour of the whole system may not be predicted with a high degree of reliability. This does not, however, mean that the system behaves in a chaotic way. It just escapes the ‘standard’ linear way of thinking and expectations (‘if.., then … ). Before embarking upon considerations of theories of location, a few words ought to be said about physical and social technologies. Physical technology refers to production techniques and energy sources that assist to transform matter and information by the use of energy into useful things and services. Economic as well as general social evolution gave birth to social technologies (or socio-economic technologies). These technologies are as important for the wellbeing of people, as are ‘hard’ physical technologies, but they do not come to the forefront of our thoughts, as is the case with ‘hard’ technologies. Social technologies organise people to do things in a more or less cooperative way. These ‘soft’ technologies include: states, towns, villages, markets, firms, money (cash and electronic), banks, government policy, rule of law, courts, police, army, group norms, culture, customs, double-entry accounting,

Theory

15

religion, even democracy, fascism, socialism, federalism, etc. They were developed over millennia to solve social problems, from the times of hunting groups and tribes over religion to transnational corporations (TNCs).19 5 Theories of location International trade and location of production The study of the location of production (spatial economics; economic geography) has a long history. In spite of general interest, there is no wide consensus in the literature about the factors that influence the location of new businesses. The problem was noted, for example, by Ohlin, who wrote that ‘the theory of international trade is nothing but internationale Standortslehre’ (teaching about international location) (Ohlin, 1933, p. 589). In the same vein, Isard wrote that ‘location cannot be explained without at the same time accounting for trade and trade cannot be explained without the simultaneous determination of locations. … trade and location are the two sides of the same coin’ (Isard, 1956, p. 207). A country’s or a region’s level of income and its structure of production (size, scope and variety of firms and industries) depend on at least two factors. The first includes a country’s endowment of primary and derived stock of factors such as land, labour, capital, infrastructure, technology, as well as production, organisation, entrepreneurial and control skills and competences. The other element is the way the country or region interacts with the external environment; in particular, the extent, level and form of its access to the international market for goods, services, capital and knowledge. Let us now turn to the principal theories of firm location. Acknowledging the arbitrary character of our classification and its limits, we shall look into the issue of how and where firms and industries locate in space, while ignoring the origin, control and ownership of the firm. The issue of foreign ownership and control of firms, i.e. TNCs, is considered in Chapter 5. A starting point in location (spatial) theory and economic geography is the national endowment of primary factors of production. Adam Smith and David Ricardo referred to the geography of production and spatial economics in an indirect way. Their theory of international trade looked at countries as just points in space.20 A country specialises in the production of a good for which it has an absolute advantage in production. Hence, according to Smith, an absolute advantage determines the geographical location of production. Ricardo’s argument is similar, but in his model a country specialises in the production of goods for which it has a comparative advantage in production. Regarding land, the fertile areas are put into cultivation first, while less productive land enters use only as demand for agricultural goods increases. The models by Smith and Ricardo depend on constant returns to scale and on the assumption that competitive prices guide spatial resource allocation. Hence, free trade allows countries to gain, because they specialise in production that they carry out comparatively well. Such an assumption is challenged by the view that there are in reality important market distortions such as: monopolies; externalities (pollution); production linkages (backward and forward); temporal relations between present choices about

Evolutionary Economic Geography

16

location and structure of production and future production opportunities; coordination failures; and controlled prices and rigid wages that do not tell the truth about social costs. As such, they send wrong price signals to entrepreneurs and may lead to the conclusion that trade liberalisation may make matters even worse! Thus, some argued that protectionist policies were required for: • the security of jobs, as was the case during the 1930s and ever since; or • the safeguard of infant industries, as during the 1950s; or • rent-shifting in oligopolistic markets, as during the 1980s; or • the safety of the environment and products (are they toxic? do they contain genetically modified substances?), as well as social (labour, human rights) standards and practices, as is currently the case.21 International trade, it ought to be recalled here, comes as a response to causes that are based on national differences in the following essentials: factor endowments (factor proportions); demand (size, sophistication and fickleness); home market effect; production functions (technology and efficiency); economies of scale; growth rates and increase in income; income distribution; market structure and competition; production and factor taxes; rates of exchange; change in tastes; agglomeration and clustering of production; trade costs; and demographic factors. In addition to these classical models, the standard neoclassical factor endowment or Heckscher–Ohlin model refers to locational fundamentals. The national endowment of factors (land, capital and labour) and their relative proportion decides the location of production. A relatively abundant national factor of production (compared with other factors) determines what a country produces and exports.22 Conversely, a country will import goods whose production requires factors that are scarce in the national economy. Considering cases of ‘localisation’ of industry in one or a few countries, Ohlin (1933, p. 133) simply argued that ‘it must be shown that costs of production on the basis of existing factor prices are lower than in other countries: certain factors are cheaper here than abroad, which accounts for that condition’. In an empirical study of trade patterns of over a hundred economies, Leamer (1984) confirmed that the factor endowment model can explain these trade patterns rather well. This model elegantly assumes that there are no specific factors and that production functions are identical in all countries. As such, it does not give the answer to the whole issue, for it cannot explain intra-industry trade. This type of trade (within an industry), which includes a large part of trade among developed countries, is not based on differences in factor endowments among them. The Heckscher–Ohlin model is more successful in explaining locational patterns in aggregate sectors (agriculture, manufacturing and services) than in disaggregated industries within, for example, manufacturing such as production and trade in passenger cars. The factor endowment model has been one of the two principal singlecause models dealing with specialisation, trade and spatial location of production. The other principal approach takes account of differences in technology, economies of scale and learning by doing. It was conceived by Krugman (1979, 1980) and later developed and expanded both by him and by a host of others. In a nutshell, a variety of causes including history, expectations, their competition and changes in technology (not only factor endowment as before) influence a country’s specialisation, spatial location of production and, hence,

Theory

17

trade patterns. Differences in technology and increasing returns to scale change productivity levels. This in turn alters the bases for specialisation, spatial location of production and trade patterns. In the case where there are both increasing returns to scale (falling costs) and transport costs, there is an incentive to concentrate production. By spatially concentrating production near its largest market, a firm can both profitfrom economies of scale and minimise transportation costs. Economies of scale have a stronger impact on the location of modern, knowledge-based production than was the case before the Industrial Revolution. Sources of economies of scale (as well as of ‘natural monopoly’) include: • High set-up or sunk costs: infixed capital, R&D and marketing. They represent costs that a firm must invest to produce the very first unit of its output. These are historic expenditures that cannot be recovered, but the ones that yield benefits in the future. • Learning effects: the more people learn how to use one Microsoft program, the easier it is for them to learn and adjust to other programs. • Coordination and network23 effects: the value of a good or a service to a consumer is linked not only to the properties of that good/service, but also to the number of others that use it. The more people use mobile phones, faxes, e-mail, Internet, Microsoft Windows, electronic banking or credit cards, the greater is the utility of that to all network users. While one cannot dispute the importance of factor proportions for the location of production and some processing of primary resources, this approach cannot explain the location of a footloose, mobile and internationally connected industry. In addition, this model does not consider market structure, economies of scale, demand conditions or trade costs. The Heckscher–Ohlin theory cannot explain the location of an industry in countries with high mobility of factors, labour in particular, such as the US and China or in countries with a broadly similar endowment of factors (France and Germany). A common feature in the standard trade theory is that it considers states while it utterly ignores firms or regions – that is, the spatial distribution of economic activity within a country. This gap is filled by spatial economics. Since the 1970s, it has been noted and highly publicised that spatial reallocation of textile manufacturing has taken place, from the developed market economies to unskilled and cheap-labour-abundant developing countries and China. The US bilateral general trade deficit with China is particularly worrying for American politicians. If one scratches the surface of this ‘Chinese deficit story’ in the US, one may discover that beneath it is a hidden reallocation of low-cost manufacturing from other Asian countries to China. China became attractive both as the location for production and as the source of supply. Why has this mass relocation of manufacturing, ‘hollowing out’ of American, European or Japanese economies and exodus of jobs to Asia, and China in particular, taken place over the past two decades? Why not earlier, say before the 1980s, even though wages were low, very low, in Asia for a long time in comparison to the developed countries? This is a complex problem that ought to refer to various issues. These include: • the policy of national openness to the location of foreign firms in China from the 1980s; • standardisation and the possibilities to geographically fragment, unbundle and offshore parts of the production process (the number of footloose tasks increased); and

Evolutionary Economic Geography

18

• a reduction in the difficulty of coordinating production over space and time zones (this is a technology created and used by the developed world). More intermediate goods, and processes became internationally tradable. Choice to producers regarding the location of production became wider. The more the choice, the better for everyone, or so say the highly recommended story in the developed world and the economic argument. Starting with fax machines in the 1970s and continuing to other means of fast and cheap communication, one may theorise that this international coordination of production became manageable and that certain agglomeration forces became weaker in the developed world. The new means of communication enable the transfer of some service jobs from the US, for instance call centres or data processing, to locations such as India or the Philippines. The opposing force that keeps jobs, good jobs, at home in the developed world is the policy or process of domestic education of the workforce, and innovation. Policy-induced changes such as the disappearance of quotas on exports of textiles to the US and EU in 2005, as a part of the Uruguay Round deal, speeded up the international reallocation of the textile industry to China. Xiansheng (south of Shanghai) is a flourishing textile export hub that profits enormously from this change. Millions of jobs throughout the world, both developed and developing, are endangered. The textile industry in the developed world was protected for over thirty years, with various tradedistorting agreements that sheltered domestic jobs. These protectionist deals were supposed to provide ‘breathing room’ and ‘advance notices’ to the endangered domestic textile-producing areas to adjust to a forthcoming liberal trading regime. Nonetheless, this protective shield did not prevent a steady fall in employment in textiles and clothing in the protected countries. The question is: how much longer does this industry in the developed world need in order to adjust? Can or should it adjust at all?24 While there was a controlled increase in imports of textiles and apparel from China from 2005, the overall textile imports in the EU, for example, changed little. This controlled increase was up to 10 per cent a year until the end of 2007 (a part of the 2001 deal to let China into the World Trade Organisation [WTO]). The increase in China’s importance as a source of imports was at the expense of other Asian and African clothing exporters. The location and volume of consumption changed little in the developed world; however, the location of production (and exports) evolved in favour of China. From 2008, the EU and China have a joint monitoring system over most types of garments entering the EU market. Protectionists are very inventive. Certain intentions by the EU and the US to reintroduce restrictions on imports of textiles from China are explicit recognitions that their respective domestic textile industries failed to adjust despite long-term protection and a decade-long advance warning that protection would go in 2005. Severe antidumping duties and obscure methods to calculate them are ill-advised, for several reasons. It ought to be kept in mind that China is among the biggest world importers (it also exports a lot, but the trade balance is positive). Both the US and the EU argue in favour of free trade. China has a comparative advantage in producing textiles and is using it, just as Kuwait did in the production of crude oil. Bearing these facts in mind, what right have the EU and the US to lecture others about the benefits of free trade? What right have they to complain that other countries are too slow to open their markets? Is it hypocritical to blame others for

Theory

19

their own home-adjustment failures? The best way to show the benefit of liberal trade to others has always been to lead by example and to be in the forefront of market opening. Import quotas in the developed world have complicated the manufacturing and purchase of clothing and textiles. It makes business sense to rationalise this in the new situation and to buy from a single location, a practice that big retailers prefer. Therefore, China, India and Pakistan may be preferred choices as locations for production and sources of supply. Laura Jones of the US Association of Importers of Textiles and Apparel commented that China does not compete just on price: indeed, it is not the lowest-cost producer of some clothing items. Its biggest advantages are its industry’s rapid response, reliability, business-like attitude and keen understanding of customer demand. It is very easy to do business there.25 Even though the Multifibre Agreements were condemned as clear protectionism, the system of carving up international markets served many developing countries well. It provided opportunities for less-productive and weaker manufacturers to stay in business. However, they must now make the adjustment to an open ‘global’ market. The marginal producers from developing countries have good reason to fear that there will be a relocation of this industry to China from their home countries. The consequence is that many of them are trying to specialise in select market niches. However, a general problem and weakness in China is a serious shortage of raw materials, including cotton. At the same time, some vertically integrated firms agglomerate in certain countries in spite of unfavourable ‘technical’ and relative factor intensities. An example may be found in firms that are in the production of rubberbased products for motor vehicles and aerospace. Indonesia, Malaysia and Thailand are the countries that are abundant both in rubber and in unskilled labour. However, the manufacturing of rubber-related products is often located in the developed European countries, Japan and the US, where it is close to the assembly of final products. Advantages of close spatial proximity to demanding and big clients are stronger and outweigh the importance of factor costs. This contradicts the theoretical predictions of the standard neoclassical trade and location model. A policy implication from this example is that trade liberalisation and economic integration of countries need not always guide the location of industry only according to the principles of comparative advantage and costs (Amiti, 2005, p. 828). In their survey of trade theory, Leamer and Levinsohn (1995, p.1363) stated that ‘one rather awkward assumption that cries out for change is that of equal numbers of commodities and factors. After all, we really don’t know how to count either.’ The existence of more goods than factors, economies of scale, multiple equilibria, trade costs and differences in technology present the principal obstacles for predicting the spatial location of production and specialisation. Harris (1954) argued that manufacturing firms tend to locate in places where they have favourable access to the market. He showed that the industrialised US regions were also the areas of high market potential (a large market attracts producers). The idea is that firms choose to locate in regions that have good market access. However, this access is good in areas in which many other firms already exist and have chosen to settle. There is a selfreinforcing, snowball or herding effect in which the size of the market attracts still

Evolutionary Economic Geography

20

more firms. Larger markets host more firms, more firms produce a larger variety of products, while competition increases productivity and reduces prices. Such areas become attractive locations as export bases. The underlining feature of this approach, as in most models of spatial economics, is that it is based on some kind of economies of scale that rely on a spatial concentration (agglomeration) of certain production activities. Such agglomeration economies represent increasing returns in spatial form. In addition, agglomeration (clustering) of certain activities in a few locations is one of the most distinctive features of production. The question with agglomeration economies is: which geographical spot is going to dominate in production? The analytical challenge and difficulty with agglomeration is that it introduces multiple equilibria. Resources are limited so firms compete for them; firms are heterogeneous (they differ in size and in their marginal production costs); consumers love variety (often the more, the better) so producers and traders need to adjust to these preferences; multiple equilibria fuel different expectations by players and shape economic space; and there are costs of trade. Analytical tools to handle these elements simultaneously with increasing returns and imperfect (monopolistic) competition were missing until the 1990s. In the case of France, Crozet et al. (2004, pp. 50–1) found that firms from Belgium, Germany and Switzerland initially preferred to locate their operations in the frontier regions, near their home market. However, once they had learned how to operate in France, it became more important to be physically closer to final demand and less relevant to be in locations that shared common attributes with the home country. Similarly, the main foreign host regions in Europe for French firms in terms of the number of investments were the eastern region of Spain, the Italian Lombardy region and south-east England (Mucchielli and Puech, 2004, p. 36). In the presence of economies of scale, monopolistic competition and transport costs, countries tend to specialise and export differentiated goods for which they have a relatively large domestic market (economic integration secures and enlarges ‘local’ markets). It is this large domestic demand or home market effect (Linder, 1961; Krugman, 1980) that induces local production, rather than domestic factor endowment. The response by firms to strong local demand for a product (and the possibility of making savings in trade costs) is so powerful that the country first produces this good for domestic consumption and then exports it. Countries have a competitive edge in the production of these goods and thus gain an advantage in foreign markets, while they import goods demanded by a minority of the home population. In this model, countries tend to produce and are net exporters of those goods for which they have a relatively large domestic market. If there are both increasing returns to scale and transport costs, there are strong incentives to locate and concentrate production of a good close to its largest market. The reason for this location of production (even agglomeration) is that one can profit from economies of scale and reduce transportation costs.26 However, in the case of diminishing returns to scale, strong home demand for a good will make it an import rather than an export item. The factor endowment or comparative advantage model would not predict the home market effect, but economic geography would expect such results. The US, Japan and Germany have – or at least had – the greatest comparative advantage in goods for which their home market is relatively big. Empirical investigation by Davies and Weinstein

Theory

21

(2003) found support for the existence of home market effects, as well as for the role and importance of increasing returns in determining the production structure in the Organisation for Economic Cooperation and Development (OECD) countries. The emergence of China as one of the principal world manufacturers (especially since the late 1990s) is somehow modifying this general observation, but one ought to note that China and India have each immense and growing domestic markets. Even though the model may seem attractive on the surface, it is based on many theoretical compromises. The model is silent on how to mobilise resources and integrate resources in a selected location. Are there intermediate goods? What is the nature and form of transaction networks? What is the structure and dynamic of the local labour market? Is the necessary technology available? Where from? Is innovation relevant? What is the role of substitution effects? What about education and learning effects? Even though the model is to a large part a ‘black box’ and narrow in its scope, it points research and policy attention to an important issue: the relevance of economies of scale and the local market size for the location of firms and industries. Home market effect and national demand may not explain all cases of business location and specialisation. For example, a small country such as Switzerland is highly specialised in research and production of pharmaceuticals and huge marine diesel engines (Sulzer) for container ships. The Swiss do not have a heavy national leaning towards diseases or hypochondria, neither is Switzerland a significant navy- or merchant fleetstrong country whose national demand would justify or explain the location and specialisation in these types of production. The size of national demand is not always translated directly into the location of production at the same place. Accumulated past knowledge and technology, various expectations about its future path, prices and income all play a role in the selection of location, together with the size and growth of the market. Domestic firms At the heart of von Thünen’s (1826) location theory are differences in land rents and land uses. This model is primarily concerned with the location of agricultural production. While Ricardo focuses on the difference in the fertility of land, von Thünen studies how different agricultural lands around a city (marketplace) bid for various uses. The model assumes a given isolated city (one consuming centre) and surrounding agricultural hinterland, homogeneous land surface, free and costless mobility of labour and identical tastes at all income levels. Land rents are highest in the city. From there, they steadily decline to zero at the outermost limit of cultivation (beyond that limit, land is a free good).27 Hence the rent in a given location totals the difference between the value of its yield minus the sum of production and transport costs. Under the above assumptions the model explains the kind of crop that would be grown at places located at different distances from the market. In other words, farmers decide on the type of production (expensive-to-transport vegetables and cheap-to-transport grain) by taking into account land rent and cost of transport. As transport costs and yields differ among crops, the result is a spontaneous development of concentric circles of production around the city. Land is allocated among different crops in a(n) (optimal) way that minimises production and transport costs of different crops.

Evolutionary Economic Geography

22

Von Thünen’s model is based on careful study of farming practices in northern Germany in the early nineteenth century. Such distribution of farmland is less obvious in the more intricate modern world, but around large urban centres there are certain remnants, such as dairy production and gardening.28 There is no answer to the question of whether vegetables can be shipped in both directions: towards the city (in exchange for cloth) and towards the outermost limits of cultivation (in exchange for grain). This early model does not consider the role played by economies of scale and the housing market (urban modelling). The model is, therefore, suitable for the analysis of a pre-railway and a pre-industrial society. The problem with von Thünen’s model is to find which good to produce in the given location. In the optimal plant site model by Weber (1909) the branch of industry is given, so the problem is to find the spatial location for production. Weber’s model takes the spatial location of markets, raw materials and population as given, and assumes that there can be only one location for production. The objective of an individual firm is to minimise the combined costs of production and delivery. When the production costs are independent of location, the locational problem relates to the minimisation of transport costs (TC) for inputs and output. This is shown in Equation 2.2, where m presents weight of goods in tons, t stands for cost of transport per ton/km, d is distance in kilometres, while i represents a particular good in transport. The key condition in the decisionmaking process about the spatial location of a firm refers to the minimal TC.29 (Equation 2.2) When key resources are used on a large scale in certain industries and are highly localised, this may affect the location of the manufacturing industry. These industries would be tempted to locate near resource sites. Hence, this case is ‘out of step’ with the spatial uniformity and the underlying principle of central-place location for some, but not all, kinds of manufacturing. The problem with the Weber approach to the location of new firms is in the initial linear and static locational needs of a firm, which may alter over time as the firm, industry, technology, consumers’ needs, income and tastes, or markets change and evolve. Weber assumes fixed and identical production functions at all locations (elasticities of substitution are zero). But firms always combine inputs and compare alternatives substitutes. This model does not take into account important location-related elements which include the cost of land and labour in the considered locations. There is also the possibility of a historical accident which can be coupled with the economies of scale, lock-in effect and agglomeration. Moses (1958) extended Weber’s work and integrated location theory with the theory of production. Coefficients in production functions were allowed to alter. This approach allowed investigation of the relation between substitution of inputs and geographical location of a firm. Hence, Moses’ approach gave different location results from Weber’s because it allowed for the existence of economies of scale. The problem with the Weber–Moses transport-only approach to the location of firms is that it does not consider market prices and profitofthefirm in the selected location. This is relevant as transport, logistics and overall trade costs often represent only a small share of total costs in most firms.

Theory

23

Geographical remoteness can be a handy scapegoat to some, who may use it to explain why peripheral regions are marginalised in economic and social terms. Hummels (1999) argues that geographical factors remain an important determinant of both trade and location. He suggests furthermore that trade composition patterns and FDI decisions can be sensitive to changes in the relative cost of different transport modes and information transfer. His proposition that international transportation costs have not declined significantly finds some support from Finger and Yeats (1976), who demonstrate that with the process of tariff liberalisation overseen by the General Agreement on Tariffs and Trade/World Trade Organisation (GATT/ WTO) the effective rate of protection in the form of transport costs exceeds that from tariff barriers for many goods. Hummels (2006) states that ‘Distance impedes trade to a surprising extent … typical estimates suggesting that doubling distances halves trade … costs are substantially rising in distance, but this effect has diminished over time’ (p. 19).30 Even though road haulage is not considered in Hummels’ article, it is interesting to note that in the period 1970–99 the value of trade in all products increased eighteen times, while the tonnage transported by sea and air combined more than doubled (pp. 21–2). Hence international trade became on average ‘lighter’ weight-wise, while transported goods happen to be dearer. Geographical remoteness is a shallow justification for a lack of competitiveness in many industries. Because of innovations in transport and communication technologies, these costs are rarely the most important determinant for the location of many businesses. In addition, the cost of ‘transport disadvantage’ on exports is already included in the local factor prices. Hence, a peripheral or geographically unfavourable location is neither an insurmountable obstacle nor a necessary condition for economic prosperity or the location of a number of firms and businesses.31 This is demonstrated in the cases of countries such as Japan, Australia, New Zealand, Finland, Ireland or Switzerland. Conversely, nor is the favourable geographical location of Somalia or Albania in itself sufficient for economic success. Consider the following theoretical example on the favourable or unattractive location of a country (or a firm) in geographical space by Leamer (2007, p. 89): Imagine that there are two countries – Japan and the United States. Japan sits on the top of a hill and the United States sits at the bottom. To get the US goods to Japan, one has to hire porters to carry the goods up the hill. But the Japanese can put their products in a chute and let gravity do the work – costlessly transporting Japanese goods down the hill to the US market. Not a level playing field, you should be thinking. Japan is clearly in the advantageous position. Not so fast, I caution the students. Who pays for the lugging the US products up the hill? Why do you presume it is the United States and not the Japanese? This should get them thinking about elasticities of supply and demand. If US goods are in short supply and are desperately desired by the Japanese, while Japanese goods are abundant and not much desired by Americans, then it is the United States at the bottom of the hill that is in the advantageous position and it is the Japanese who pay for the lugging of the goods up the hill.

Evolutionary Economic Geography

24

Christaller (1933) attempted to clarify and explain the rationale for the number, size and spatial distribution of cities (in southern Germany). The inductive analysis of this geographer is based on the idea of market threshold and transport distance. He suggested that cities form a hierarchy of central places. This hinges on the supposition that larger cities can sustain a wider variety of activities relative to smaller (low-order) cities and villages. Economies of scale are the source of such uneven distribution of production. The consequence of Christaller’s model is that if there are economies of scale, then the size of the market matters for the location of business. The concern here, therefore, is about the most ‘national’ product, below which are other products. The hidden idea in this model is based on the minimisation of transport costs by rational consumers who make multipurpose trips. In spite of its obvious value for the analysis of urban growth and distribution of services, this rigid application of the impact of market size ignores the consequences of unequal distribution of natural resources, changes in technology and negative externalities that come from agglomeration. A coherent general equilibrium model found some justification for the central-place theory (Krugman, 1993a, p. 298). Developing further the issue of how the economy fits into space, the economist Lösch (1940) starts with a useful, but most unrealistic, assumption that there is a perfectly even distribution of raw materials and population. He starts his deductive consideration of the homogeneous economic landscape (flat world) with a ‘local’ good and seeks to minimise transport costs for a given density of central places. This is not about central place conglomeration, but rather about a different problem: how sellers of identical goods would distribute themselves over geographic space and how firms selling similar but differentiated goods would distribute themselves over characteristic space. In such a space, centres specialise in different products, hence there is a diversification of economic landscape. An efficient pattern of central places would have the shape of nested hexagonal (honeycomb) market areas with no empty corners.32 This means that certain economic activities can be done only at a restricted number of locations. It was subsequently demonstrated that there is a wide range of geographical configurations of firms. The spatial arrangements that can satisfy the equilibrium condition include squares, rectangles and regular and irregular hexagons (Eaton and Lipsey, 1976, p. 91). Even though based on unrealistic assumptions, the model of central places need not be disregarded out of hand. In any case, it would seem that Christaller and Lösch both deal with planning problems, rather than with considerations of market results. The theory of central places points to the factors that need to be examined during the decision-making process about the location of an industry or a firm. These factors are sources of supply, intersections of traffic routes and the centre of gravity.33 In order to reduce inaccuracies in such a ‘technical’ process (suitable for the centrally planned system), it is necessary to consider additional elements that are in the parcel of a marketbased economy, including actions of other functionally related firms, competitors, consumers and government policies. The classical German location theory dealt with the locational decisions of firms which are in essence reduced to two issues: homogeneous distribution of natural resources over a flat space and optimum cost of transport. This literature was obsessed with the geometrical shape of market areas in an idealised landscape or with the optimal production site with given resources and markets. It ignored the crucial issue of market

Theory

25

structure and competition. This was ‘doing things in the wrong order, worrying about the details of a secondary problem before making progress in the main issue’ (Krugman, 1992, p. 5). Organisational issues such as institutional reality and policies were put aside. Institution-free theoretical models avoid the problem of the impact of various policies on the location or reallocation of firms and industries. Industrial concentration may be explained in theory by proximity to inputs, localised non-pecuniary externalities and the effects of market size in industries sensitive to economies of scale. If isolated, these forces produce clustered production geography. If combined, these factors may offset each other. Brülhart (1998a) recalled the fact that what economic geographers call ‘industry concentration’, trade economists call ‘interindustry specialisation’. A low level of intra-industry trade (IIT) in industries that are subject to economies of scale may indicate the existence of centripetal forces and industrial concentration in space. Conversely, a high level of IIT could point to a dispersion of an industry. As far as the EU was concerned in the 1980–90 period, Brülhart found that there was no further concentration of already clustered industries that are subject to increasing returns in the central regions; there was further concentration of textile-related industries at the periphery; and there were certain indicators of spread of ‘high-technology’ industries towards the EU periphery. Perroux (1950, 1955, 1961) introduced a predominantly intuitive concept of growth poles in spatial economics. The idea was discussed in the context of controversy between balanced and unbalanced regional growth during the 1950s. Geographical agglomeration, significant production linkages (with the key industry) and strong human contacts are necessary for the growth of a pole. A firm is located in a space consisting of poles. Each pole has both centrifugal and centripetal forces. Hence, each pole has its zone of influence (it both attracts and repels firms) and interacts with other poles. If, however, a certain pole does not have a degree of flexibility and adaptability to new technologies and changes in the market, it will stagnate and decline. For instance, France selected eight urban areas and bolstered their growth during the 1960s in the expectation that such a policy would counteract the growth of Paris. Hence, the policy of spatial concentration of investments was diluted over time as it was coupled with political snags. Many other national ‘areas’ exerted pressure to be included in the select group of geographical poles for special government treatment. The evaluation of the policy of growth poles is beset by difficulties, as it was not vigorously implemented in practice. Even though there is no firm jury about the past policy of growth poles, the French government (again) identified six industrial clusters34 in 2005. Further sixty-one ‘poles of competitiveness’ were also identified (perhaps for political reasons). The ‘poles’ were allocated Є1.5 billion (double the previous amount) in various incentives (grants and tax breaks). The intention is to liberate the potentials of the French economy, to stimulate R&D, to link public research bodies with universities and create new products and services, and to keep these firms in France, as well as to boost growth. The problem with this policy may be that scarce resources may be spread across too many recipients. However, if the resources are concentrated on six national clusters of excellence, this may herald a new French regional policy approach: assistance to top clusters and a departure from the traditional concentrated assistance to the most depressed areas. Brülhart and Trionfetti (2004) analysed the outcome of home-biased procurement policies on the location of manufacturing. They found evidence in the EU that, on

Evolutionary Economic Geography

26

average, a country would specialise in the production of goods for which it has relatively large home-biased procurement (‘pull effect’). A government’s preference for domestic bidders (discrimination against foreign ones) causes a shift in profits from foreign to home-based firms. In addition, they found empirical support for the ‘spread effect’: ‘industries that are subject to a relatively large share of public expenditure tend to be less concentrated across EU countries’ (p. 877). Biased procurement received increased international attention. This was not only in the EU where the Single European Market was supposed to reduce the impact, even eliminate this NTB, but also in the WTO under the auspices of the Uruguay Round. Still, fully open public procurement remained one of the principal remaining obstacles35 to a fully fledged Single European Market. Paelinck and Polèse (1999) tried to pass certain region-related integration lessons from the EU to the North American Free Trade Agreement (NAFTA). Basically, a region’s gain/loss from continental integration depends on: its distance from ‘the continental core’ (economic heartland) and on its endowment of ‘scale-scope, competitive and externality elements’. In simple terms, the location of production is expected to move over time towards the nation’s principal trading partner. This would favour regions that are located close to the continental core. Applied to the case of Canada in NAFTA, the national economic core is contiguous to the continental core. Most of the Canadian economic activity and population is concentrated in a 150km-wide (or narrow) band along the border with the US. In the case of Mexico, the national core (Mexico City) is located at some distance from the continental core (the US). Integration within NAFTA could increase certain tensions between Mexico City and more dynamic regions at the north of the country, along the US border, towards which migrate certain (parts of) industries previously concentrated in Mexico City. Multiple equilibria create a situation for welfare ranking and set up many temptations for politicians and bureaucrats to try to pick winners. These temptations appear whenever something new arrives that is commercially viable. The selection problem, however, remains unresolved as the entire system is unstable. Hence a small historical accident, a tipping point (chance, arbitrariness, a serendipitous event, a small difference in timing) remains the unique deciding factor that can sway the final outcome (Arthur, 2002, p. 6)36 (see Annex I). If governments want to tip markets towards a preferred solution, then it is timing and, to an extent, instruments that are crucial. There are few and relatively short moments of enhanced locational freedom for new industries that set in motion new industrial trajectories, hence there is only a narrow window of locational opportunity during which the policy maybeeffective.37 The light of locational opportunity is rare and has a narrow beam. Otherwise, governments with their limited knowledge, instruments and resources will try to stabilise a ‘naturally’ unstable and evolving process artificially. [This] ‘window of locational opportunity view’ at times appears to conflate ex ante unpredictability with ex post inexplicability, and it is possible to advance a different interpretation of initial events and the ‘birth’ of new path-dependent industrial trajectories that allows local context (and hence local prior path-dependent development) a more significant role in determining where (and when) new techno-economic paths emerge. (Martin and Sunley, 2006, pp. 425–6)

Theory

27

In a number of cases there may be a relative variety of potential locations for the establishment of business. Often the winning location wins by the skin of its teeth. Selection of the location may be done, in theory, either by market forces or by a powerful ‘ruler’. Clients vote with their orders in a market economy, while the powerful ruler may think about his or her own interest first. Lobbying and corruption may have a significant impact on the decision of the ruler. Many of these rulers in Europe build sumptuous palaces that may be found from Spain over France to Russia. Markets started increasing their influence over locational decisions from about the 1700s. However, powerful rulers should not be taken to mean only monarchs and the nobility. They exist today in big companies or ministries. Even in relatively small firms, decisions and selections are made by various bosses who guard selection privileges in their fiefdom. Neoclassical equilibrium theory Neoclassical equilibrium theory provided a convincing and useful initial research and analysis model in economics. This ahistorical model presents the economic market system in its idealised, fluent and simple form. This theory is based on perfect competition; maximisation under certainty or computable and insurable risk;38 constant returns to scale; invariable technology; reversible choices and policy actions; small transitory and seamless adjustment costs; and stationary equilibrium. Structural changes and development take place on the bases of additive forces and exogenous change in technology and demography. This theory fares rather well in relatively simple situations. In such cases prices, their changes and markets freely convey all relevant business information, they provoke (smooth) adjustments and they bring the system into equilibrium (stationary state). If there are differences among countries in terms of technology, growth or institutions, a relatively smooth process of convergence irons this out on its own. The final result (equilibrium) is clear, expected and highly predictable.39 Once it is reached, there is no further reaction. On these grounds intervention (economic policy) is not welcome, as it introduces new and/or reinforces existing market distortions.40 The usual simple, general and only policy advice (intervention) for all market economies and for all times is: remove market imperfections. This policy recommendation, however, falls into the trap of one-size-fits-all advice at all times. Context-specific issues are neglected and the policy interventions often result in economic disasters as, for example, was the case in a number of the International Monetary Fund’s (IMF) interventions in some unlucky Latin American countries in the 1980s. The policy advice was not based on what actually works in practice, but rather on what the IMF officials would like to see operate in line with the neoclassical theory. ‘The standard IMF procedure before visiting a client country is to write a draft report first. The visit is only intended to fine-tune the report and its recommendations, and to catch any glaring mistakes’ (Stiglitz, 2002, p. 47). Different national institutions and their capabilities, various national levels of development and established specialisations were neglected by the Fund (not to mention social fairness). It was simply assumed that the private sector would swiftly replace almost any government activity. ‘The IMF simply assumed that markets arise quickly to meet every need, when in fact, many government activities arise because markets have failed to provide essential services’ (Stiglitz, 2002,

Evolutionary Economic Geography

28

p. 55; emphasis original). It has to be recognised that the neoclassical market equilibrium forces do operate in practice (to an extent), however, this process is rather selective and sometimes quite slow. As such it may not always be politically acceptable. Hence, intervention may be necessary and welcome. The very things considered by neoclassical equilibrium theory as market failures, on the one hand, are the moving spirit of the evolutionary and endogenous growth and development processes and models, on the other. As such, policy implications by the two economic schools are completely different. While neoclassical equilibrium theory advises the removal of market imperfections and obstacles to progress, the evolutionary theory finds in these imperfections crucial sources of growth and economic progress. These imperfections ought to be promoted and streamlined, not repressed. The policy advice of evolutionary economic theory depends both on the given context and on time. It is also related to the ability of the target country institutions to carry out this recommendation successfully. Policy advice that is suitable for country A at a given time t may be totally inappropriate for country B at the same time, while the same advice in country A at time t+1 may give miserable or harmful results. There are certain economic activities that have the social and legal responsibility to provide continuous service. These are usually linked with utilities, health, education and infrastructure. The supply of these services may not be allowed to cease if the providers get into financial troubles. A bankrupt water utility company may not be permitted to stop supplying water to the city; it would be unacceptable for a failed maternity ward suddenly to send away women in advanced pregnancy. There are quite a few specific and important ‘businesses’, some of them quite large, which may not easily fit into the liberal market business discipline. For some, it would be too embarrassing for them to fail, like Cambridge University or the Louvre, or too complex, like the Eurotunnel, or too big, like Chrysler in 1979 or Alitalia over decades, or too important, like Lockheed in 1971. For the most part, neoclassical equilibrium economic theory has its base and inspiration in rigorous Newtonian physical concepts of motion and energy: various gravitating forces balance each other, and they produce unique and stable equilibria in which all changes cease. The final strategic outcome is certain: one just needs to arrive to it (but this is ‘only’ a tactical detail). The neoclassical system is ergodic, i.e. possible perturbations of the system are absorbed over time and the system returns to the (initial) equilibrium. If competition is perfect, then homogeneous firms are not active rivals in this rather flat economic landscape. Each one is a price or technology taker so the whole economy can be easily aggregated. The problem of coordination among players is absent (the price system solves it fully) and there is no need for government policy. In such a situation, one has good chances of predicting the future course of economic events in a more or less smooth, slightly incremental, linear and mechanical way. The desirable market characteristics in the standard neoclassical economic theory model include: the absence of market power so that price taking is the typical situation; prices are equal to opportunity costs and do not, therefore, allow for any pure profits; rents associated with market power of oligopolies and monopolies or other forms of market power are eliminated; sources of non-convexities such as scale effects and high entry costs are minimal or non-existent.

Theory

29

In the structuralist–evolutionary view, the things that drive the economy towards desirable results are the very characteristics that are undesirable sources of market imperfections in neoclassical economics. Although the special case of an entrenched monopoly that does not innovate is regarded as undesirable, most other market ‘imperfections’ are the very forces that drive economic development. Perfectly competitive industries rarely innovate. It is rather oligopolies that do the most … Innovation creates asymmetric information that creates market power, which in turn creates the profits that drive the system. Really large profits are the carrot that induces agents to attempt leaps into the unknown and to make many more modest decisions under conditions of uncertainty. (Lipsey et al., 2005, pp. 45–6) The possibility of gaining economic rents41 is, therefore, perhaps the most powerful incentive that drives economic agents to invest and innovate in an evolutionary (nonequilibrium) world. There are still those who argue in favour of the neoclassical outcome. The three great eras of globalisation, according to Friedman, contributed to the flattening of the world and gave certain arguments to the ‘death of distance’ school of thought. Globalisation 1.0 (1492–1800) shrank the world from a size large a size medium; globalisation 2.0 (1800–2000) shrank it from medium to small; while globalisation 3.0 (from around 2000 ‘is shrinking the world from a size small to a size tiny and flattening the playing field at the same time’ (Friedman, 2006, p. 10). Hence, according to this view, geographical distance is neither a prevailing feature of the world economy nor an important economic barrier. A flat and shrinking geographical space is preferred and necessary for the mathematical modelling of the economy and location of firms and industries by those searching for the solution within equilibrium. We shall see later that the actual shape of the world is not flat and that distance still plays a role for the location of production and trade, although perhaps not as much as it used to play in the past. Practical people of affairs often distrust traditional neoclassical equilibrium economics and its models ‘because it seems to describe the behaviour of incredibly smart people in unbelievably simple situations’ while the actual world is facing with the situation of ‘how believably simple people cope with incredibly complex situations’ (Leijonhufvud, 1996, p. 40; emphasis original). To obtain a closed-form solution to a problem, the question to be examined ought to be based on a relatively simple structure with a number of restrictive assumptions. Neoclassical equilibrium theory does not take into account various market anomalies such as whims, price bubbles, market volatility and crashes. It does not ‘see’ the invisible market hand in the same way as the evolutionary and complex analysis aspires to ‘see’ it. As far as economic policy and its evidence are concerned, Lipsey et al. recall: So much for the theory. The other major input into policy judgment is evidence. In the trade policy debate there is ample evidence of the power of appropriately used trade restrictions and infant industry assistance. As a general observation, not a single country in the West industrialized under completely free trade. The closest was probably Britain. But that

Evolutionary Economic Geography

30

country gave key protection to its textile industry in the eighteenth century when it banned the importation of Indian cotton goods. It also placed a series of restrictions on manufacturing activity in its colonies, and forced exports from its colonies bound for anywhere in Europe to flow through English ports. Virtually all other countries including Germany, France, the USA, and the former British Dominions sheltered their emerging manufacturing industries behind trade restrictions. These were eventually reduced, and some were lifted completely, but only after the industries had developed for at least several decades and often for more than a century. (Lipsey et al., 2005, pp. 511–12; emphasis original). … it is policy imperialism to argue that the poor must accept the standards of environment and labour protection that the rich ones can now afford (but could not when they were poor). (Lipsey, 2004, p. 12) There is certain room for astute intervention. However, this does not give carte blanche for any policy, as no intervention in the economy may be superior to lousy intervention. Physics and economics After several unsuccessful professional careers Léon Walras42 followed his father’s advice to employ calculus in order to put economics on scientific foundations, as was the case with physics. A range of relevant independent variables ought to be put in a model of general equilibrium. Even though a late bloomer in economics, his book Elements of Pure Economics (1872) was Walras’ masterpiece that pioneered the general equilibrium theory. Before him, economics was not a mathematical field. Earlier economists were dealing with a market for one or two goods (assuming that the rest would remain more or less the same). Walras wanted to deal within an economy with everything at once. Many economists before Walras regarded themselves primarily as philosophers and used very little or no mathematics and statistical data. Walras introduced into classical economics (1) differential calculus and (2) the general equilibrium point as principal workhorses.43 Mathematical physics, rather than mathematical biology, entered economic analysis. In the neoclassical economic models, the economy is the solution to a series of equilibrium conditions. Presented in a simple way: resources, labour, capital, demand, supply, energy, information, consumers, producers, prices and the government are put together in a mathematical package to be solved. As an academic field, economics gained numerical accuracy and contributed to a certain predictability of economic events, but this came at a very high and long-lasting price: detachment in certain cases from real life. Chemistry and physics deal with complex systems, but these can be artificially simplified, controlled and repeated in labs. Unlike in biology and economics, these are nonadaptive systems. Discoveries in physics may be directly applied to other similar situations elsewhere within this field, while this linear option is not always possible in biology and open economies. Biological and social systems are much more complex and unpredictable than those in physics and chemistry.

Theory

31

Assumptions necessary for the calculation and solution of mathematical problems related to market equilibrium started off and continued to be so restrictive that they rendered conclusions unrealistic and impractical. One of these unrealistic assumptions is that people behave in a perfectly rational way (so that it is predictable). But recall a wellknown joke: A young economist tells to his old colleague: ‘Look, there is a Є50 note on the pavement.’ The old economist retorts: ‘Nonsense, if it were true, someone would have already picked it up by now.’ The moral of this story is that if there is an external shock or a sudden opportunity that puts the economic system out of its steady state (equilibrium), the neoclassical system self-equilibrates itself (almost) instantly. An example: suppose that you have some extra money and you want to do something with it. But what? You calculate: should you repay a part of your mortgage or go on a cruise in the Greek islands for the first time in your life next summer? You take into account your remaining debt, rate of interest, expected earnings in the future, your age and health, the impact of drought on the price of food and all other relevant information … and after some pondering a rational decision is reached. There is no mistake. You decide (I hope) to go on a Greek cruise. Two weeks. This rational and deductive reasoning is fine when there is a clearly defined problem and a full set of relevant information.44 Now consider another situation related to rational expectations and decision-making. Arthur (1995, 1999) named it The Bar Problem. The bar, El Farol, offers Irish music on Thursday nights and it has, of course, limited space. Imagine now that: One hundred people must decide independently each week whether to show up at their favourite bar (El Farol in Santa Fe, say). The rule is that if a person predicts that more than 60 (say) will attend, he will avoid the crowds and stay home; if he predicts fewer than 60 he will go. This seems innocuous; but it destroys the possibility of long-run shared expectations. If all believe few will go, then all will go, thus invalidation these expectations. And if all believe many will go, no one will go, invalidating those expectations. Predictions of how many will attend depend on others’ predictions, and others’ predictions of others’ predictions. Once again there is no rational means to arrive at deduced a-priory predictions. (Arthur, 1995, p. 21, emphasis original) This example presents a deep difficulty in rational economic expectations. If you try to be logical in this case, this type of rational reasoning may lead and keep you in a decisional wasteland. Even if you have movement-eager legs (curiosity), you do not have a theoretical bottom to stand on and move. Similar coordination and expectations problems may be found in decisions about going to a restaurant during lunchtime, a dance hall, a sports stadium or choosing a resort for holidays. One may be tempted to say about certain places that they are so popular and crowded that no one goes there any more.

Evolutionary Economic Geography

32

Nonetheless, these are only trivial examples regarding a real problem that faces decisionmakers in complex and changing reality about the location of their business. People may easily act in an unpredictable fashion. Their behaviour may often be whimsical and they make mistakes. Hence there is a degree of randomness in their actions. Despite this, people generally seek and reward cooperation, and avoid and punish cheaters and free riders in both formal and informal ways. The degree of cooperation depends on the number of players and the frequency of their encounters. Inductive reasoning is more flexible and appropriate for ambiguous situations with an incomplete set of background information. Coming back to the previous example, imagine that you have a bit of money and that you are on a lovely afternoon stroll down the main street. You see a poster showing the Greek islands in the window of a travel agency. The evolutionary and complex system development of subsequent events would be: ‘Hmm … Greek cruise? Quite an attractive offer. My sweetheart’s birthday is next month. I’ll buy a travel package for the two of us, now.’ While there is certain intuition behind the general equilibrium theory and models (cause and the likely effect), this is not always the case with evolutionary, non-linear and complex systems. That is the reason many scholars and policy-makers may have difficulties reaching a decision or policy recommendation based on these non-equilibrium systems. Economic decisions are often based on incomplete information and uncertainty, hence they may not be ‘rational’ and may well be wrong. Is it correct and justifiable to try to solve abstract economic problems that have little to do with reality? In addition, in the real world certain firms, for instance consultants or accountants, rarely use their potential to the full. They keep always some in reserve just in case there is a sway in the market for their services. Hence, economic agents are not as rational as neoclassical equilibrium theory would like them to be. ‘Some critics argue that this borrowing of equilibrium from physics was crucial scientific misstep that has had lasting consequences for the field’ (Beinhocker, 2006, p. 32). This observation is even more important if one recalls that physics went through a deep change at the turn of the twentieth century, just after the time when Walras ‘borrowed’ physical concepts of gravitating bodies and equilibrium, and when he introduced them into economics. New concepts including complexity, symmetry breaking, high-energy states, quantum theory, relativity and chaos theories entered physics. In addition, these equilibrium concepts have dominated neoclassical equilibrium economics even though they were at odds with how social system operate and evolve. An overhaul of economic theory is necessary. Evolutionary and complexity theories provide potential alternatives along this reform path. The field of economics is notoriously slow to open itself to ideas that are different. It has become usual practice that most path-breaking articles in economics are rejected several times before they find a home in a journal.45 Perhaps academic curiosity may lead certain economists to explore achievements by Ilya Prigogine (1917–2003), who won the Nobel Prize for Chemistry in 1977 for his contribution to our knowledge in the area of dissipative structures and their role in thermodynamic systems in disequilibrium, selforganising systems, formation of complexity in biological entities and the irreversible impact of time in natural sciences. Another possible intellectual voyage may be to

Theory

33

contributions by Murray Gell-Mann (b.1929), winner of the Nobel Prize for Physics in 1969, for his explanation of difficult and puzzling aspects of elementary particles. Identical firms may act differently in equal situations because some are risk-lovers (and prefer to ‘gamble’ with risky investments) while others may be risk-averse. Some people and some firms may become sometimes altruistic and at other times selfish. This is difficult and risky to predict. ‘Even if we were as rational as possible and had all the information we could want, the computational complexity of the economy is such that the future would happen before we would have time to predict it’ (Beinhocker, 2006, p, 323). As a ‘technical’ academic field and fully interlaced with mathematics, neoclassical equilibrium theory ought to obey the first law46 of thermodynamics: energy is neither created nor destroyed, it is only transferred (conservation of energy principle). The change in the internal energy of a system (∆U) equals to the heat added to the system (Q) minus the work done by the system (W) (Equation 2.3). ∆U=Q– W (Equation 2.3) If the first law of thermodynamics were transferred into the field of economics, it would label economics as a study of the way social systems transfer scarce resources into goods and services and distribute output among people.47 There is, however, the second law of thermodynamics. It refers to the everincreasing state of disorder or entropy48 in an isolated system: a colder unit is not able to transfer heat to a warmer unit. Therefore, there is a fraction of the total energy that is not available to do work. A change in entropy (∆S) equals the change in heat (∆Q) over time (T) (Equation 2.4). (Equation 2.4) Entropy provides information about the evolution of an isolated system over time. If one takes a look at the same object in two different moments in time and one of these observations is more ‘disordered’ than the other, one may infer that the observation that finds more disorder refers to the moment that came later in time: flowers blossom, strawberries and cream rot, facades crumble, roofs leak, babies grow, skin dries and faces get wrinkles. Entropy assists in observing past, present and possible future. Simply said, entropy gives time an arrow. Physics distinguishes between two types of systems: open and closed. A closed system has no contact or exchanges with the outside world. According to the first law, within this system matter may be converted into energy, while energy may be converted into different forms: the total amount is always constant. Entropy increases over time in this system, and it moves to a minimum energy equilibrium. At the end of the day the system arrives at a steady state. Predictability of behaviour in closed systems is rather high. An open system, on the contrary, has exchanges of matter and energy with the surroundings. The earth constantly receives energy from the sun. This enables life, decay, death and various transformations of matter and conversion of energy. These movements and activities release heat that is, together with entropy, given out back into the general system. Such a complex and adaptable system has a rather low level of predictability

Evolutionary Economic Geography

34

regarding its path of behaviour over time. John Holland, a pioneer in the area of nonlinear science and complex systems said that ‘if the system ever does reach equilibrium, it isn’t just stable. It’s dead’ (Beinhocker, 2006, p. 70). Therefore and in general, the lower the entropy of a thing, the higher its value (average house vs average car). Uncertainty forces entrepreneurs to experiment, to evolve (not necessarily always for the better), to search for new opportunities or to adjust to them if they arrive on their own (e.g. shortage of liquid fuels). The idea of equilibrium (rest) is contrary to the notion of perpetual change (creation, differentiation, selection, adaptation, innovation, retention, amplification, replication and constant search for creative–destructive optimal solutions … ) brought by the evolutionary view of how an economy operates. Karl Marx’s analysis of social phenomena is dynamic, but only up to a point. He argued that everything is in a continuous process of change. This – dynamic – part of his model emphasises relentless, at times even violent, struggle among social classes and revolutions. The longer term outcome of these frictions is normally social progress. The ultimate goal of this process of change is communism. Once this objective is reached, struggle ceases and the system settles down (static part). In reality, one most often comes across complex systems that are somewhere in between utterly predictable and purely random. Nonetheless, a majority of systems in the real world act in a non-linear way.49 Still, this non-linear behaviour of a system (economic or otherwise) is usually more complex than chaotic. Therefore, those involved in long-term weather forecasts and economic predictions have a rather low reputation. No wonder – they deal with huge, n-factor, complex, dynamic and non-linear problems. Therefore, rather than forecasting, a more modest task in our academic voyage in this book is just to identify and discover elements of the system and how they interact and operate regarding the location of firms, clusters and industries. Considering the economy as a closed equilibrium system that balances itself, as is the case with traditional neoclassical theory, is a misclassification on a grand scale. If it were true, entropy would over time transform it into something that has less order and complexity, into something that has unremarkable stasis. The economy is on the contrary an open, complex, adaptive and disequilibrium system. Changes towards a new qualitative state (rather than towards a steady state) are perpetual and never ending. ‘The growth of economic activity from Stone Age until now has been one long story of fighting entropy on a grand scale – something that could only happen in an open disequilibrium system’ (Beinhocker, 2006, p. 71). Energy flowing from the sun assists in continuous order creation and maintenance that ‘fight’ against the other force of entropy, which introduces order destruction (pollution is an example). Even though the economy is a social phenomenon, it has certain features of an open physical system. Hence it may not fully escape the two physical laws. Building and maintaining a road or a bridge is a constant fight against entropy. If someone cuts the supply of water, energy and/or food in an economy, as it is the case during and after earthquakes, tsunamis, hurricanes, fire, floods or wars for example, entropy has free rein during those times and in those areas. It always takes some energy to both make and unmake things. Every morning when we wake up (at least the readers of this book), we go to the bathroom to wash and spend some time and energy to try to ‘repair the overnight entropy damage’ on our skin, hair, ears, neck … And take a look at the facades of buildings. They change over time. Bridges and roads also crack over time. The most

Theory

35

sophisticated and solid fortifications, if not maintained, are often destroyed over the long run, not by enemy firepower but by water, humidity, plants, land movement … (entropy). The process is rather slow, but it is firmly and visibly present. Although the above critique of neoclassical equilibrium economics points to important problems that question many of the standard neoclassical claims, this is not to say, of course, that traditional economics that used ‘wrong’ mathematics was harmful in reality. At certain times and based on specific assumptions, theoretical explanations and policy suggestions by the neoclassical theory improved the livelihood of hundreds of millions of people throughout the world. We would like to point that (1) entropy has an impact on traditional neoclassical equilibrium theory and that (2) the evolutionary and complexity theories may do a better job. General equilibrium calculations should not be the principal tools for studying the economy and guiding medium and long-term policy advice. While most economists remained in the theoretical world of static equilibrium for a very long time, the inspiring (Newtonian) science of physics, as well as chemistry and biology, became involved in the study of complex and dynamic systems that are constantly away from rest in equilibrium. This turn started in sciences and continued from the 1950s. A bee on its own may not always be all that thrilling a subject for study. But if one observes thousands of bees together, they create communities that live, thrive, communicate, share labour (specialise), adapt, defend, dust flowers (assist in fruitproduction), and produce wax and honey. Similar research excitement may be found in evolution regarding changes in climate. It is never ‘settled’. These, and many other systems, are complex, dynamic, adaptive and out of equilibrium. The neoclassical stable equilibrium theory (the outcome is certain) is remarkably ineffective and potentially misleading in coping with more demanding situations that include shocks such as a general change in technology, endogenous ones in particular. For example, the arrival of new goods within electronics and materials such as semiconductors and computers improved ‘old things and services’, eased their production and assisted in the making of entirely new goods and services in ways that have little relation to the original change in prices. They also made possible a wide spread of location of production of goods and services to the places that were not technically feasible in the pre-Internet era. Or, earlier, the arrival of the steam engine replaced waterwheels in the early nineteenth century in Britain. Entrepreneurs were free to locate their factories wherever they expected the most favourable economic advantages and returns. This was most often away from the countryside, with fast-moving water, towards new industrial cities. ‘No decrease in the price of water power could have brought about this change, since water power must be consumed where it is generated and the possible sites for its generation are given by technical conditions’ (Lipsey et al., 2005, p. 408). Standard neoclassical equilibrium theory may need at least a partial adjustment along lines taking into account the following developments that changed the spatial location of certain economic activities: • Skills: a huge labour force in the emerging economies including educated engineers may not always be up to the standards expected in the developed countries. However, the emerging economies are catching up fast along these lines with the developed countries. This permits them to start production of relatively simple products that, over time, improve and increase technological sophistication.50

Evolutionary Economic Geography

36

• High-technology production: this is no longer the privilege of the developed world. China has featured very highly in the global production and export of various hightechnology products from the 1990s. One has to observe, however, that in many cases the Chinese just assemble these products from imported inputs. Nonetheless, in the light of a large-scale shift of global manufacturing towards Asia, the currently developed countries may grow and keep their economic position only by constantly keeping ahead of potential competitors in inventing profitable new technology in production, as well as in management. A few humble words about possible evolution in the future. Society at large and business in particular owe to the physics of the twentieth century technologies such as motor cars, aircraft, nuclear energy, satellites, computers, mobile phones, global positioning devices and the Internet. Certain powers over nature have made some parts of life easier and faster, but these powers have not always been benign (atomic bombs; genetically modified food). James Chadwick (Cambridge, England) and Hans Falkenhagen (Rostock, Germany) independently discovered the neutron in 1932. This discovery triggered the development and use of atomic bombs in 1945, just thirteen years later. What physics was during the twentieth century for society and business, biology may become during the twenty-first. Many huge problems that trouble society are in essence biological: climate change (the emission of CO2 into the air is faster than the ability of plants to process and crack it); ageing populations; diseases (cancer, allergies and infections such as HIV/ AIDS or bird flu) and the rising cost of health care. All this may demand some intervention through biology and biotechnology. Genetics may provide answers to the operation of what is called, in contemporary parlance, the ‘biological Internet’ in cells, as there are many proteins in living creatures that are doing ‘who knows what’. A biological or biotechnological Chadwick/Falkenhagen-type ‘Internet search engine’ may be necessary. A tap on this process may provide guides towards solutions to many of the problems that face humanity. The evolutionary strand in economics Evolution is a process of unfolding, self-transformation, self-organisation after oscillations and the appearance in orderly succession of events over a long period of time. This process is irrevocable and the historical CD, song or card is played only once. A glance at relations between economics and evolutionary biology shows that these were known and recognised long ago. However, these interactions were put aside by many researchers in the field of economics. Thorsten Veblen (1898) worried about the failure of (neoclassical) equilibrium economic theory to explain economic changes in a satisfactory way. More influenced by Darwin’s biological view of evolution rather than Newton’s technical rigour, Veblen introduced term evolution into economics. He did not, however, provide much more than that. Alchian (1950) was one of those rare economists who were writing about economics partly in biological terms. Joseph Schumpeter (1911) rejected any connection between economics and Darwin’s ideas. His claim was that the principal business energy and vigour comes from the continuous endogenous creation of novelty by heterogeneous entrepreneurs. Everyone who tries out something new in an economy is an entrepreneur. However, not all of their

Theory

37

actions have real success and a widespread impact. For Schumpeter, the principal moving force in an economy is variety generation, rather than Darwin’s selection of species. Schumpeter’s view refers more to radical51 than to incremental innovations. Entrepreneurs, in their search for new opportunities relentlessly introduce and maintain disequilibrating forces in the markets that can never stay in equilibrium. Or, in the words of Schumpeter (1939, p. 70): we will not postulate the existence of states of equilibrium where none exist, but only where the system is actually mowing toward one. When, for instance, existing states are in the act of being disturbed, say, by a war financed by government fiat, or by a ‘mania’ of railroad building, there is very little sense in speaking of an ideal equilibrium coexisting with all that disequilibrium. It seems much more natural to say that while such a factor acts there is no equilibrium at all. In this example, Newton’s mechanical concepts of equilibrium, as transplanted into economics, were turned upside down. There is no equilibrium in an economy. If somehow one arrives in the economy, it is never more than a temporary and transitory moment – not a state. Hence general equilibrium should be a splendid, but only a partial, teaching tool about the operation of an economy, rather than an end point at which all changes cease. The norm in the evolutionary economics is disequilibrium. Or, in Schumpeter’s words about the process of ‘creative destruction’: ‘The opening up of new markets … that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one’ (Schumpeter, 1987, p. 83; emphasis original). A problem may be found in the fact that the economics–biology approach was left on research margins for a long time. A notable revival of general interest in the evolutionary economic theory and its new life was brought by Nelson and Winter. They devoted three chapters (12 to 14) of their book An Evolutionary Theory of Economic Change (1982) to the ‘Schumpeterian competition’. Schumpeter’s hypothesis argued that the existence of large firms with considerable market power is the price that the society has to pay for technological progress. Nelson and Winter inverted this hypothesis and claimed instead that the existence of monopolies is the consequence, rather than the prerequisite, for active innovation. In addition, various teams and branches of corporations most often took the role of an individual entrepreneur the possession of ‘technical knowledge’ is an attribute of the firm as a whole, as an organized entity, and is not reducible to what any single individual knows, or even to any simple aggregation on the various competences and capabilities of all the various individuals, equipment, and installations of the firm. (Nelson and Winter, 1982, p. 63) Earlier, referring to Darwin’s thinking, Marshall (1890, pp. 300–1) wrote: Since that time biology has more than repaired her debt; and economists have in their turn owed much to the many profound analogies which have been discovered between social and especially industrial organization on the one side, and the physical organization of the higher animals on the other. In a few cases indeed the apparent analogies disappeared on closer inquiry: but many of those which seemed at first sight

Evolutionary Economic Geography

38

most fanciful, have gradually been supplemented by others, and have at least established their claim to illustrate a fundamental unity of action between the laws of nature in the physical and in the moral world. This central unity is set forth in the general rule, to which there are not very many exceptions, that the development of the organism, whether social or physical, involves a great subdivision of functions between its separate parts on the one hand, and on the other a more intimate connection between them. Each part gets to be less and less self-sufficient, to depend for its wellbeing more and more on other parts, so that no change can take place in any part of a highly developed organisms without affecting others also. This pointed ages ago to the fact that a (large) part of research inspiration for the economists may and ought to be found more in evolutionary biology than in physical mechanics. The snag is that biological systems are more complex than mechanical ones. Evolutionary biology introduces dynamism into the research and analysis agenda. This is, however, harder and more uncertain to grasp than is the case with static mechanical systems and social mathematics. Compared with neoclassical equilibrium theory, the complexity and evolutionary theory of economic behaviour is another general and conceptual look at how the world works (Table 2.1). This theory is concerned with: • uncertainty against which one cannot get insurance; • (micro) analysis of endogenous technological change and its spillovers; and • non-maximising behaviour of profit-oriented firms. Neoclassical equilibrium economists neglected or deliberately omitted the fact that says that: where foresight is uncertain, ‘profit maximization’ is meaningless as a guide to specific action. (Alchian, 1950, p. 211, emphasis original) The evolutionary approach is about much more than allocation and is fundamentally not about cost-minimization; it is about the forces that allow the parameters of cost minimization to be altered … And in terms of mechanism … evolutionary economics opens up the field of untraded interdependencies which does not figure in transactions-based approaches. (Storper, 1995, p. 207) When agents are operating under uncertainty, it is unclear what maximizing behaviour even means … maximization is not a definable concept in the face of uncertainty. (Lipsey et al. 2005, pp. 38 and 47, emphasis original) In such situations, welfare comparisons have little meaning. That is another difference between the evolutionary and the neoclassical equilibrium model. Even though the evolutionary model makes note of other theoretical models, it looks at the economy

Theory

39

through non-equilibrium spectacles in which, to use mathematical terms, ‘minus minus’ does not always make a plus. Choices are Table 2.1 Two basic theoretical concepts of how an economy operates Aspect

Neoclassical equilibrium model

Assumptions Given factor endowments, ‘first nature’ conditions Technology given (exogenously) and available Inputs are perfectly divisible No entry barriers Static and linear model Maximisation under certainty or insurable risk (fully informed individuals make rational choices), hence there is weak uncertainty Better-informed markets (Internet, 24/7 news coverage) reduce price gaps and volatility Trade based on costs, i.e. supply potentials Considers trade in goods (i.e. wine for cloth) Factors are mobile within a country, but not between countries Tariffs may be the only traderelated costs Equilibrium is normally achievable, unique and optimal Theoretical ideas: sound, clear, smooth and ‘mechanically’ straightforward

Evolutionary model ‘First nature’ important to an extent, but manmade activities (‘second nature’) to improve and extend the ‘first nature’ are more important Economies of scale High fixed costs (entry and exit barriers) Innovation in production, organisation, distribution and control Technology constantly changing, endogenous and available at cost (financial and time) Economic activity is lumpy Trade is based on demand, i.e. consumers ‘love variety’ Intellectual property rights Dynamic, nonlinear and evolutionary disequilibrium (equilibrium brings stasis of the system which means a system’s death) the system is everchanging and never-ending Accumulation and upgrading of factors which magnify output potential (cumulative causation and acceleration) Path dependence and non-ergodicity Considers trade in goods, services and specific tasks (fragmentation and unbundling of production process and ‘offshoring’ of routine tasks) Both national and international mobility of factors is constrained Nature of production, technology, trade and consumption changes fast because of multiple equilibria, hence the target (aim) that the theory tries to fathom and explain is constantly moving Existence of trade cost Choices based on imperfect information, groping under strong and endemic uncertainty Better-informed markets (Internet, 24/7 news coverage) increase price volatility as there are more technical opportunities for arbitrage Multiple and potentially ex post non-optimal equilibria Non-maximisation behaviour of profitoriented firms Seeking, selection and adaptation under strong

Evolutionary Economic Geography

40

uncertainty (no possibility of getting insurance) Historical accidents and chances play a role Theoretical ideas: an incomplete and unsystematic collection of various ideas and common themes that discount the weight of tariffs and trade creation and trade diversion effects Micro and macroeconomic planes of the economy

Separate and clear

Mixed

Trade barriers

Transparent Tariffs and quotas

Non transparent Standards and their implementation Customs and administrative procedures Government intervention: NTBs, public procurement, taxes, subsidies, public monopolies, local content rules…

Specialisation

Absolute advantages Comparative advantages Homogeneous goods

Man-made comparative advantages (periodically shortlived) Goods and services are differentiated Widening in varieties of goods and services High-quality goods with aftersales service Clusters and agglomerations1

Competition on the market for goods (and services)

Perfect and only through market mechanism (price-taking is the typical situation, firms have no market power) Many small suppliers First best situation (end-state equilibrium) Through prices New technologies are often regarded as the fall in price of the service provided by the old technology. Old things become cheaper.

Imperfect, networks are important A select number of suppliers that may influence prices within their (narrow) market niche Second-best situation Price and non-price (innovation, advertising, after-sales service…) A new technology such as semiconductors enables the production of entirely new things and services Continuously evolving process

Theory

41

Competition among Passive, determined by factor locations to attract firms endowments and industries Countries are discrete economic points with irrelevant location in space

Active, relentless, hence there is no equilibrium economic landscape over longer period of time Location is relevant

Size of firms

Identical, i.e. there is a representative agent (social planner)

Heterogeneous in size and productivity (technology), hence aggregation is highly questionable

Economic policy (trade, industry…)

Liberal: remove all market failures in all circumstances and at all times Intervention not welcome as there is the ‘ invisible hand of the market Exceptionally ‘ infant industry protection Reversible

Interventionist, but selective and (presumably) diminishing and adjusting over time Intervention is not heavy-handed, but rather ‘ nudging-handed Intervention is context- and timespecific

Trade pattern

Inter-industry Highly specialised Set of traded goods is both complete and fixed

Intra-industry Not totally specialised (full specialisation is only within a given and narrow product-classification niche) Trade pattern in developed countries is as wide as ever

Adjustment costs

In theory: none (all markets clear at once) In practice: significant

Weak if related to product differentiation and changes in varieties within the same commodity group Significant if related to sunk costs, economies of scale, innovation, launching of an utterly new production or a cluster

Effectiveness

Effective at resources allocation in equilibrium

Effective at innovation and selection of fit projects in disequilibrium

Welfare effect of trade

Favourable for all participants Higher steady-state equilibrium comes (almost) instantly

Highly uncertain Some may gain a lot, many others may lose a lot Always changing Higher steady state (equilibrium) comes after unpredictably long period of time and is only temporary and transitory Equilibria are punctuated (they are quasi-equilibria)

Johnson–Krauss ‘law’2

Applies

Capital–labour ratio Given and stable

Does not apply because of irreversibilities that can perpetuate themselves Changing

Evolutionary Economic Geography

42

Coordination failure Temporary and short-lived

Present

Openness to ideas outside economics and mathematics

None

Wide and often crucial

Physics and thermodynamics

Closed system that could be explained and understood by the use of algebra

Open, always evolving and unfolding system that could be partially understood by the use of complex systems analysis and evolutionary process (creation, differentiation, selection [including mistakes], adaptation, innovation, retention, amplification, replication and constant search for creative–destructive optimal solutions) which is different from the biological process, but is still evolutionary

Treatment of mind in choice-making

A black box

Reliance on neuroeconomics and psychology

Notes 1 Agglomeration (clustering) produces rents which hold firms and factors in certain place, even if policy action would lead to a geographical move. There is, however, a threshold. Once costs to firms and mobile factors outweigh the agglomeration rents, firms and mobile factors would move elsewhere. Here lies the principal and sharp difference with the standard neoclassical equilibrium model in which small changes in policy lead to small economic responses. Evolutionary models predict dramatic non-linear effects in such circumstances. A small change at a crucial point in time (once the threshold is reached) can lead to unexpected and shocking outcomes (a locational ‘ catastrophe ). Non-linear reaction to policy measures makes it hard to predict the actual effects of a given policy change. 2 If, for instance, a country subsidises domestic production and exports, then it might gain an advantage over foreign rivals, but only temporarily. Anything that provides a country with an advantage (or disadvantage) in exporting in the short run will cause the adjustment of the exchange rate or factor prices in the long run.

strongly irreversible and they are not-ergodic. Economy is essentially a dynamic and generally incremental system in the long term. Economic players endogenously create changes such as technology and adapt to them; they also innovate and learn, they are subject to strong endemic uncertainty and (huge and continuous) adjustment costs, and they often cluster in space. Everything is in this evolutionary framework an experiment. Nothing is steady and final. Growth and development are, in this model, the outcome of dynamic bottom-up relations between many and various factors in the evolutionary process. Therefore, it is impossible to predict outcomes precisely from an initial position even if all starting elements are known and players are rational. However, some vague predictions are still possible. Nonetheless, evolutionary economics can, to an extent, handle endogenous changes in technology; its spillovers; multiple equilibria and selection; market and nonmarket institutions (organised interests); learning, innovation and knowledge; economies of scale; as well as diversity of players.52 But every approach has its nemesis. This improved ‘handling’ of the variety of important elements came at a price. The cost comes in terms of increased complexity and certain harm to transparency and tractability in comparison with the ‘clean’ social mathematics and static neoclassical equilibrium

Theory

43

theory and its (simple) clear-cut conclusions and policy recommendations. The challenge for the future is to increase the empirical importance and link of the evolutionary model with more formal models (if this is possible). Formalisation may be more suitable for cases with risk that may be calculated, than for evolutionary situations with strong endemic uncertainty and innovation. One has to recall first that mainstream economics deals with decisionmaking within given structures. Evolutionary economics is concerned with decision-making and longterm changes of the structure. It is necessary to better understand the situation when the existing local system loses its dynamic efficiency and becomes a source of rigidity. A successful transformation ought to refer not only to finding a new economic base and direction, but also to various other aspects that include education and culture. An evolutionary economic model is charged with complex, changing and uncertain situations and multiple equilibria. Players are only imperfectly informed: they may not know or be aware in advance of the whole array of possible solutions and results that their current choice and action may produce in the future. Even if different players (firms and/or individuals) have similar or identical endowments of ‘hard’ material and intangible resources, they may easily each make a different choice in an identical uncertain situation when they have to choose the course of action. ‘Given the uncertainty, neither individual’s choice can be said to be irrational’ (Lipsey et al., 2005, p. 31). However, quite another matter is which one will be successful in economic terms in the future, i.e. which one will ‘get it right’ and be rewarded, and how many will ‘get it wrong’ and be penalised by the evolutionary and selection hand of the market and general environment. This complex evolutionary model may look less explicit regarding explanations and economic policy intervention than is the neoclassical one. The evolutionary model, however, offers additional choices for the government policy intervention.53 The search process involving the actions of agents in a lumpy, structuralist– evolutionary (S–E) world should be viewed as ‘informed’ rather than pure ‘blind groping’ because agents have some information, but also face issues of historical accident and uncertain chance. Each player may have certain knowledge, experience (good and bad), preferences and love or aversion for risk which back certain decisions. Players are heterogeneous (there is no neoclassical ‘representative agent’, i.e. the social planner), equilibria are punctuated (i.e. they are quasi-equilibria) and rivalry takes place. This rivalry is revealed most often through innovations by which firms seek temporary comparative advantages: in the evolutionary model nothing is either absolute or permanent in the long term. Hence, Lipsey et al. (2005, p. 35) wrote that: As a result, long-run equilibrium analysis of the competitive process is argued to be not just irrelevant but misleading because firms that are competing through innovations will cause technology to change endogenously long before any long-run tendency based on fixed technology and tastes is manifested in observed behaviour. The great problem with this type of theory is that no one knows how to aggregate the behaviour of rivalrous, price-setting firms into a theory of the macro behaviour of the whole economy. We argue, however, that if that is the way the world is, there is no point in pretending otherwise.

Evolutionary Economic Geography

44

Eaton and Lipsey (1997) rejected the neoclassical timeless and spaceless competitive vision of the economy. Three phenomena (the core of the Eaton–Lipsey approach), lumpiness (inseparability), specificity and diversity, are not easily accommodated in the neoclassical theoretical framework for three reasons: • In neoclassical theory all inputs are perfectly divisible. Reality is different. Economic activity is extremely lumpy in geographical terms.54 The location of production is not evenly spread across the globe, but rather, it is concentrated. ‘History matters in the sense that economic behaviour at any point in time is determined by the current and past values of the relevant economic variables’ (Eaton and Lipsey, 1997, p. x).55 • In the neoclassical production function inputs are non-specific. They could be used to produce any product. In the real world, both physical and human capital is often activity-specific. • There is also a diversity of tastes and preferences in decisions by individuals. Consumers in similar situations make different decisions about consumption of, for example, cereals, drinks, painkillers, shirts or ice cream. In the same way workers in similar economic situations make different career choices. Eaton and Lipsey also introduced the ‘address model’ in which goods are described by their address in space. This address is determined by product characteristics: computers may have their ‘address’ (location) in a space that is depicted by features such as speed and memory. The address model can explain these ‘awkward fact’ of product differentiation. Unlike in the neoclassical stable equilibrium model with a certain outcome, there is a more important call for economic policy in the evolutionary framework. Targeted intervention is welcome in some matters and areas that encourage innovation and technological change. Policy advice depends on the actual context and time in each case. As there is no unique optimum allocation of resources, expected values may not be calculated even with a modest degree of confidence. Hence, policy advices on most issues (trade, technology, investment, taxes … ) must be based ‘on a mixture of theory, measurement and subjective judgement’ (Lipsey et al., 2005, p. 515).56 Econometric modelling can only assist sound judgement, eyes and ears: it can never be their full replacement. To illustrate these statements, let us consider the following example: Business men do not always ‘calculate’ before they make decisions, and they do not always ‘decide’ before they act. For they think that they know their business well enough without having to make repeated calculations; and their actions are frequently routine. But routine is based on principles which were once considered and decided upon and have then been frequently applied with decreasing need for conscious choices. The feeling that calculations are not always necessary is usually based upon an ability to size up a situation without reducing its dimensions to definite numerical values. (Machlup, 1946, pp. 524–5)

Theory

45

What sort of considerations are behind the routine decision of the driver of an automobile to overtake a truck proceeding ahead of him at slower speed? What factors influence his decision? Assume that he is faced with the alternative of either slowing down and staying behind the truck or of passing it before a car which is approaching from the opposite direction will have reached the spot. As an experienced driver he somehow takes into account (a) the speed at which the truck is going, (b) the remaining distance between himself and the truck, (c) the speed at which he is proceeding, (d) the possible acceleration of his speed, (e) the distance between him and the car approaching from the opposite direction, (f) the speed at which that car is approaching; and probably also the condition of the road (concrete or dirt, wet or dry, straight or winding, level or uphill), the degree of visibility (light or dark, clear or foggy), the condition of the tires and brakes of his car, and – let us hope – his own condition (fresh or tired, sober or alcoholised) permitting him to judge the enumerated factors. Clearly, the driver of the automobile will not ‘measure’ the variables; he will not ‘calculate’ the time needed for the vehicles to cover the estimated distances at the estimated rates of speed; and, of course, none of the ‘estimates’ will be expressed in numerical values. Even so, without measurements, numerical estimates or calculations, he will in a routine way do the indicated ‘sizing-up’ of the total situation. He will not break it down into its elements. Yet a ‘theory of overtaking’ would have to include all these elements (and perhaps others besides) and would have to state how changes in any of the factors were likely to affect the decisions or actions of the driver. The ‘extreme difficulty of calculating’, the fact that ‘it would be utterly impractical’ to attempt to work out and ascertain the exact magnitudes of the variables which the theorist alleges to be significant … (because the action has become routine) and which perhaps he would never be able to perform in scientific exactness … Yet I can hardly believe that any sensible person would deny the relevance of the enumerated variables … The business man who equates marginal net revenue productivity and marginal factor cost when he decides how many to employ need not engage in higher mathematics, geometry or clairvoyance. Ordinarily he would not even consult with his accountant or efficiency expert in order to arrive at his decision; he would not make any tests or formal calculations; he would simply rely on his sense or his ‘feel’ of the situation. There is nothing very exact about this sort of estimate. On the basis of hundreds of previous experiences of a similar nature the business man would ‘just know’, in a vague and rough way, whether or not it would pay him to hire more men. (Machlup, 1946, pp. 534–5; emphasis original) So what have we learned from this observation? Orthodoxy treats the skillful behavior of the businessman as maximizing choice, and ‘choice’ carries connotations of ‘deliberation’. We, on the other hand, emphasize the automaticity of skillful behavior and the suppression of choice that this involves. In skillful behavior, behavioral options are selected, but they are not deliberately chosen. This observation directs attention to the process by which skills are learned, the predictions for the effective exercise of skill, and the possibilities for gross error through automatic selection of the wrong option. (Nelson and Winter, 1982, p. 94; emphasis original)

Evolutionary Economic Geography

46

When evolutionary mechanisms such as chance and increasing returns to scale are involved in the development of new economic activities, Lambooy and Boschma (2001) recalled that policy-makers might have several contradictory options for action. On the one hand, chance is linked with multiple equilibria for the location of firms or industries. Regional policy may not easily foresee or plan the possible development path. Suboptimal locationrelated solutions in cases with multiple equilibria must not be excluded. On the other, geographical space may play a minor part in the location of new business. Hence, there is room and opportunity for policy action. Policymakers can and should intervene to create a favourable local business environment (to secure a large market, provide investment funds, education, etc.). Still, there is no guarantee that this would be enough to attract and ‘lock-in’ the location of new firms and industries. In many cases new products, services and technologies both are developed and settle in unexpected locations, not necessarily in the ones that are the origin of previous locked-in old technologies and other inventions. 6 Complex systems and evolutionary economics Neoclassical equilibrium economic theory studies consistent patterns of behaviour that lead the economic system into stable state (equilibrium). This steady state means that there is no further movement or reaction to any action, external or self-grown internally. Complexity and evolutionary economics deals with actions, reactions, endogenous change, competition among various expectations, strategies and whims of economic actors. It is also concerned with the phenomenon of how these fit into the aggregate disequilibrium and ever-changing picture. As such, complexity and evolutionary economics ‘is not an adjunct to standard economic theory but theory at a more general, out-of-equilibrium level’ (Arthur, 1999, p. 108). This helps us fathom, at least partly at an aggregate level, issues such as the location in space and evolution of clusters, cities, persistence of poverty, appearance and demise of firms or the emergence of monopolies in a way that may assist us to be prepared to deal more effectively with these issues in the future. Nonetheless, these approaches are still a long research agenda with a plethora of new ideas rather than a rounded-up theory. Since its creation in 1984, the Santa Fe Institute (Santa Fe, New Mexico) has been perhaps the most active institution in the area of complex systems and evolutionary economics. One should not forget the role and influence of the International Institute for Applied Systems Analysis (IIASA) in Laxenburg, Austria, in this area of research too. Brian Arthur (1989, 1990a, 1994a, 1994b), associated with both Institutes, argued that certain models of production geography give weight to differences in factor endowment, transport costs, rents and competition. In such cases, the pattern of production locations is an equilibrium outcome. Hence, in these models history does not matter: the locational system is determinate and predictable (Arthur, 1994a, pp. 49–50). However, if one takes increasing returns and multiple equilibria into consideration, the new dynamic model of selection and adoption has four properties that create serious challenges and difficulties in analysis and policy-making (Arthur, 1989, pp. 116–17). These properties are nonpredictability, absence of optimality, inflexibility and path dependence. One may also add

Theory

47

time delays to Arthur’s four properties. Let us now consider these five characteristics in turn. Multiple and rapidly shifting equilibria (non-predictability) Meaningful multiple equilibria exist when firms make independent profitmaximising decisions about output, prices and location. This means that there is a chance for the existence of more than one possible outcome. Marshall (1890) was aware of the problem related to the application of equilibrium notion in cases with increasing returns. He noted that certain products might be attractive at the beginning principally to the rich because of their relatively high price. After a certain time it occurs to someone that it is a thing which the masses would like to have if it were brought within their range … Other producers follow suit, and after a time instead of a few hundreds being sold weekly at so many shillings, tens of thousands are sold for an equal number of pence. It is then possible, at all events theoretically, that the price and the amount produced may jump from one position of stable equilibrium to another. (p. 424) This was a theoretical underpinning for the mass production and appearance of Ford’s Model T. Sources of this multiplicity in a non-linear dynamic system are found in pecuniary externalities that originate in (large) fixed costs,57 increasing returns and imperfect competition. This result or these results vilify and undermine neoclassical welfare theorems and introduce a possibility for the role of government in the selection process to influence and, if possible, improve the result. This is a highly subversive development regarding the neoclassical equilibrium model that wants only one equilibrium outcome. If several outcomes are possible, then neoclassical equilibrium theory wants to know exactly what they are. ‘Besides positions of stable equilibrium, there are theoretically at least positions of unstable equilibrium: but their only significance lies in the fact that they are the dividing boundaries between two positions of stable equilibria’ (Marshall, 1890, p. 424). The question is, which equilibrium gets established? Will it be stable in the long run? Since a particular geographical space has only a limited influence on the location of new economic activity (especially footloose industries), what are the options for policymakers in this situation? Krugman (1991) reasoned, on the one hand, that there was a belief that the choice was basically resolved by history. Past events set preconditions that move the economy from one steady state to another. This reasoning, found in the traditional literature, argued that history mattered because of increasing returns to scale, lumpiness (inseparability), activity-specific knowledge, sunk costs and capital goods. There is, on the other hand, a view that the choice of equilibrium is determined by expectations. This observation is based on the belief ‘that there is a decisive element of self-fulfilling prophecy’ (p. 652). The relative importance of history and expectations depends on the structure of the economy and on the costs of adjustment.

Evolutionary Economic Geography

48

Ex ante knowledge, experience and preferences of firms, as well as potentials of technologies, may not be sufficient to predict the ‘market outcome’ regarding the (possible equilibrium) location of new activities. The outcome is indeterminate as equilibria are not locally unique.58 There is a coordination failure59 because of multiple equilibria. Such a dynamic or evolutionary system may always be out of equilibrium. Hence, when there is much uncertainty about the location of firms, especially new ones (because of their creativity, unpredictability and chance), comparative statics is largely a futile exercise in such circumstances. In addition, multiple equilibria may introduce the possibility of a locational ‘catastrophe’ (when all firms relocate and agglomerate in only one region). Therefore, the distribution of economic activity in space is potentially quite unstable and unpredictable in the (very) long term. The problem of indecision and ambiguity is old and known, but its solution is not always easy to find. Jean Buridan (1300?–58), a noted French scholar and philosopher of the late Middle Ages, considered one’s will. A well-known example of a paradox, ascribed to him as ‘Buridan’sass’, described an (unfortunate) hungry donkey that was in between two identical and equally attractive and accessible loads of hay (alternatives) who starved to death because it could not choose between them. One’s willpower may delay the choice. The situation and choices have to be analysed and assessed in order to opt for the greater good. Even now, if offered five different brands of cornflakes in a supermarket, a customer may decide about the purchase in an easier and faster way than if offered 200 different brands at the same time (in the second case the purchase is usually not made). The original idea for this paradox is found in Aristotle’s De Caelo (On the Heavens). Aristotle asked a question: how can a dog choose between two equally tempting meals that are simultaneously offered to him? Another expression of a similar puzzle between the rock and a hard place (frying pan and fire, devil and the deep blue sea) is known as ‘Morton’s fork’. It was named after John Morton, a powerful Lord Chancellor in 1487 during the rule of King Henry VII. Morton used to visit noblemen of the time and would then judged their tax for the coming year. If the hospitality was ‘economical’, Morton reasoned that the host was saving money, hence he could afford a ‘large gift’ to the King. If, however, the hospitality was sumptuous, the host was evidently wealthy and could afford a ‘generous gift’ to the King. These arguments were the two prongs of the ‘fork’. No matter what the noblemen chose to do and regardless of whether they were rich or poor, their ‘tax faith’ was the same. To round up the issue, one may recall that the task of a good tax collector is to pluck geese in such a way as to obtain the maximum possible feathers with the minimum possible noise. Venables (1999) recalled that the Ricardian model of comparative advantage predicts that there is a unique partition of industries in the world. Countries specialise in this model according to their comparative advantage. However, agglomeration forces operate in the real world. This situation produces a quite different set of outcomes than those predicted by the Ricardo-inspired models of trade and location. The spatial distribution of industries among countries is not unique. Moreover, it need not necessarily be in line with the comparative advantages of the location in question. Multiple equilibria arise from a general coordination failure. Hence, policy instruments need to be thought out in a way to surmount this failure. ‘Subsidies designed to overcome the coordination failure might be relatively large to attract the first entrant, and could then perhaps be successively smaller for further entrants’ (Venables, 1999, p. 506).

Theory

49

The international textile and clothing industry has been adjusting to the closed or quota-fragmented ‘global’ textiles market since the 1960s. When exports from one developing country hit the quota ceiling, the textile producers moved to another country to profit from its quota, which is how this ‘quota-jumping’ brought and located textile manufacturing for the first time to countries such as Mauritius. This isolated island state in the Indian Ocean protected its economy for a century with high tariffs. Nonetheless, its exports continued to be composed of agricultural goods. The government created a free economic zone in 1970, after which the garment firms from Hong Kong moved in. There was no textile production in Mauritius before the location of firms from Hong Kong. The islanders could have afforded sewing machines themselves and there was no problem obtaining literature (knowledge) about the garment industry, but this production did not start before the firms from Hong Kong located in Mauritius. The locals did not know how to use common knowledge for the production of garments. The Hong Kong firms brought this general knowledge on how to use it for pecuniary purposes. This raises a new question: when the low-technology knowledge is freely available and not used for commercial purposes because of coordination failures (perhaps), what can one expect to be the case with the freely available high-technology knowledge? Equilibrium theory is based on restrictive assumptions which include fixed and freely available technology and resources. It is confined to very special circumstances, but the power of equilibrium theory is that it has rather clear explanations, policy advice and predictions. Evolutionary theory refers to a complex situation with multiple and rapidly shifting equilibria. If this is too complex, unfortunately one may lose many strings and the ability to follow and understand what an abstract model does and suggests as the way forward. If an economic system is in equilibrium, that is for evolutionary theory only a transitory situation or a temporary ‘chance’. Hence this theory has often ad hoc and less firm explanations, weak predictions and policy advices than is the case with neoclassical equilibrium theory. Even though one may learn a lot from ‘simple’ equilibrium theory models, this is still confined to an illusionary situation. Market forces are weak in medical and military technologies. Hence, the ‘market’ selection mechanism is strongly mixed with and influenced by political and professional pressure and authority. Both the standard neoclassical equilibrium and the evolutionary theories have great difficulties in explaining such problems. They are areas where there is no real market, or where market selection is strongly mixed with political or professional influences. The problem in theorising here clearly lies not in the evolutionary form, but in the complexity of the subject matter. (Nelson, 1995, p. 85) Certain intellectuals may, for example, prefer Bach to punk music. Hence they may be contemptuous of pure market forces as a trustworthy judge of inter-generational enduring product quality, value and relevance.

Evolutionary Economic Geography

50

Potential inefficiency (absence of optimality) Choice and actions by players under risk and strong endemic uncertainty follow different, often more unpredictable courses than is the case in an idealised static neoclassical world with a stationary equilibrium as the final goal. Alternative technologies may compete passively (like certain species in biology) or actively and consciously. Active and strategic competition relies on sponsorship such as aggressive advertisement campaigns in which each rival emphasises his or her own superiority. Increasing returns (i-activities) may uplift the development of technology with inferior long-run aptitude as firms make irreversible investments under uncertainty.60 Perhaps adaptation in the medium and long term, rather than maximisation, may be a response to unforeseen events and changing environment. Biological evolution has almost unlimited possibilities, but the probability that a certain path is selected is quite another story. The path choices depend on what works and what doesn’t. Those that ‘work’, that are fit to survive and that replicate well go on to the next evolutionary step. They compete in the environment for their space with others and other species that have also successfully passed through the same evolutionary ‘filter’ process on their own. The ending point is not predestined or planned. ‘For the LORD knows the way of the righteous, but the way of the wicked will perish’ (Psalm 1:6). Let us imagine that an economic agent (a firm) is trying to improve its fitness on the market. Let us also suppose that the market is a three-dimensional landscape with peaks and valleys. Peaks are of various heights, some similar, others not, and they are unevenly distributed across space. The task is to move the location of this firm up in this imaginary (market) space. Rather than looking for a global optimum position for this firm, the task is to find the highest peak we can at a given moment. Constraints in the environment create conditions for a fitness function. Before we start our hiking trip, we are going to introduce three additional challenges to our search. First, we have to imagine that it is a pitch-black, moonless night. You can sense if your wanderings are taking you higher or lower, but that is about it. This represents evolution’s inability to ‘look ahead’ in any sense; all evolution can do is try something to see if it works. Second, this landscape has a dangerous feature to it. Kauffman uses the image of a poisonous fog that clings to the low-fitness valleys and crevices of the landscape. If you go too low in your wanderings, you will descend into the fog, choke, and die. The fog represents natural selection – at some point if your fitness is too low, you will be selected out. Third, the landscape is not static, but is changing all the time. As the environment changes, the fitness function changes, and therefore what is a high-fitness peak today might not be a high peak tomorrow. As the environment shifts, the landscape bucks and heaves as low-fitness valleys are thrust up into new mountains while formerly high peaks collapse into lowlands below the poison fog. At any point in time, some regions of the landscape may be stable, others may be shifting gradually like the

Theory

51

movement of tectonic plates, while still others might be highly active, the equivalent of fault lines full of dramatic upheavals, earthquakes, and eruptions. Some of the changes to the landscape, such as an asteroid’s hitting the earth and altering the climate, represent random changes to the environment. But much of the bucking and heaving is the result of the evolution of species themselves. Previously we noted that all species exist in complex webs of relationships with other species: predators, prey, symbiotic, parasitic relationships, and so forth. As species coevolve, engage in arms races between offense and defense and between cooperating and competing, an evolutionary change in one species can set off a chain reaction in the fitness of other species. (Beinhocker, 2006, p. 207) The changing landscape may be thought of as a resistance to predators or distribution of inhabitable areas because of changes in climate. The three strategies to explore this complex and evolutionary landscape are proposed by Beinhocker (2006, pp. 208–12). The first strategy from a random starting point is an adaptive walk. You take a step in a random direction and if it takes you up, you move there; if not, you return your leg to the starting position before you take another step elsewhere (Figure 2.1). This strategy is quite effective in exploration of individual peaks, but its limitation is that once you climb the local maximum, you may not be aware that there is a higher peak just across the valley but that you have to go down first to get to it. You will never find it. The second strategy is a random jump. Imagine that you find yourself in a random location in the situation just described as adaptive walk, but that you have a powerful pogo stick with which you may jump in a random direction over a random distance (Figure 2.2). The attractive advantage of this strategy over adaptive walk is that you are not tied to the local maxima; you may hop over valleys and you may get to higher peaks. The disadvantage of this strategy is that you have a good chance of finding yourself in a death valley. Peaks are, after all, peaks, while valleys are all around. Adaptive walk may at least keep you somewhere up, out of valleys with poisonous fog (until the geography of the landscape changes). Therefore, random jump is a riskier strategy. Third is a strategy that mixes the two previous strategies (Figure 2.3). Adaptive walk lets us come to a local peak, while random jumps give an opportunity to avoid being stuck on a local peak. Now imagine the same situation with another twist. Instead of one ‘hiker’ who explores this landscape, imagine a multitude of them at the same time, each with his own strategy. ‘Bets’ are dispersed. Some of them will lose, while there is certain chance that some of them will climb the highest peaks without being stuck on local highlands or suffocated in valleys. The evolutionary process in biology and in economics ensures that there are agents (hikers) spread out across space. They try out various possibilities, they cross bridges and jump, they differentiate, select, amplify, mutate; and the fittest survive, multiply and go on. As the landscape changes, some peaks and highlands collapse, so only the ones that are on the top, the fittest ‘scouts’, rebuild the population. Therefore, bets ought to be spread out across the landscape. There may come a sudden moment when those who were in the valley come to the top, sometimes rather fast. When times are good, more

Evolutionary Economic Geography

52

resources may be devoted to the exploitation of the situation. However, when times are bad, more resources may be devoted to the exploration of new ways out of trouble. Evolution is a process that ‘gambles’, but ultimately it plays the odds quite well. In the evolutionary randomness, there is some order in the longer term. Energy is continuously used to fight entropy and to introduce and keep certain order through adaptations (fit designs). Evolution is a process of continuous innovation. The recursive nature of the algorithm never stops. This is essential, given the constantly changing nature of the landscape. There may be periods of more or less active search as evolution balances exploration and exploitation, but the search is never complete. The system has no equilibrium – in evolutionary systems, stasis is a recipe for extinction. In effect, evolution says, ‘I will try lots of things and see what works and do more of what works and less of what doesn’t’. (Beinhocker, 2006, p. 216; emphasis original) There were many trials to find a standard good design for various things. Bad ways always outnumber the good ones. The design of a bicycle may be illuminating: there were various sizes of wheels, transmission and steering. The core standard for a bicycle might have been arrived at, but tinkering ' continues, variations are constantly appearing

Figure 2.1 Adaptive walk Source: Beinhocker (2006, p. 208)

Theory

53

Figure 2.2 Random jumps Source: Beinhocker (2006, p. 209)

Figure 2.3 Combining adaptive walk and random jumps Source: Beinhocker (2006, p. 210) on the market. Aircraft have been passing through continuous reshaping and remodelling. One does not dare to pass a judgement on the settling of a long-lasting standard for garments and shoes. Female, of course. This may sound somewhat pessimistic about our longer-term prospects of facing, predicting, outsmarting and even shaping the future within the evolutionary framework. If evolution and complexity are smarter than we are in the medium and long term, we may nonetheless be able to do two things: first, create and equip ourselves with institutions (e.g. social security) that may help cushion the possible negative evolutionary events; and second, take actions, such as the allocation of resources in education and knowledge-creation in certain directions, that may support the future evolutionary developments and fight entropy. In any case, the future is no longer what it used to be in the neoclassical equilibrium world, while the past lasts longer. A famous example of being stuck on a local plateau relates to the QWERTY arrangement for typewriter keyboards. Its name refers to the first six top-line letters on the keyboard. Christopher Latham Sholes, a Milwaukee mechanician, designed this keyboard at the end of the 1860s. He was more interested in a distribution of keys to

Evolutionary Economic Geography

54

prevent jams rather than in fast strokes.61 In fact, the QWERTY distribution of letters on keys was designed to slow down typing in order to prevent mechanical jams. Augustus Dvorak, a professor at the University of Washington, designed the most comfortable, user-friendly and efficient keyboard during the 1920s and patented it in 1936.62 The Dvorak keyboard allegedly offers the possibility of significantly increasing the speed of typing. However, this keyboard still languishes in obscurity. Even though it has repeatedly been shown that the QWERTY distribution of letters is suboptimal, this system continues to be the standard for keyboards. This is a clear case of market failure where an inferior product ‘locks in’ and takes the market. Liebowitz and Margolis (1999; 2001, p. 13) sharply criticised the pathdependence claims and lock-in effects in this case. A standard story holds that the most widespread and locked-in QWERTY keyboard is much less efficient than the Dvorak keyboard, which was developed more than half a century later. Liebowitz and Margolis showed that the Dvorak keyboard offered no advantage over QWERTY and that claims for Dvorak’s superiority were mainly based on bogus research by someone with a financial interest in the success of the Dvorak keyboard. Who is right in the QWERTY case? The lock in thesis (David, 1984) or its critique? In this case with uncertainty and multiple equilibria, the QWERTY may have two advantages: the first-mover’s advantage and potentially no inferiority to other existing or forthcoming solutions. Foresight, word of mouth among consumers, consumer journals, moneyback guarantees and other marketing efforts all work to prevent lock-in. If everyone knew that a new product was sufficiently better, wouldn’t someone find a way to help us switch? Liebowitz and Margolis argued that it was extremely likely that someone would, especially since in doing so there would be chances to ‘make money’. (Someone would pick up the Є50 banknote that was on the pavement.) Competitive advantages in hightechnology products are short-lived and temporary. Instead of antitrust actions, policymakers ought to rely on competitive entrepreneurship (strongly linked with innovation) within the market system. The market evolutionary process often chooses what is good at replicating itself, not what is good by itself. Hence it may be always desirable to keep more than one technology alive. This may avoid problems related to monopoly and assist in having some variety standing by in reserve if things go wrong with the dominant technology, against a possible future ‘technological Bhopal or Chernobyl’. A superior technology may have bad luck in the situation with uncertainty in gaining few early adherents. For example, the American nuclear industry is dominated by lightwater reactors because of the adoption of such a reactor to propel the first nuclear submarine in 1954. Engineering literature asserts, however, that a gas-cooled reactor would have been a superior choice (Arthur, 1989, p. 126). If the claim by engineers is that Sony’s Betamax is a technically superior system for videotape recorders, then the market choice of the VHS (about 1980) did not represent the best outcome.63 Similar arguments could be used for the initial triumph of user-unfriendly DOS over Macintosh during the mid-1980s.64 VHS and Microsoft did not prevail on the market because of the superior quality of products based on their technology, but rather because of their ubiquity. Liebowitz and Margolis (2001) considered these alleged lock-in types of market ‘failure’. They argued that there were no failures at all and that ‘good products have won’ (p. 237).

Theory

55

A new market battle started in 2006, for the standard format of a new generation of high-definition DVDs, between Sony’s Blu-ray format and Toshiba’s HD-DVD. Sony has Philips, Pioneer and Dell on its side, while Toshiba has Microsoft and Intel. Sony learned a lesson from the Betamax commercial failure. This time the company built a broad coalition in the consumer electronics industry prior to the launch of the Blu-ray format and it believes that it has most of the Hollywood studios on its side (Sony, Disney and 20th Century Fox, although Universal Pictures is on Toshiba’s side). The Sony camp also believes that it has a technically superior product. The problem for Sony’s player is its cost of $1,000, twice as much as Toshiba’s. Another anxiety is that film studios and technology partners may easily switch camps. Yet another fear is consumers’ hesitation to bet on one of the offered product technologies. They may hesitate with their purchase, as they may be afraid to make the ‘wrong’ choice and suffer the disadvantages that go with it. Retailers will have to stock both products; studios will have to produce in both formats. And if consumers do not buy either, that is a nightmare scenario for both producers and retailers. In fact, consumers were quite indifferent regarding the format. They just did not want to buy the loser, i.e. the format which would become obsolete within a year. A new Sony PlayStation 3 (about $500 in 2007) is equipped with the possibility of playing Blu-ray format films, and the big question is whether this Trojan horse in game-players’ homes will make them use this platform to play films. If one of the two rival systems fails to make a lead, this could mean a fiasco for either of them. If at the end of the day one wins, then the moral of the story is that the location of production will be in the hands of one of the Japanese (domestic) companies. Both Sony and Toshiba lowered prices for their products significantly towards Christmas 2007. The intention was to tilt market choice of the format in their own respective favour. Sony’s Blu-ray had a narrow lead in the titles sold in the US, while it had a clear lead in Britain, France, Germany and Japan. Warner Brothers, one of the world’s largest producers of film and television entertainment, gave exclusive support to the Blu-ray format. This may have tilted the market victory towards Sony at the start of 2008 and may also boost Sony’s confidence after the market failure of its Betamax video tapes. It may also increase its profit, as the winner gets a virtual monopoly that offers possibilities for higher fees. If this happens, consumers may turn out to be losers.65 Competition is a superb device that may greatly assist in the development of new technologies. However, merging mew technologies into a single standard that supports compatibility and is open and general is the finest way to commercialise new technologies on the market. Consumers, confident that the new standard will not disappear fast, would decide on their purchase earlier and faster. Firms ought to compete to supply output that is compatible with the new standard, rather than competing with (perhaps) unfriendly standards. However, the possible Blu-ray format victory may have little medium-term significance. New discs are a marginal technology, just as (the now forgotten) MiniDiscs were an interim step between audiocassettes and compact discs for music. Microsoft was a thriving company in 1987. This was a good six years after it won the IBM contract for MS-DOS. In line with evolutionary competition among expectations, MS-DOS was at the end of its market life. IBM, Sun Microsystems, AT&T and others were developing new user-friendly operating systems. Apple was also a dangerous

Evolutionary Economic Geography

56

competitor. Bill Gates, the boss of Microsoft, had various options in this situation: he could gamble and develop a new operating system (Windows), join the well-funded research efforts of potential competitors, or sell Microsoft. Beinhocker (2006, pp. 335–7) argued that Gates followed a more interesting strategy. Microsoft followed six business experiments simultaneously: • It continued investing in MS-DOS. Users were used to this system and were quite wary about switching to something else. • It joined forces with IBM to create an operating system. A leaf was taken from the film The Godfather Part II, where Michael Corleone said: ‘Keep your friends close, but your enemies closer.’ • It entered into discussions about joining efforts to develop further Unix, just to see what was going on there. • It purchased a major stake in the company Santa Cruz Operation. This was the biggest seller of Unix systems. If Unix takes off, Microsoft is there. • It made applications for Apple Macintosh. • It invested heavily in a new product of its own: Windows. This would combine the best of all worlds: while compatible with MS-DOS, it was as multitasking as Unix and as friendly as Macintosh. Bill Gates created a mixture of strategic business options. He was searching in the evolutionary business space using both adaptive walk and random jumps. If one option fails, there is a population of competing others that may work and assist in adapting to the new situation. ‘Journalists cried that Microsoft had no strategy and was confused and adrift: they wondered when Gates was going to make up his mind’ (Beinhocker, 2006, p. 337). The general lessons for management from this business engineering strategy by Microsoft are fourfold: • A context for a business strategy ought to be present. • There needs to be a process of differentiation among various strategies: should one exploit what is now a successful product or service on the market? Should one invest in innovation and create something superior or different? There ought to be a certain population and spread of business bets and risk. Not all firms may pursue such a strategy; it is too costly and uncertain. But recall that Microsoft was the smallest and most resource-limited company during the second half of the 1980s when it entered this race with others. In addition, not all options in the portfolio got an equal share of relevance and attention. Options branched out, but they were all within the same industry. Microsoft was not creating a business conglomerate that was operating in various industries, as was the fashion during the 1960s and 1970s. • The selection of strategies ought to reflect the situation in the market. If market signals are blurred, infighting to protect each project may be strongest. If there are certain signals, then rather than fighting tooth and nail for each fiefdom, there ought to be certain flexibility to transfer resources to the promising projects. • A business strategy that is potentially successful in the future ought to be amplified, even though it may not be immediately as profitable as the current, established one. Various options ought to be open as long as it is reasonable. There are three important aspects in successful amplification: ‘knowing where the “nectar” is in the portfolio of experiments, having some spare “bees” around to swarm onto attractive opportunities,

Theory

57

and having an ability to shift from one “field” to the next as the situation changes’ (Beinhocker, 2006, p. 345). The theoretical dogma in neoclassical equilibrium economic theory is that free markets automatically balance themselves. Supply and demand end up in equilibrium. The best product is selected, while the allocation of resources is most efficient. This view of how the world operates regards the QWERTY example as an anomaly. Otherwise, the neoclassical equilibrium model cannot work. The problem is, however, that there are too many QWERTY-type examples (Annex II). They demonstrate that small events at critical times in a complex economy have crucial and long-term implications on business and its location in space. This is exactly the case that made sense of many otherwise unexplained or neglected occurrences in economics. Inflexibility (lock-in effect) Once an outcome such as a dominant technology or a location begins to surface, it turns out to be ‘locked in’ (stuck) and it persists for a long period of time. This ‘overcommitment’ to a particular technology or a location may be the consequence of: • increasing returns based on huge sunk costs, • strong externalities (positive feedbacks), • an early government procurement done before other possible outcomes were genuinely considered and evaluated, or • a risk-loving firm marketing its new product with a low introductory price or a moneyback guarantee, just in order to create a critical mass of clients necessary to tip demand and new standards its way.66 However, this matters if choice is irreversible, which is often the case over a certain longer period of time. In order to replace an entrenched product or technology, the new one must be very much superior to the existing and accepted one. It must significantly increase convenience in use to consumers, and it has to be available at a price that takes into account three factors: • culture (openness to change); • investment of capital and time (learning) into the new good or service (once people learn to use something, they do not want to switch easily to something else); and • convenience in use. Once locked in, the system may be in a ‘conditional evolutionary equilibrium’. This transitory ‘stable equilibrium’ may not be necessarily optimal. However, even if it is optimal at a given moment in time, it is subject to redefinition by the system’s internal (improvements in something in existence) and external (radical change in technology) dynamics of change (evolution) over time. Even the locked-in QWERTY may be challenged by the new technology that converts voice (even thoughts) into a typewritten text. As far as commitment to a mission or business is concerned, one may wish to consider the decision by Hernán Cortés in 1519. After unloading his eleven ships in Veracruz (now Mexico), Cortés ordered them all to be burned on the spot. Even though they did

Evolutionary Economic Geography

58

not know what was waiting for them in the new land, his 600 men with fifteen horses (never seen before by the natives) and several cannons had ‘no other option’ but to go inland. Retreat was impossible. Courage and love of risk may be also necessary. ‘John Gutfreund, chairman of Solomon Brothers, one of the most aggressive investment banks of the 1980s, said successful traders must wake up each morning “ready to bite the ass off a bear”’ (Kay, 2004, p. 315). A few examples. VCRs were introduced during the 1970s. A new superior and marketable DVD technology arrived to replace VCRs after three (long) decades. In other areas, such as software, Microsoft’s Word replaced in a much faster way early textprocessing programs such as WordStar, Chi Writer or WordPerfect because it was so much more convenient and powerful that this justified the trouble to invest time, money and effort in switching and upgrading. Earlier, even though electricity convincingly proved its superiority over coal, steam and smoke, it took it more than three decades to provide half of the mechanical power to firms in the US and to replace old steampowered factories with new, electricity-driven ones. The lock-in effect may have two sides. On the positive and initial side, firms have the benefits of specialisation and economies of scale. This business vigour may last sometimes even for decades. However, on the negative side, complacency with the past and current success may introduce inflexibility, lack of adaptability and slowness to changes in technology, competition and demand. Such firms may turn ‘fat and lazy’ in a slow-growing market. The case in question is with the British paints and chemicals group Imperial Chemicals Industries (ICI), founded in 1926 and acquired by its competitor the Dutch Akzo Nobel in 2007. A location, industry or a firm is trapped in rigid specialisation and over-committed to a certain way of doing things (skills, habits and routines). Earlier engines of growth turn out to be weaknesses and hindrances to a continued growth. An external shock such as a radical change in technology, materials or the rise of new competitors may be tackled by self-protective cost-cutting or qualityreduction measures. There may be redundancies and people will go away. Even if firms are willing to adjust in this way, this only fortifies the ‘negative lock-in’. The problem in such situations is not always the lack of change, but rather its substance. An alternative strategy may move firms to a new and more competitive stage by innovation and a change in technology. People, in particular innovative people, need to be kept, and new ones attracted to come to firms and regions. Openness to continuous learning (a strongly path-dependent activity) is also necessary, not only within a firm’s own area of technical expertise but also from other related and totally unrelated areas (new materials, for instance). Continuous changes in markets and technology create a steady tension linked with old path destruction, as well as the need for and prospect of searching and creating new paths. Established locations can be vulnerable to ‘technological lock-in’ in certain cases. New ideas may need new space. Certain firms may be intent on escaping and breaking constraints that impose a path-dependent legacy upon them, not only in technical and production dimensions but also regarding various social and political compromises done in certain locations. For example, ‘When IBM developed its own personal computer, the company located its fledgling PC capacity in Boca Ratton, Florida, way outside of the manufacturing agglomeration in the North-east Corridor’ (Audretsch, 1998, p.24). Another example: vacuum-tube-based electronic products (radio and TV equipment)

Theory

59

were located on the east coast of the US. Then came technological miniturisation of components which replaced vacuum tubes, first with transistors and, later, with siliconbased semiconductors. The new semiconductor industry was not technologically and spatially attached to its ‘parent’ industry in the east; instead, it was established in Silicon Valley (west coast) as it searched and found its own supply chain of knowledge and other inputs (Storper, 1995, p. 209). These are examples of bifurcation points in the spatial location of firms and industries. A ‘clean sheet’ found in new and ‘green’ locations may not always be a business organisation panacea. New and flat production hierarchies may be hard to implement in certain new locations for business. For instance, carmakers found it hard to get line workers in India to maintain and clean their workstations (an element of Toyota-style worker involvement) ‘since sweeping floors has traditionally been a task assigned to the lowest, “untouchable” caste of society’ (Sturgeon and Florida, 2004, p. 66). Certain industrial regions may become victims of their own past success. Institutional sclerosis, vested interests of large firms, labour unions and public authorities may oppose changes and adjustment to new circumstances. For example, the German Ruhr region was led into the ‘trap of rigid specialisation’ (Grabher, 1993, p. 275). The Belgian Walloon region is in the same situation. Strong and relatively successful regional organisational strictures may become, over time, a liability. Once certain skills, habits and routines have been learned, it is tough to unlearn them. This is the weakness of strong ties, in particular if these skills, habits and routines were successful in the past (the competency trap). They act like genes in living organisms, storing information in the organisational memory, and passing it on among the members of the same firm or from this firm to a newly created firm or a partner firm in a cluster or region. Such a system may be open to ‘spontaneous mutations’, i.e. innovation. However, these innovations may be only incremental rather than radical. Nonetheless, certain strong competencies may decay over time if not maintained and refreshed. At one point, this lock-in was highly profitable because of high specialisation, increasing returns of scale and competitive advantages. However, vested interests may strongly oppose change, even though past expertise becomes both redundant and a burden that prevents progress and adjustment over time.67 New dynamic faculties and institutions, open and adjustable to new ideas, their implementation and change, are necessary to break with the lock-in situation and to change the old development trajectory (Boschma, 2004, p. 1004). It is sometimes overlooked that in certain cases new footloose industries often create their own conditions for growth. For instance, new environment-related industries are emerging in the Ruhr (the connection may be the solution of the pollution problem). Former coal and steel giants such as Krupp, Thyssen and RAG diversified and invested in environment-related technology, plant engineering and control services. It is estimated that these new activities include about two-thirds of these companies (Hospers, 2004, p.152). Elsewhere, there is little technological continuity between the textile machinery complex in New England of the nineteenth century and the current electronics cluster (Boschma and Lambooy, 1999b, p.394). Regional competitiveness does not have a clear and widely accepted meaning in the literature. What are the similarities, differences and other dimensions in competition along which various regions, cities or clusters compete? Can the established competition concepts applied to firms be sensibly applied to regions too? Do reactions to

Evolutionary Economic Geography

60

opportunities and adjustment to challenges brought by competition differ between firms and regions? Do they behave differently in the short, medium and long term? As there is no agreed solution to these problems, it comes as no surprise that both academic and policy debate about this issue is associated with many puzzles. Linked with this is an open evolutionary question: would the increased speed of new knowledge creation boost the possibility of more easily liberating those regions that are ‘locked in’? This process is described graphically in Figure 2.4. While new knowledge may initially boost a region’s economy, the same knowledge may provide a hindrance to progress in the future. The evolutionary model argues that a constant inflow or creation of new knowledge is

Figure 2.4 From positive to negative lock-in regional path-dependent economic development Source: Martin and Sunley (2006, p. 416) necessary for progress. We shall refer to this issue below, while considering the Red Queen race. In spite of certain initial and temporary costs, regions and firms may successfully and gradually reinvent themselves. Clear examples among countries that found ‘escape routes’ may be found in Ireland and Finland, among regions in Massachusetts (Boston) or the Ruhr,68 while a firm such as General Electric has been reinventing itself for over a century. From a manufacturing conglomerate that produced durable goods, the company turned to technology, commercial finance and media. Hence, in this new situation the function at the right-hand side (not shown) in Figure 2.4 has an upwardsloping (curled) shape. This is, of course, only a possibility (not a sure outcome) that may serve simply as an example to others. The productivity/application curves of new technologies over time are described graphically in Figures 2.5–2.7. Figure 2.5 presents four distinct phases in the life of a

Theory

61

technology from its introduction (Phase 1) to its maturity (Phase 4). Phase 5, decline and eventual demise (not presented), does not necessarily and always follow maturity (Phase 4) in all cases. Figure 2.6 shows two arbitrary and discontinuous technological jumps. Technology A is replaced by an irregular innovatory leap forward that creates technology B. The new technology needs some time to improve and its life curve is similar to developments presented in Figure 2.5. Even though the change is discontinuous, it is relatively sustainable as one technology follows the other. Then comes a new technological breakthrough, technology C (which may take off without a long adjustment period). The situation often found in real life, in which several technologies may exist simultaneously, is described in Figure 2.7. These types of changes are disruptive. Therefore, discontinuities may not be easily discernable at an aggregate level. A newly arrived technology (or product) may not perform initially to its full potential at an early phase in its evolution. Hence two or more technologies (products) may coexist for a certain period of time until the market ‘filters out’ the ones that have a weak market acceptance. Examples of this case include the simultaneous existence (at least for some time) of steam, liquid fuel and electric locomotives; computer diskettes, CD ROMs and flash memories (USBs); film and digital cameras; or CDs and iPod music. At any moment in time, the two (or more) technologies will each be most efficient in their own niches. For example, because water power is free, it was economical to use water power in some textile factories right up to the early twentieth century. One has to keep in mind that various technologies may be strongly hierarchical. This means that some may not be able to operate without others. Few modern technologies may run without electricity, while others (communication) may not work without computers. Figure 2.8 presents the transition in energy use to power machinery from steam to electricity in the US in the period 1869–1939. Up to 1899, more than 95 per cent of

Evolutionary Economic Geography

62

Figure 2.5 Productivity/applications curves for a general purpose technology. Curve A stands for two quite distinct curves with different scales on the y-axis. It can be regarded as either a productivity curve or applications curve. Source: Lipsey et al. (2005, p. 433)

Theory

63

Figure 2.6 Productivity/applications curves for a succession of general purpose technologies. The curves show the transition from A, which is still in Phase 3, to B and than from B, which has already reached Phase 4, to C. Source: Lipsey et al. (2005, p. 437)

Evolutionary Economic Geography

64

Figure 2.7 Productivity curves for different patterns of two successive general purpose technologies. The incumbent A is replaced by several alternatives (B, C, D) that pass the various selection criteria. Source: Lipsey et al. (2005, p. 438)

Theory

65

Figure 2.8 Diffusion of mechanical power in the United States manufacturing firms. (Percentage of total horsepower from the three sources of mechanical drive in manufacturing, 1969-1939) Source: Devine (1983, p. 351) and Atkeson and Kehoe (2007, p. 67) mechanical power came from water and steam. Since then, electricity has gradually replaced these two primary sources. In 1929, electricity provided over 75 per cent of mechanical power. Three mechanical power technologies existed side by side for a relatively long time. Electricity started to be distributed to manufacturers from 1869. It took electricity fifty years from its introduction to provide half of the mechanical power. Firms had to invest a lot to reorganise and to make the full use of the new source of power that permitted a natural sequence of production process. They had to replace the previous shaft-type organisation of production and to replace overhead mechanical transmission of power. The removal of a labyrinth of belts also improved the work environment (ventilation, illumination and cleanness of factories). However, if one starts observation from 1899, when the share of electricity was 5 per cent, it took it only twenty years to provide half of the power consumed by manufacturing firms in 1919. Path dependence, strong history and non-ergodicity69 There are small, unpredictable, surprise, random and arbitrary events (chances, accidents, first-mover advantages, serendipity) at a micro level during a crucial time in the past that may decisively and irreversibly challenge the established order and change our behaviour. These small and sometimes spontaneous past events in space ‘chose’ from various locational possibilities around which crystallisation began. In the real and second-

Evolutionary Economic Geography

66

best situation and with increasing returns to scale in a non-ergodic system, these small events became magnified and self-reinforced, tipping the system towards a certain outcome. This inertia may set in motion mutations and routine in economic structures, at least for some time in the future.70 A few examples. When IBM wanted to launch its PC and enter the fastgrowing market for Apple’s products in 1980, they approached a small company in Bellevue, Washington. IBM wanted to consider the possibility that a small and flexible company called Microsoft might write a programme for its PC. During the initial meeting, IBM asked Bill Gates for his advice on the operating system the new PC ought to run. Gates’ suggestion was that IBM ought to talk to Gary Kildall of Digital Research, who had an operating system that became standard for hobbyist microcomputers. Kildall was doubtful about IBM’s visit, and when they came to meet him, he went hotair ballooning while his wife and lawyer, whom he had instructed not to sign anything, met IBM’s team. The frustrated IBM team turned back to Gates, asking whether he would be interested in the operating-system project. Gates accepted the proposal. (IBM could have bought Microsoft, but did not want to. As a market leader in the production of mainframe computers, IBM might have come under government pressure to break up. Hence IBM unbundled the software business.) While licensing the software product, Gates retained the right to sell his Microsoft Disk Operating System (MSDOS) to non-IBM producers. As IBM was more interested in hardware than software, the deal seemed fine. The rest is history: a product complementary to a computer (i.e. software) turned out to drive the market and to set the relevant standard. What would have happened if Kildall had not gone hotair ballooning? What if Microsoft had not gained IBM’s agreement to sell its products to other producers (Beinhocker, 2006, p. 327)? The effect of such small and seemingly irrelevant events for the long-term ‘big picture’ at the time when they take place is impossible to predict. This impact is often hard to recognise during the time when these events occur. Later on, they turn out to be a punctuating, a bifurcation point that decisively turns the evolution of an industry, even of a society. Or another example, When downward pressure on the Thai baht started 10 years ago, nobody expected what followed – its devaluation in early July. That seemingly small event generated a financial tsunami that engulfed most of east Asia and overwhelmed Indonesia, Malaysia, the Philippines, South Korea and Thailand. Exchange rates collapsed, financial systems went bankrupt, governments teetered on the edge of default and economies succumbed to deep recessions. Officials from the International Monetary Fund raced from one crisis-hit country to the next. In its last movements, the crisis went global, overwhelming Russia in August 1998 and Brazil in early 1999.71 Earlier, Annie Oakley was a woman who worked for Buffalo Bill’s Wild West Show. She was famous for shooting dimes tossed in midair, corks from bottles and ash from a cigarette held in the mouth of her husband. While she was touring Europe in 1890, Wilhelm, the Crown Prince of Germany (the future Kaiser Wilhelm II), saw her show several times. He invited Annie to shoot a cigarette held in his own mouth, but Annie had

Theory

67

Wilhelm hold the cigarette in his hand, not his mouth. Many ask: if Annie’s hand had trembled and she had shot Wilhelm, would the First World War ever have taken place?72 Some may remark that the above is not a good example as the Kaiser did not have a lot to do with war policy; some, but not a lot. A better example may be: what if Neville Chamberlain had not given in at Munich (1938) and Adolf Hitler had given the order to invade and fight Czechoslovakia and the plot carefully prepared by the generals to unseat him had gone forward (it was to be timed for the moment Hitler gave the order which he never had to do because of the Munich Agreement)? These small events/facts, like genes in biological evolution, may irreversibly decide behaviour and the direction (new path) towards the final outcome.73 These events are not normalised, or averaged away, or forgotten by the dynamics of the system, or shaken free of new history. They are carried on into the future by the accumulation of knowledge, experience, technology and institutions. Hence, a set of these past events together (history) favours a certain bias in the development path.74 This bias may limit or widen the potential choice of outcomes.75 This choice of the outcome and the development path that leads to it need not be either economically optimal or organisationally rational (as seen in the QWERTY example).76 Local deposits of natural resources, huge sunk costs, technological legacy, learning effects, specialisation, increasing returns to scale, networks and location-specific institutions are the principal sources of path dependence.77 However, the configuration of the past and current interaction in space and time between these ‘sources’ and firms introduce, after a certain minimum threshold, the irreversible element into the system, at least for some time in the future. There is no homeostasis (habitual return to the initial equilibrium), so the historical tape, apologies, historical CD is played ‘only once’. This is a demanding, discrete, uncertain, new path-creation, non-stationary, differentiation, selection, adaptation and amplification process in which new knowledge is build on previously accumulated knowledge and experience, layer by layer. The results of this process are unique and without predestined outcomes. Such an evolutionary (i.e. selforganising, long-term and progressively changing) view surpasses clear-cut orthodox, instantaneous, ‘mechanical’ and linear economic models in the long-run promise of understanding and (partially) influencing events. One of biggest problems in the evolutionary model is that we know little about how it works. How does this fresh uniqueness come about,78 and where it will be located? Why does it occur? Well, it may be the consequence of human ingenuity or the entrepreneurial spirit, an outcome of investment in knowledge (R&D) or serendipity. The outcome of this process may be known, but its origins are in a penumbra. Why does it work in some locations, industries and firms and not in others? When will/may it take place? What is the threshold (if it exists)? Is there a correct or preferred timing for the start of the new industrial and technological trajectory? What is its speed and might? What is the replication and reproduction mechanism? Does it exist? Certain unresolved issues associated with path dependence (strong history) are summarised in Table 2.2 by Martin and Sunley (2006). The issues raised may act in part as an agenda for research in this area.

Evolutionary Economic Geography

68

Time delays Neoclassical equilibrium theory imagines that changes in the economic system are very fast, if not instantaneous. The real world is, however, full of imperfections such as a lack of inventories or sophisticated personnel, or insufficient production capacity. They all have an impact on the time delay between a policy action and the desired or expected feedback. Time delays are inherent in dynamic systems that have multiple equilibria. Imagine that you are in a town that is new to you and that you check into a hotel that you do not know. Suppose that you want to take a shower in your hotel room. You turn the handle and you wait a while between your action and the feedback from the water temperature. Once it starts flowing, it may be too cold so you turn the handle in the direction of the red arrow. After a certain time delay, the water is too hot (the temperature overshoots the level you want). You then turn the handle in the opposite direction, and what happens? The water may be too cold (undershooting). Eventually, after some time you figure it out, the oscillations become smaller and, eventually, you have the temperature at the desired Table 2.2 Unresolved issues associated with path dependence Problem/limitation

Some key questions

Are there different types, degrees and causes of path dependence?

Under what circumstances does path dependence arise? Is it possible to separate causes from consequences? Are there strong and weak forms of path dependence or ‘historicity’ in the economy? Do different elements of the economy exhibit differing degrees of path dependence? Is path dependence a pervasive feature of the economy?

Problems surrounding the meaning and nature of ‘lock-in’

What, precisely is ‘lock-in’? Does path dependence differ from ‘lockin’?Is‘lock-in’ a necessary/inevitable feature of path dependence? Is ‘lock-in’ necessarily associated with inefficient outcomes? Can ‘lockin’ have positive features and outcomes? Under what conditions does positive ‘lock-in’ become negative ‘lock-in’? Is it possible to avoid ‘lock-in’?

Lack of discussion of path creation

How do (new) paths emerge? How does novelty relate to path dependence? How are new paths selected? What is the relationship between path dependence, path destruction and new path creation? Are new paths latent in old paths?

Does path dependence need In what ways is path dependence intentionally created by actors, or an a theory of human agency? unintentional emergent effect at system level? How do agents interact with, reproduce and transform the path dependence structures within which they are embedded? How do agents create new paths? In what sense is path dependence an evolutionary process?

What sort of evolution is implied by path dependence? Do all evolutionary processes imply path dependent, but not vice versa? What is the relative significance of incremental versus radical change? Is the notion of economic adaptation consistent with path dependence?

Source: Martin and Sunley (2006, p. 404).

Theory

69

level (Beinhocker, 2006, p. 101). Nature has similar oscillation and balancing features. For example, the number of lions depends on the number of ‘available’ antelopes and zebras. The more zebras and antelopes, the more food for lions. The population of lionsmay increase in number. An expanded number of lions reduces the number of zebras and antelopes, which in turn reduces the number of lions (they have less food). The fewer the lions, the higher the number of zebras and antelopes … or animals that find food in other animals’ eggs or pups, and so on. If there is sudden surge in demand for a firm’s product and if this firm works at full capacity, management may decide to invest in a new plant. This plant is built and starts operations after a certain period of time. But when this happens, the demand for the product may turn out to be temporary and one-off only. The firm is sunk with excess capacity and debt. Some period for reflection and analysis (time delay) may be necessary before an investment is undertaken. However, if this period for reflection is excessive, other fastmoving firms may enter this business and skim the market milk. A forwardlooking firm may wish to have a rather flexible production capacity (not too small and not excessively big), to try to link clients with longer-term contracts, to keep an eye on the current production capacity in its industry and how much is under construction, before deciding about new investment. History and geography matter in this evolutionary model. What changes is only the nature of the impact. Where we are today is the result of were we were yesterday and what has happened (how we behaved) in the past. The system’s dynamics, increasing returns to scale, multiple equilibria, lock-in effect and path dependence generate theoretical and practical limits to predicting the future spatial location of an industry with a high degree of certitude. Certain opinions about the future developments are possible, but only within the context of some hypothesis. Therefore, there are troubles with proofs in economic models. They are proofs about the economic model, not about the real world. This world needs to be observed by economists. At least now and then. We shall continue to do that in this book by the means of examples. 7 Locational possibilities The probability of the location of an industry resulting from a historical accident is shown in highly stylised, graphical form in Figure 2.9. If the distribution of potential locations of an industry in an imaginary space is concave (Panel a), with a single minimum and a corresponding single outcome, the location is not influenced by a historical chance. This type of distribution is exemplified by the mining and sometimes steel-making industries, which are usually located close to their source of raw materials. If, however, the distribution is convex (Panel b), with two minima, then there are two potential outcomes, each resulting in a different location which may depend on the historical chance.79 The

Evolutionary Economic Geography

70

Figure 2.9 Concave (a), convex (b) and spherical (c) distributions of potential industry locations third case represents a sphere in which there are n solutions for the location of footloose and mobile industries (Panel c). Hence, multiple equilibria make policy analysis conceptually difficult and uncertain. This type of distribution is exemplified by corner shops, bakeries and petrol stations. However, firms in many industries need to be close to one another (that is, they tend to agglomerate and create clusters, towns and cities), not only to be close to common suppliers of inputs, but also to foster competition and to facilitate exchange of information, experience and knowledge, which can be hampered if firms are spatially dispersed. If agglomeration forces based on increasing returns are unbounded, then a single geographical location monopolises the industry. Which region is selected depends on its spatial attractiveness and the historical accident of firm entry, in particular early preferences of first entrants. If, however, agglomeration forces that come from increasing returns are bounded, then various regions may share the industry as if agglomeration economies were absent (Arthur, 1990b, p. 249). Locations with a large number of firms cast an ‘agglomeration shadow’ in which little or no settlement takes place. This causes separation of an industry. Agglomeration occurs at a certain level of trade costs, while at another level of these costs the spread of activities takes place. With bounded agglomeration economies, neighbouring locations cannot share the industry, but sufficiently separated regions can. In France, for example, Lyons lies between Paris and Marseille. Bounded agglomeration economies caused separation and dispersion. Again, ‘which locations gain the industry and which become orphaned is a matter of historical accident’ (ibid., p. 247). One of the most convenient, but one of the most unrealistic, theoretical assumptions is that firms are identical. In reality, firms are heterogeneous, not only regarding their size and marginal production costs, but also regarding their trade, investment and R&D behaviour. Firms move to be close to relatively large markets, not only because of clients and demand, but also because of supply of inputs and marketing of output. Firms are more productive in the economic core than in the periphery, because they have lower marginal costs and they tend to sell more. So, if a national regional policy subsidises reallocation of firms towards regions with a certain disadvantage (periphery), the implication is that the least productive firms from core regions (business agglomerations) may be attracted first by these (modest) subsidies. The opportunity costs of these lowproductivity firms leaving the agglomerated region are smallest. The selection effect of such a policy is that the most productive firms will move to (or stay in) the core regions,

Theory

71

while the least productive firms will go to the periphery as they have least to lose from the move from the core region. Modest subsidies have very little effect on regional welfare. They attract few firms and all of them are inefficient. Hence such a regional policy is ineffective in improving the competitiveness of remote regions (Baldwin and Okubo, 2005). Entrepreneurs (sometimes seen as ‘maniacs with a vision’) often have genuine ideas, but many of them do not have the necessary knowledge about how to run a business. Some of them may know, but cannot persuade those with funds to support their projects. Why is this so? Reasons include a number of cultural factors (risk-lovers or risk-averse investors) and the lack of knowledge and understanding to recognise and seize the opportunity. Certain minimal material wealth is always necessary for sheer survival, but wealth exists also in the form of knowledge and understanding. One of the principal defining features of humanity is the search for understanding, as much as the pursuit of happiness80 and the quest for beauty. An example by Arthur (2002, p. 14) is quite instructive: a century ago a man went to a Belfast bank and asked for a loan to start a company there. He explained to the bank managers that his invention was a highly productive mechanical horseless device that would replace horses in ploughing. This was too risky for the bank, as it could not understand the technicalities and it could not foresee that the enterprise would be commercially viable. So this man, whose name was Ferguson, went to Canada. He met another man by the name of Massey who understood Ferguson’s idea. They teamed up, and the rest of the story is generally known. The potential location of this business went from Ireland to Canada. Massey Ferguson became one of the major world producers of tractors and agriculture-related machinery. The moral of this story is that one needs to recognise, understand and have guts to seize the opportunity. Carpe diem (seize the opportunity now). Let us consider a few other examples. Soichiro Honda (1906–91) grew up in Hamamatsu (Japan) in a modest family; his father was a blacksmith and also in the bicycle-repair business, while his mother was a weaver. At the age of fifteen and with little formal education, Honda went to Tokyo to look for work. He became an apprentice in a garage. In 1928, at the age of twentytwo, he returned home and opened a garage of his own. Following the Second World War and disasters in the Japanese economy and society, Honda recognised and seized business opportunities. Japan was starved of money and fuel, but not of business opportunities. The Japanese needed basic and cheap means of transport. Industrious Honda created a firm in 1948. Even though it was housed in a simple wooden shack, its name was pretentious: the Honda Research Institute Co. Ltd. Honda simply attached a small engine to a bicycle, with great commercial success. This created a cheap and efficient means of transport that did not create an enormous import demand for fuel. The rest of the story is generally known. Towards the end of his career, when asked about his achievements in business Honda used to say: ‘My success is based on many failures.’ This is an indicator that Honda was ready and willing to make many experiments. (Recall here the adaptive walk and random jumps story.) Michael Dell dropped out of college at the age of nineteen to start a firm named PC’s Limited with just $1,000. This start-up took place at his dormitory room at the University of Texas at Austin in 1984. Dell thought that consumers’ needs ought to be better understood than they were by firms, such as IBM, making and marketing standardised

Evolutionary Economic Geography

72

PCs. Specific consumer needs ought to be offered specific and effective solutions. He advertised his services in national computer magazines. His idea (out of thin air) after he recognised and seized a market opportunity was rewarded with a great commercial success. The company was subsequently renamed Dell Computer Corporation in 1988 and became one of the leading world producers of computers. Big IBM with enormous research, production and marketing potential was, perhaps, too big to seize the new marketing idea. Huge IBM had an enormous scale of possibilities because of its network that reached far and wide. But this was also IBM’s weakness. Entrenched interests within company’s fiefdom lowered the degree of freedom for action. IBM could have easily started mail order sales. However, a network of IBM’s retailers would have rebelled, and sales would probably have fallen. Hence, IBM fell victim to a ‘complexity catastrophe’. Huge opportunities within an established network (benefit side) reduce the degree of freedom in action (cost side of the story). The ‘carpe diem’ stories are magnificently presented in Painting 1 from the school of Andrea Mantegna: Opportunity and Penitence. It is most instructive for economists and people in business. Opportunity is represented here as a volatile woman. She is very light, almost weightless, and she is flying all around (note a pair of wings on her left foot and her dress). She is also unstable, as symbolised by her position on the ‘globe’. She ‘stands’ on only one foot and outside her gravity centre. She has hair only on her forehead. The back of her head is clean-shaven. When she, the opportunity, approaches you, you have (1) to recognise her and (2) to grab her hair. At that moment, and from the front! When she is gone, as is the case represented in this painting, she may not be snatched any more. One may not catch her hair from her back. What remains is penitence, represented in this painting by a man. He (penitence) is stable and stands firmly with both feet on the ground. He recognised the opportunity from her back. Too late. She’s gone! Probably for ever, to the arms of someone who can recognise her and who can ‘snatch’ her. To someone who may spot and appreciate the chance provided when she is approaching. Another woman in this painting (who stands firmly above the disappointed man) tries to calm him down. What remains after this episode is wisdom (if we are lucky and smart). Applying this, for instance, to our example of Ferguson tractors: the Irish did not recognise and appreciate Ferguson’soffer of a project. Hence, the location of this business went to Canada. In order to profit from rare and exceptional opportunities, one needs to be endowed with both good knowledge and foresight and, yes, luck. The Italian government decided in the late 1950s and early 1960s to postpone the introduction of colour-TV broadcasting. This political decision created ‘a competitive disadvantage on the domestic electronic industry that was never later made up, with wide negative external effects on the entire technological trajectory’ (Camagni, 2002, p. 2400). Conversely, an early introduction of environmentally friendly legislation in countries such as Germany (concerning gas emissions) provoked the creation and development of

Theory

73

Painting 1 School of Andrea Mantegna, Opportunity and Penitence (cca. 1503-06) (Occasio et Poenitentia, Museo della Città Palazzo San Sebastiano, Mantova). environmentally friendly technology. Spatial positioning technology may be in the same league as earlier essential electronic technologies such as transistors, computers or mobile phones. Would the continuation of the beleaguered Galileo project in Europe keep this region in the select league for this technology or not? Would the potential chance in this industry be seized by Europe? On the other side, there are firms that are known to be good at spotting new trends. Toyota is an example. This company noticed and recognised the emerging market for hybrid cars. It invested in the development and production of this newly demanded good. Let us round off this topic with a quotation from William Shakespeare’s Julius Caesar:81 There is a tide in the affairs of men, Which, taken at the flood, leads on to fortune; Omitted, all the voyage of their life Is bound in shallows and in miseries. On such a full sea are we now afloat, And we must take the current when it serves, Or lose our ventures.

Evolutionary Economic Geography

74

8 Spatial units Economic and geographical space may be classified in several ways. These spatial units may include households, firms (plants), clusters, industrial districts, networks, cities, regions, states, regional integration arrangements among countries, continents and the world. Our considerations will be restricted here to attempts to unpack, at least partly, only clusters, cities and regions, even though these units do not have a coherent internal structure themselves. A system with a high degree of order and stability has a low level of internal connections. If one if its component fails, the system continues to operate without any major disruption and damage. If a system has a high degree of connections among its internal parts, then a change in one of its components may have a strong impact on the whole. If everything is (strongly) connected to everything else, then this unstable system is volatile and may behave in a chaotic way. Complex systems are somewhere in between these two systems, between order and chaos, and comprise elements of both. There is a certain longer-term stability in the location and existence of spatial units. Clusters or cities do not appear or disappear overnight. Regional specialisation has a certain longerterm stability. Yet this is not the outcome of a low level of connections among components. Hence, internal networks, weak and strong ties, order, stability, selforganisation and chaos ask for a deep and multilayer economic and geographical analysis (Martin and Sunley, 2007, p. 588–9). Clusters Clusters are a form of economic organisation of business that survives. Hence, clusters are an efficient or more efficient form of organisation in certain lines of business than are other forms. Industrial clusters often determine what a country exports. Their success often establishes the competitive position of a country’s goods and services on the international market. Prior to the Industrial Revolution, high transport costs splintered production into many units throughout geographical space. The effects of the Industrial Revolution reduced the cost of transport, but brought high fixed costs of production (and economies of scale) that contributed to the development of new and the expansion of existing agglomerations. In any case, clusters are often innovation incubators and depend and thrive on demand for their output when trade is (in general) liberal. Underlying forces A promising point of departure for the analysis of clusters is to consider two basic forces which, through their tension and interplay, influence the spatial distribution (concentration or spread) of firms and industries: the centripetal and centrifugal forces. These forces also help to explain the agglomeration and dispersion forces of the ‘core–periphery’ development model. A core region is one that has the capability and potential to create, attract and employ resources and ideas. A periphery region is one that has its economic path determined chiefly by the developments in the core region(s). Let us consider the two background forces:

Theory

75

• Centripetal forces promote agglomeration (distance matters). They concentrate production and employment in specific geographical areas. These forces tend to be stronger during the early phases of an industry’s life cycle because of tacit knowledge and nonexistent standards (among other elements). Knowledge and experience is highly contextual and hard to codify. Therefore, spatial proximity with face-to-face contacts and strong personal relations are necessary for staying in the business game. These ‘snowball’, ‘me-too’ or herding forces are based on certain complementarities among diverse players and they include: market size (concentration of firms and consumers), economies of scale, forward and backward linkages in production and distribution, trade costs, increasing returns in transport, existence of suppliers, limited spread of information and embodied knowledge, as well as a ‘thick’ labour market (especially for certain skills so that employers can find workers and workers can find jobs). Workers can and do invest in industry- and firm-specific productivityincreasing skills, rather than in general ones that they may not need. Centripetal forces take the production system towards spatial equilibrium. • Centrifugal forces (distance is dead) push the other way and test whether the equilibrium is stable. These dispersion (comparative advantage) forces discourage further spatial concentration of business. They favour a geographical spread of firms and include: spatial factor immobility, regional or international wage differentials, relative amount of land rents, competition for factors and consumers, commuting costs, pollution, congestion, traffic accidents, crime, allergies, infectious and other diseases, lack of drinking water, sewage system and waste disposal.82 The final result on the location of business in space depends on the balance between these two forces; on what factors are mobile and what immobile; on barriers to the reallocation of resources; on demand and its change; and on public policy. In essence, three outcomes are possible: • Economic activity may spread so that each region, industry and firm becomes more specialised in a certain activity. This type of clustering does not mean polarisation, as is exemplified in North America. • Activity may agglomerate in core regions, leaving others without production potentials and without people. • A long-run polarisation may split the country or an economic grouping into advanced regions with high incomes and low unemployment and depressed regions with low income and high unemployment (Braunerhjelm et al., 2000, pp. 29–30). This final result may turn out to be the creation of a set of growth poles and a set of growth sinks. One needs to distinguish, therefore, between ‘good’ and ‘bad’ agglomerations. In theory, an addition of a firm to a location reduces the profitability of the location through competition in goods and factor markets. That would provide incentives for a geographical spread of firms. Conversely, if an increase in the number of firms at a location raises the return to other firms in the region, then agglomeration effects will operate. This depends on linkages in the operation of firms and imperfect competition. In practice, backward and forward linkages in production are strong in financial services and in chemicals. Hence, economic integration would provide further incentives for the agglomeration of businesses that have strong input–output links and create an economic geography of the EU similar to the one in the US. There is also a possibility that linkages

Evolutionary Economic Geography

76

are strong not only within narrow industrial groups, but also across industries. If that is the case, agglomeration effects in the core area(s) would be further strengthened. There would be implications for the demand of labour. If that factor is not mobile, the process would increase wage differentials and create regional disparity in wages. Instead of producing factor price equalisation, economic integration and agglomeration, forces may cause widening of differences in factor prices. Fujita and Thisse (2003) stated that ‘the main effect of interregional and international integration is likely to increase economic efficiency within the space-economy’ (p. 121). However, globalisation defined as opening and deepening of world markets may be accompanied by an increase in the severity of the regional problem (distribution of the economic activity) between the dynamic and prosperous core and the lagging periphery. However, everybody can be better off in the Pareto sense. Agglomeration supports and generates additional growth that ‘spills over’ the integrated area in which the rich can get richer, and so may the poor but without ever catching up. Absolute discrepancies may easily widen across individuals (all are gainers compared to the starting position, but in unequal proportions). These widening welfare gaps call for certain remedial policies (as is the case in the EU), even though that may damage growth and overall welfare. Fujita and Thisse conclude that it is not so clear that agglomeration, growth and equity do conflict: even the people residing in the periphery are better off in the core–periphery structure than under dispersion. There is a conflict only when a fairly narrow interpretation of justice, i.e. egalitarianism, is considered, since the unskilled workers living in the core region are better off than those in the periphery. At this stage of the debate we do not have much to say: the answer depends on societal values. (p. 144) The rural interior of China is home to three-quarters of the country’s population. In many cases this population lives in pre-industrial conditions. A relatively balanced development of China is one of the principal public and social challenges in the country. Meaning and motives A cluster is a relatively large group (a dense critical mass) of functionally related specialised firms in a particular relatively small geographical area. Inseparably linked with this is a system of communication (formal and informal), collective learning, knowledge, coordination capacities, cost and risk sharing, skills and competences, as well as specialised institutions (standards-setting bodies, schools, universities, research institutes, trade associations, networks, rules and practices). This functional relation among firms may be downstream (suppliers), horizontal (competitors and collaborators), upstream (clients) and/or through the circulation of accumulated knowledge, skills and competences by means of personnel turnover. Localised spillovers among the participating actors are the chief glue that keeps a cluster together.83 Clusters, as spatially limited business incubators, are a part of the new industrial order called ‘alliance capitalism’, in which continuous collaboration and networking with

Theory

77

related knowledge and experience-rich firms are sources of competitiveness. It is also associated with the complex and dynamic ‘new economy’ based on knowledge; information, organisation and control technologies and competences, computers and continuous innovation, as firms that operate in clusters are usually quite innovative. In addition, a vibrant dialogue is constantly taking place among all players in a cluster. The reason why this view fails to define a clear geographical boundary for a cluster is because the structure and assets of a cluster may change over time. At the same time it is hard to define the specific industrial boundary of a cluster, as this may overlook the complex and changing relationship among different industries that support one another’s existence and competitiveness. A concentration of functionally related business activities within a relatively small area (agglomeration; thick market effects; benefits of co-location; and non-ergodicity) provides firms with collective gains that would not be available if the firms operated in remote places or in isolation. These shared benefits or externalities are different from those that are created within and available to a single firm. In essence, clusters create economies that are external to individual firms, but internal to a network of firms in a cluster. These economic gains are exchanged and enriched in a non-market way. Hence, a great deal of an individual firm’s competitive advantage is outside it, but inside the location where it operates (e.g. relations among suppliers or gains from various types of partnership). A firm locates in an area where there are firms from the same or related industry (in a cluster) because: • it has production links with other firms; • it may benefit from the already existing pool of suppliers; • there are services such as finance, information, consulting and maintenance; • there may be a pool of trained and experienced labour; • firms may reduce the cost of transport; • there may be a concentration of consumers (proximity to the major growing markets is often the most important reason for the selection of a particular location); • there are competitors that may stimulate innovation in technology; and • firms may jointly negotiate contracts with transporters and organise export promotion boards and the like. Manchester (England) in the nineteenth century provides an example of such a cluster specialising in textile manufacturing. Subsequently, the manufacture and repair of looms, and bleaching, dyeing and finishing facilities also located there. In the course of time, the town set up a technical college in order to train people for the manufacture of machinery, textile design and other skills related to the local industry. All this was supported by marketing organisations such as the Cotton Exchange (Smith, 1981, pp. 60–1). It was a similar story with Reutlingen in Germany. However, even the most recent industries rely on the ‘old concentration rule’ of geographical clustering. This is why there is a Silicon Valley in California or a Route 128 in Boston. The ingredients of success in these two clusters include a close proximity and interaction among entrepreneurs, some of the best universities, venture capitalists, as well as high labour mobility and immigration of highly qualified personnel. To emulate that elsewhere in the short term would be quite tough as these ingredients and their combination may not be replicated easily.

Evolutionary Economic Geography

78

Another example of similar developments in the manufacturing of basic surgical instruments and certain sports goods can be found in the town of Sialkot (Pakistan). This cluster is the world leader in its export of surgical instruments and hand-sewn footballs. This success, as well as that of all other clusters, is based on two factors: first, a demanddriven approach and, second, competition based on collective efficiency (Schmitz, 1998, p. 6). Yet another illustration of a joint action can be found in the surgical instruments manufacturing cluster in Tuttlingen (Germany). This cluster produced joint publications on various topics of common interest, including booklets on: quality requirements in the surgical instrument-making industry; a guide on the correct use of the instruments; and a guide providing customers arriving in Tuttlingen with a list and profile of firms that produce such instruments. Krugman (1996b) drew an interesting analogy between clusters and the natural evolution process: The general attitude of evolutionary theorists seems to be that Nature can often find surprising pathways to places you would have thought unreachable by small steps; that over a few hundred thousand generations a slightly light-sensitive patch of skin can become an eye that appears to be perfectly designed … [Imagine] a group of frogs sitting at the edge of a circular pond, from which a snake may emerge … and that the snake will grab and eat the nearest frog. Where will the frog sit? … if there are two groups of frogs around the pool, each group has an equal chance of being targeted, and so does each frog within each group – which means that the chance of being eaten is less if you are a frog in the larger group. Thus if you are a frog trying to maximize your choice of survival, you will want to be a part of the larger group; and the equilibrium must involve clumping of all the frogs as close together as possible. Does this remind you of the principle of agglomeration? Unlike changes in biology, realworld successful entrepreneurs are smart (the importance of knowledge and education) and often radically change their behaviour within a short period of time in response to challenges, risks and opportunities for business and profit. Technical advances in transportation, information technology, organisation and control of production and distribution reduce production time and costs. They ‘do away’ with distance. ‘Globalisation’ (integrated international production) creates possibilities for the spatial fragmentation and dispersion of production and ownership. In such circumstances, already established locations face tougher times anchoring incomegenerating activities. At the same time, strong centripetal forces ensure that certain spots remain ‘sticky places in slippery space’ (Markusen, 1996, p. 293). Spatial barriers for the location of businesses are becoming less important, but capital is more sensitive to the choice of location. Until the 1960s, it seemed to some that it might be more efficient to move people to jobs by migration than otherwise. The footloose element of many modern industries supports production in relatively small and flexible units. Hence, there is an arbitrary and uncertain element (multiple equilibria) in the location of firms that have a footloose industrial character. Locations that are close to consumers save in transport costs of final

Theory

79

output, while other locations may save in the cost of production. Because of costs of inputs, economies of scale and forwarding outlays, one business may favour peripheral locations, while another’s preference may be for locations nearest to the consumers. Agglomeration and the spatial clustering of firms and industries are motivated by efficiency considerations within a network. They are based on specialisation and economies of scale in production and transportation, as well as on transaction costs. For example, the Massachusetts General Hospital in Boston is among the largest in the world in terms of R&D funds. Industrial managers are attracted to the R&D department of the hospital because all the specialists and all the needed knowledge are ‘within 20 minutes’ walking’ (Lambooy, 1997, p. 298). This cluster of more than 400 firms linked to medical devices is invisible to an outsider as it is concealed in business categories such as electronic equipment and plastics products.84 Agglomeration applies not only to private businesses, but also to certain public institutions. Bearing in mind the complexity of action and need for a fast response to changing and uncertain events, Simon Jenkins has noted that ‘Britain is a centralised state. Two-thirds of the top 100 powerful people in Britain work within 300ft of the Prime Minister and Chancellor of the Exchequer’.85 Marshall (1890) considered the issue of why it is beneficial for producers from the same industry to locate together. He offered three basic reasons why producers concentrate and form a cluster: • Knowledge spillovers: proximity eases exchange and spread of information. This increases efficiency in production through technical, organisational and control improvements. • Advantages of thick markets for specialised skills: firms may easily find the necessary labour and workers may get promotion or work if the current employer does poorly. Risk for both of them is reduced. There is no need to create the human capital from scratch. The ‘mysteries of trade’ (craft secrets) are no mysteries at all as they are ‘in the air’ and children learn many of them unconsciously from their parents (vertical cultural transmission).86 Knowledge, experience and values are stored.87 In addition, tacit and complex knowledge and shared culture of beliefs and practices are difficult to transfer without labour mobility. • Backward and forward linkages associated with large markets: links may clarify and explain a part of the concentration story only if there are economies of scale. Otherwise, a firm would set up a separate production unit to serve each distinct market. A concentrated industry provides a market for specialised local suppliers of components, as well as private and public suppliers of various services (institutional thickness). These economies, external to individual firms in a cluster, but internal to a network of firms in a cluster, are essential, but not sufficient to explain the basic reason for the strength of firms in a cluster. Conscientiously pursued joint action and horizontal and vertical cooperation by firms in a cluster (sharing equipment, developing a new product, various consortia) enhance collective efficiency and improve the competitive advantage of the participating firms. When there is a problem (market failure), it can be resolved by government intervention and/or by private self-help, i.e. by collective action or institutions (as was the case in Sialkot, Pakistan) (Schmitz, 1999, pp. 468–70). However,

Evolutionary Economic Geography

80

relations between firms in a cluster may take different forms. Take a look, for example, at shoe-producing clusters in Italy and Mexico. While cooperation is more common in Italy, market rules (competition) prevail in Mexico. Upgrading of skills of the labour force is an important consideration for firms in a cluster. They may wonder why to invest in this upgrading when labour, if it moves, may transfer these skills to other firms in a cluster that did not pay for this upgrading. Therefore, costs of training of labour in a cluster may be the task of cluster authorities and/or labour itself. If a location for a certain type of business is isolated (a single textile firm in an iron-producing cluster), then such a (textile) firm may be generous regarding training costs of its staff. There are, however, other reasons for the spatial clustering of firms and industries, additional to the ones advanced by Marshall. They include the following: • The past arm’s-length and hierarchical organisation of a firm can be replaced by a flexible network of business organisation in a cluster, because there is a changed situation in competition and technology. Interfirm competition is based on innovation. • Firms may reduce transaction costs. • The presence of one firm/industry creates a direct or indirect market for another firm/industry. One industry’s output is used exclusively as another industry’s input (tyres and cars). • Firms use common services (marketing, storage, accounting, repair, transport). • Two industries use a common resource. • Firms use common social infrastructure (schools, health, roads). • Firms may create entry barriers (such as the very hard entry into the Geneva watchmakers’ guild). • Firms may intend to build a shared brand (for example, Solingen knives, Parma cheese and ham; Californian wine). • Labour markets of two unrelated industries may be complementary: one industry (metal or car assembly) uses male labour, while another (textiles or food processing) employs female labour. • The more firms there are that can use certain new infrastructure, the more likely it is that that infrastructure will be built. Each user in a cluster assists other users in this situation, hence the fixed cost of an infrastructure project (such as educational and training centres, power plant, railway or airport) can be recovered and the project can come close to profitability. These are relatively standard projects, so the government can readily be involved as specific local knowledge is not always essential for the commercial success. The presence of a relatively large concentration of demanding consumers in an agglomeration, coupled with technologies with economies of scale and agglomeration forces, can bring certain economic advantages through the interrelationship of functionally connected as well as competing firms. Competition forces firms to employ state-of-the-art technology, hence there are no low-technology industries but only lowtechnology firms within various industries. The clustering of firms selling identical goods (competitors), such as ice cream on the beach, was thought to be welfare-inefficient from a social standpoint (Hotelling, 1929, p. 53). However, the clustering of firms often occurs in response to consumers’ desires to

Theory

81

make comparisons between goods (for example, in shopping malls). Retailing exists because consumers are small and sometimes both ignorant and immobile. Consumers usually visit more than one store prior to making a purchase. Department stores often demand that certain space is made available in shopping malls to small shoe and clothing retailers, to assure clients that the offered goods are balanced by other competitive offers (that is, that a minimum differentiation exists) for comparison. Clustering of identical firms can serve a socially useful purpose because they lower transportation costs (Eaton and Lipsey, 1979a, pp. 422–3). The same holds for the clustering of heterogeneous firms. A rational consumer would connect purchasing activities such as search, purchase and transport. Multipurpose shopping assists in the lowering of shopping costs (Eaton and Lipsey, 1982, p. 58). If the firms are similar, they seek comparable if not identical features for the spatial location of business. If this is true, then the outcome may be a cluster in such an industry. For example, the US film industry started more than a century ago in California, around Los Angeles. The entertainment companies were all looking for comparable locational features such as dry weather and excellent daylight conditions. Although modern filming technology does not depend on natural light or weather conditions at all, the clustering of the film and entertainment industry in California continued its existence in the same location. Homogeneous initial needs of firms created the film cluster rather than the input–output production relations. Hence, functional production links and agglomeration economies are not on their own a necessary condition to create a cluster. Spread of information and knowledge In spite of a few global TV stations such as the BBC, CNN or Al Jazeera and newspapers such as the Financial Times, the International Herald Tribune or certain electronic media, most of the stories that editors publish or broadcast are still local. The local proximity of firms in the same industry increases both the visibility of the course of action of competitors and the speed of the spread of information. ‘Popular luncheon spots are patronized by executives from several companies, who eye each other and trade the latest gossip. Information flows with enormous speed’ (Porter, 1990a, p. 120). This ‘lunch club’ or ‘cafeteria effect’ breaks the spatial boundaries to information flow and gives incentives for the creation of matching improvements by other firms in the cluster. It also offers a partial confirmation that there is a spatial limit to knowledge spillovers.88 Information is simple structured and unstructured data (facts). It may be freely available and transmittable at constant cost, notwithstanding distance. Information could spread in the eighteenth century by the speed of horses and the work of the ‘Friday market’ or street-corner drummers (the Internet of the time). Modern telecommunications transmit information by the speed of light. Because of the current physical laws, one may not expect an increase in this speed in the future until these laws are replaced by some new ones. Timely, correct and cheap information is becoming a crucial input in the decisionmaking process. A banker who is handling large funds over his PC terminal is a long way from the British general, Sir Edward Pakenham, who lost the Battle of New Orleans and his life on 8 January 1815, fifteen days after the Treaty of Ghent ended the war, but

Evolutionary Economic Geography

82

several days before the frigate arrived at his headquarters bringing the news about the end of the war (Lipsey, 1992b, p. 288). Knowledge is, however, the structured capability to link and process data (information) and ideas, to analyse, evaluate and understand them and, possibly, to deepen and extend them in order to do something useful with them. Evolution and learning create knowledge. ‘Wealth is knowledge and its origin is evolution’ (Beinhocker, 2006, p. 318). Tacit knowledge can be acquired by experience. It is something that is within a knowledgeable person (stockpiled in human capital), but something that one cannot put in a manual. The charge and trouble for the spread of embodied knowledge,89 in particular tacit, complex, changeable and ‘sticky’, increases with the geographical distance.90 This is what makes an economy a complex system. Applied to markets and business, knowledge helps us to solve problems. Potts (2001, p. 422) argued: the market process is an experiment in knowledge: the creation of unforeseen compounds out of ephemeral elements that become perfectly obvious only after the event. This is what innovation is and does and is why it is not and cannot be a closed mechanism. A localised knowledge spillover is an issue that still lacks a solid theoretical background and empirical foundation. Breschi and Lissoni (2001) argue that the content of this ‘black box’ may be rather ambiguous as there is a blend between spread and tacitness. The concept may be too narrow to include a wide variety and mix of knowledge creation and transmission mechanisms that are involved in spreading ideas and experience in a relatively limited space. These spatially limited knowledge externalities enable firms from that area to introduce innovations earlier and faster than competitors that are located outside such a cluster. One analytical problem is the definition of area under enquiry. Administrative boundaries are often different from statistical, business, technological and knowledge boundaries. The evidence on the existence and impact of externalities may often be indirect. As such, its interpretation may raise certain suspicions. Workers can move locally from one firm to another for reasons that have nothing to do with technological spillovers. Then, certain firms may prevent labour and knowledge mobility and contract key technical superstars not to engage in any (similar) work anywhere for at least two years following an innovation or departure from the company. Firms want to appropriate the benefits of commercial achievements for themselves. They want to prevent imitation (spillovers) of the profitable fruits of their uncertain and costly investments, for as long as is legally and technically possible. The high speed of information diffusion is one of the major strengths of clusters. Entrepreneurs often prefer to enter or stay in a cluster even though they may be able to appropriate a higher return on their current innovation elsewhere. The reason for this preference is that the firms in the cluster are not only providers of information but also recipients (Schmitz, 1999, p. 475). In fields where technology changes often and fast, personal contact may be the preferred way of communication rather than the less immediate sources such as professional journals, fairs and conferences. ‘Human capital accumulation is a social activity, involving groups of people in a way that has no

Theory

83

counterpart in the accumulation of physical capital’ (Lucas, 1988, p. 19; original emphasis). This all supports progress in technology and creation of knowledge. Personal contacts and informality present a powerful and efficient social communication technology. They are essential for the exchange of complex information and tacit knowledge as they facilitate interpersonal transmission of imperfect and uncodifiable data (something that cannot be put into symbols: linguistic, mathematical, visual or vocal) and experience. Has anybody learned to cook a good French or Chinese dish merely by reading a cookbook? Or has anyone initiated and maintained a PC simply by reading a 800-page manual? One may learn Greek grammar and words, but it is not possible to make reasonable use of metaphors in an appropriate context without a face-toface contact with the teacher. Personal contacts also ease learning and may create and sustain motivation. A few additional observations: Tacit knowledge, as opposed to information … can only be transmitted informally, and typically demands direct and repeated contact. The role of tacit knowledge … is presumably the greatest during the early stages of the industry life cycle, before product standards have been established and a dominant design has emerged. (Audretsch, 1998, p. 23; original emphasis) Companies that have gone furthest towards linking their global operations electronically report an increase, not a decline, in the face-to-face contact needed to keep the firms running well: with old methods of command in ruins, the social glue of personal relations matters more than ever. (‘Does it matter where you are?’ The Economist, 30 July 1994, p. 11) Ford … operated its own planes for twice-daily flights between the UK and Germany in order to keep in close touch with its two main subsidiaries in Europe. This confirms that the importance of face-to-face communication is true not only for Japanese MNEs, but also for other MNEs. (Ando, 2005, p. 85)91 In order to ease communication among staff and make the working space more agreeable, certain companies introduced ‘casual Fridays’: employees come to work dressed casually, not in suits and ties. To follow this up, Deloitte and Intel came up with a new idea in 2007: they proposed Fridays without e-mails. Employees need to come to the office and move around: they need to meet others and talk to each other face to face. Repetition of certain uncodified and complex actions or operations may translate them into routine and, later, into codified and transmissible standards. Thus, an initially localised business operation may be spread elsewhere in space where physical proximity of actors is not necessary. However, a spurt of innovations and their evolution may endanger such potentially general spread and keep the exploitation of the newly created opportunities in the places where they started.

Evolutionary Economic Geography

84

Achievements in science and other information may be freely available in various papers, journals or via the Internet, but the problem is that only a handful of professionals have the key knowledge and experience to digest and understand them. The matter may often be too nuanced and complicated for the non-specialists. A large part of science and technology is a ‘private good’ because it is tacit. Close proximity and a durable social network (behavioural cohesion, handshakes and body language)92 shorten the learning curve; informality in contacts and local embeddedness is essential to identify user requirements, proper installation, operation or service of the product.93 We know more than we are capable of writing or telling in words. The phrase ‘you know what I mean’ among members of a professional group in face-to-face communication can be very illuminating. The same holds in relations between suppliers and their clients. Here come select reasons: This is because social networks are a form of durable social capital, created (and maintained) through a combination of social history and ongoing collective action … economic networks represent a form of club which is neither completely open nor fully closed. Access to the club will depend on past experience and routine interaction as well as on investments of effort in developing personal relations and trust. (Gordon and McCann, 2000, pp. 520–1) In Britain, the mechanism of financial regulation was for decades described as ‘the raising of the eyebrows of the governor of the Bank of England’. This was a metaphor for informal but powerful expressions of regulatory disapproval. The system was made possible by the common social background of market participants. It ceased to be sustainable when the City of London became a more democratic institution, and when globalization brought into the marketplace foreigners who did not understand what these signals meant, or that failure to observe them would have adverse consequences. So Britain acquired a rule book and an analogue of the Securities and Exchange Commission. (Kay, 2004, p. 14) Two hundred representatives of creative companies were asked in 1999 at a plenary symposium: ‘If there was one single thing that Scottish Enterprise could do to promote the growth of this cluster, what should this be?’ Their answer was: ‘Keep us talking to each other’. (OECD, 2005a, p. 50; original emphasis) What matters in such situations is not only the spatial proximity between the participants, but also their relational closeness together with the sense and logic of belonging to the success of a common project. Therefore, operation of firms within clusters makes firms relatively less sensitive to local taxes. This provides the local authorities with unique opportunities to tax them more heavily than would otherwise be the case.

Theory

85

Contrary to the belief and trust in the ‘magic of the market’, clusters facilitate nonmarket forms of transmission of knowledge and experience. These interactions may not be easily, empirically and statistically distinguished from one another. Therefore, it may be too ambitious to try to fathom and measure all factors that motivate and shape agglomeration tendencies. Certain clusters do not form around common output, but rather around common production and knowledge-seeking techniques. These agglomerations ease a diversity of ‘attacks’ on common problems where each firm can learn from the successes and failures of others. Innovation and institutions Papageorgiou (1979) demonstrated that agglomerated firms can achieve higher aggregate profit and lower prices per unit of output and generate more demand than more dispersed firms. In addition, firms located in a cluster rich in knowledge tended both to introduce more innovations and to grow faster than more isolated firms (Beaudry et al., 2002, p. 191; Audretsch and Dohse, 2004, p. 23). Relative to the rest of the economy, clusters are likely to have higher productivity and higher profitability, and they may create more new jobs and new firms. On the cost side, land rents tend to increase in and around clusters and so do pollution and congestion. Labour costs can inflate, while institutions and specialisation may suffer from a path-dependence and lock-in effect which may be unconducive to adaptation to changes in technology. This adaptation may not be directly related to the specialisation in the cluster at a given moment in time but may easily become relevant because of changes in the structure of general demand. While close spatial proximity between actors may set favourable conditions for interactive learning, imitation and innovation, this is not without perils. Boschma (2005) warns that the negative impact of proximity on innovation may be found in the lock-in effect consisting of routine, the complacency trap and a lack of openness and flexibility. This is related to the following issues: • Comprehension Knowledge building and its extension often ask for different and complementary sources that may trigger new ideas and approaches towards a problem. It also asks for an exchange of information and ideas, as well as for feedback. • Organisation Too much hard hierarchy and organisational closeness may not always recognise and reward (by a bureaucracy) useful ideas. • Social embeddedness Social proximity is necessary and friendly with interactive learning and innovation. The actors share trust based on friendship, common education, experience and/or kinship. This assists in learning and in the exchange of tacit knowledge. But too much closeness creates emotional bonds of loyalty such as friendship and empathy that may introduce negative effects of opportunism. • Geography There is certain agreement in the literature that spatial proximity is relevant for innovation. Knowledge externalities are spatially bounded. Short distances bring people together. This is important for the exchange of tacit knowledge. But too much proximity may not sit comfortably with openness for new and fresh ideas. • Institutions (factors of collective behavioural order) The availability, quality and enforcement of general-purpose public goods such as institutions (rules of the game; organised interests) are very important elements that streamline coordination and consistency of actions in an evolutionary environment.

Evolutionary Economic Geography

86

Institutions support the creation and ‘proper’ operation of market forces, economic growth and development. Institutions and their enforcement can be of two types: • Formal or hard (laws and taxes): these are arrived at after long consideration. • Informal or soft (trust, confidence, habits, customs, routines, culture and group norms): these arrive as a spontaneous product of continuous and repetitive interactions among individuals over a long period of time. Institutions provide ‘stability glue’ for collective action as they reduce uncertainty and lower transactions costs. In fact, strong institutions (conventions) as a form of untraded interdependencies offer perhaps the principal vehicle to mitigate uncertainty at a local level: • Too little institutional proximity is harmful because of the lack of social cohesion and common values. • Too much may introduce inertia, institutional sclerosis and lock-in effects. • Institutional thickness may ease and enhance operations of local firms. It provides stability, reliability and predictability. However, if this thickness is ‘too thick’ and rigid, it may act as an obstacle to interactions, learning and growth. Existing institutions have strong links with the ones prevailing ‘yesterday’ and, often, with the ones from the previous century. One of the reasons why the western world ‘grew rich’ relatively fast, in particular from the eighteenth century, was that it broke ties from its old institutions. When there are radical technological changes, the pressure mounts to alter old institutions or even to break away from them. Creative destruction ought to take place. Hence, forgetting the old learned things may be as important for evolutionary development as learning new ones. It is obvious that both too much and too little proximity are detrimental to learning and innovation. Hence, a dynamic context demands a fine balance between the two aspects of spatial proximity. ‘To function properly, proximity requires some distance, but not too great, between actors and organizations’ (Boschma, 2005, p. 71). Organisation and externalities The principal key to economic success in clusters of related firms in northern Italy (Padania) is their ultra-specialisation and family-run type of business. Relatives support one another in business so there is no need for supervision; they constantly upgrade skills and they have opportunities for invention. Common values and objectives, as well as local consensus, are preserved. This reduces the possibility for business sclerosis, so common in large, vertically organised firms. For example, in Lumezzane (near Brescia) which produces two-thirds of the national output of bottle-openers, one family specialises strictly in the production of the corkscrews, another has expertise in the covers of the bottle-openers, and so on. A similar situation can be found in Cadore (near the Austrian border) regarding the production of spectacles. In general, company strategy is conducive to rapid changes, customised products and niche marketing. A summary of the properties of such industrial districts would include: • a high level of division of labour between the firms and closer inputoutput relations;

Theory

87

• a high level of specialisation which stimulates accumulation of knowledge and introduction of new technologies; • a high level of skills among workers as the result of a very long-term accumulation of knowledge at a local level; • a large number of local competitors, leading to the adoption of ‘trial and error behaviour’ and a fast imitation process; • efficient local informal (and formal) systems for the exchange of information; and • an increased emphasis on face-to-face relations. The existence of the following elements could also be added:

(Garofoli, 1991a, p. 52)

• entrepreneurs; • innovators; • financial institutions with venture capital; and • demanding clients. The growing importance of intangible assets, particularly intellectual capital together with complexities and changeability of information and uncodified knowledge, increases the importance of knowledge externalities (spillovers) and frequent face-to-face contacts among the relevant players. They need to exchange uncodifiable knowledge, to be involved in interactive learning, to receive feedback and suggestions, as well as having a dialogue about risk and the changing situation in the market and technology. Clear-cut, routine, simple activities based on codified, standardised and teachable knowledge can be managed and controlled from a distance. They can easily be located in regions with lower wages. However, successful handling of more elaborate activities still requires a high degree of exchange of information and face-to-face contacts. This is highly important in complex activities with little or no codified knowledge and where market changes and uncertainty require a fast response (contracts are incomplete). ‘Knowledge is the most important resource and learning is the most important process’ (Pinch and Henry, 1999, p. 820). Knowledge externalities are among the key reasons for the existence and success of clusters. Hence, clusters may also be seen as networks for information-gathering, factprocessing and network-creation places. ‘Almost every internationally successful Italian industry has several if not hundreds of domestic competitors. Frequently, they are all located in one or two towns … Where domestic rivalry is absent Italian firms rarely succeed internationally’ (Porter, 1990a, p. 447). Is FIAT a dangerous competitor to Toyota or VW? Toyota in Japan, for example, is cultivating suppliers in its own back yard. Toyota’s independent suppliers are on average 94km away from its assembly plants. They make nine deliveries a day. Lean production and just-in-time delivery (‘soft’ organisational technology) may be as significant in raising productivity as ‘hard’ technologies in actual manufacturing. By contrast, in North America, GM’s suppliers are on average 683km away from plants they serve, to which they make fewer than two deliveries a day. One of the results is that Toyota and its suppliers keep inventories that are a quarter the size of GM’s.94 Just-in-time delivery and inventory system is based on trust. However, this increases the cost of transport in the final price of the product as it requires frequent, flexible and punctual deliveries. While a just-in-time delivery and logistics method may

Evolutionary Economic Geography

88

be technically possible, organisationally feasible and financially desirable for a firm, an area that demands further research includes the environmental and social costs of this type of logistics. The just-in-time delivery method (as opposed to just-in-case inventory system) is often linked with half-loads and more transport journeys than the more traditional delivery methods. This has an impact on the environment and congestion in traffic.95 The existence of competitors in the vicinity serves a useful business and social purpose. If one imagines a cluster in the shape of an input–output matrix for information, then the direct and indirect functional relations are diffused through rows and columns. Competition increases productivity, which is the key ingredient of prosperity. Firms do not pursue every opportunity to ruin rivals. This is contrary to the standard competition story in every introductory economics textbook. Even though Siemens is a dominant firm in the Munich high-technology cluster, it does not threaten or absorb SMEs in the cluster, but rather develops ties with SMEs that are characterised by collaboration (Sternberg and Tamásy, 1999, p. 375). Such willingness for collaboration with SMEs may also be the consequence of anti-monopoly laws. Small and medium-sized enterprises may be sources of fresh ideas and subcontractors to giants. Such symbiotic relations are also observed and known in nature. Crocodiles come up on to the shore and, when necessary, open their mouths. Small teeth-cleaning birds fly in and enter their mouths to do the cleaning. The crocodiles are happy because they get their mouths (teeth) cleaned; the birds are glad as they can feed themselves from the food remains in the crocodiles’ mouths. The crocodiles could easily close their mouths and devour these small birds, but they don’t. Both creatures profit from this natural symbiosis. Every firm in a cluster perceives its survival, growth and success in terms of collective growth and collective success. These firms learn and prosper together. The learning process in a cluster is not only interactive but also cumulative, as it persists over time (in a virtuous circle). Once something is learned it seldom ceases to exist;96 many people may have and enjoy the knowledge at the same time because it is non-rivalrous (many people may use the formula E = mc2 at the same time, for nothing and without consuming the formula); experience and discovery build on experience and discovery (everyone profits from the past improvements and changes, but not everyone benefits immediately from the current ones). Knowledge is an increasing returns asset. Turnover of labour, technical staff and management among these firms reinforces the transfer of tacit and compound knowledge, cross-fertilising research, the collective learning process and regional competitive advantage. The underlying general operations within a cluster can be seen from further examples: New York City’s garment district, financial district, diamond district, advertising district and many more are as much intellectual centers as is Columbia or New York University. The specific ideas exchanged in these centers differ, of course, from those exchanged in academic circles, but the process is much the same. To an outsider, it even looks the same: A collection of people in similar activities, each emphasizing his own originality and uniqueness. (Lucas, 1988, p. 38; original emphasis)

Theory

89

tacitness in relationship – routines and behaviour of ‘uncertain imitability’–which cannot be replicated because no-one, not even the participants themselves, fully comprehend their nature. (Kay, 2006, p. 164) Talented people tend to cluster together. They invigorate each other’s brainpower. In certain cases they search for and find partners or spouses of similar ‘intellectual calibre’ who will potentially add to their own human capital: actors marry other actors (sometimes several times over a lifetime), lawyers wed lawyers … Continuous and straight dealings among players allow tacit, complex and accumulated knowledge to be exchanged within their network. Social closeness in such cases may be equally relevant to spatial proximity, if not more so. Digital decided to close its mainframe computer factory in Galway (Ireland) in 1994. This decision was something like the ‘end of the world’ for the town at that time. Nonetheless, Galway found a way to ‘reinvent’ itself and to weather this economic storm. Today the area is thriving, with a cluster of medical devices companies. Many of the former Digital employees started firms that make these gadgets. This whole area moved steadily up the value-added chain.97 The ‘cluster spirit’ assists in bridging the gap among firms, resources, R&D and markets. It helps the group to put its assets in function and on the market. New clusters may grow spontaneously out of old clusters. At a time and age of ‘globalisation’ when transport costs are allegedly declining in the final price of traded products, when most trade barriers erode and when there is a declining consumer loyalty to domestic brands, access to a large domestic market is no longer an unquestioned benefit. If this is all compounded with international economic integration, what is the rationale for the existence and development of clusters? What is the advantage of having N clustered firms of size S involved in similar or related activities that are not transferable to a single firm of size S×N doing the same? Maskell (2001) answered that enhanced knowledge creation following from the ongoing sequence of variation, monitoring, comparison, selection and imitation of identified superior solutions is the essence why N similar firms of size S are not equal to one firm of size S×N doing the same. (p. 930) An evolutionary (or experimental) economy involves competing visions and approaches among firms. However, such competing ideas and styles within a single firm, unless delicately managed, can easily court trouble because of entrenched fiefdoms in difference departments. Firms learn from each other by watching, discussing, comparing and competing with different solutions to their common problems. Firms in a cluster set what is done and what is learned, but institutions arrange how things are done and how the learning process takes place. Different activities need different institutional set-ups and outcomes. For example, Finland is generously endowed with timber. The country has relatively highly trained labour and first-class designers. In spite of these advantages, Finland was losing out in production of wooden furniture to countries that do not have such

Evolutionary Economic Geography

90

advantages. The manufacturers of pulp and paper shaped Finnish institutions in the wood-processing industry. This industry is characterised by long production runs; high capital intensity; no design; a stable set of big clients; long-term contracts; a structured system of production and maintenance; and a large number of low and semi-skilled workers. While Finnish institutions strongly support the paper and pulp industry, they are unfriendly to the production of wooden furniture. Denmark (a country that imports timber from Finland, Russia, Sweden and Poland) developed a thriving and renowned industry that produces wooden furniture. Hence, different production activities (e.g. fashion garments or finance) need different institutions to support them, not only in the same region or country, but also in an internationally integrated area. University The existence of a good university or research institute is not enough by itself for the overall economic success of an area. Boston (Route 128) and San Francisco (Silicon Valley) have excellent universities and innovative environments, but they also have socially attractive living, working and business environments. In spite of admirable universities in upstate New York (Cornell in Ithaca and the University of Rochester) and government assistance in attracting business there into science parks, these efforts were not rewarded with noticeable economic success. Elsewhere, during the period of central planning in the former Soviet Union, Akademgorodok (near Novosibirsk) was essentially a town full of research institutes financed by the Soviet Academy of Sciences. It remained an ‘isolated island’ as it has not developed functional links with industries. During the centrally planned period, the Soviet enterprises depended primarily on the respective ministries and the central plan. However, this is changing. Russia intends to bank on past success and accumulated engineering talent. It is, after all, the country that put the first satellite and the first man in space. There is a saying now in Akademgorodok: ‘If you want simple programming work done, give it to India. If you want complicated and serious tasks done, give it to China. But if you want to solve an unsolvable task, give it to the Russians.’98 New Russian brands are oriented more towards innovative technologies than towards consumer markets. The cluster of high-technology firms around Cambridge (England) offers a different story. The common tacit code of behaviour among such firms in the region includes trust and cooperation. The cluster is based on two local ‘collective agents’. They are Cambridge University and consulting firms in R&D. In contrast to most other British universities which have formally regulated ties with the industry, Cambridge has rather liberal rules governing such links. In fact, faculty members are allowed to work part-time in the private sector. All this has had a strong and positive spillover effect on the regional cluster of high-technology firms. Sophia-Antipolis was developed in a vacant space at the end of the 1960s in the south of France. It did not develop around an established university, as was the case in Cambridge. Much time and effort, such as the establishment of the University of Nice, was necessary to create the necessary links among the high-technology firms in the Sophia-Antipolis cluster. In general, it is hard to find any plan of action or best practice to promote a cluster that could have instantly recognisable success. However, there is a rare general agreement on

Theory

91

one point: to have commercial success, one needs first to build on and improve already existing competences in the target location. This does not mean that the earlier production has to be continued. However, something (vaguely) related to it may be the starting point. Later on, these local competences may need to be altered, sometimes fundamentally. Cluster policy Some may argue that the ‘cluster policy’ may be another name for the old and often unsuccessful public policy of ‘picking winners’. Porter, however, noted that: Clusters often emerge and begin to grow naturally. Government policy had little to do with the beginning of Silicon Valley or the concentration of mechanical firms around Modena, Italy. Once a cluster begins to form, however, government at all levels can play a role in reinforcing it. Perhaps the most beneficial way is through investments to create specialized factors, such as university technical institutes, training centres, data banks, and specialized infrastructure. (Porter, 1990a, p. 655) In fact, many governments impeded the ‘natural’ development of clusters in the interests of regional policy, and subsidised firms to locate in areas with high unemployment and without supporting infrastructure (rather than to encourage migration of labour from these regions). Until the 1960s, it was thought that production costs were not sensitive to regional location. It is therefore not surprising that a number of firms located in such areas became ‘white elephants’ that were demanding and dependent on subsidies to continue operations. The importance of endogenous technological change came to be appreciated only in the following decades, so the policy of regional decentralisation became suspect (Bekar and Lipsey, 2001, p. 1). However, universities have always been established in cities, where there are many children and students. Only relatively recently have universities become established in the countryside. Governments and large TNCs often have nothing to do with the establishment of clusters such as Silicon Valley, clusters in Italy or the City of London financial district. These clusters were created out of thin air and largely proceeded on autopilot.99 In most other cases, the actual impact of the government policy was unclear. ‘New technological trajectories often tend to emerge spontaneously and unexpectedly in space due to the importance of chance events, human agency and increasing returns’ (Boschma, 2004, p. 1010). Longer-term success or failure of a cluster may rest in the factors that act beyond the influence of those regions. Hence, there is a shadow of doubt about the real influence of local agency on economic performance. However, (local) governments may well encourage and support the growth of clusters through the provision of education and training, infrastructure, information, attraction of vibrant and innovative people with positive business attitude, tax policy and public procurement, as well as through sponsoring R&D and the creation of core competences.100 Underperformance of certain new clusters or new policies in old clusters may be partially the result of unreal

Evolutionary Economic Geography

92

expectations. Managing clusters is a formidable, demanding, long-term and chancy business. It is a task that may not give positive material results in the short and even in the medium term as it relies on for building trust and collaboration among actors. The cluster-policy goal should not be and cannot be to grant a silver bullet now, but rather to assist the region in its long-term economic positioning in time and space. If one wants to intervene and to assist in the creation of a cluster, then one ought, perhaps, to start to build on some of the existing local specialities and competences, rather than creating entirely new ones as the first step. This does not mean that the ‘same old thing’ ought to be perpetuated indefinitely because of necessary adaptations to the industrial structure and adjustment to the changing situation. However, something (vaguely) related to it or derived from it may be a good start. The old local structure ought to be ‘disembedded’, but certain and useful old local knowledge, skills and experience needs to be rebottled and re-used. If the location in question has accumulated knowledge in the production of fertilisers, it would be sensible to start the production of pharmaceuticals rather than TV sets, for which there is no accumulated local knowledge and experience (although this may be created at a later stage). Lipsey et al. (2005, pp. 43–4) state an illuminating example from the US: There was a massive influx of firms to produce the radios that fuelled the expansion of stations. Over 600 radio-producing firms were established during the period 1923–26, but only 18 survived until 1934. By 1934, the evolutionary hand had done what no individual could have done in 1923 – to sort out those who had what it took to survive from those who did not. This experience also provides another illustration of path dependence since, as Chandler points out, ‘the most successful survivors were existing producers of electrical devices that had built their functional capabilities before the coming of the new radio technologies’. Indeed, there were only two successful start-ups in the list of survivors; the rest had developed their skills previously in such industries as auto batteries, auto ignition systems, telephone equipment, light bulbs, and other electrical equipment. A more recent story of the evolutionary hand in action concerns telecommunications where billions of dollars were invested in the mistaken belief that the public would take to new telecommunications products in volumes that were not forthcoming. Entrepreneurs find (idea), create (innovation) or respond (to demand) and adapt (commercialise) to business opportunities. The private sector needs to lead the process of a cluster creation, while the government should play a flexible catalytic role that gives grounds and incentives to the self-reinforcing entrepreneurial activity, as a one-size policy does not fit all clusters or all regions at all times. Uniform development policies do not travel well across time and space. This evolutionary approach is in sharp contrast with the neoclassical model that argues in favour of a general and widely transferable (hands-free) market approach. The evolutionary model of understanding of the world argues that there may not be a mechanical transfer of policy instruments from one cluster (or region) to another. Local characteristics ought to be tackled by tools that suit local circumstances. Positive experiences from successful clusters ought to provide more a

Theory

93

source of inspiration (perhaps a vague benchmark) and less a prescription for general action. The emergence of a cluster may (1) be based on a historical accident such as the existence of certain skills or raw materials, or a crossroads or serendipitous event (such as a discovery in physics, mechanics or chemistry). (2) Its evolution may enter into the second phase with the agglomeration of firms and the arrival of externalities that may have cumulative and accelerating features. (3) The next step in the growth of a cluster may include the creation of supporting bodies such as business associations and knowledgerelated institutions (schools and research centres). (4) The fourth stage includes the spread of non-market relations. Firms and institutions create and maintain the spirit of strong non-market collaboration which is external to individual firms, but internal to the cluster. Such clusters may be highly specialised and may have success with their output on the market. (5) However, they may easily become victims of their own success. They may fall into a ‘trap of rigid specialisation’ in the final period of their life. They may be locked into their own technology, specialised output and ethos (competency trap). (6) If there is another, competing and new good, service or technology elsewhere, which is in demand, then the cluster may decline and disappear. In any case, a cluster is mature when it has at least two properties: (1) a ‘thick labour market’ that allows both labour and employers (firms) many opportunities; and (2) the ability to withstand and adapt technology, demand and other shocks. While Boston and Silicon Valley are areas with vibrant life and business culture, Washington DC is a rather sleepy bureaucratic city. However, it has certain positive features such as people with a higher than average education level and income, as well as a tradition of federally funded R&D. Entrepreneurs in this area self-organised in the early 1980s and started rather ordinary business projects with personal funds (rather than venture capital). The projects included maintenance of computers and their networks, as well as production of medical test kits and reagents for biotechnology. Some of the early start-up firms became successful and the local critical industrial mass grew. Some of these firms went public or were bought by larger ones. Their original owners made large fortunes. This fed the ‘incubator’ of entrepreneurial human capital. Regional universities were alert to the vibrant industrial activity and started offering graduate courses that fitted well with the needs of firms. Johns Hopkins University (Baltimore, Maryland) started offering courses in biotechnology in Silver Spring (50 miles away from the main campus). Similarly, Virginia Tech University (Blacksburg) opened a campus in Falls Church (northern Virginia, some 250 miles from the main campus) within the information technology cluster. Donors enabled universities to recruit top professors (Feldman et al., 2005). Finance has always asked for certain business concentration in a defined geographical space. This was not only because of the flow of sophisticated information, but also because of the existence of economies of scale and the reduction in transaction costs. This all had to be supported by strong and accountable institutions. If one banking cluster declined, much of its legacy was transferred elsewhere. The evolutionary geographical chain in finance is obvious in the case of banking. In spite of the absence of a harbour, Florence started modern banking in the fourteenth century, or even earlier. The Florentine merchant bankers set operations at home and abroad. Through their operations in England, they were able to purchase wool for the textile industry in Florence, to collected

Evolutionary Economic Geography

94

papal contributions, to finance wars and to collect customs duties that were given to them as guarantees for their loans. The Black Death of 1348 reduced banking operations in Florence. Then Genoa and Venice appeared as the world banking capitals in the fifteenth century. Later, financial business moved from Antwerp in the sixteenth century over to Amsterdam in the seventeenth century, London in the eighteenth century and New York in the nineteenth. These were no random events. They were based on factors that include importance in trade, innovations in financial instruments and strong institutions (rule of law; honouring debts; legitimate, stable and credible governments), as well as a strong public bank. These geographical relocations of the banking business were also caused by political events and choices. For instance, England wanted to raise substantial funds to pay for its active trade and foreign policy that was in direct competition with France. Strong legal instruments were developed to honour debt and reduce credit risk (Fratianni, 2007). In the time and age of ‘globalisation’ when worldwide communication has become instantaneous, quite safe and rather cheap, it may be bizarre and puzzling to find that the international financial services are highly concentrated in very small geographical areas within London and New York, as well as Singapore and Hong Kong. Well, behind this concentration is the necessity to establish and keep trust and penalise opportunism in a network of relational contracts … We are all more inclined to trust people we know; a view which is partly based on instinct and emotion, partly on our capacity to make our own judgements (we are also inclined to mistrust some people we know), and partly on a rational calculation that people are less likely to cheat us if by doing so they sacrifice a social reputation as well as a commercial one. (Kay, 2006, p. 107) The English education and class system created and stored a relatively homogeneous background and rather unique culture and values within the City of London. The financial district’s spirit and a network of social and business relations are reinforced by powerful professional sanctions within the group against unbecoming behaviour. Hence, mechanical attempts to replicate these competitive advantages elsewhere over a relatively short period of time have had rather meagre success. Another advantage of London is that it is located in the time zone that permits it to deal with financial markets both in Asia and in the US within the same working day. In addition, London has been attracting more and more financial business from New York over the past two decades because of the US litigious climate and the Sarbanes–Oxley Act (the Public Company Accounting Reform and Investor Protection Act of 2002, following major corporate accounting and financial scandals such as Enron, Tyco International and Peregrine Systems). A favourable legal and light-touch regulatory climate, together with the openness of the British economy, is quite attractive for financial business in London. This may be the principal reason why the City is suspicious about the new and forthcoming panEuropean regulation in services. The City resists this regulation to the point that it may, for the first time ever, be against continued British participation in the EU. The reason for this stance is that the EU is becoming too involved in pan-European regulation of finance.101

Theory

95

Martin and Sunley (2003) listed advantages and disadvantages that may be found in clusters (Table 2.3). Table 2.3 Clusters costs and benefits Claimed advantages

Potential disadvantages

Higher innovation

Technological isomorphism

Higher growth

Labour cost inflation

Higher productivity

Inflation of land and housing costs

Increased profitability

Widening of income disparities

Increased competitiveness Over-specialisation Higher new firm formation Institutional and industrial lock-in High job growth

Local congestion and environment pressure

Source: Martin and Sunley (2003, p. 27)

They argued that ‘In terms of recommendations on best practice, however, there is a consensus that cluster promotion policies are unlikely to succeed in creating clusters ab initio … Rather, they should somehow attempt to build on the potential already present in a particular economy’ (p. 23). However, they warn and advise that the local authorities ought to encourage productivity improvements in all local firms Ellison and Glaeser without necessarily committing to a cluster mind-set. The danger of a cluster-based approach to policy is that it detracts from the need to take a more holistic view of regional development … even cluster enthusiasts find it enormously difficult to point to any examples of deliberate cluster promotion programmes that have been unambiguously successful. (p. 28) (1997) discussed the prevalence of Silicon Valley-style clustering of individual manufacturing firms in the US. The authors created a model in which natural advantages, localised industry-specific spillovers and pure random chance all contribute to spatial agglomeration of production. The empirical evidence provided strong confirmation that localised spillovers are an important factor in the shaping of spatial concentration and coagglomeration of industries. Certain governments devote substantial resources to attract foreign investors. The general desirability of such outlays depends in part on the strength of wide-ranging agglomeration effects and spillovers. The Gini concentration index does not distinguish between firms of different size. This shortcoming is put right by the Ellison–Glaeser index.102 This index controls for differences in the size distribution of plants and for variation in the size of the geographical area for which data is available. Therefore, this facilitates and increases confidence in cross-industry and cross-country comparisons. Their finding was that almost all industries were more concentrated than a random dart-throwing103 model would predict. This reaffirmed earlier observations about agglomeration of production.

Evolutionary Economic Geography

96

However, the level of agglomeration varies considerably across industries, as does the tendency of an industry to co-agglomerate with other industries. In addition, Krugman (1998a, p. 15) observed that ‘models are not at all easy to calibrate to actual data; in general, the tendency toward agglomeration is stronger in the models than it seems to be in the real economy!’ Garofoli (2002) built on his earlier work and looked at local development in the context of Europe. His leitmotif was that the ‘high road’ to development is based on quality products and innovation is the way for local and regional development in Europe. This development may be exogenously controlled and managed, hence it depends on ‘others’. Conversely, endogenous development relies on the local capacity to learn, innovate, produce and extend ‘collective intelligence’ (spillovers). Clusters benefit from an accumulation of knowledge; reproduction of location-specific resources; collective efficiency; and the creation of dynamic competitive advantages. The longterm competitiveness of products made in depressed regions such as the Mezzogiorno cannot be based and sustained only on relatively low costs of production (the ‘low road’ to development). An opportunity and challenge is to engage in the ‘high road’ to development, based on new skills, knowledge, competences, technology and innovation, which bring and keep better jobs and pay higher wages in clusters of small and mediumsized dynamic firms. Garofoli recommended this ‘high road’ to the whole of Europe in the face of global competitive challenge. Recap Even though the concept of clusters has many seductive elements and can be associated with a large set of positive images, there is much about it that creates problems and asks for a degree of caution. Are clusters an economic policy panacea? Or are they a chaotic concept? The general idea of clusters was introduced by Marshall, well over a century ago. It is not new. Relatively new in economics is the discussion about the relevance of international economic integration (some equate it with ‘globalisation’) and market enlargement on the location of economic activity. Some argue that the location of firms in space is increasingly becoming irrelevant because of globalisation (footloose firms/industries coupled with a sharp fall in transport and communication costs). Others present the opposite view: competitive advantages are strongly concentrated because of highly localised accumulated skills, knowledge, experience, institutions, suppliers and rivals. One of the problems with the cluster concept is a lack of a widely accepted definition of a cluster. Clusters are often referred to as industrial districts, new industrial spaces, territorial production complexes, neo-Marshallian nodes, regional innovation milieux, network regions and learning regions. Various things are put under the ‘umbrella’ of a single, all-embracing notion. This ‘definitional and conceptual flexibility’, as well as its use in a cavalier manner, could potentially make it a ‘chaotic concept’. Certain quantitative studies rely upon complicated statistical acrobatics and mathematical virtuosity, while others refer to clusters in an arbitrary way. Public policy and administrative areas in which the authorities implement policy and collect statistics are usually not the same as clusters, which is why it is hard to determine the exact economic significance of clusters with a high degree of reliability. Industrial and spatial boundaries

Theory

97

are not clearly defined and mapped, which makes policy influence and its assessment a demanding and uncertain task. Clusters as an organisational form of industrial business demonstrate a certain age in a relatively mature economy such as Italy is. Hence, the Italian experience with clusters is highly relevant for economies throughout the world. If national economies are becoming globally integrated and linked though a complex supply (and demand) chain, then in order to profitfrom increasing returns to scale, the size of firms becomes important (small size is a constraint). However, consumers are becoming more and more demanding and picky with an increase in income. They demand custommade goods (in small batches). Therefore, there is still a bright future for local SMEs that adjust to demand even in the ‘global’ economy. Clusters will continue their lives, even though they may be relocated from one country or continent to another. What are the implications for policy towards clusters? Well, one thing is clear: there is no general best practice cluster policy that is valid for each and every situation at all times. In a way, cluster policy is also creative art. However, new and successful clusters may appear with a certain degree of success. The examples of Ireland, Finland, Texas and the coastal regions of China and India provide evidence for this assertion. Old and successful clusters may continue their lives (finance in London), while old and declining clusters may be revamped, as is the case in the German Ruhr region or in Boston. Economic integration (regional free trade) and liberal treatment of trade and FDI may also assist. Many general country-wide and non-directly spatial policies, such as education and training, may have a stronger and more important impact on cluster creation and life than openly spatial and cluster-related policies. Conclusions and policy implications Certain general policy advices and guidelines regarding clusters may be found in Porter (2000b). Successful cluster ventures have several common features: • a shared understanding of competitiveness based on productivity and innovation (not on a low wages, low taxes and devalued currency); • concentration on removing obstacles to upgrading clusters; • cluster boundaries should not necessarily reflect administrative borders, but rather mirror economic realities; • the general involvement of cluster participants and associated institutions in the operation of clusters ought to be wide and deep; • the private sector ought to lead. Private firms usually have a superior sense regarding identifying obstacles, constraints and opportunities compared with other participants; • cluster ventures ought to be as non-partisan as possible and should remain independent of any party or administration’s political agenda. If a new government enters office, earlier promises may be withdrawn (elimination of uncertainty); • close attention to personal relationships: a good and deep communication among the participants is essential for successful cluster ventures. There ought to be firm action based on the desire to achieve results; • the upgrading of clusters ought to be a long-term process. Participants and institutions in successful clusters have vibrant, continuous and long-term relations and common interests.

Evolutionary Economic Geography

98

Porter (2000b) concluded that although global sourcing mitigates disadvantages, it does not create advantages. Moreover, distant sourcing normally is a second-best solution compared to accessing a competitive local cluster in terms of productivity and innovation. Paradoxically, the most enduring competitive advantages in a global economy seem to be local. (p. 31) In spite of trendy talk about the ‘global economy’ and the diminishing role played by specific geographical locations for business, lasting competitive advantages are often heavily local. ‘The more the world economy becomes complex, knowledge based, and dynamic, the more this is true’ (Porter, 1998, p. 90). These gains come from the clustering of highly specialised knowledge and skills, as well as the existence of rivals, sophisticated customers and institutions. One has to recall, however, the creation and existence of successful firms outside clusters. Microsoft, founded in Albuquerque (New Mexico) in 1975, relocated to Seattle in 1979. At that time Seattle did not have a cluster of software firms. The software cluster was at that time in Silicon Valley. Also, Switzerland is far from seas and shipbuilding clusters, but it is specialised in huge marine diesel engines (Sulzer) for container ships. Therefore, clusters are just one of the possible areas and forms for the spatial location and organisation of business. Cities According to an extreme and unqualified version of the neoclassical equilibrium theory, cities should not exist. All economic activity ought to be spread more or less evenly across space, i.e. flat geographical space. This extreme version requires constant returns, no fixed costs, no agglomeration of resource inputs, no differences in the productivity of land, etc. Nonetheless, neoclassical economists were not such fools. Fixed costs are in the standard theory of the firm. Yet people and businesses are not spread evenly across space. Certain agglomerations are facts of life and business. Increasing returns in production may be among the principal man-made reasons for this development. It is compounded by the convenience and advantage of spatial proximity for relations among various agents. Finally, the time dimension is also relevant. Accumulation of knowledge takes time, and it also takes time to transfer it to the future generation. It is possible that this accumulation may decay at a slower rate in time if it is concentrated than if it is widely spread in space. It is now supposed that it was the Sumerian civilisation that created the first cities, in Mesopotamia (now Iraq) about 3000 BC. The evolution and appearance of first citystates such as Eridu, Ur and Uruk was closely linked with economic activity based on irrigation and commerce. The advantages of this type of social organisation technology (cities) spread elsewhere in the area and beyond. Satellite photographs and research provide evidence that Angkor (now in Cambodia) was the largest pre-industrial city in the world. Its period lasted over six centuries (802– 1431). Angkor’s closest rival was Tikal, a Mayan city (now in Guatemala). The Khmer

Theory

99

kings constructed a new capital following independence from Java in 800. The region was convenient also for the construction of a big reservoir, principally for the irrigation of rice fields but also for religious purposes. The population of this great city was about a million and its area was roughly the same as the area of modern Los Angeles. The end of Angkor came in 1431 after its defeat and ransacking by the Thais. Apart from Angkor Wat, which remained an active Buddhist shrine, the remaining citizens abandoned the city. There are several theories on the abandonment of Angkor. Some think that a heavy defeat by the Thais was the cause. Others argue that wars exhausted the government’s ability and funds to engage in public works to rebuild and maintain the existing irrigation system, so necessary for the production of rice on which the population depended for work and subsistence. Yet others think that natural disasters such as floods, shortages of water, earthquakes or diseases were the causes or contributors to the demise of Angkor. Even great cities may have their ups and downs over time. Cities are settlements that generate their own growth (Jacobs, 1969, p. 141). A city may also be thought of in economic terms (with certain qualifications) as the absence of space among people, firms and institutions, as well as the presence of invisible communication channels. Diversity and a certain tolerance are also an integral part of cities. Even though rents are, in general, higher in cities than in the countryside; even though certain streets in some older cities are in general narrower than is the case in the rural areas; and even though there is more pollution and less agreeable green space than in the countryside, people create, maintain and flock into cities. It may be pertinent to recall Plato’s discussion about the location of cities or citystates, written in 360 BC in The Laws (Book IV) (Plato, 1980, p. 159): Then the state will have tolerably healthy prospects of becoming virtuous. If it were going to be founded near the sea and have good harbours, and were deficient in a great number of crops instead of growing everything itself, then a very great saviour indeed and lawgivers of divine stature would be needed to stop sophisticated and vicious characters developing on a grand scale: such a state would simply invite it. As it is, we can take comfort in those eighty stades. Even so, it lies nearer the sea than it should, and you say that it is rather well off for harbours, which makes matters worse; but let’s be thankful for small mercies. For a country to have the sea nearby is pleasant enough for the purpose of everyday life, but in fact it is a ‘salty-sharp and bitter neighbour’ in more senses than one. It fills the land with wholesaling and retailing, breeds shifty and deceitful habits in a man’s soul, and makes the citizens distrustful and hostile, not only among themselves, but also in their dealings with the world outside. Still, the fact that the land produces everything will be some consolation for these disadvantages, and it is obvious in any case that even if it does grow every crop, its ruggedness will stop it doing so in any quantity; if it yielded a surplus that could be exported in bulk, the state would be swamped with the gold and silver money it received in return – and this, if a state means to develop just and noble habits, is pretty nearly the worst thing that could happen to it, all things considered (as we said, if we remember, earlier in our discussion).

Evolutionary Economic Geography

100

The principal workhorse of the new economic geography is that spatial proximity of mobile factors is favourable for their higher productivity. Predictability in transactions is increased, while delivery time and transport costs are reduced. This is linked with the idea of agglomeration (creation of cities and clustering) and dispersion (spread) of economic activities. The equilibrium spatial location of activities is the result of the conflict and interplay between centripetal and centrifugal forces. A crucial question is: what happens to firm A’sprofit when another firm B comes to the same location and starts business? Does this profit increase or decrease? If the profitoffirm A increases, or if profit increases in both firms, then agglomeration forces operate in this location. If profit falls, then spread forces predominate, the market becomes overcrowded and firms ought to be dispersed in space. Mobile factors and the ones that are sophisticated (educated) profit most from this process, while the principal costs fall on those factors that are immobile or partially mobile. Urbanisation is one of the important features of modern society (Figure 2.10).104 One of the reasons for this development is a declining share of labour relative to capital in agriculture. This is particularly obvious in the developed countries. In general, both labour and capital are heavily concentrated in cities. Footloose industries are less reliant

Figure 2.10 Urban and rural population of the world, 1950-2030 Source: UN World Urbanisation Prospects: The 2005 Revision (http://www.un.org/esa/population/publication s/WUP2005/2005wup.htm, accessed on 8 January 2008) on natural inputs than other types of manufacturing and service industries. In some cases they may place more emphasis on the proximity to final markets, which further adds to the urbanisation trend. In any case, an obvious general trend throughout the world is the movement of the population towards the booming coastal urban regions.

Theory

101

More than half the world’s population lives in cities. However, a third of that is within urban slums with poor housing, inadequate services, and crime. Often slum dwellers migrate from poor countries to slums in developed countries. Management of slums throughout the world presents one of the most demanding policy challenges. The number of cities in the world with a population of over 10 million was two in 1950. This number was quite stable until 1980, when it increased to four. However, it rose to ten in 1990, seventeen in 2000 and twenty in 2005; and projections are that this will increase to twenty-two in 2015.105 These data refer to the largest cities; however, a very similar trend throughout the world is also obvious in general terms in relation to many cities that are of smaller size (below 10 million inhabitants). Hence, in spite of strong agglomeration diseconomies and enormous problems, there are robust agglomeration gains (economies of scale and proximity–productivity effects) that support further location of economic activity in cities. Firms and mobile workers move to and locate in areas where they may reap the benefits of higher productivity. This creates in turn a positive feedback, cumulative causation, acceleration and amplification of the proximity– productivity effect.106 Hence mobile firms and workers ‘tend to further raise productivity, creating an uneven distribution of activity and spatial income disparities’ (Venables, 2006, p. 5). This is a ‘lumpy’ process that inevitably leaves someone behind. In a survey of the literature, Rosenthal and Strange (2004, p. 2133) report that doubling the city size may increase productivity by 3 to 8 per cent. External economies exist in cases in which the size and variety of urban environment adds to productivity. This is a considerable effect, as ‘moving from a city of 100,000 inhabitants to one of 10 million is predicted to increased productivity by more than 40 per cent’ (Venables, 2006, p. 12). If a new activity is about to start somewhere in space and if the potential locations A and B for this activity are identical, then the location that gets or attracts this business obtains a head start and advantage over the other. This is the reason why many places enter into ‘location tournaments’ in order to attract new firms and industries. Another incentive for urbanisation comes from the dependency of many manufacturing and service operations on access to public services, in particular in those countries that are known for intervention. Because of this, lobbyists cluster in politically significant cities such as Brussels or Washington DC, because this is the geographical location where big decisions are made, procurement contracts awarded and subsidies distributed. Both labour and capital become over time more and more concentrated in cities and reinforce unevenness in the spatial distribution of economic activity. Markets are strongly imperfect in this case; consequently, there exists a need to spread economic activity and population in space. The rationale for policy intervention is there, but we continue to be ignorant on how to manage this situation. What exactly ought to be done? what needs to go where? how? when? and what may work in the desired direction over time and space? The question of why many economic activities are agglomerated in a relatively small number of places (typically cities) raises a number of complex issues. For example, a consumer organises his or her own multiple shopping itinerary so as to minimise the total cost of purchases, including transport costs. This is a particularly difficult combination problem to solve, while the final equilibrium becomes very problematic. In any case, the reasons for agglomeration include (1) externalities under perfect competition, (2)

Evolutionary Economic Geography

102

increasing returns to scale under monopolistic competition and (3) spatial competition under strategic interaction (Fujita and Thisse, 1996). Even though rents may be higher in cities than in the countryside, people are paying higher rents in the cities basically to be closer to other people. The same holds for many firms: externalities create centripetal forces; there is a functionally indivisible production (strategic complementarity) in certain types of production; firms want to be closer to their principal market; and agglomeration is a cumulative, accelerating and self-enforcing process. In the dim and distant past, the creation of wealth largely depended on the local availability of natural resources. A mix of geographical, economic and historical factors shaped industrial development and the location of firms, industries and cities. However, as the economy evolved, wealth creation came to depend more on the accumulated physical assets (mainly equipment and finance).107 The prosperity of the modern economy depends not only on the physical resources, but also and increasingly on intangible assets such as knowledge, information-processing potential, and organisation, distribution and control capabilities. Over the past century, cities that were based on educated and skilled labour have grown faster than comparable cities with less embodied human capital. Firms in such cities are more productive than elsewhere where labour is less skilled. In addition, cities based on skilled labour and management adapt better to shocks. Accumulated knowledge and its transfer through education and spillovers, as well as talent, allows easier adaptation to changes, challenges and opportunities. In addition, the lifestyle in such cities may be more attractive to some people than elsewhere. Some threequarters of the US population live in the 100km-wide coastal belt which includes the Great Lakes and the St Lawrence Seaway. Only 1.9 per cent of the land area of the United States was built-up or paved by 1992. Yet, despite the wide availability of open space, almost all recent development is less than one kilometre away from earlier development. Not only does the proximity of earlier development matter, but so does its density … One cannot make sense of this sort of numbers, of the extent to which people cluster together in cities and towns, without considering some form of agglomeration economies or localised aggregate increasing returns. (Duranton and Puga, 2004, p. 2065) Regarding the location of cities, clusters and industries, spatial economics distinguishes between first- and second-nature geography. First nature refers to exogenous conditions. It deals with the preference for certain locations over others because of their natural endowments of factors (fresh water, deposits of coal or minerals) or proximity to rivers, estuaries, coasts or natural harbours. This was particularly relevant for the location of production before and during the First Industrial Revolution. Second-nature geography refers to man-made activities which improve and extend the ‘first nature’. This deals with economies of scale and interaction among economic agents. Agglomeration tendencies in modern industries may be much more influenced by second-nature forces than is the case with first-nature (natural) advantages.

Theory

103

Even though cities were initially created to reduce transport and communication costs for people and their goods, services and ideas (to eliminate physical space between people and firms), many other factors were quite accidental. An Italian saint prompted the creation of Brasilia, Brazil’s capital, in 1883. He promised great bounties if the interior of the country, mostly the Amazon jungle, was developed. In 1891, a new constitution called for the move. A foundation stone was laid in 1922, but it was not until 1957 that the new capital city project got under way 900km inland. This was a utopian experiment in modern urbanism that was intended to stimulate a quantum leap in the development process. What resulted was a pedestrian-unfriendly city that lacks a human touch. It is packed with (monstrous) concrete government buildings in various states of decay. Looking elsewhere, Washington DC was created in the 1790s when President George Washington chose a swamp for the permanent seat of the US government (somewhere between the North and the South). ‘It’s not quite so bad now, but … once the city’s cubicle-dwellers flee to the suburbs, Washington can feel as empty as a bad highschool dance.’108 The most attractive spatial locations for setting up cities over long centuries included: • crossroads; • estuaries; • safe natural harbours (with a rich hinterland) on the sea and navigable rivers;109 • defendable hilltops; • places with ample fresh water (for drinking and power) and rich soil; and • dry cool highlands (free of malaria-carrying mosquitoes). In the past, the location of a firm or a city was influenced both by the endowment of immobile local resources and by flows of mobile factors.110 Once the development of a business activity starts in an area and if the economic system is flexible, this area attracts other business activities to the region. This model of uneven development is based on the possibility of meaningful multiple equilibria in the presence of external economies. The point is evident: with external economies, the return on resources in a particular industrial activity is higher when more resources are committed to it (the proximity–productivity effect; Krugman, 1991a, p. 651). The property of a modern firm is high mobility in its search for profitable opportunities, not only within its region or country, but also internationally. For a footloose firm, the advantages of one location in relation to others are much more man-made than subject to resource endowments. Many cities used to have economic advantages (production and market access) for a long time just because they were located in the US. However, this advantage is losing its strength and importance in a ‘globalised’ world. Similarly, footwear production is moving from Italian clusters into Romania (and elsewhere). Design and sophisticated production is kept in Italy, while much of the ‘raw’ production is done in Romania, often in Italian-owned firms. According to the analysis of spatial economics by Fujita et al. (1999, p. 131), cities exist because firms locate at a cusp in the market potential function made by a concentration of other firms. A growing and spreading agricultural population makes it beneficial for producers to develop new cities. The size of the cities differs (there is a hierarchy) because of differences in industrial externalities and transport costs. Finally, natural advantages (such as the existence of harbours) help as they create natural cusps in

Evolutionary Economic Geography

104

market potential. Arthur (1994b, p. 109) argued that the observed pattern of cities cannot be explained only by economic determinants without considering chance events. Cities also grew up where immigrants with certain skills settled, where Captain Cook found an anchorage, where politicians decided to build railways and canals or where trains stopped for the night. Foundation mechanisms Cities and urban agglomerations are based on three micro-economic mechanisms (Duranton and Puga, 2004): • Sharing Sharing indivisible facilities (production facilities, markets and public goods); sharing the gains of a wider variety of inputs and suppliers; sharing the gains of narrower specialisation and economies of scale; and sharing risks. • Matching Urban agglomerations with a large, even increasing number of heterogeneous agents (firms and workers with various skills, as well as intermediate goods) increase both choice and the probability of matching. The incomplete information problem is potentially reduced in cities.111 There is also the possibility of higher-quality matching among agents. If the economic environment is uncertain and changing fast, cities offer a variety of matching gains. This diversity within cities (as opposed to specialisation) is linked with the interaction between urban location costs and the cost of trade. Diversity of opportunities (jobs and life amenities) is an important feature for a city to attract talent and human capital. • Learning Cities bring together a large number of people with their ideas. Knowledge is created, accumulated, stored, extended and diffused. Proximity among players reduces their communication costs. In the consideration of the composition of industrial systems in cities, their growth and innovation, the professional literature often refers to three basic views: • Localisation economies (specialisation) (Marshall, 1890; Arrow, 1962; Romer, 1986) Most urban and agglomeration externalities arise from the transfer of knowledge and various spillovers within an industry. Local specialisation, increasing returns, intense competition and knowledge spillovers among firms in the same industry are the leitmotif according to this approach. These intra-industry transfers are obvious in Silicon Valley or the City of London financial district or within the Basle pharmaceutical industry or in Chicago regarding transport. Barriers to knowledge spillovers are quite low, while competition is high.112 • Urbanisation externalities (diversity) (Jacobs, 1969) Most externalities take place across different industries.113 A diversified and heterogeneous industrial structure within a city may make it innovate more and grow faster than is the case with specialised cities. Sometimes serendipitous meetings and contacts among people belonging to different but open networks in a diversified city may make this happen (cross-fertilisation of ideas). These Jacobs’ externalities are relevant in rapidly changing hightechnology industries. As diverse businesses are not in direct competition with each other, they may be quite open to new ideas and willing to interact and cooperate with other businesses. This feature may be stronger than is the

Theory

105

case in the Marshall–Arrow–Romer situation, where firms may perceive other firms as their rivals. • Competition (Porter, 1990a) The power of competition within cities intensifies the value of various externalities. This supports innovation. Most cities continue to grow and sprawl in spite of both ‘globalisation’ of economic activity and an alleged general reduction in the cost of transport and communication. Empirical evidence about the predominance of one or another view may be rather inconsistent. Results are based on the availability and comparability of data, particularly longer-term time series; general or industry-specific effects; manufacturing that produces tradable goods (production can be concentrated) and services, many of which have a nontradable character (they have to be spread and produced close to consumers); the maturity or development phase of each industry; external shocks; and developments in costs of transport. Nonetheless, there are quite convincing observations about reality, to which we now turn. Benjamin Disraeli, Karl Marx, Friedrich Engels and their contemporary observers were impressed by the city of Manchester and the stunning efficiency of its immense textile mills during the 1840s. Manchester was the most advanced city at that time and was the envy of the world. Birmingham, on the other hand, with few relatively large manufacturers, no obvious business specialism and its small-scale household trades in guns, steel, glass, leather, toys and jewellery seemed to be outmoded. Nobody was nominating Birmingham as the city of the future. However, over time, highly specialised Manchester fell into stagnation, while diversified Birmingham became the city of the future. Today, only two cities in all of Britain remain economically vigorous and prosperous. One is London. The second is Birmingham. The others have stagnated one by one, much as Manchester did, like so many lights going out. British town planners, ironically, have regarded London and Birmingham as problems, because they are places in which much new work is added to old and thus cities that persist in growing. (Jacobs, 1969, p. 89) At the start of the twenty-first century, specialised Manchester remained stagnating and vulnerable to obsolescence, while diversified Birmingham and London remained thriving. Also, ‘London is the safest place to base a company, even if living in some of its inner suburbs requires reckless courage and Kevlar undergarments. Follow the money, in other words.’114 Although Singapore had invested nearly twice as much as Hong Kong for nearly twenty-five years, it had grown no faster. The implication was that its leaders had been too eager to chase each new fad: microelectronics, computers, financial services, biotech. For all the frenetic upgrading, Singapore’s citizens were worse off than their counterparts in highly diversified Hong Kong (the story of Manchester and Birmingham once again!) (Warsh, 2006, p. 320). In addition to the diversity story, Jacobs (1969, p. 51) gave an example of how one kind of work in a diversified city is leading to another (even within the same general industry). She illustrated her argument by the use of a pleasing story about the invention,

Evolutionary Economic Geography

106

creation and development of the manufacturing of bras in New York City. Ida Rosenthal was making dresses in New York on her own in a small shop during the early 1920s. Dissatisfied with the way her dresses were hanging over her clients’ various corsets, Ida decided to improve the fit. She experimented with various changes and improvements in the underwear of the time and the result was the first brassiere. Each customer who bought one of Ida’s dresses got a custommade brassiere. Bra-making was only an accessory activity to her principal and older business of dressmaking. Nonetheless, the new underwear item became more and more popular on its own. Ida found a partner and together they started producing, wholesaling and distributing brassieres. The new business activity stood and thrived on its own. Apart from New York City’s garment district, there are also other districts such as those for finance, diamonds, advertising, publishing, education … full of potential competitors and entrepreneurs who are observing what others are doing and are ready and willing try out new methods and products. New possibilities and changes, such as the use of the Internet, permit modification in working time and place of work for select professions. Certain key employees can work at home in the environment that they enjoy best (presumably non-congested countryside). In this situation, certain office space is liberated in cities which may be partly transformed into entertainment areas. New country towns may be big enough to create employment on their own. This may be, perhaps, a superior route for the future, rather than endlessly expanding the existing urban areas. If commuting becomes less trendy and acceptable because of congestion, perhaps new towns may be developed, provided that there are suitable locations. Examples can be found in the creation of Milton Keynes in England in 1970, or Stevenage and Corby during the 1950s.115 Others, such as Crewe, appeared from nowhere. Crewe was a village with just a few hundred residents. Two centuries later, it had grown into a town with a population of over 70,000. One should never forget the attraction of a location for social life to people, in particular to creative people. They also need a favourable local natural and social milieu that may inspire them for innovation in production and management, as well as that keeping them in the area. They need to meet peers, not only for business or research ventures, but also for family reasons and an agreeable social life. Florida (2002) put emphasis on the ‘people climate’ that ought to be a complement to the traditional ‘business climate’ in a location. Size and sprawl Why are cities of different sizes? Henderson (1974) argued that there is a trade-off between economies of agglomeration of industries specifictothat city and general diseconomies (negative externalities), such as costs related to commuting and high rents which (apart from pollution) do not depend on the structure of the local industry. The optimum city size depends on the maximum welfare of participants in the economy. It does not make much sense to put industries without spillovers (e.g. steel production and publishing) in the same city. However, it does make sense to locate the steel and textile industries in close proximity as they employ, respectively, mostly male and mostly female labour. Needs and potentials of dual-industry and dualcareer couples ought to be taken into account. Cities need to be specialised in one or a few industries with related

Theory

107

external economies.116 These external economies, however, vary a great deal across industries. There are still certain coagglomeration tendencies for different industries to locate near each other. Because of strong links internal to the industry, a financial centre may do best if it includes virtually all national financial institutions. The same is not true for the textile or food industry. Hence, the optimal size of a city depends on its role (Fujita et al., 1999, p. 20). Although Henderson dealt with spatial aspects, this type of model is surprisingly ‘aspatial’, because it says little about the actual location of a city, either in space or relative to other cities. In addition, national trade policy can influence the size of a city, as exemplified by Mexico City. An unintended by-product of the Mexican importsubstitution policy was the expansion of the capital city because of production linkages and economies of scale. Once Mexico started to open its economy during the 1980s, there was a relocation of certain firms away from Mexico City, mainly towards the northern frontier. Zipf’s Law refers to a special static case of the relation between city population size and its rank among other cities. If this power law with exponent 1 holds, then the largest city is k times as large as the kth largest city. This implies that the largest city is twice as large as the second largest, three times as large as the third one, and so on. The law depends on the existence of labour mobility. The distribution of US city sizes is close to the rank-size rule, while this relation is less obvious in Europe (Midelfart et al., 2003, pp. 860–1). The Second World War had a profound and lasting impact on the city size distribution in Germany. ‘The overall city size distribution does not adhere to Zipf’sLaw’ (Bosker et al., 2006, p. 26). Desire to realise a new lifestyle in a suburban environment, outside the inner city, stimulates city spread. Urban sprawl takes place when the rate of landuse conversion exceeds the rate of population growth. This means that urban sprawl is a product of changing lifestyle, rather than of a growing population. A century ago people were walking to their work. Today most of them use public transport or drive cars (especially in the US). Sprawling cities require more utilities, more roads and more energy and water supplies. Citizens travel more, pollute more and contribute to traffic congestion and climate changes. Cheap oil during the 1950s and a large part of the 1960s (and through the 1980s) made the Americans invest massively in roads and cars, as well as in new real-estate developments in the suburbs. These developments made them free from public transport and enabled them settle where they chose. They became dependent on roads and cars. Many of them live miles and miles away from everything: their jobs, golf courses, shopping malls, schools, railway stations … They were not, generally, driven, by circumstances to live in congested areas. Many areas in the US have a low density of population that makes the existence of public transport impracticable and very expensive for users and taxpayers.

Evolutionary Economic Geography

108

Examples NEW ORLEANS Before we turn to the third type of spatial units (regions) considered in this book, let us look at select examples of the location and growth of cities. The importance and impact of the location of a city may be well described with the example of New Orleans. The War of 1812–15 was fought between the US and Britain (together with its colonies in North America). The Battle of New Orleans in January 1815 was one of the key moments in the American history. Had the British won this battle, perhaps they would have not given it back to the Americans. France sold its territory of Louisiana to the Americans in 1803 for $15 million. This ‘old’ French Louisiana included at that time land that covers almost a quarter of the current US territory. The old Louisiana covered, in general, the American lands that drain into the Missouri and Mississippi rivers, including New Orleans. Without this city, the whole purchase would have been of no value to the US as the city controls huge upstream and downstream river traffic. The Germans knew this very well, as some of their U-boat operations took place near New Orleans during the Second World War. The Soviets must have had similar ideas during the Cold War. The paramount importance of the port of New Orleans for the US economy is obvious, not only because it is among the largest ports in the world, but also because of the structure of its throughput. Through this port the US exports (downstream) agricultural products (corn, soya beans … ), so essential to the world, while it imports (upstream) crude oil, chemicals, concrete, iron (for the auto industry) … so vital for manufacturing and farming in the central US. The importance of the Mississippi and Missouri (and the port of New Orleans) is not only that river transport is cheap, but also that there are no good substitutes, even in the longer run. The US transport system does not have a sufficient number of rail wagons or lorries to carry these enormous quantities for long distances. Hurricane Katrina (2005) destroyed a big chunk of New Orleans and its port. However, all is being restored, at a huge cost. The simple reason for this rebuilding, in spite of high and real hurricane dangers, is that a city ‘has’ to be there.117 ANTWERP Antwerp is a safe natural harbour located in Flanders at the mouth of the Scheldt, close to the North Sea. Because of its favourable and important geographical location, Flanders has been a battleground in many wars. Thanks to its good geographical location, this city has specialised in trade in salt, fish, grain, English wool and textiles since the eleventh century. The economic decline of nearby Bruges as an important commercial and banking town began in the 1470s. The authorities in Bruges turned hostile towards banks after a number of bankruptcies, notably demanding that foreign bankers leave the town (or resettle in Antwerp) between 1484 and 1488. The fifteenth century was the golden age for Antwerp. Its location was supreme for the distribution of spices (pepper and cinnamon) throughout Europe after the Portuguese discovered the sea route to India, and Antwerp became the world trade centre. Hundreds of ships were docking daily, while

Theory

109

many traders and bankers had their branch offices in this highly open, international and tolerant city (which attracted a large Jewish community). Banking also flourished and trade on the bourse was not only for spot supplies, but also for future deliveries. By 1560, Antwerp’s population had grown to over 100,000, making it the second largest European city (after Paris). Protestants preached and practised the humble values of modesty, frugality, hard work, literacy and cleanness. They were against the luxurious habits of the clerical nobility. The bad harvest of 1565 and other economic difficulties contributed to a general discontent. A crowd stormed the church in the town of Hondschoote in Flanders (now in France) in August 1565. Ripples from this relatively small event immediately spread north, and in 1566 the Reformation revolution erupted in Flanders. Antwerp Cathedral was ransacked and defiled by the Protestant iconoclasts. Philip II, a devoted Spanish Catholic, sent the Duke of Alva to the region to calm the situation. Repression, massacres and the exile of Protestants, known as the ‘Spanish fury’, followed. This, coupled with the Roman Catholic Inquisition, provoked an eighty-year war (1567–1648), with a naval blockade of Antwerp for most of this time. The population of Antwerp declined to 42,000 during the period 1560–90. Eventually, the Treaty of Münster (1648) closed the Scheldt to traffic (but not the Rhine). This destroyed Antwerp’s trading business and advantages, already declining because of the naval blockade. These external events made the location of trade and finance move to Amsterdam and Genoa. Amsterdam became so prosperous on its own and developed into the principal trading point, so that the local merchants doubted the wisdom (because of cost/benefit concerns) of recovering the south of the region. The Scheldt was reopened in 1795 when the French took the city. In 1803 Napoleon recognised Antwerp’s strategic location as a ‘cocked pistol aimed at England’. He developed a port and the first dock was named the Bonaparte Dock. Although sporadically interrupted by subsequent wars, Antwerp developed into an important Belgian trading, transport and industrial city in the following centuries. VENICE There are very few secure natural harbours in the Adriatic Sea. One such favourable geographical location is found in Venice. This advantage made Venice a natural crossroads for trade between Europe on the one hand and the Mediterranean region, the Middle and Far East on the other. Venice (La Serenissima)profited from its nearmonopoly on trade with the East for centuries. It became a city of commerce and art,118 as well as of sensual and visual pleasure (rather than of intellectual pursuits). Carnivals were perpetual. Venice was living well, but ‘working little’. The power and greatness of this consumer society started to wane after the discovery of alternative routes to China around the Cape of Good Hope by Vasco da Gama, the Portuguese navigator, in 1498. This just added to the effects of the discovery of the New World in 1492. The fading of the importance and prominence of Venice as a centre for business, trade, politics and culture lasted until 1797, when the last of her rulers, the Doge of Venice, surrendered her independence to Napoleon. Nonetheless, there is no doubt that, over those three centuries of ‘decay’ (and well beyond), Venice has been one of the most attractive, interesting, amusing and exciting places in the world. It still is. The stream of tourists from the early twentieth century has restored it to colourful economic life based on services.

Evolutionary Economic Geography

110

External shocks External shocks may bring important changes to the life of a city. However, one may not wish to give those shocks an absolute value. For example, the nuclear bombs dropped on Hiroshima and Nagasaki in 1945 had no longrun impact on the size of these two cities. ‘Within the space of just 20 years, they recovered from the devastation to return to their former place in the constellation of the cities’ (Davies and Weinstein, 2002, p. 1283). Similarly, disastrous bombardments of Tokyo with firebombs at the same time, despite enormous damage and casualties, did not have a very long-term impact on the position of that city in the ‘hierarchy’ of Japanese cities. Likewise, the impact of the bombardment of many German cities during the Second World War on the place of these cities in the economy and its growth was significant, but temporary (Brakman et al., 2002, p. 28). Also, during the sixteenth century in Italy, many cities such as Florence, Pisa, Ravenna, Rome, Siena, Verona and others suffered profound demographical shocks because of wars and plague. Nonetheless, the cities recovered in their importance. Their location was resilient and stable in the face of external change and shocks. Sometimes there may be a stronger resilience to shocks in the location-specific fundamentals than one might expect a priori.119 Regions Introduction A region may be more easily discerned than defined. The definition of a region depends on the problem that one encounters. A region may be seen by some as a geographical phenomenon with distinct spatial borders with others. In addition, a region disposes of a bundle of assets that are soundly connected within this region’s (internal) economy. This observation may apply without much dispute to relatively small island states. Conversely, others may argue that a region is an unclear concept, maintaining that it has a degree of openness120 within a country or inside a group of integrated countries and that the internal economic cohesion of a region’s fortune (assets) may be overemphasised. A region also has political, governmental and administrative features and is an ethnic and social concept with human and cultural characteristics and mores.121 ‘Increasingly it makes sense in practical terms to think of economic Europe as consisting not of a collection of internally homogeneous countries, but of a collection of regions, of varying degrees of peripherality from a centre located somewhere around Brussels’ (Krugman and Venables, 1990, p. 59). A region as an economic concept is defined by its endowment factors, their combinations and mobility. North (1955, p. 257) suggested that ‘the unifying cohesion to a region, over and beyond geographic similarities, is its development around a common export base. It is this that makes it economically unified and ties the fortunes of the area together.’ Therefore, a region can be defined in theory as a geographical area that consists of adjoining particles with similar unit incomes and more interdependence of incomes than between regions (Bird, 1972, p. 272). Untraded interdependencies among economic particles assist to a large extent to identify a region, but these intra-regional links are not the only reasons for a region to

Theory

111

exist and grow. Politics or military reasons may determine which regions grow and how they do so. A region of a country or an economic union may in theory be thought of as an open economy. It may be that regions are difficult to define as ‘essential’ economic units, but the fact is that a process of ‘regional institutionalisation’ of policy intervention and responsibility appears to be underway that is endowing politically and administratively defined regions with some degree of functional economic meaning. (Martin, 2006, p. 163) Let us approach this issue by looking first at the regional problem. Regional problem There are many causes of regional disequilibria. According to Vanhove and Klaasen (1987, pp. 2–7), they include the following: • market rigidities (such as relatively low mobility of factors); • geographical factors; • differences in the availability of resources; • education of management and training of labour; • regional economic structure (in some regions industries with ageing technologies, in others industries with modern technologies); • institutional and policy-related factors such as the centralisation of public institutions (Paris is an obvious example); and • national wage-setting in spite of differences in productivity and labour market conditions across different national regions. Regional disparities adopt numerous forms. They include not only differences in income levels per capita, rates of growth or rates of unemployment, economies of scale, externalities, output and consumption structures, productivity and access to public services, but also age structure, population density and pattern of migration. The controversy regarding the definition of the ‘regional problem’ remains. However, several factors together may provide an insight into the issue. These elements include: • Different regions may grow at uneven rates for a long period of time. This provokes government intervention to reduce that problem. • Intervention may aim to equalise consumption or GDP per capita among different regions. • A government may wish to create equal access of the population to an adequate level of public goods and services. • Public authorities may be interested in having a spatially stable distribution of economic activities and population in order to avoid negative externalities. Regional policy in the developed marked economies is basically aimed at influencing economic adjustment in the four (theoretical) types of regions:

Evolutionary Economic Geography

112

• Regions in which agriculture accounts for a relatively high share of production and employment. These are usually underdeveloped rural areas with relatively low levels of income, high levels of unemployment and poorly developed infrastructure. • Regions whose former prosperity was founded on industries that are now in decline, such as coal, steel, shipbuilding or textiles. These are the regions that failed to keep pace with changes in technology and were unable to withstand external competition (in some cases because of earlier excessive protection). In the case of recession, labour in these regions is the first to be made redundant. • Regions with a high concentration of manufacturing and resulting congestion and pollution problems. There are benefits from the joint use of goods and services available in these areas. However, regional policy may try to reduce the existing congestion (concentration) and pollution and/or prevent their further increase. • Frontier regions, which are far from the strongest areas (poles of growth) of a country’s or union’s economic activity. Jobs can be located in regions where there are people with suitable qualifications, training and experience. Therefore, regions with unskilled labour cannot be expected to attract any significant number of industries that use modern technologies. By definition, these industries require trained labour and educated management. The location of business near large and/or growing markets saves transport costs. If returns to scale were constant, as neoclassical equilibrium theory assumes, this could tend to equalise factor owners’ rewards in different regions. With economies of scale and other market imperfections (the situation with multiple equilibria), these tendencies may increase rather than decrease regional disequilibria. Convergence and divergence Convergence in the level of development among different regions is not a self-sustaining process. It is more desirable during an economic recovery and times of prosperity than during recession. Lagging regions (countries) usually have a higher proportion of both ‘sensitive industries’ and public enterprises than prosperous ones. Therefore, they may be hit harder by a recession and budgetary restrictions (no subsidies and public purchases) than other regions. There are two basic theoretical approaches regarding this issue: the convergence and divergence schools. The convergence or neoclassical school of thought regarding regional matters does not consider either market imperfections, such as economies of scale, sunk costs and externalities, or institutions that set the organisational environment. This school argues that the ‘regional problem’ and the spatial location of economic activity are not problems at all. Free markets (trade included) and unimpeded factor mobility would produce a smooth spatial dispersion of people, skills, technology and economic activity. Of course, in the long term this would equalise factor earnings and living standards in all regions and countries. Apart from transport costs, it should not matter where a tradable good or service is produced. Hence, the spatial location of output is no more than an operational detail. This is because transport, communication and other trade costs have been declining over time because of innovation and increases in productivity. The peripheral regions and countries are expected to gain from trade liberalisation and integration in

Theory

113

terms of an increased relocation of industries and trade. Ultimately, there would be a full equalisation in factor prices. The neoclassical one-sector growth models based on exogenous technological progress anticipate economic convergence among regions. This expectation based on comparative advantages depends on the starting conditions (including the endowment of factors). However, different regions may have different long-run growth rates because of a different starting situation. As technological progress in this model is exogenous, developing countries should grow faster than developed countries. If, however, technological progress is endogenous, then the convergence school does not give a clear prediction about growth rates and patterns. This liberal (non-interventionist) attitude in the field of regional policy in the market economies had a certain validity until the economic crisis of the 1930s. With free trade, perfect allocation and use of resources,122 intervention in regional policy is not necessary because markets clear and bring the economy to equilibrium. Such a convergence approach was not validated by time and experience. Even empirical evidence is not conclusive.123 In spite of expectations about convergence, even laissez-faire governments may consider it necessary and beneficial to intervene as the adjustment process may take too long to be politically acceptable. A successful convergence process, catching up through a spread of new technologies and certain imitation, requires the existence of new social and institutional abilities for the absorption and spread of novelties among the ‘followers’. ‘This explains why catching up and convergence are not automatic and common outcomes’ (Castellacci, 2007, p. 597). The necessary conditions for successful convergence and the catching-up process are different because of the specific situation in each country. However, these conditions include at least the following two, together with their combinations and mutual support: • Technological similarity: how deeply embedded are the old technologies? How difficult might it be to jump (transfer resources from old) into the new production concept? The size of the local market (relevant for the economies of scale), access to external markets, local tastes, growth and potential growth of the economy, flexibility in the labour force and laws, as well as the general macroeconomic and political situation also play important roles. • Social potentials: education level and flexibility; production, management and entrepreneurial skills and experience; links between firms and centres for the creation of new technologies and education; and supporting institutions in business and education. The evolutionary position is that national differences explain why certain countries that constantly change perform better in the catching-up process with leaders, while leaders, if they reinvent themselves (i.e. change), may keep their lead, at least for some time in the future. National institutions ought to support this evolutionary ever-changing and neverending process. There is no steady state. Evolutionary economic geography (spatial economics) emphasises potentially growing disparities, rather than convergence, among regions and countries. This divergence school of thought offers theories that refer to the growing discrepancy among regions. The focus is on a much nearer time horizon compared with the neoclassical equilibrium

Evolutionary Economic Geography

114

school. Explanations are based principally on consumers’ love for variety, economies of scale, high sunk costs, endogenous change in technology, spillovers, heterogeneous firms and imperfect competition. Together with the new geography of production and the new trade theory, divergence theories consider market imperfections and institutions that set rules which streamline liberties and behaviour.124 This moves theoretical concepts closer to reality. The list of new evolutionary assumptions includes economies of scale, high fixed costs, trade and adjustment costs, entry and exit barriers,125 constrained mobility of factors, transport costs, externalities, multiple equilibria, path dependence, the multiplier effect, cumulative causation, acceleration, non-ergodicity and inflexibility. If applied to intermediate production, these forces can combine and influence one another in such a complex way that their behaviour becomes fickle. They can produce most unpredictable outcomes regarding locations of firms and industries. In many cases, history, chance and competition among expectations will set the final outcome. An evolutionary transformation of a background region may be observed in the case of Bavaria. The creation of the German Zollverein integrated thirty-nine German states in 1834. They were of different sizes and they were at diverse levels of development. Bavaria’s economy was principally based on agriculture. Other partner states in the Zollverein were lagging behind the British level of industrial development, but were far ahead of Bavaria. Towards the end of the nineteenth century, Bavaria attracted lowtechnology manufacturing that profited from the local low wages. However, during the early twentieth century, when the second industrial revolution took place (principally the production of vehicles), Bavaria became the heart of German car manufacturing. Bavaria did not have to struggle with the decline of heavy industry, unlike the Ruhr in northwestern Germany which still bears the ‘scars’ of the restructuring process. Being poor and disadvantaged may be transformed into an advantage, as this example shows (the eastern part of Germany, after reunification, is not yet following this Bavarian example). Bavaria’s current high level of development should not be attributed exclusively to German integration. Investments in the education of engineers, researchers and managers, training of labour, capable public administrators and private managers, and entrepreneurship are the principal explanations for this development. However, none of these factors would have brought about the current scale of success had Bavaria stayed out of the Zollverein and, later, the EU, and remained a relatively small and poor farming-based economy. 9 History and competition among expectations Background As already noted, the Heckscher–Ohlin theory cannot explain why industries locate in regions with a high mobility of factors (US and China) or in countries with a broadly similar endowment of factors (France and Germany). Patterns of regional specialisation and location of firms and industries are often created by a historical accident. In 1894, Karl Marx wrote about capitalist production in Das Capital, Vol. III, Ch. 48, III: But in reality this sphere is the sphere of competition, which, considered in each individual case, is dominated by chance; where, then, the inner law, which prevails in

Theory

115

these accidents and regulates them, is only visible when these accidents are grouped together in large numbers, where it remains, therefore, invisible and unintelligible to the individual agents in production.126 Ohlin noted that ‘Chance plays a significant part in determining the localisation of industry … Adifferent distribution of inventions would have caused a different localisation’ (Ohlin, 1933, p. 137). More recently, Krugman wrote: ‘I at least am convinced that there is a strong arbitrary, accidental component to international specialization; but not everyone agrees, and the limitations of the data make a decisive test difficult’ (Krugman, 1992, p. 9). As already noted, economics is a field known to be especially slow to open itself to new ideas that are different from the ones accepted by the mainstream. In the evolutionary world ‘equilibrium, if it exists at all, is not unique and historical accidents matter’ (Lipsey et al., 2005, p. 41). In any case, evolutionary dynamics regarding the spatial location of firms in the future is based to a large extent on chance (and human agency). The greatest research and policy difficulty and challenge is that the evolutionary course is rather unpredictable (at the start). However, this ‘primary chaos’ regarding the location of firms is settled through evolutionary mechanisms that ‘normalise’ locationrelated behaviour of firms into a configuration of (initial) spatial concentration. Without chaos, there may perhaps be no progress; without order, there may perhaps be no survival. Chaos and order may be two sides of the same evolutionary development coin. When there are multiple equilibria, various spots for the location of business are substitutable ex ante (‘putty’). After the investment is made, these locations are not easily substitutable for quite some time (‘clay’). Once a business is established at a specific geographical location, it is then ‘locked in’ through learning, path dependence, sunk costs, circular and cumulative causation127 effects (for examples of this putty–clay locational model, see Annex II). In this sense ‘history matters in a way that it does not in neo-classical theory’ (Eaton and Lipsey, 1997, p. xxv). ‘The firm’s present position, and alternatives for the future, depend on its past history and its present location. For it, history matters’ (Lipsey et al., 2005, p. 54; original emphasis). Two questions are relevant here: • Are there inherent differences among locations that create predestination for certain activities? • How can a small historical accident, a chance (something that is beyond the prior knowledge or control of an investor), alter the economic fate of an industry, region or country? On the one hand, there is a belief that the choice about the location of firms and industries is basically resolved by history. Past events set preconditions that move the economy from one steady state to another. This reasoning, found in the traditional literature, argues that history matters because of increasing returns and lumpiness (inseparability) and activity-specific knowledge and capital goods (Eaton and Lipsey, 1997, pp. x–xi). Various locations (regions, cities, clusters) may relentlessly compete to attract the same capital, firms, industries and labour, as well as possibly struggling to penetrate the same markets.128 Some locations are more successful in attracting business and people than others. They are more ‘competitive’,but this local ability evolves and changes over time.

Evolutionary Economic Geography

116

The ‘lock-in idea’ is also behind promotional campaigns or ‘location tournaments’ (David, 1984) among countries and/or regions to attract international footloose capital (TNCs). The impression is that once an activity starts in a place, it is self-perpetuating and related businesses gravitate towards it. However, this ‘self-reinforcing aspect of foreign investment begins to operate only after a certain development threshold has been reached’ (Wheeler and Mody, 1992, p. 71). That is, spillovers on the local economy are not a direct and automatic consequence of the location of a TNC. Gains that come from these externalities can be achieved only if the local firms have taken on, or can be motivated to take on, new (foreign) technologies and skills. The structure of the local economy, attitude of the private sector and policies of the government that promote change play a key role. In addition, there is evidence in the case of the US foreign investors that ‘past investment in the country was a strong predictor of new investment. That persistence was attributed to the favourable effects of agglomeration’ (Mody and Srinivasan, 1998, pp. 780–1). History mattered in this case. Regions may compete among each other, but their competition is different from the one that exists among firms. While firms may appear and disappear from the market or move from one region to another or be absorbed by other firms, regions are there to stay in defined geographical space. A number of regional assets are not tradable. Regions, of course, differ from one another in the endowment of natural and changing created assets, wealth, or growth rates, but they do compete. They compete with their local institutions (rules, laws, regulations, property rights, habits, practices, customs and routines) to attract investment and talented people. These institutions set the pace of transformation and upgrading of regional economies. Institutions that have built-in flexibility and responsiveness to adapt to novelty and change offer higher chances for a regional economic success over a longer period of time. Regions also compete to serve similar markets, as is the case with Antwerp, Rotterdam and Hamburg which compete in the transport business. However, regions may not disappear or move elsewhere, as firms can and do. There is, on the one hand, a proposal that history sets the equilibrium location of firms and industries, while on the other hand is a view that the choice of equilibrium is determined by expectations. This observation is based on the belief ‘that there is a decisive element of self-fulfilling prophecy’ (Krugman, 1991a, p. 652). Let us consider both these views in turn. Historical lock-in effect The non-linear probability theory can predict with some certitude the behaviour of systems subject to increasing returns. Suppose that balls of different colours are added on a table. The probability that the next ball will have a specific colour depends on the current proportion of colours of balls. Increasing returns occur when a red ball is more likely to be added when there is already a high proportion of red balls (Arthur, 1990a, p. 98). Equilibrium depends on the initial point and later arrivals. Hence, history as a series of (random) arrivals sets the final result. The national rate of growth of capital stock (without FDI and foreign loans) depends on home savings and investment.129 Suppose now that one region/country initially accumulates more capital than the other. In the following period both regions grow, but

Theory

117

the one with more capital grows faster than the one with less capital. As manufacturing capital grows, the relative prices of manufacturing goods fall. After a certain period of time, there is a point where the lagging region’s industry cannot compete internationally and it begins to shrink. Once this process begins, the new theory of trade and strategic industrial policy suggests that nothing can stop agglomeration for a long time. Economies of scale may drive prices down in the capital-abundant region and at the same time the lagging region’s manufacturing industry disappears. In this model, relatively small beginnings can have large and irreversible final consequences for the manufacturing structure of a country, its trade and the competitiveness of its output (Krugman, 1990a, pp. 99–100). The dynamics of capital accumulation ensure that the region that starts with a higher capital stock (than the other regions) ends up with a dominant industrial position. If this is reinforced by a learning process and cumulative causation (strong internal production links where extension of one activity increases the profitability of others), then the existing pattern of comparative advantage is reinforced over time even if the overall structure of the economy has changed. This process-dependent development adds new layers of firms and industries to the inherited production structure. If output is concentrated within a relatively small area, firms can benefit from economies of scale and linkages (growth of one activity increases the profitability of others). If this area is close to a larger market, there are additional benefits in the form of lower trade (including transport) costs. Hence, in this example, the current state of the economy influences, even determines, its future shape. The investment decisions and trade policy of a country in the current period will have a lasting impact on the shape and direction of the national economy in future periods. For example, at the end of the nineteenth century, Argentina and Sweden were relatively comparable low-income farmingbased economies. At about the same time, Argentina invested in the education of lawyers and priests, while Sweden invested in the education of engineers. The impact of such choices (coupled with an important problem of weak governance in Argentina) on the material standard of living of the two countries is obvious. Lipsey et al. (2005, p. 414) contributed to the debate about the crucial importance of educational policy: For example, people may be uncertain as to what changes are needed in the educational system. What is happening to the required skill mix and what kind of new skills are needed and what type of old skills will become redundant? Such debates raged towards the end of the nineteenth century with Britain and Germany taking very different decisions. The Germans decided to establish trade schools where those who did not go on to formal higher education received an excellent training in some trade. The British concentrated their resources on higher education for a small elite and devoted fewer resources for training the working class in the skills needed in the new factories. This gave Germany a comparative advantage in the production of state-of-the-art but standard consumer durables – an advantage that persisted for over 100 years.

Evolutionary Economic Geography

118

Once the structure of an economy becomes unsustainable, there are certain critical branching points (bifurcations) at which the qualitative behaviour of the economy changes (off-path change). New production geography either evolves or is triggered. The long-term and dynamic picture of evolution of a relatively open market economy is not like a ball that smoothly rolls down a slope in a bowl with a predictable direction and speed (towards a new and pre-determined steady state). It is more like a cube or, better, a polyhedron that ‘rolls’down a slope: full of unexpected developments, novelties, turbulences, accommodating adaptations to a dangerous environment, amplifications, ‘births and deaths’, replications and sudden turns in uncharted directions. Quite a bumpy ride! Hence, the location strategy of a firm is like a lizard’s separable tail: predators (competitors) may have it, but the animal without it may continue its life without any major disruption for some time in the future.130 It is similar in the living world too, but contrary to Darwin’s opinion about biological evolution as a continuous smooth selection process. Fossil findings prove that this smooth process was not something that happened in reality. Consider the following opinion by Boulding (1991, pp. 13 and 15): Evolution is not a continuous process as Darwin thought, but is punctuated by catastrophes and other improbable events. The transition from one geological age to the next always seems to represent some sort of catastrophe, which led to a large extinction of species and therefore the opening up of a variety of new empty niches, some of which were filled, and, of course, others not. Thus, if it had not been for the catastrophe, whatever it was, that exterminated dinosaurs, it is very doubtful whether the human race would be here today … The mammals at that time seem to have been rather small … On a smaller scale, the defeat of France by Germany in 1871 seems to have led to Paris becoming the cultural capital of the world for two generations, while the great cultural age of the German people seemed to be over. Thus history-dependent development can follow different and complex paths, it has many bifurcation points, hence it is unpredictable both in theory and in practice. A long period of relative stability is broken up by an event of change (equilibrium, if it exists, is punctuated, i.e. it is quasi-equilibrium).131 Smooth and calm periods, during which one may observe growth and organisation in a system, alternate with stormy and rough times.132 A temporary stability of the system is only relative and stratified. The principal source of changes in this indeterminate133 business world is the entrepreneur who creates and tries out novelties. Many of these novelties have little to do with changes in prices or demand, hence the neoclassical theory is a rather weak workhorse for this type of analysis. The neoclassical equilibrium model deals mainly with prices and their changes. So new technologies are often regarded as the fall in price of the service provided by the old technology. Old things become cheaper. This may be correct, but it covers only a part of the story. Another and perhaps more important part is that a new technology such as semiconductors enables the production of entirely new goods and services.

Theory

119

There are many people who have difficulties understanding how an economy works. Even professional economists have serious differences of opinion. Hence, a number of lay people may be prone to herding behaviour. They may be tempted to follow forecasters and opinion-makers who ‘know the future’. These may lead the ‘herd’ of investors and consumers to a selfreinforcing optimistic and cheerful path, but they may also easily lead them into a self-fulfilling investment and spending abyss. Changes, in particular in technology, have been (many would argue) the principal sources of growth since the Middle Ages. For example, when water-wheels were the principal source of power, the location of many production lines was along rivers and was spatially limited by the length of the drive shaft. The steam engine permitted a huge spatial reallocation of production towards cities but, here again, the length of the shaft restricted the size of the plant. Electricity permitted flexible location of production and the creation of a number of new goods and services. Can you imagine a steam-powered mobile phone? How handy and useful would it be? Changes also affected cities. Nearly all big cities are ports. However, it is a long time since New York or Boston ceased to be primarily harbour cities. Even though Britain’s days as a maritime power are long gone, London keeps strongly its place as one of the primary international centres for marine insurance and ship-broking. There are other examples of punctuated equilibrium. In biology, for instance, certain species such as cockroaches,134 crocodiles, crabs, turtles and monotreme animals remain static for a very long period of time. Many others appear, disappear, evolve and change without a predefined stopping point as long as there is an inflow of energy from the sun or, perhaps, from some other source. There are, however, crucial, relatively brief and unpredictable moments (bifurcations) when new species arrive and old ones disappear. The change is usually from ‘outside’. These include a change in climate 64 million years ago caused by a meteor colliding with the earth,135 which provoked enormous volcanic activity and fires throughout the earth, with a 300m-high tsunami; water and the atmosphere were polluted, while smoke and dust prevented the penetration of heat from the sun. Hence a very large part of life on earth was frozen, burned, starved or suffocated over many months.136 This change in climate caused the extinction of the dinosaurs.137 External temperature decides the sex of reptiles: above a certain temperature, a male reptile shows up from an egg, while below it a female is the outcome. The change in the climate might have determined that certain reptiles were all of the same sex, so that the biological continuation of their species was impossible. While the external shock story may probably hold in the case of dinosaurs, there are fossil records that some other species vanished without external shocks and probably owing to internal evolutionary dynamics. Evolution in the natural world has some similarities with particular social relations and with business. First, certain social progress can be annihilated. The case in question is when west European civilisation almost totally collapsed when the (nomadic) barbarians smashed almost everything they found. The intruding barbarians138 were predominantly but not exclusively Goths, Gepids and Huns. They poured into and ransacked the Roman Europe (composed of western, central and southern Europe) from northern Europe and the East, while the Arabs attacked from the south-west (the Battle of Poitiers of 732). In economic terms this was an external shock like the impact of the meteor that uprooted the dinosaurs. These nomadic tribes that turned western Europe upside down carried with

Evolutionary Economic Geography

120

them infectious diseases such as plague. This was coupled with their high charge of destructive energy towards everything that they could not understand. And there was a lot of it. The inherited Roman culture that also influenced lands long abandoned by the Roman legions, particularly regarding education (intergenerational transmission of knowledge), was deeply damaged in the period from the fourth to the ninth centuries AD. At the same time, the Greek-speaking Byzantine Empire (330–1453) was thriving, and we owe it much for the protection of western civilisation, including Roman law. It could not, however, come to the rescue of the western ‘brethren’ because it had its ‘own’ barbarians (Muslims) who were intruding from the East. Civilisation in what we know now as western Europe collapsed for some two centuries. This situation lasted in western Europe until the reign of Charlemagne (742–814). The social evolution of the old Franks on two branches, i.e. on what we now know as Germany and France, is intriguing, but beyond the scope of our current subject. (Which language did Charlemagne speak? Do the French and the Germans agree about this?) Many social constructions can be undone, and sometimes quite fast. Several states in Europe dissolved, for instance, during the 1990s. The evolution of another social construction in Europe needs to be observed closely: the point in question is the European Monetary System and the euro. Apart from wars, this was perhaps the biggest political event in Europe during the twentieth century, besides the October Revolution in 1917. Europe is, however, full of examples of monetary unions that collapsed even though they were based on gold. The current one is based on a credible promise – until now. It remains to be seen how the euro and its members weather the storm. The first one will come quite soon, about 2015, when huge pressure on publicly financed pensions is going to put a strain on the public budget deficit that has a cap of 3 per cent of GDP. The babyboom generation will retire en masse from 2010 and demand its state-guaranteed pensions. Second, in business, changes are rather obvious. For instance, the computer-related industry and market are at present dominated not by the ‘old’ and transformed giant IBM (established in 1888), but by a relatively recently created giant, Microsoft (founded in 1974). Others, such as Google (of 1998), are rising fast. More than a century ago, Alfred Marshall spelled out the idea of ‘backward-looking dynamics’ (or ‘external economies’ in modern jargon). In his analysis, factors of production are moving towards those industries in which they earn the highest current rate of return. If there are several meaningful equilibria in which the returns would be equalised, then the initial conditions determine the final location-related outcome. History matters, together with factor endowment, tastes and technology (Krugman, 1991a, pp. 653–4). Marshall also described the concentration of specialised industries in particular localities in the following way: When an industry has once chosen a locality for itself, it is likely to stay there long: so great are the advantages which people following the same skilled trade get from near neighbourhood to one another … if one man starts a new idea it is taken up by others and combined with suggestions of their own; and thus becomes the source of yet more new ideas. (Marshall, 1890, p. 332)

Theory

121

Contemporary jargon refers to these processes as ‘externalities of innovation’.139 Somewhat similar ‘first-mover advantages’ (to use the term from the analysis of competition) may be found in biology, as many elements of the past can be found long after their functions have been lost. For example, [E]mbryonic birds and mammals still have gill arches, which have been useless for 400 million years. Why are the vestiges not eliminated by natural selection? The usual answer is that baupläne and the vestiges are developed early in the embryo and hence are more difficult to modify than features that develop later. The genetic programs controlling embryonic development were formed in the early days and have been frozen ever since. (Auyang, 1998, p. 195) A similar example can be found in the appendix in the human body. However, a question immediately comes to mind: why cannot newly created adaptive features also be frozen? Why do they evolve further? One of the chief causes for the localisation of industries can be found in the physical conditions of the area. Metals producers located either near mines or close to sources of abundant and cheap energy. First-rate grit (for grindstones) was found near Sheffield, England, therefore it also became the place for the manufacture of cutlery. Another specific reason for the localisation of production has been the presence of a royal court. In the Middle Ages many rulers in Europe were constantly changing their place of residence (partly for sanitary reasons). They frequently invited artisans from distant places and settled them in a group near the royal court. Once the court left, if the town survived, in many cases it continued the development of a specialised industry (Marshall, 1890, pp. 329–30). Thus, cities often began to develop around royal courts and bishops’ sees (where there was a concentration of consumers).140 Later on, the administrative, defence and educational dimensions of towns and cities reinforced their function as consumers, but industrial production also started evolving in the cities. Today, many cities produce more value-added than they consume. Recent research has provided some evidence that general government policy can indeed influence the location of firms and industries. From 1947, a number of manufacturing industries in the US migrated to the southern states because of the wish to escape the unions. This was, however, only one of the reasons for the change in the location of certain businesses in the US. Others included innovation and change in transport (substitution of truck haulage for rail freight, which contributed to a spread in production), as well as the advent of air-conditioning that ‘made the climate in the South relatively more attractive than the climate in the North’ (Holmes, 1998, p. 670). The Southern Sunbelt (sun and sprawl) was quite attractive to some compared with the Northern Snowbelt (cold and dense) residence patterns and lifestyle. Destler (2005, pp. 282–6 and 305–6) reported interesting observations regarding the background ideology in US policy-making. The general past pattern in the process was that most Democrats were on the ‘left’ side, while most Republicans were on the ‘right’ side of the policy-making equation. However, there was a certain and obvious overlap between the two sides in the period 1969–70 (Figure 2.11, upper panel). Three decades

Evolutionary Economic Geography

122

later (1999–2000), there was almost no common ground (Figure 2.11, lower panel). What is the reason for this shift? 1 One possibility is that the Americans became more polarised. The presidential elections in 2000 and 2004 both gave a very narrow majority to the winner. One of the reasons for the 1969–70 ideological overlap between Democrats and Republicans in the Congress was the presence of many southern Democrats from the states where that party had had dominance since the Civil War. ‘They were typically more conservative than many Northern Republicans. But the enfranchisement of blacks through a movement in which national Democrats were particularly prominent drove conservative Southern Democrats (and their children) to the Republican side’ (Destler, 2005, p. 285). 2 Another possibility is redistricting following the Supreme Court’s ruling that districts within a state have to have substantially equal populations. District lines need to be redrawn after each decennial census. The politicians are redrawing these districts with an eye to their own survival. This created more ‘safe’ districts with both Republican and Democratic majorities, and fewer marginal and ‘swing’ areas where candidates fight for the centre (median voter).

Figure 2.11 Ideological distribution of the parties in the US House, 1969–70 and 1999– 2000. Note: Ideological scores are first-dimension DW-NOMINATE scores. The figure shows density plot of House members’ ideological score by party. Source: S. Binder, 2003, pp. 24–5 3 The third reason or possibility, and perhaps the most interesting for our evolutionary analysis, is a change in technology, in particular jet air travel. At times when travel

Theory

123

back to their states and districts was a timeconsuming and exhausting venture, members of the Congress were living relatively close to each other in one area of Washington DC through much of the year (weekends included). They were spatially close to each other. They were also socialising with one another and they could often find (for instance, in a golf club) persons from the other party who they preferred to some from their own. Once jet planes enabled legislators to return ‘home’ for the weekend, they used this facility to solidify their local political support. They were pushing for the local interests in Washington DC between Tuesdays and Thursdays. Life in Washington became a stopover and a movement from one meeting to another. There remained little time to rest and cultivate personal relations with others and to take a wider look at issues, i.e. to create centrist coalitions across party lines. This has important implications for trade and industrial policy in the US, as well as for the location of firms and industries, as ‘political victory margins’ became unpredictable and razor-thin. This makes trade policy hostage to any protectionist interests that hold the decisive, marginal votes. Whether the product be steel, textiles, sugar, or orange juice, there is a bloc of House members ever ready to abandon ship if the administration seeks to negotiate, for example, a fullledged Free Trade Area of the Americas, or thoroughgoing agricultural reforms in the Doha Round … So stay tuned. (Destler, 2005, p. 306) These razor-thin decisions and marginal swing votes sometimes have a profound international impact on the spatial location and expansion or reduction in production. Firms may also use their leverage to reduce costs of production by threatening the local labour unions that they might move the business elsewhere. Siemens, for example, planned to move two plants from Germany to Hungary in 2004. In return for it not doing so, the employees agreed to work forty hours a week (instead of the customary thirtyfive) with no extra pay.141 Siemens was not the only firm to try to fight the inflexible German labour market, cut costs and enhance the competitiveness of its output. At the same time DaimlerChrysler also threatened to move some production from Germany to South Africa. In certain cases, offshoring of certain jobs (presumably low-skilled ones) from the home country abroad may protect other jobs at the home location. This is the case when the alternative is that all jobs from a firm or industry go abroad. Firms cluster together in order to benefit, among other reasons, from the availability of a close network of suppliers. They usually, but not necessarily, cluster in locations with a large local demand. This demand will be large in the areas where most producers choose to locate (a process of circular interdependence or cumulative causation). ‘There is a degree of indeterminacy in the location of activities – firms locate where they do because of the presence of other firms, not because of underlying characteristics of the location’ (Venables, 1996a, p. 57). For example, ‘Japanese business firms operating in Germany have an unexplainable attraction to Düsseldorf rather than Frankfurt’ (Beckmann, 1999, p. 61). In addition, there are about 600 tanneries in Arzignano (Vicenza, Italy), most of which employ just a few dozen workers. However, the region produces 40 per cent of Europe’s leather supply. The area around Vicenza not only soaks, dyes, stretches, stamps,

Evolutionary Economic Geography

124

cuts and ships material used for Gucci handbags, Louis Vuitton luggage, Nike sneakers and BMW car seats, but also produces gold chains, clothing and machine tools, many of them for export.142 The major problem for a small or a medium-sized enterprise is often not in its ‘smallness’, but rather in its isolation, which is eased in a cluster. In an indeterminate situation regarding the spatial location of firms when there are multiple equilibria, should one search elsewhere for answers to the questions on the geography of production? Should it be outside standard economics? Should it be in nonlinear dynamics (popularly known as ‘chaos theory’), which examines unstable behaviour with multiple dynamic equilibria, lock-in effects, bifurcations, off-path changes and extreme sensitivity to initial conditions; or in self-organisation (spontaneous appearance of order after oscillations)143 so common in complex systems; or in the new evolutionary economics that has open frontiers, maintains an interdisciplinary dialogue and considers the role of history, technical change, institutions and development of human capital; or in evolutionary biology (evolution and hybridisation);144 or elsewhere? Organisms are vulnerable, but they are not passive; they dig holes, build dams, modify the environment, and create their own niches of survival. Thus organisms and their environment interact, and together they form an ecological system that is better treated as a whole. (Auyang, 1998, p. 61) Biological and economic evolution have in common similarities such as creation, differentiation, selection (including mistakes), adaptation, innovation, retention, amplification, replication and a constant search for creative– destructive optimal solutions; humans, however, respond and create novelties intentionally and in certain cases react collectively. This purposive action makes the evolutionary process faster than is the case in biology. Similarly, the effects of externalisation are also pronounced in economics. Consumers’ tastes and institutions, both public and private, interact, evolve and alter over time. For example, consumers in similar situations make different choices (recall the diversity of breakfast cereals, painkillers, cigarettes, chocolate bars, T-shirts, shoes, cars, skis, perfumes, cameras or bicycles). Hence, complex systems keep on moving from one pseudo-stable situation to another. Behavioural patterns of living organisms may be quite flexible and accommodating. Organisms may switch from one source of food to another or adapt to a different climate (dry, wet, hot, cold … ) or form linkages and networks (flock together to protect heat or fly with a reduced friction). But an organism can also mutate. When continuously exposed to a certain aggressive force (heat, humidity, pressure, light, predators … ), if this organism survives, its offspring and the following generations may transform in such a way that this ‘aggression’ is converted into a survival advantage for the organism. Therefore, history matters in the evolutionary model, but only up to a point. Competition among expectations in the business world, to which we now turn, also plays a role. Competition among expectations Resources move gradually from one location and/or industry to another in response to differences in current earnings. If this shift is only gradual, then there must be certain

Theory

125

barriers that increase the cost of the move. If there are costs, then the resource owners will be interested in expected returns in the future, rather than only in current returns. However, future returns also depend on the decisions of owners of other factors and their expectations about future earnings.145 For instance, when a new technology is introduced, multiple equilibria cause competition among expectations of potential action from competitors. In this model, expectations (rather than history) determine the future shape of the economy and the location of production. A river may, over centuries, enrich its bed and reinforce the work of natural forces. However, a strong earthquake or a tsunami (a bifurcation point) may instantaneously ruin such a long history and create a new geographical shape. Applying this to economics, such was the case with the mechanical and, later, electrical cash registers that were made obsolete by digital ones. Subsequently, optical scanners that read bar codes replaced digital cash registers. These changes were all based on totally different technologies. A coalmining town may cease to exist after the arrival of affordable liquid fuels (innovation-driven shock). Photo cameras based on celluloid films were ‘attacked’ over decades with various alternatives to the standard 35mm-film format. They failed. However, in about 2002 digital cameras based on totally different technology dealt a decisive blow to the widespread use of 35mmfilm cameras. Similarly, the demand for rail transport began to fade away once cars and lorries appeared after the arrival of new transport technologies. This was reinforced by the arrival of air travel (another new transport technology), particularly cheap air travel. The appearance of mobile phones had a strong and negative impact on the expansion or even longer-term existence of public phones. The underlining competitive advantage of a firm or a location may remain intact; however, the markets, consumer preferences, technology and expectations may evolve in a way that renders these original advantages less and less relevant. The British banks expanded over the nineteenth century when they mobilised the small savings of individuals to lend those funds to expanding firms. Bank managers’ knowledge of the local market was crucial in this business, at that time and for a good part of the twentieth century. Great banking halls and ostentatious head offices gave the impression of permanence to outsiders. However, the growing securities146 market breached the clear link between deposit-taking and lending. In addition, credit card business evolved to become the source of banking profits for the foreseeable future (Kay, 2006, pp. 79 and 83). Those who founded their banking business on the basis of their past achievements were ‘overtaken’ by those who expected and ‘bet’ on a change i.e. these entered the securities and credit card business. Other examples may be found in the almost overnight disappearance of the market for movie cameras after the appearance of video cameras; the vanishing of 35mm-film cameras and film-makers after the arrival of digital cameras; the redressing of the market for mechanical watches after the invention of digital and quartz ones; a shift from dotmatrix to bubble-jet and laser printers; or fibre-optics that evolved independently of the telecommunication technology. Those who expected and invested in the winning technology (‘got it right’) kept the production in their old locations. Alternatively, if the winners relocated to other locations, such as China, they kept the ownership and control of the production process in their head office and let others elsewhere produce and potentially pollute their local environment and/or increase congestion there.

Evolutionary Economic Geography

126

A good encyclopaedia has always been a must if parents want their children to be well educated. The door-to-door salesman selling the Encyclopaedia Britannica has been replaced by Microsoft’s cheap CD-ROM encyclopaedia Encarta, first produced in 1993. Wikipedia, the Internet encyclopaedia, is finding its place with many users. In addition, Amazon is taking business from real-world retail booksellers. Faxes and e-mails damaged and significantly diminished the rents available to postal services. Google enters into a part of Microsoft’s market, while elsewhere YouTube is finding strong inroads as a preferred Internet video search engine. For now. Just as you climb to the top within an industry and become complacent about your geographical and hierarchical location of business because you ‘got it right’, Annie’s hand may shiver (you may still remember Annie Oakley, shooting at a cigarette held by the future Kaiser Wilhelm II) … and out of thin air somewhere someone else may appear who is fitter than you are in the evolutionary world. The shelf-life of each advantage in the evolutionary world is ephemeral and finite. If this advantage works well for you, you have to keep in mind that it is only an experiment, more or less successful in the fight against entropy. Hence, there may always be a pressure and need to try out something new or better. One of the most powerful companies ever, Westinghouse Electric, met its demise in 1999; earlier, the British East India Company dissolved in 1874 after a long and splendid business life. Evolution changed the economic environment while these companies’ adjustments were insufficient to fit into the newly evolved situation. These business giants were not adjustable enough, or not innovative enough, or not fast enough in comparison with their rivals. The Red Queen, from Lewis Carroll’s Through the Looking-Glass,147 said, ‘Now, HERE, you see, it takes all the running YOU can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!’ There is no such thing as winning the Red Queen race in the evolutionary world in the longer run. If there is any suggestion to a businessman, it is to run, run always, do it faster than the others who are competing, as there is no finish line. ‘In a flat world … you can’t hide’ (Friedman, 2006, p. 185). In a frightening flat world one may say to the children: when I was growing up, my parents used to say to me, ‘Tom, finish your dinner – people in China and India are starving.’ My advice to you is: ‘Girls, finish your homework – people in China and India are starving for your jobs’…In a flat world there is no such a thing as an American job. There is just a job, and in more cases than ever before it will go to the best, smartest, more productive or cheapest worker – wherever he or she resides. (Friedman, 2006, p. 277) The above hypothesis emphasises insecurity. It also reflects the possibility that workers will increasingly be paid what they are worth on the international market. According to this view, someone or something such as a new technology will come from somewhere and take your job or business. This may be added to another general hypothesis,148 concerning why certain countries keep on growing wealthier than others. This emphasis on insecurity should not come as a surprise because many economists worry for a living.

Theory

127

In addition, the flat world hypothesis is in sharp contrast to what we observe: a continuous urbanisation and growth of mega-cities. The concern about preserving achievements pushes certain societies to continuously reinvent themselves. This is done through education and innovation. The US during the twentieth century presents a clear example. In addition, the general insecurity and death of distance hypothesis ought to be countered with the view that, instead of staying in the same place, most countries grow, though not at the same rate. One also has to recall that only a limited share of the domestic workforce is in direct completion with unskilled workers in China or India. Most service workers do not fall into this group. In any case, continuous education just provides the chance of employability rather than employment. Bill Gates, a college drop-out, created Microsoft, the world’s most powerful software company, without too much formal education. Modestly educated Soichiro Honda did a similar thing regarding motor bikes. However, these are extreme and rare aberrations rather than a general rule to follow. The Swiss may have, on average, some of the highest wages in the world. This is particularly obvious among the medium- and lower-educated segment of the labour force. Why is it so? In general terms, the Swiss benefitfroma mostly tacit, but rather detailed and well-organised society, institutions and division of labour. Everyone’s capacities and potentials may be used almost to the full. The economy is so open for trade and investment as to permit a rather open exchange with the world (it is not producing for local subsistence). The lower-level-educated segment of the labour force (highly paid in terms of almost any current international comparison) benefits from rents (supernormal profits) earned by Swiss corporations such as Novartis or Nestlé, or by Swiss banks which have the highest world reputation for prudence and discretion, or from specialist and innovative chemical, engineering or other business skills. International demand for these Swiss goods and services is high, strong and continuous. Earnings (high) in these Swiss corporations and industries set a reference for earnings in other parts of the Swiss economy. This may be necessary and fair if one wants to have both bakers and car mechanics, as well as pharmaceutical engineers, in Switzerland. This is why the Swiss primary school teachers, butchers and bakers, for instance, are among the best-paid in the world within their profession (Kay, 2004, pp. 303–4). The poor countries are not poor because the rich countries are developed and rich. However, the poor countries could be even poorer if the developed countries were poorer. Natural resources play some part in the wellbeing of a country, but the most important resource for prosperity is the human capital (health, knowledge and skills). A new product (or service) has to be vastly superior to a well-established and embedded product, be it in convenience of use, price, size, shape and/or speed, in order to dislodge the entrenched product or service from the market. Innovation is threatening the existing firms from every side; hence the lock-in effect is not for ever. Being a top dog at a given time t is no guarantee at all that this position will be kept at time t+1 and beyond. Achievement and continuous safeguarding of top position and competence is linked in an evolutionary framework with sustained learning, adjustment, innovation and change. High business concentration and reliance of a cluster or a town on a single line of production can leave it highly vulnerable to external shifts in demand and technology. Rational entrepreneurs should profit from the ‘natural course’ of business evolution. Being a key player and staying on top is not a state, but a process.149

Evolutionary Economic Geography

128

The evolutionary shift in the spatial location of production (of knowledge, in this case) may be found in the case of the Cavendish Laboratory at Cambridge, England. This was the place for inventions in atomic physics during the first three decades of the twentieth century. Whatever one wanted to know about this field, be it theory, mathematics, equipment or techniques, one could find it there. Arthur (2007, p. 285) wrote: But it does not last forever. Sooner or later knowings become mere knowledge and leak out; phenomena and principles become understood and are mined elsewhere; craft becomes codified into instruction and appears in textbooks; and people leave over time and spread their expertise geographically. Leadership in technology needs to be constantly renewed. A clear example of how competition among expectations shapes the national geography of future production can be found if one considers the case of debt in the developing countries.150 In the 1980s, these countries found themselves unable to repay huge debts owed to the big international banks. Had they all gone into default together, they would have threatened the global economic system. US Secretary of the Treasury Nicholas Brady developed a scheme for refinancing this debt, with the IMF playing the role of the financial policeman of developing countries by linking lines of credit to severe austerity programmes. This was taken to be a victory for the IMF and international institutions. It could have been, but only in the short run. During the 1970s, the world saw a massive increase in the price of oil and other internationally traded commodities. The conventional wisdom (expectation) was that commodity prices would only go higher. The Club of Rome and other similar observers of history (and future) pointed out that the world was running out of scarce resources (this was their expectation). The world was compared to a space ship: resources were being exhausted by growing populations and intensified industrial use. It followed that the price of commodities such as oil, copper and wheat could only increase (a similar situation evolved in the market in 2008). If this were the case, a rational expectation was that the best business area to invest in was commodities. Correct? The primary producers of commodities were developing countries that lacked manufacturing capability but controlled natural resources so much demanded by the industrial world. Any sane investor in the 1970s knew that investing in industries that purchased raw materials was foolish, while investing in production of raw materials was smart. So everyone (or very many), particularly the international banking community and the World Bank, began investing billions of dollars in ventures designed to produce raw materials in developing countries from Mexico to Nigeria to the Philippines. As the crucial commodity, namely oil, was expected to cost 40, 50 or even 100 dollars a barrel in the future, the cost of production was not a critical element in decision-making about investment. The price of commodities was going up and it was important to get into this business early. All of the technocrats simply knew this, and the entire international economic system became skewed towards investing and lending to commodity producers in the developing countries. As a consequence of the above process, the inevitable happened. It turned out that while the world may have a finite amount of oil, iron or copper, there were still huge

Theory

129

untapped reserves and progress to be made in technology (which saves on the use of resources). In addition, there was a shift in the production structure in the developed world from a metal-based bashing production towards services. Together with the development of substitutes, this shift towards services curbed the demand for commodities. When the megaprojects in the developing world began to operate, the price of commodities collapsed. When prices fell below the cost of production, projects became unsustainable and bankruptcy ensued. The outcome was the debt crisis. Nicholas Brady and the IMF stepped in to rectify the trouble. The debt crisis, arising from a belief in commodity scarcity (expectation), led to an avalanche of investment decisions that has left a legacy of misery. The simplistic and linear projection of the future in which commodity producers dominated industrial commodity consumers was rendered false by the collapse in commodity prices and made irrelevant by another phenomenon in the early 1980s. This new phenomenon is related to Microsoft and similar companies that emerged from nowhere.151 Microsoft, and the endless number of other software and related companies that appeared during the 1980s (many of them disappeared later), altered the equation that had obsessed the World Bank and most other serious economic thinkers. The emergence of computing technologies and ‘brain imports’ in the US meant that it was possible to increase economic growth without having a similar increase in commodity consumption. Microsoft and creators of ideas, after all, produce wealth without consuming commodities in proportion to growth. The new economy based on knowledge, information, organisation and control technologies, computers and continuous innovation may increase total factor productivity as inputs in production are combined more efficiently. Given that inputs may produce more output, so the production frontier shifts outwards. Some even thought that this would mean an end of the business cycles. The Internet, for example, may create value as it reduces time and cost for finding and transmitting information (similar in principle to the reduction in transport costs brought about by the advent of railways). Therefore, the new economy may increase investors’ hopes. But too many hopes may create ‘irrational exuberance’. When translated into plain English, this cryptic term used by Alan Greenspan in 1996 means that growing markets may become dangerously overvalued (over-investment in telecommunication networks). After this ‘new gold rush’, stock markets fell sharply in 2002. This had a negative impact on earlier great hopes linked with the new economy. Perhaps too much was expected from the new information and communication technology. New technologies may rejuvenate the growth process; they do not necessarily accelerate it … Typically, several decades are required for a GPT to make a major impact – and that impact may then stretch over more than a century … electricity is the prime example of this trend. (Lipsey et al., 2005, pp. 112–13).152 The overall impact of the new economy based on the new information technology is rather mixed. Does the new economy then look much like the ‘old economy’? It is also important to remember that the case for the New Economy rests not just on faster productivity growth, but also on the remarkable

Evolutionary Economic Geography

130

behaviour of inflation and employment, and a level of overall economic stability not seen since the 1960s. (Temple, 2002, p. 249) In spite of obvious benefits brought by the new economy, such as increased access and flow of information, speedy electronic transactions, increased transparency and lower prices, the dark side of the information economy increases risks. These nemeses refer to and include serious problems such as identity theft, financial fraud and a potential risk in the loss of privacy. A sharp acceleration in productivity growth153 and increased competition based on knowledge and innovation is at the heart of the ‘new economy of permanent prosperity’.154 In this situation, it was thought in about the mid-1990s that ‘old rules’ no longer applied. The economy could be closer to full employment without acceleration in inflation. The new situation may have a greater impact on the production of services than on goods. Investments by US firms in shorter-lived assets such as information and communication technology (supposedly important contributors to an increase in productivity) were very high during 1995–2000. After 2000, these investments shrank, but productivity growth accelerated. What is the reason for such growth? Are studies about productivity flawed? These studies may arbitrarily include or exclude depreciation. They may also assume that new computers have a direct and full impact on production from day one, but in reality it takes quite some time to master new equipment.155 Or, are invisible and intangible activities such as formal and informal computerrelated training and reorganisation of business methods more important than measurable investments in new equipment that show only the tip of the iceberg? New intangible assets helped American companies produce more output with fewer production workers. As far as Europe is concerned, it had the worst performance in industries that are heavy users of information and communication technology, especially the retail trade (where US productivity is growing much more strongly, as is evidenced by retail chains such as Wal-Mart).156 In addition, US energy consumption in 1997 and 1998 was almost unchanged, while the US economy grew by 9 per cent during those two years.157 Those economies that are successful in reducing energy consumption per unit of output (expectation of energy shortage) can, comparatively, benefit greatly from increases in the price of energy in the future; competitors that have not made similar adjustments will suffer. The extraordinary growth of the US economy has many causes. There is no doubt that the persistent growth of productivity in the US is due to the improvement in efficiency introduced by computing and ‘brain imports’. Even Federal Reserve Chairman Alan Greenspan has acknowledged this (during his term in office), while also recognising that it is hard to calculate the impact. This much is apparent. At a time when productivity should be falling, inflation and interest rates soaring and the economy moving towards recession, productivity (driven by the effects of computing, as well as new ideas) is maintaining a long-term growth trend.158 In addition, the American economy became much more energy-efficient. Demand for electricity, for instance, grew roughly 1 per cent a year over the first decade of the twenty-first century. This was far below the roughly 3 per cent annual rate of growth for the economy. Over the preceding decade, the annual rate of growth for power stood at 2.5 per cent.159

Theory

131

At the start of the 1980s, Japan produced cars and cameras at lower prices, smaller in size and of better quality than the US versions. Two decades later the ability to produce cars, copiers and cameras is much less rewarding profit-wise than the ability to write software. The structure of production has changed almost beyond recognition. The hardware that runs a web server is much less valuable than the non-material intellectual property that resides on the server. Perhaps the Japanese industrial targeting (expectation) was superior during the age of fax machines, while the US liberal (unplanned) economic system and alert financial markets that force firms to adjust rapidly to shifting markets and changing technology are more appropriate for the age of the Internet (Krugman, 2000, p. 174). To put it bluntly and in terms of expectations, the Japanese bet on hardware, while the Americans bet on software. The decoupling of value from physical production, its shift to intellectual production, is a millennial shift whose full meaning will not unfold for many generations. The firms that will define the next twenty years of economic history and production geography are probably completely unknown to anyone today. Uncertainty and risk are very high and the way forward is not always clear. Therefore, it is very difficult to predict with a high degree of confidence where the world’s economy is really going. Profound changes that occur little by little (such as the transformation in the economy and in our lives that was brought about by Microsoft) may be noticed only long afterwards. Therefore, the greatest danger and mistake is in ‘linear thinking’: the belief and expectation that what exists today will also exist tomorrow, but with a higher degree of intensity. Consider what the world looked like in 1980. No scholar ever dreamed or imagined then how the general situation would turn out in reality in 1991 or 2000. Or go back to 1900–20, or almost any other twenty-year-period, in particular during the twentieth century, and you would be able to observe these changes in retrospect. 10 War and the location of firms Most of the land area on earth has no military history. There are many places that are as inhospitable to the traveller (or a settler) as to the soldier. These vast geographical areas are either too high or too cold or too hot or too waterless for the conduct of military operations. Battles not only tend to recur on sites close to each other – the ‘cockpit of Europe’ in northern Belgium is one such area, the ‘Quadrilateral’ between Mantua, Verona, Peschiera and Legnano in northern Italy another – but have also frequently been fought on exactly the same spot over a very long period of history. The most arresting example is Adrianople, now Edirne, in European Turkey, where fifteen battles or sieges have been recorded, the first in AD 323 and the last in July 1913. (Keegan, 1993, pp. 69–71) The reason for a noted concentration of military operations in certain geographical spots is not necessarily the material loot (wealth of the place), but its specific location in space

Evolutionary Economic Geography

132

from which the controlling authority may efficiently project its power. Warriors seldom quarrel with geography. They try to use their skills and equipment to profit from topography.160 Sometimes they assist it, as the Chinese did with the Great Wall to prevent the annual intrusion of northern Mongolian nomads into the cultivated and irrigated lands in the south. The ferocious intruders looted in the south in order to obtain provisions for the winter. Hence the Chinese built the wall. The same was done in Europe by the ancient Romans and their limes. Unexpected political events (chance, accidents) such as wars, their outbreak, intensity, length, resolution and overall impact cannot be predicted with a high degree of certainty. Despite high costs and disasters, wars and war-linked efforts decisively influence the development, expansion, location, dislocation, spread and demise of certain industries and economic activities. There are numerous examples that support such arguments. The area, for instance, that is now known as Italy was swamped with wars over long centuries. Yet in spite of war-related disasters, this area of Europe gave birth to the Renaissance, Leonardo, Michelangelo, Botticelli, Giotto, Giorgione, Mantegna, Dante, Petrarca, Galileo and many others. Let us consider some other examples: • As a latecomer in the colonial era and, later, because of post-First World War external sanctions, Germany had the ‘incentive’ (was forced by events) to produce various chemicals as substitutes for natural inputs that were not available either in its colonies or through ‘normal’ trade. Selfsufficiency has often been of vital national importance. The German chemical industry was based on domestic deposits of salt and coal, as well as skills. And trouble and necessity too. There was a shortage of soap in Germany during the First World War. This gave the incentive to chemical engineers to invent the first washing powders in 1916. Germany’s success was remarkable. The country developed world-leading chemical industries and all the related products that go with them, such as pumps and sophisticated precision measurement and control instruments. • After the First World War there was a recession in Seattle (Washington). The economy of the region was based on fishing, production of timber and ship/boat-building. At about the same time, demand for aircraft (made of wood) started to emerge. Seattle had unemployed boat-builders and other inputs: workers had the skills to make wooden boats, they could easily build a fully covered boat (the body of an aeroplane), and they also knew how to fix a propeller – all of which led to the emergence of Boeing. The US recognised the huge potential of airmail on a large scale, large purchases of military aircraft and high demand for fast travel, which in turn provided domestic producers with early incentives to lower production and learning costs per aircraft as compared with foreign competitors. • The first commercially successful motor scooter, the Vespa, was produced by the Italian company Piaggio in 1946, although the idea for it was conceived during the Second World War.161 During the war, Piaggio, at Pontandera, near Pisa, made aircraft engines. After RAF bombing destroyed the factories, warehouses and roads at the site, it became difficult and tiring to get around the site on foot. The company owner, Enrico Piaggio, asked one of his engineers, Corradino d’Ascanio (who, earlier, constructed the first fully operational helicopter in 1930) to design a simple and economical two-wheel personal run-around vehicle. D’Ascanio built the prototype using his imagination, the leftovers from small twostroke motors used to start the

Theory

133

aircraft engines, aircraft wheels (note the shape of wheels on a Vespa when you next see one) and whatever else he could find in the warehouse (or what remained of it), such as metal sheets that were used for wings. The vehicle also incorporated a shield on each side to protect the rider’s legs from injury. The prototype was ready within a few weeks and, once in production, the Vespa became an immediate hit. It was extremely popular and fashionable in the 1950s and 1960s and was rediscovered by another generation in the 1990s because of increasing traffic congestion in cities. • The Swiss chemical and heavy industries enjoyed certain benefits during the two world wars. On the one hand, as German industry did not have access to the international market, Swiss entrepreneurs filled the gap where they could. On the other hand, the Swiss benefited from the inflow of funds, talents and invalidation of German patents by the Allies. • US Caterpillar relocated abroad during the Second World War in order to service machines that were used by the US Army. After the war, the machinery that remained abroad had to be serviced. Hence, Caterpillar became an international company on a larger scale. Coca-Cola also followed the US Army during the Second World War as a morale booster for the soldiers, and also remained abroad. Mars Bars were distributed to armed forces all around the world during the war.162 • Sialkot is a provincial town in the Punjab, Pakistan. It had a military garrison and a mission hospital during the British period a century ago. This created a demand for the repair and, later on, production of surgical instruments, tennis rackets and footballs. The area had had a local tradition of producing swords and daggers for several centuries. When the military technology changed, the need for surgical instruments at the local hospital provoked a demand for scalpels. Later on, at around the time of the Second World War, the supply of surgical instruments became a necessary adjunct to the war. Technology and experts were brought from Britain in order to support the production of those goods. Currently, a core of about 300 family-run firms (almost all have fewer than twenty employees) in Sialkot produce a range of 2,000 surgical instruments. Together with firms in Germany, they are major world exporters of such instruments. As Sialkot has no airport and as it is located more than 1,500 km from the nearest sea port, a private self-help initiative created the Sialkot Dry Port Trust which ‘brought the port to Sialkot’. The trust offers a range of collective services such as customs clearance, warehousing and transport. The success of this, as well as all other clusters, is based on two factors: first, the demand-driven approach and, second, competition based on collective efficiency. However, the situation in Sialkot is not perfect: the infrastructure is still rather undeveloped; the power supply is inadequate; communications are poor; and roads are ankle-deep in mud during the monsoon season. In addition, health and safety standards are insufficient and child labour is a serious problem in Sialkot’s football-making industry (Nadvi, 1998). The leading world cluster for the production of surgical instruments, in particular if measured by the value of output, is in Tuttlingen (Germany). While Sialkot serves the market segment for mature surgical instruments, Tuttlingen operates at the technological frontier: it manufactures new products such as minimal invasive instruments, endoscopes and surgical implants. There are no innovative activities in Sialkot and this cluster relies on passive extra-cluster transfer of knowledge, principally from Tuttlingen. Firms in the Tuttlingen cluster not only perform their

Evolutionary Economic Geography

134

own R&D activities, but link with specialised firms that are outside the cluster. This is most obvious in the field of optics (not the basic Tuttlingen specialty) (Giuliani, 2005, p. 283). • Italian motor car racing companies such as Alfa Romeo, Ferrari and Lancia dominated the market in the period immediately after the Second World War. Within a few years, however, a cluster of small firms from around Oxford became another dominant world player. Several fortuitous events prompted this evolution. First, there was a huge surplus of abandoned airfields in southern England after the war. Second, Mercedes withdrew from motor racing in 1955 after an accident in Le Mans in which 183 spectators were either killed or injured. This left room for the British racing-car builders. Third, large and vertically integrated manufacturers such as Ferrari and Porsche built cars for their own racing teams, while the British were selling cars to anyone who wanted to buy them. Even though the British were not winning races, they dominated the starting grids with the sheer number of cars. Fourth was the ban on cigarette advertising on British TV in 1965. Tobacco companies became sponsors of motor racing, doling out vast sums of money in return for having their logos displayed on racing cars. Pinch and Henry (1999) argue that the success of the British constructors may have been influenced by a few accidents, but only marginally. Other more important causes, such as accumulated knowledge and its circulation through the companies by means of transfer of personnel, as well as skilled labour, are the true origins of the British success. The British have a tradition of expertise in engines and lubricants, aerodynamics and composite materials (the Ferrari engineers were initially concentrating on engine power). In particular, the national aerospace industry was bigger and more sophisticated after the war than that in Italy. According to Pinch and Henry, this would probably have encouraged the production of racing cars even without the Mercedes crash or the money from cigarette manufacturers. • Another example of the impact of war on the change in the spatial distribution of production emanates from the end of the Cold War around 1990, when huge defence cuts badly damaged the military aerospace cluster in southern California. An increase in the military spending following the war in Iraq (from 2003 and beyond) also has an impact on the maintaining and expansion of the US military industry in specific locations. • Vienna is renowned for its splendid cafés. These café bars began to be opened and developed following Turkish occupation of parts of Austria and sieges of Vienna (1529–1683). Once the Turks were pushed back, the Austrians found huge quantities of coffee in the former Turkish camps. And the coffee-bar business started … • The Peace of Westphalia refers to two treaties concluded in Münster and Osnabrück in 1648. The first Treaty ended the Thirty Years War (1618–48), while the second finished the Eighty Years War (1568–1648). These two locations, just 50km apart in what is now Germany, were necessary because Protestant and Roman Catholic persons in command refused to meet each other. France favoured Münster because of its Roman Catholic background, while Sweden chose Osnabrück for its Protestant milieu. The leader of the Holy Roman Empire in either location was Count Maximilian von Trautmansdorff. The Thirty Years War took place for the most part on the territory of today’s Germany. It was a savage religious war between Protestants and Roman Catholics

Theory

135

with a wide use of mercenaries. Greedy and brutal mercenaries perpetrated enormous atrocities on the civilian population and property. After a siege (typical war technology at the time) in Magdeburg in 1631, out of 30,000 inhabitants, only 5,000 survived the massacre. Typhus and dysentery also spread. The result was that Germany’s population was reduced by a quarter.163 Among the results of the Peace of Westphalia was the recognition of the principle cuius regio, eius religio, i.e. each prince had the right to determine the religion in his own state. The Papacy had to accept the fact that it must live with its rival Christian churches. France emerged as the dominant European power, but not for long. England developed a new form of constitutional monarchy and soon overtook both France in power and the Netherlands in commerce and wealth. In any case, following the Peace of Westphalia, the future was to be found in selfreliance, rather than in obedience and unquestioning faith. This opening had an important impact on the location of ideas and production in Europe. The European scene was set and open to the creation of free and new ideas stared by Galileo Galilei and further developed by Isaac Newton. 11 Conclusions A century or even half a century ago, economic theory looked at the location of firms and industries in a rather technical and deterministic way. Subjects and causes were separated from objects and effects, and everything ended in equilibrium. Parts of modern spatial economics and the new geography of production consider the location of firms in an evolutionary way, where the state of equilibrium is only a temporary special case as everything is in an unstable organic process of change. Where a firm or an industry should locate is not a question that has a straightforward answer. Rich nations fear the relocation of firms and industries to low-wage nations; poor nations worry that the production of goods and services will migrate to developed countries; small countries are concerned that businesses will move to large countries; countries that are not integrated with others are afraid that production will migrate to the integrated countries; while all countries are troubled because of the grand reallocation of manufacturing to China164 and certain services to India, especially following the Asian financial crises of 1997. In an age of business ‘globalisation’, one might expect the importance of firm location to be diminishing. Some argue that the world is flat or that distance is dead in terms of production and business because of the astonishing changes in communications and computing. ‘Time zones and language groups, rather than mileage, will come to define distance’ (Cairncross, 2001, p. 5). But this weightless society or frictionless economy may be relevant only for certain manufacturing and services activities that are simple, that could be split, that are routine and that have codified (standardised) and teachable knowledge. Trade costs – that is, all costs linked with getting the final good to the consumer (tariffs, quotas, NTBs, transport, storage, distribution, information, contract enforcement, insurance, exchange and banking) – are alive and kicking. The many components of trade costs may rise or fall in a different proportion from distance costs. Trade costs matter not

Evolutionary Economic Geography

136

only for economic geography and economic policy, but also for theory as a number of puzzles in international economy rest on their existence. In addition, strictly local factors such as high-quality labour and sophisticated management, respectable educational facilities, transport infrastructure, certain taxes and a supply chain of goods and supporting services are often controlled by local communities, not national governments. The location in space still matters. Geography and local proximity (clusters) of firms that produce similar, competing and/or related products together with supporting institutions still matters for complex activities, with little or no codified knowledge, and where market changes and uncertainty require a rapid reaction. Economies of scale, activity-specific backward and forward linkages (indivisible production, strategic complementarity), accumulated knowledge, innovation, lockin effects, the existence of sophisticated customers and a fall in transportation costs play relevant roles in the protection of clusters and the absolute advantage of certain locations. ‘Global’ competitiveness often depends on highly concentrated ‘local’ knowledge, capabilities and a common tacit code of behaviour which can be found in a spatial concentration (a cluster) of firms.165 Hence, successful managers must often think regionally and act locally. The neglect of spatial economics (geography of production) in mainstream economic theory was not because this research field was uninteresting, but rather because the issues were traditionally regarded as intractable. Ricardo’s comparative advantage model provided insufficient answers to the problem. New research tools such as increasing returns, production linkages (presence of intermediate goods and services), multiple equilibria (with centrifugal and centripetal forces), evolutionary approach and imperfect competition were introduced into the field of spatial economics in the early 1990s. These methodological tools also helped to explain why firms form clusters. Such new developments did not mean the birth of a new subject. Spatial economics has always been important. Even though the ever-increasing demand for quantitative vigour is making such considerations more difficult, the introduction of new analytical tools assisted spatial economics in finding its proper place in mainstream economics and becoming a hot research topic. Evolutionary economic geography differs from the traditional model in several important dimensions. The new evolutionary model makes the case that production specialisation in a given locality is based not only on certain local comparative advantages, but also on a self-reinforcing lock-in effect, path dependence, accumulated knowledge, agglomeration, clustering and linkages (indivisible production). In addition, while the traditional models reason that a reduction in trade costs among locations may favour local specialisation, the new evolutionary economic geography claims that the effect on local specialisation is ambiguous. The final outcome is time, context and industry specific and depends on the functional intra-industry production linkages, market structure, consumer preferences (homogeneity of tastes) factor market (availability and mobility of factors and flexibility in prices) and expectations. In general, the choice of location for a new investment or a new firm that produces tradable goods and services depends on a complicated interplay of at least three elements: factor intensity; transport intensity; and each in relation to the already established firms and their activities and (re)actions. Remote places do not necessarily need to be poor

Theory

137

locations for firms as their remoteness is already reflected in their factor prices (Venables and Limão, 2002, pp. 260–1). Locational disadvantages may often have an effect on the lowering of per capita income. Landlocked countries in general have transport costs about 50 per cent higher than otherwise similar coastal countries (Venables and Limão, 2002, p. 240).166 Access to the coast may raise per capita income by over 20 per cent (Redding and Venables, 2004, p. 78). While this may be the case for certain developing countries, most notably in Africa, such an expectation does not have a general value and application. Switzerland or Austria are clear proofs and counter-examples. As a reaction to trade liberalisation and economic integration, various industries follow different paths. Depending on the functional intra-industry production links, some industries concentrate (reinforcing absolute advantages of certain locations), while others spread. This is contrary to the expectations of neoclassical equilibrium theory, which predicts that all industries will be affected in the same way. Different forces, therefore, propel such changes in the spatial structure of production. Strong functional intraindustry linkages (a high share of intermediate inputs from the same industry and/or the need for a large pool of highly skilled labour and researchers) stimulate agglomeration. Where these functional linkages are weak, this acts as an incentive to spread production. Ultimately, one can reason that there is an important lacuna in our understanding of evolutionary spatial economics. It is little wonder that there are many disagreements and only several theoretical agreements: • Starting a new activity in a given location requires building on and improving already existing accumulated expertise and specialisation of that location. One may need to stick to the production process that one understands in general terms. This bottom-up specialisation needs to be extended, but this does not at all mean that the earlier production structure has to be continued. Something (vaguely) related to it or derived from it may be the starting point which, later on, needs to be altered, sometimes fundamentally. • Using policy intervention to influence the location of firms is most effective only very early in the process. Evolutionary economics argues that very little intention is necessary to produce a successful locational result. There are few and relatively short moments of enhanced locational freedom for new industries that set in motion new industrial trajectories. If there is a choice of potential locations for the establishment of business, the winner location often wins by the skin of its teeth. Hence, there is only a narrow window of locational option through which can pass a short, rare and slim beam of opportunity light and during which the policy may be effective.167 Market forces may be the strongest ‘socio-economic technology’ for resource allocation in the situation of disequilibrium in the medium and longer term. • Local education and training has a positive impact on the location of firms and industries. • Depending on industry, economies of scale and functional linkages in production and its servicing (strategic complementarity) have certain impact on the location of firms and industries. • Policy should be reviewed periodically and have elements of flexibility. • Industrial success or failure of a region or a policy sometimes depends on uncontrollable factors and events that are located outside regions themselves.168

Evolutionary Economic Geography

138

Regions and firms ought first and foremost to look inwards. They ought to know or locate their past and current ‘core competences’, unmatched and unmatchable assets (rather than first search for what they would like in the future). Then the regions and firms should know how well they are matched with current and expected future business activities and market environment. The real core competences and unique firm-specific assets are normally individual in nature and few in number. Regions and firms need to know what necessary complementary assets have to be present. Are the core competences the regionor firm-specific technology (sparkling wine) or brand name (Moët et Chandon champagne)?169 Core competences are normally established over a (longer) period of time, rather than designed or transferred. Fizzy-wine technology is freely available to anyone, but brand name is not. Nonetheless, the Australians and Americans are making strong inroads in the sparkling wine market. They learn and attack this wine-market segment fast. The analysis of evolutionary spatial economics depends on special assumptions. It is often a detective-type study and a collection of particular cases. Generality is often abandoned in favour of discovery because this may be one of the few ways to make a complex problem partly tractable. This is not surprising, as evolutionary spatial economics is simultaneously charged with some of the most complex and difficult problems in economic theory: multiple equilibria, economies of scale, externalities, imperfect competition (which kind?), linkages, path dependence, endemic uncertainty and expectations. Nonetheless, many useful things can be learned from exceptional cases and simplified stories. These examples do not serve the purpose of emulating reality in its entirety, but rather serve as metaphors to illuminate certain important background forces that are at work. Living organisms in biology and firms in business have capabilities that are more or less adapted to their environment. These capabilities emerged from the process of interaction with the environment. Evolution influences living organisms, while the same evolutionary role in the case of firms, industries, clusters and regions is performed by human agency and competitive market process. Notes

1 Trade theory has not come to grips with multiactivity firms and multiplant production, as was the case with the theory of industrial organisation. Heterogeneous size and technology structure of firms has not been considered either. 2 The word ‘new’ ought to be qualified here. It refers to the new light that is shed on known, but scattered and underappreciated, previous research work. 3 Partial solutions to crucial research and policy questions were searched in urban economics (benefits from concentration of economic activity and costs from such agglomeration); industrial organisation (product differentiation and inter-industry links); new economic geography (prominence is given to trade, while land markets are ignored); and theory of endogenous growth. 4 If pushed to the limit, this ‘cultural turn’ in a pluralistic society may evolve into an aggressive and naive effort to make less harsh the basic reason for the existence of private market-based firms: to earn profit. 5 Probably the only people left who think that economics deserves a Nobel Prize are economists. It confirms their conceit that they’re doing ‘science’ rather than the less tidy task of observing the world and trying to make sense of it. This, after all, is done by mere

Theory

139

historians, political scientists, anthropologists, sociologists, and (heaven forbid) even journalists. Economists are loath to admit that they belong in such raffish company.

(R. Samuelson, The New Republic, 3 December 1990)

The search for ‘sharp prediction’–the mantra of the modern scientific economist who seeks to replicate the successes of physics for social science – is doomed to failure. The best that economists and their clients can do is identify qualitative regularities and patterns in events. (J. Kay, ‘The fruitless search for exact knowledge’, Financial Times, 16 October 2007)

6 Institutions are taken here to mean organised interests. They are ordered collections of rules, laws, regulations, property rights, habits, practices, customs and routines that are enforced and that enable, guide and normalise relations and behaviour among various parties. These institutions may be formal (such as laws) that are arrived at after long considerations and informal (trust, culture, habits, group norms) that are a spontaneous product of continuous and repetitive interactions among individuals over a long period of time. Institutions shape not only relations among economic agents, but also their ability to alter and improve existing institutions (e.g. laws) needed for adjustments to newly arrived or created situations. 7 Complexity theory deals with computation and the required resources to solve a given problem. It studies both time, in terms of how many steps are necessary to arrive at a solution, and, in a computer, space in terms of the necessary memory. Similarly, complex system analysis is a multidisciplinary approach to the problem-solving process. 8 Location matters not only in economics. The first lesson in geopolitical theory teaches us the same thing. The international political behaviour of states is rooted in and influenced by their geographical location. Britain, as an island, has a freedom in geopolitical actions that comes from its location in space. This windfall advantage is not available to countries that are ‘boxed in space’, such as Switzerland or Germany. Their options for external action are more constrained than is the case with Britain. 9 Economic transactions across space always involve positive costs. 10 Comparative advantages and geography (among other factors) set the equilibrium location of firms under fixed exchange rates. A flexible exchange rate may introduce a new element that influences, i.e. disturbs decisions about location of business. This ‘departure’ is linked with efficiency losses that stem from decisions resulting in going away from comparative advantages and geography towards policy-related factors such as a flexible exchange rate regime. A flexible exchange rate system (partially) adjusts real economic shocks to exports in a country. Firms in that country experience a lower change in their foreign sales than their foreign competitors in the short (and perhaps medium) term. This outcome does not take place under fixed exchange rates. Under the fixed exchange rate regime, a shock affects all firms producing the same good in all locations. Countries are, on average and according to this view, more specialised under a flexible than a fixed exchange rate regime. 11 This increases the need for printed matter. 12 Income distribution also affects the location of production. If, for example, income is distributed to a small segment of the population (landowners or owners of capital) who spend it on imports of luxury goods and services, there may not be a spread of development of industries in their country. If, however, income is distributed in favour of a large number of families that demand domestically produced goods and services, this may have a positive impact on the location of firms and industries closer to the domestic market.

Property rights are closely linked with the distribution of income. If these rights are distributed in a top-down way, as was the case when Europeans occupied lands in Latin America, the deprived constantly tried to upset such a distribution. This

Evolutionary Economic Geography

140

polarised societies on the haves and have-nots. The consequences of such polarisation are still felt in our time in terms of lack of access to sufficient financing by local firms. The Latin American manufacturing industry is in tough competition with producers in east Asia that count on generous and supportive national governments. If, however, the colonial distribution was bottom up, as was the case in many English-speaking countries, those who actually worked on the land had the strongest claims on property rights. While the English-speaking colonisation of America and Australia supported democracy, certain consensusbuilding and fast economic progress, the Hispano-Portuguese colonisation of Latin America backed dictatorship and polarisation. The concern for the newly democratic Russia from the 1990s is that it may have certain elements that may turn it towards the Latin American outcome. 13 Ageing population and retirement patterns influence the location of business and trade. The old demand health-related goods and services. For example, some Germans retire to Spain to enjoy a milder climate and a lower cost of living, as well as their pensions and life savings. As such, they create new demand in Spain for the local supply of services such as financial advice, health care, catering, cleaning … 14 Who would have predicted in 1985 how the world turned out to be in 1990 or in 2003 and beyond? Particular issues of concern are: technology; energy; wars; redressing of political systems, states and their borders; demography … 15 Human agency is the collective ability of human beings to be creative, to choose and to impose those preferences on the world. Human agency stands out against natural forces and deterministic processes. 16 ‘Self-organisation is not a physio-chemical analogy but a general principle in systems that process energy, matter and information. Economic self-organisation is not the same as biological self-organisation, despite the fact that they share common properties’ (Foster, 2000, p. 325). This different and purposeful behaviour is because of an important role of knowledge and foresight (Foster, 2004, p. 11). 17 A neighbourhood effect arises when one agent’s presence affects the utility of another agent. 18 The precise meaning of ‘lock-in’ is not fully clear. Why are certain locations locked into one type of development trajectory while other similar locations are not? Why do certain locations ‘unlock’ themselves, change their development path and (continuously) reinvent themselves? If this is linked to a special locational genius, what is it? Can it be replicated elsewhere? How? 19 Royal Dutch Shell (founded in 1907) is a company with 112,000 employees worldwide and operations in more than 140 countries in 2008. Most of these people are unlikely ever to meet, but Shell is organised in such a way that it operates successfully. The glue that links these people in the pursuit of a common goal is the social technology. 20 The initial growth theory was founded on closed economy models. Like the classical equilibrium trade theory, it has also an aspatial character. 21 The old protectionist rationale of national security is not dying easily or quickly, as was the case with the Dubai Port World case in 2006, as well as others. A recent protectionists’ invention is the carbon border tax. This is a new tariff on imports. It is set according to the quantity of greenhouse gases released during the production process of a good. The message is: if you do not have our clean (and expensive) technology in your location, well, we shall tax you for the damage that you do to the environment. 22 The assumption is that there are no factor movements. Otherwise, the factor abundance proposition has no meaning. 23 Networks are very old and enduring social institutions. They also include families, clans, casts, tribes, gangs, nations …

Theory

141

24 Quotas have inflated clothing prices by creating scarcity and ‘rents’, price premiums that act as taxes on trade. The annual cost to US consumers has been put at $70bn and has fallen hardest on poor families, which spend a relatively large share of income on clothing. Each job saved by quotas in the US industry is estimated to have cost consumers an average of $170,000. (G. de Jonquières, ‘Clothes on the line’, Financial Times, 19 July 2004). 25 G. de Jonquières, ‘Clothes on the line’, Financial Times, 19 July 2004. 26 Another implication of the home market model is that wages decline with distance from the principal national cluster. 27 Does the world have end points in this model? 28 Europe imports a huge amount of flowers from Kenya. This production of flowers in East Africa and their flight to Amsterdam, Berlin, London, Paris and elsewhere may easily leave a much smaller carbon footprint on the environment than production of flowers in heated greenhouses throughout Europe. 29 In the case with one market and two deposits of resources, the optimal firm location would fall inside the triangular area which links these three different spots. 30 Similarly, Grossman (1998) reported that ‘two regions separated by 500 miles will, all also equal, trade more than 2.67 times as much as two regions separated by 1,000 miles’ (p. 30). 31 ‘Instead, a “peripheral” region is the one which has few sizable markets nearby, whereas a “core” region is the one which is located close to such markets’ (Begg and Mayes, 1994, p. 91). South-east England is a core region in spite the fact that it is at the geographical edge. However, Kansas, located in the middle of the US, can be treated in economic terms as a periphery, while New York or San Francisco, for example, may be treated as a core region. 32 If one imagines a geometrically even distribution of centres across a flat and homogeneous surface and the corresponding circles (representing ranges of goods) around them, then if one wants to cover all the space with circles, there would be an overlap between the two adjoining circles. If one draws a straight line between the points where the two adjoining circles intersect, one would get a hexagonal market space of identical size around each centre without empty corners, as consumers would purchase goods from the cheapest (nearest) producer. 33 Paris and Madrid have a relatively central geographical location in France and Spain, respectively. The same holds for Munich in Bavaria. In 1998, the capital of Kazakhstan was transferred from Almaty to Astana precisely because the latter city lies at the intersection point of major north–south and east–west transport routes.

One of the reasons why Peugeot-Citroën announced in 2003 that it intends to invest and to locate a large (Є700 million) car-assembly factory in Trnava, Slovakia (in preference to alternative locations in Poland, the Czech Republic and Hungary) is that this country has good access to transport links and that it is ‘in the centre of Europe’ (R. Anderson, M. Arnold and J. Reed, ‘Peugeot to build new plant in Slovakia’, Financial Times, 15 January 2003). The factory, with an annual production capacity of 300,000 cars, started production in 2006. The management intends to be closer to fast-growing markets in the ‘new’ EU countries from central Europe and to profit from lower costs of operation, i.e. from wage rates that are a quarter of those paid in the ‘old’ EU(15) for similar industrial operations. In 2004, Hyundai (KIA) selected Slovakia over Poland (known for the aggressive reputation of its labour unions) to build a Є825 million car-assembly plant in Žilina. This is Hyundai (KIA)’s first car-assembly plant in Europe. It will assemble 300,000 cars a year. The reasons for the decision to locate in Slovakia include: wage costs (a quarter of the prevailing rates in western Europe); car ownership levels half those of western Europe; laws that made the labour market

Evolutionary Economic Geography

142

flexible; reduced corporate and income taxes; an existing network of suppliers (expected to expand further); and membership of the EU. As Volkswagen already produces cars in Slovakia (Bratislava; 238,000 cars in 2006), this ‘tiny’ (5.4 million people) central European country is becoming one of the European motor giants. The three Slovak car plants could produce over 550,000 cars a year from 2006, making it the highest per capita car production in the world (see also R. Anderson and J. Cienski, ‘Hyundai picks Slovakia for new Є700m car assembly plant’, Financial Times, 3 March 2004). Slovakia may produce about 900,000 cars a year from 2008. The automotive industry contributes a quarter to the manufacturing output and a third to exports (this will increase to a half once the two plants are completed). The problem may be in the increasing imbalance in the Slovak industrial structure. Although Slovakia may be the victim of its own success, some of the neighbouring countries (Poland, for example) might have preferred to be faced with such ‘troubles’ than to lose out to Slovakia. A looming problem in the central and east European countries is an increase in wages above the improvements in productivity. Having joined the EU, workers in the new member countries want to achieve the ‘old’ EU(15) living standards as soon as possible. This will jeopardise the central and east European countries’ attraction on account of their relatively low wages. Such an improvement may take a decade, or even a generation, to accomplish. Similarly, the mobile phone producers such as Motorola or Nokia started production in India in 2005, not only because of lower costs of operation, but also because of a rapidly expanding market for mobile phones in this country. India had 52 million mobile phone subscribers in 2005, and this is expected to expand to more than 300 million in 2009 (J. Johnson, ‘Back to the future: India is gaining belated credibility as an emergent export titan’, Financial Times,30 November 2005). 34 The six world-leading clusters that have more than 5,000 researchers in their area of competence and that are worthy of decisive government funding are: neurosciences and complex systems (Ile-de-France); aeronautics (Toulouse); space (Bordeaux); health (Lyons); nanotechnology (Grenoble); and secure communication systems (Provence-Alpes-Côte d’Azur) (P. Hollinger, ‘France identifies six target areas of growth’, Financial Times, 13 July 2005). 35 The other one relates to services. 36 Alchian (1950) considered, among other issues, the role of chance and luck in achieving economic success. 37 Certain lawyers may have similar expertise and talent. But a small difference in the knowledge and experience of the ‘right’ lawyer and his/her choice (tipping point) by a party in a legal process may bring a crucial difference in the result of a legal case. Therefore, if one prefers averages together with stability and does not go for or select extremes, one may have a relatively steady life, but may at the same time forget about economic rents (supernormal profit). 38 Risk is a situation in which one may ascertain and calculate possible outcomes with a reasonable degree of probability. Hence, most often an insurance market will appear. Risk may be spread among those who are willing to share it and bear it (the reinsurance market). Risk could be also passed on from those who know more about it to those who know less

Theory

143

about it. Uncertainty is an irresolvable situation in which it is impossible to establish either all the possible outcomes or the probability of reaching the known outcomes. One may insure against certain risks, but not against uncertainty. Uncertainties may arise in instances such as relations between buyers and suppliers; between labour and management; between the public and private sectors; between current and future technology … Some of these uncertainties may be sometimes mitigated or, from time to time, even resolved through institutions, i. e. formal and informal conventions. 39 If markets are so efficient, why are there more Ferraris on the streets of Monaco than on the streets of Marseilles or Nairobi? 40 The convergence hypothesis may find some support in certain sub- and crosssectional samples; however, it lacks unconditional and general empirical support. 41 Economic rents refer to the proceeds of the producers and exporters in excess of what is necessary to cover the costs of production and trade, and to yield an average return on investment. These rents represent supernormal profits that are returns to scarcities such as unique capabilities of firms or individuals. Rents are due to barriers to entry such as large sunk costs, economies of scale, externalities, advertising, regulatory policies, distribution and service networks, asymmetric information (adverse selection), as well as consumer loyalty to a certain brand or a person (a singer, actor or sportsman). 42 Léon Walras (1834–1910) dearly wanted to enter the prestigious French École polytechique. As his mathematical proficiency was insufficient, he was rejected twice. He ended up at a mining school, but did not have success as an engineer. Later he tried his luck as a novelist and art critic, but to no avail. Then, a depressed Léon talked to his father, Auguste Walras, an economist, about what to do next in life. His father, a schoolmate of the famous French economist Augustine Cournot and a schoolmaster, told Léon during a walk in 1858 that economics needed to be based on a scientific theory. 43 In 1893 Walras selected Wilfredo Pareto (1848–1923) to take over his chair at the Lausanne School of Economics. Pareto was, like Walras, trained as an engineer and worked for railways in Italy. He later developed an interest in economics. As a late bloomer in economics, Pareto published his magnum opus Manual of Political Economy (1906) that took him forever out of Walras’ shadow. All economists learn that the Pareto optimal equilibrium point is a situation in which no one can make a move without making someone else worse off. This is the finest social outcome in a free society and Walras’ world of complete information. An evolutionary economy is Pareto sub-optimal as it is in constant disequilibrium. 44 A bad sign in a company is if everyone on the executive board has grey hair. It is equally bad when nobody does (as was the case with many dot-coms that went bust during 1990s) (Beinhocker, 2006, p. 361). 45 It took Brian Arthur six years and fourteen rewrites of his working paper ‘Competing technologies, increasing returns and lock-in by historical events’ before it was published by The Economic Journal in 1989 (‘From QWERTY to Microsoft’, Options (IIASA), winter 2007, p. 20). 46 A scientific law is a set of proposals that specifies conditions for a certain result. If the conditions are met, the result will occur. 47 But is economics about scarcity and diminishing returns? Are there balancing forces of abundance? What about people, knowledge, ideas, labour, capital and technology? 48 Entropy is a state of disorder, confusion and disorganisation. 49 Aerobic organisms transform oxygen into carbon dioxide, while plants do the reverse. Human activity in transport and production emits extra carbon dioxide into the atmosphere that potentially has an impact on the change in climate and the environment as plants cannot always process this ‘extra’ in time. 50 Capital–labour ratio determines relative returns between two factors of production. Political changes and economic liberalisation from the early 1990s brought the entry into the ‘global’

Evolutionary Economic Geography

144

economy of the emerging economies such as China, India, Brazil and Russia with their huge labour force (and relatively little capital per worker). Their functional integration with the global production, investment and trading system shifted the worldwide capital–labour ratio in favour of capital (and against labour). The ‘world labour force’ doubled. 51 This view of the turbulent, as opposed to moderate, smooth and gradual, character of changes makes Schumpeter closer to Marx than to neoclassical economists. 52 For example, firms may be of different size and may employ different technology and business strategies. Certain players may have a highly ambiguous and vague purpose such as research institutes and universities, while others, such as firms, may have moderately ambiguous goals such as earning profit. 53 This was the case in all countries in different stages of their development. Even today in the most ‘liberal’ market countries the most advanced technologies have been developed with substantial and continuous assistance from the government. There are broad and strong reasons for the government to assist in the general R&D process over and beyond the level available through a pure and free-market system. Unassisted markets may not be the best economic guides in situations with a lot of risk and uncertainty. But this does not give either carte blanche for a full-blown intervention by the state authorities. 54 Lumpiness (indivisibility) presents a kind of a bottleneck in a smooth production and geographical space. Many products are often lumpy, as one cannot make money from only a half of them. Examples include: an electric bulb, a radio signal, a bridge, a TV set, a mobile phone or national defence. 55 ‘equilibrium if it exists at all, is not unique and historical accidents matter’ (Lipsey et al. 2005, p. 41). 56 Regarding the appearance of occasional turmoil in financial markets and hard econometric models, here comes an instructive view:

A glance at recent financial history shows that this type of ‘rare’ event is not so unusual at all. Back in 1998, for example, a key reason for the nearimplosion of Long Term Capital Management was that the fund’s economic whizzkids who included some Nobel prize-winning economists had devised model-based trading strategies that turned sour when markets moved in unforeseen ways … ‘People say these are one-in-a-100,000-years events but they seem to happen every year,’ says Satyajit Das, a consultant to hedge funds and investment banks. ‘This episode should make people ask questions about models I think it could lead to a real reassessment’… Consequently, the lesson that some bankers and policymakers may yet draw from this month’s events is arguably the simplest: namely that it could be time to reinsert humans into the trading process. One way to do this would be to improve risk management functions; another key step is to ensure that banks have the ability to switch off the models during times of stress. ‘Anyone using a human brain, rather than algorithms, could have made money in recent days in the equity markets,’ chuckles one banker. One unexpected winner from this month’s market

turmoil, in other words, may be a newfound appreciation of human intuition and old-fashioned common sense.

(G. Tett and A. Gangahar, ‘Why computer models proved unequal to market turmoil’, Financial Times, 14 August 2007)

Theory

145

Friedrich Hayek once said that he knew few people who had made money from acting on economic forecasts, but a good many who had made it from selling them. It is difficult for those in any profession to stand out against the spirit of the times. (S. Brittan, ‘It is time to jettison the forecasts’, Financial Times, 20 December 2007)

57 Just as a reminder, fixed costs deal with minimum outlays that are necessary to start a business; these are all costs before the first piece can be sold (machines, shelter, energy, permissions, material, publicity … ). Variable costs change (rise and fall) in proportion with the volume of output (materials, energy, labour). Average costs are the sum of all fixed and variable costs divided by the total quantity of output. Marginal cost is the cost of producing the last or the most recent unit of product. Opportunity cost is what you missed by not doing something else. 58 Murphy et al. (1989) discussed multiple equilibria with respect to the process of industrialisation, postulating that a certain critical mass of industries may have to be involved in order for industrialisation to be successful. A parallel could be drawn with location theory, and a role for government policy to aid the ‘takeoff’ of a particular location, and encourage firms to move to that location. 59 The coordination problem may be illustrated with the example of the small village of Palanpur in rural India.

Palanpur farmers sow their winter crops several weeks after the date at which yields would be maximized. The farmers do not doubt that earlier planting would give them larger harvest, but no one, the farmer explained, is willing to be the first to plant, as the seeds on any one plot would be quickly eaten by birds. I asked if a large group of farmers, perhaps relatives, had ever agreed to sow earlier, all planting on the same day to minimize the losses. ‘If we knew how to do that,’ he said, looking up from his hoe at me, ‘we would not be poor.’ (Bowles, 2004, p. 24) The solution of the coordination problem (through rewards for cooperation and a system of sanctions for cheaters) may be effective if the number of repeated rounds of play is infinite. Cooperation may mean a positive sum game for each player as uncertainties may be reduced to a manageable level. It may be much harder to elicit such cooperation if the envisaged number of rounds is finite, e.g. four rounds or the lifetime of each player (finite, but unknown). 60 Positive feedback economics (increasing returns) may also find parallels in nonlinear physics. For example, ferromagnetic materials consist of mutually reinforcing elements. Small perturbations, at critical times, influence which outcome is selected (bifurcation point), and the chosen outcome may have higher energy (that is, be less favourable) than other possible end states (Arthur, 1990a, p. 99).

Another example of these mutually reinforcing links that accelerate changes moving the system away from equilibrium is when a microphone is too close to a loudspeaker. Sound bounces back between the microphone and loudspeaker and amplifies itself. Or: the more operating telephones there are in the world, the more useful they potentially become to everyone. The higher the consumer confidence,

the higher the expenditure: this translates into higher employment which increases income, which in turn stimulates consumer confidence.

Evolutionary Economic Geography

146

A negative feedback effect has opposite effects from the ones just described. If a change happens, negative feedback forces drag the system back towards the starting or desired (equilibrium) point. When temperature in a boiler falls below the adjusted level, a thermostat turns on a heater to raise the temperature until it reaches the desired level, at which it switches itself off. 61 Veblen referred to a similar problem in 1915 when he wrote about the undersized rail wagons in Britain. ‘All this does not mean that the British have sinned against the canons of technology. It is only that they are paying the penalty for having been thrown into the lead and so having shown the way’ (Veblen, 2003, p. 53). Later on Germany came along and used ‘the English technological lore’ (p. 53) in a superior way. The Germans were able to start from a relatively clean technological account. They were not locked into this technology, like the British. 62 The Dvorak keyboard may be seen at many www sites. For example: http://www.tifaq.com/keyboards/dvorak-keyboards.html (accessed on 8 January 2008). 63 Sony expected that its reputation and favourable opinions by the experts about Betamax were just enough to impose this standard and to capture the consumer market. Sony came into the market first, and expected that people would primarily use camcorders to make home movies (birthday parties, weddings, family reunions, holidays … ). Sony was wrong. VHS (JVC) had one important advantage over Betamax: its longer playing time. VHS devices filled homes, chiefly for watching pre-recorded films. For this Betamax was inferior. The market referendum made Sony abandon Betamax. 64 Even though it was user-unfriendly, DOS initially succeeded on the market over Macintosh because of the following merits. First, it was cheaper. Second, it was easier to create programs for DOS computers than for Macintosh, which had copy-protected programs; diskettes had to be kept in the computer so only one program could run. Third, DOS updates on the screen were instantaneous, while Macintosh users had to wait a while for changes to appear on the screen. However, the switch from DOS to Windows was rather quick in the early 1990s. Even though it was costly, the advantages of the new system were so great that they more than compensated for the trouble. 65 J. Chaffin, ‘Warner plumps for Blu-ray in format wars’, Financial Times,4 January 2008; ‘Betamax’s revenge’, Financial Times, 7 January 2008. 66 This agglomeration potential and advantage is highest in the early days of the existence of a technology or a product. As some inputs or parts of technology become standardised over time, this initial agglomeration potential and advantage fades away. 67 The British cotton textile industry in Lancashire, successful in the nineteenth century, fell victim to the rise of the same industry in the US and Asia during the twentieth century.

In this and many other cases, the primary cause of decline is the inability or slowness of a region’s industries to respond to the rise of major new competitors and technologies elsewhere. According to the path dependence model, external shocks of this sort should eventually stimulate a region’s economy to ‘break free’ from its ‘locked-in’ path of development and move to another, more competitive configuration. Instead, however, as many examples testify, such shocks may merely set of a spiral of negative ‘lock-in’ effects whereby the region’s firms react in terms of defensive cost-cutting and quality reduction rather than by innovative investment or moving into new productive fields, so that it is not simply a lack of change that is the problem. In such cases of interregional competitive exclusion and ‘lock-out’, the occurrence of crisis and loss of competitiveness can be a gradual

Theory

147

process extending over several decades rather than a dramatic collapse and reconfiguration during critical juncture. (Martin and Sunley, 2006, p. 417)

68 Akron (Ohio) was famous worldwide for its tyre-making. Goodyear, Firestone, BF Goodrich, General Tire and Uniroyal once all had their head office in Akron. The location of tyremaking in this industrial town was quite favourable as it was just two hours’ drive from Detroit. However, tyre-making was transferred during the 1970s to lower-cost locations in the southern part of the US. In addition, firms wanted to get rid of the sometimes rigid trade unions that were strong in the north. Even though this was a serious blow to the economy of the Akron area, the location reinvented and upgraded itself in a related field of polymer engineering and production. 69 An ergodic system (a pendulum; water in a glass [if it is not spilled out] on a table) ultimately returns to its original steady state, regardless of the disturbances between the starting and ending points in time. This system cannot shake free of its history and yields a path-dependent outcome. Put loosely, the system is predictable. However, for predictions about the future of a non-ergodic system one ought to have certain information and knowledge about its past. 70 This hypersensitivity to small events makes the mathematical model (equations) impractical for predictions about the future behaviour. The problem is too complicated. Nonetheless, fathoming facts on how the system behaves and operates may be a great and useful achievement in itself. 71 M. Wolf, ‘The lessons Asians learnt from their financial crisis’, Financial Times, 23 May 2007. 72 http://www.lkwdpl.org/WIHOHIO/oakl-ann.htm (accessed on 8 January 2008). 73 There were no strong a priori reasons why Britain selected the left side of the road for driving while the European continent and the US chose the right. But they did. Other countries opted for the side of either their closest neighbour or their colonial master. Path dependency captured the rest. Once this locks in, it is costly and perhaps hazardous to change. 74 Finland was a resource-based (depended) economy. Within a relatively short period of time, towards the end of the twentieth century, Finland ‘created’ its path towards a knowledgebased economy. Sweden did the same, but many decades before Finland. 75 If a country has a poor educational system and record, then the disposable pool of labour and management may severely limit the possible development choices in the short and medium term. It is a different story when this system is good. 76 With perfect competition and no increasing returns to scale, institutions do not matter.

Institutions, both formal and informal (such as routines, conventions and traditions) are well known to change slowly over time. Institutions are both the product and a key factor shaping social agency: they provide the stability and predictability needed for social and economic actions and transactions whilst incrementally responding to and incorporating the outcomes of those actions and transactions … institutions thus tend to inherit the legacy of their past. (Martin and Sunley, 2006, p. 402) Institutions are given in the short term (they are exogenous), but they respond to the changes in the economy, they change in the long term and become endogenous. 77 Without these ingredients the impact of history may be rather weak.

Evolutionary Economic Geography

148

78 ‘paths emerge from a chaotic swirl of random events’ (Martin and Sunley, 2006, p. 424). This may be totally correct, but it does not give clear guidance to a policy-maker who is facing a real problem at a given time. 79 In order to keep transport costs down, the location of production of steel needs to be either close to iron ore deposits or near final consumers. Centralisation of production of steel in one place is hard and potentially unnecessary as there are also technical limits to the economies of scale. 80 The US Declaration of Independence (1776) states that unalienable rights of ‘men’ are among others ‘Life, Liberty and the pursuit of Happiness’. 81 William Shakespeare (1998, Act 4, Scene 3, p. 77). 82 In specific circumstances (the single driving variable is the cost of transport) one may construct a world in which there is an evolutionary (self-organising, long-term and progressive change) path for the world economy in which inequality of nations and the division of world into primary and industrial producers first spontaneously emerges and then dissolves. 83 Tacit knowledge and experience in a cluster is very important for localised spillovers. That may not be transferred easily and without costs. Global standardised knowledge and spillovers in the form of patents may be transferred often freely across firms and regions. 84 A driving time of about 45 minutes is the area where labour market effects are the strongest (Venables, 2006, p. 10). 85 S. Jenkins, ‘Cherie mauled by the tiger Tony couldn’t tame’, The Times,11 December 2002.

Regarding concentration of power, John Dalberg-Acton, first Baron Acton, has an often-quoted saying: ‘And remember, where you have a concentration of power in a few hands, all too frequently men with the mentality of gangsters get control. History has proven that. All power corrupts; absolute power corrupts absolutely’. 86 Learning from teachers is called oblique transmission, while learning from the members of one’s professional group is horizontal transmission. 87 When life expectancy is only thirty years, then the society is devoting most of its resources to passing the current knowledge from one generation to the next. Once life expectancy increases, there are resources in the society that may be devoted to an increase in the stock of knowledge (Boulding, 1991, p. 14). 88 It is difficult to trace knowledge flows, because they are invisible and do not leave a paper trail by which they can be measured and tracked. Hence, a theorist may assume anything about them (Krugman, 1992, pp. 53–4). Certain traces may be found, however, such as quotations in professional journals. Nonetheless, the evidence about these externalities is largely indirect. 89 Embodied knowledge is human capital that did not outlive the individual, while disembodied knowledge (non-rivalrous good) is the one that did (Warsh, 2006, p. 291). 90 Audretsch and Feldman (1996; 2004) found evidence that the US industries in which new knowledge spillovers are relevant (semiconductors and computers in California; pharmaceuticals in New Jersey) have a greater propensity to cluster innovative activity than industries where these spillovers are less important. Bottazzi and Peri (2002) reported that the benefits of R&D in generating innovation were extremely localised. 91 MNE – multinational enterprise. 92 Developed personal networks may not be key determinants for the selection of a location of business. However, these networks are of vital importance during the initial phases when the firm is established. Networks provide advantageous access to information, learning, cooperation, suppliers, clients … 93 Personal contacts, face-to-face communication, proximity and trust with a partner are emphasised by Kleinknecht and ter Wengel (1998, pp. 645–6), Porter and Sölvell (1998,

Theory

149

pp. 445–6), Sternberg and Tamásy (1999, p. 374), Gordon and McCann (2000, p. 520), Porter (2000a, p. 262), Crafts and Venables (2001, p. 32), Henderson et al. (2001), Venables (2001, p. 24), Audretsch and Feldman (2004), Kay (2004, p. 14), Ando (2005, p. 85), Boschma (2005, p. 69) and many others, but were also well known earlier to Perroux (1955, p. 317 and 1961, p. 152). 94 A. Taylor, ‘How Toyota defies gravity’, Fortune, 8 December 1997, p. 43. 95 It has been estimated that empty trucks travelled about 60 billion kilometres in the EU at an approximate annual cost of Є45 billion. Better transport organisation could radically reduce these figures (S. Coss, ‘Fair pricing key to avoid gridlock’, European Voice, 29 January 1998, p.15). 96 Fires in the libraries in ancient Greece, Alexandria, Rome and elsewhere may have rendered certain knowledge inaccessible and lost. 97 R. Atkins and J. Brown, ‘A stability trap?’ Financial Times, 27 May 2005. 98 ‘A Silicon Valley in Siberia?’ The Cockatoo, No. 17, August 2007, p. 4. 99 The First Industrial Revolution proceeded largely without any systematic intervention from government. Parliament did not help to build a modern financial sector, fund transportation infrastructure in a widespread and coordinated way, provide additional incentives to adopt new technologies, manage fiscal or monetary policy with growth in mind, nor help to train a new labour force with appropriate skills for a new industrial nation. Nonetheless, Britain industrialized relatively quickly.

Still, the use of the steam engine, and the rise of urban factories, created pressing policy issues. Many of these were local and specific to single uses of the engine rather than generic to steam power itself. Some were so large as to require broader policy changes. These broad policy issues are so numerous and diverse that we only illustrate them here with two cases. When it entered factories, the steam engine gave employment to men, women, and children, bringing them and their low wages to the attention of middle-class reformers who had been largely unaware of the rural poverty that had existed for millennia. Laws were passed to govern child labour and working conditions. Britain also saw the rise of powerful unions and the many laws governing them. (Lipsey et al., 2005, p. 186)

100 Public policy and administrative areas in which the authorities collect statistics are usually not the same as clusters, which is why it is hard to determine the exact economic significance of clusters with a high degree of reliability. 101 The Northern Rock (a British bank) financial fiasco of September 2007 weakened the case for domestic control and regulation of the British financial industry. Following a liquidity crisis, there was a run on the bank during which depositors withdrew about £1 billion a day over a week or so. The case for national regulation may be weakened in Britain, but what is the alternative? A heavy-handed continental-style regulation and control from Brussels? This may be subject to greater political influence that may be the case in Britain. Many in the City of London would argue that the alternative situation would be worse. 102 The assumptions on which the Gini and Ellison–Glaeser concentration indices are based are different. The Gini index deals with a concentration of firms from an industry in an area against the distribution of firms from all industries in the same area. The Ellison–Glaeser index checks the deviation of plants in an area against a hypothetical random location of firms in the same area (firms independently and randomly chose their location). An industry is concentrated if the actual concentration considerably diverges from a random distribution of firms.

Evolutionary Economic Geography

150

103 Think of plants as darts to be thrown out in space. Initially, nature chooses to put some ‘darts’ together in clusters. These are plants that are functionally linked and that will always locate next to each other. Spillovers are seized by the fraction of plants that must co-locate. In the second round, these darts (clusters) are thrown randomly at the dartboard to choose a location. 104 Urbanisation occurred before advances in modern medicine. Hence, urbanisation contributed to the worsening of health in cities. It is puzzling: why then was there a mass migration of rural population into cities? 105 UN World Urbanisation Prospects: The 2005 Revision. http://www.un.org/esa/population/publications/WUP2005/2005wup.htm (accessed on 8 January 2008). 106 Economies of scale in larger firms permit lower average costs of production and lower prices of output. Larger markets provide conditions for the supply of a larger quantity and variety of goods. 107 In the past, the colonial powers obtained natural resources from their colonies. In many cases they prevented the development of manufacturing industry in the colonised countries in order to secure those ‘outer markets’ for the export of manufactured goods from the colonial master. Thus local competition was eliminated, and if any development of manufacturing activity took place it was usually in port cities and was limited to primary processing. 108 K. Iskyan, ‘Trading places – what drives a country to switch its capital city?’, Slate, 1 March 2004. 109 New York and Philadelphia are examples, and even Boston, although its hinterland is smaller. However, why are there no major port cities in the upper south of the east part of the US? The reason is found in the colonial regulation of trade. The Navigation Act of 1660 (and subsequent Acts) regulated and reserved exports of colonial articles such as tobacco, sugar, cotton and indigo to England only. Similar restrictions did not apply to trade in wheat, corn, pork, beef, fish and other items. Merchants handled colonial trade from London, rather than from Chesapeake (Virginia). However, the Philadelphia, New York and Boston merchants were free to enter business and export noncolonial commodities to wherever they wanted. They met in coffee shops to exchange information about business opportunities. When Alexandria and Baltimore emerged in the upper south in the 1750s, their activities were based on wheat, rather than tobacco (Kim and Margo, 2004, pp. 2996–7). 110 Relevant factors that influence agglomeration tendencies for producers include the availability of raw materials, energy, labour and capital, while for consumers they include the availability of jobs and education, as well as favourable climate and surroundings. 111 Technology may also reduce in part the information ‘search and find’ problem. If something or someone cannot be ‘Googled’, well, to certain searchers that information or target object or person may not exist. 112 There are, however, regions that once were highly specialised industrial leaders (shipbuilding in Europe), but over time turned into problem areas. A current high local specialisation is not in itself a guarantee of perpetual economic success in the future. 113 Winchester was the secular and Canterbury the religious capital of England. Only when London became the largest and economically most diversified commercial and industrial hub of England after the eleventh century did it become the de facto and, later, formal capital of the country. Berlin was not even the capital of its province (Brandenburg) until after it became the largest and economically most diversified commercial and industrial city in Prussia. A similar story is found in the case of Paris, which was just one among several royal residences. Orleans was a more imposing city until the twelfth century. Paris became the real capital only after it turned out to be the largest and economically most diversified commercial and industrial city (Jacobs, 1969, p. 143).

Theory

151

114 J. Guthrie, ‘How the old corporate tortoise wins the race’, Financial Times,15 February 2007. 115 In spite of huge subsidies, a number of towns such as Wittenberge in the eastern part of Germany are ‘dying’. They have been unable to stimulate the revival of the local economy (K. Connolly, ‘Vision of unity fades in German east’, Daily Telegraph, 11 September 2004). 116 The most dynamic cities (and regions), even though they may seem to have a diversified industrial structure, in fact have a variety of highly specialised clusters of activity. 117 George Friedman, ‘New Orleans: a geopolitical prize’, Stratfor, 1 September 2005. 118 Painting is the art of silence and halted movement. Within this art form, Venice put emphasis on colour as well as free and spontaneous diluted shapes, rather than on drawing, which is the most important pra-element of any painting (Giorgione and Bellini apart). In this consumer society, many painters worked for dealers and painted quadri dozziniali. These were ordinary, low-quality serial paintings portraying saints. Many were used as cheap substitutes for tapestries. A similar type of artistic production and dealership developed in Rome around the Piazza Navona and catered to demand from pilgrims. (The Vatican and rich patrician families had totally different artistic tastes and demand.) In Florence, conversely, art was scholarly and idealistic, realistic and self-confident, with firm shapes. Paintings were commissioned by churches, wealthy merchants for their palaces, guilds, civic buildings and charity hospitals. As wealthy consumers of art were many and various, there was no single taste-setting body. Hence, artists had to be inventive in order to attract and satisfy the changing demand of their patrons. When times were good, the Florentine bankers also imported paintings from the Low Countries. As well as in Venice, quite a few important paintings can be found in museums along the routes used by merchants: from Antwerp to Cologne, Munich, Vienna, Venice, Milan and Florence (and Rome). Another branch of the location of valuable classical paintings follows the road from Antwerp via Paris to Madrid. 119 A division of Germany following the Second World War transformed German cities that were in the middle of the country into border towns and cities. Population declined in those border cities in West Germany. Fear of armed conflict, combined with loss of access to markets (on both sides of the border) and sources of supply, left these regions and cities less attractive to footloose firms and investors. 120 Openness, in particular its international dimension, is a tricky concept as it is unclear how to aggregate it over different goods. While tariffs are evident, NTBs are not, particularly if one tries to convert them into tariff equivalents. Doubts about the application of tariffs and antidumping actions increase uncertainty and affect openness and its perception. Discriminatory exchange rates and the black market just compound the analytical trouble. 121 According to the Assembly of European Regions Declaration on Regionalism in Europe,

The region is the territorial body of public law established at the level immediately below that of the State and endowed with political self-government. The region shall be recognised in the national constitution or in legislation which guarantees its autonomy, identity, powers and organisational structures. (See http://www.a-e-r.org/aboutaer/members/1.html?no_cache=1&sword_list%5B%5D=established [accessed on 8 January 2008])

122 The tacit assumption is that there are no transport and adjustment costs. 123 See Vertova (2006, pp. 7–8). 124 Institutional organisation, social regulation and political intervention may have a significant influence on the location of production. 125 Entry and exit barriers include: huge sunk costs (in project investment and advertising), economies of scale, product differentiation (for example, the local market demands a specific

Evolutionary Economic Geography

152

brand-name drink), access to distribution channels, R&D, regulation (product quality), marketing, restriction of access to complementary assets and structures (such as a computer reservation system for tickets), reactions of competitors such as predatory pricing, exclusionary pricing, as well as trade, competition and industrial policies. 126 See http://www.marxists.org/archive/marx/works/1894-c3/ch48.htm (accessed on 8 January 2008). 127 Cumulative causation mixes causes and effects of an event. They are combined in a chain reaction that is increasingly circular, snowballing, herding or perpetually accumulative. This type of self-reinforcement has different labels in economics which include: economies of scale, path dependency, virtuous and vicious circles, as well as threshold effects. The sources of this process are large sunk costs, learning, and network and coordination effects. Cumulative causation is reinforced by the following four elements: favourable demand conditions, good factor-related setting, strong market competition and fine access to supporting and related industries. 128 If this competition among locations to attract firms is relentless, then there is no equilibrium economic landscape. The neoclassical equilibrium model is silent about this evolutionary possibility and reality. 129 General sources of economic growth include market size (division of labour, economies of scale and innovation); investment (an increase in capital per worker); and technological change. Technology is taken to mean knowledge of everything (products, processes and organisation) that can create economic value (Lipsey et al., 2005, p. 10). In addition, Arthur (2007, p. 276) defined technology as ‘a means to fulfil a human purpose’ (original emphasis). This purpose may be explicit and direct: to power an aircraft. It may be multiple, changing or cloudy. A PC has no single and precise purpose. 130 Schumpeter (1983, pp. 62–3) commented that ‘static analysis is not only unable to predict the consequences of discontinuous changes in the traditional way of doing things; it can neither explain the occurrence of such productive revolutions nor the phenomena which accompany them.’ 131 There are relatively long periods when the system does not change very much followed by relatively short periods in which many things change in a profound and rapid way. 132 These rough times (like earthquakes or epidemics) have an interesting property: the stronger they are, the rarer they become. An exception, hopefully temporary, may be summer forest fires. 133 Invariance is the attribute of God. 134 Cockroaches are among the oldest living and most unchanged creatures. Fossil remains have proved that these insects have not changed over the past 250 million years or more. It is likely that they may remain as they are, with little change, perhaps for another 100 million years. They are sophisticated living creatures as they can withstand freezing, a radioactive environment and several months of famine. After external ‘perturbations’, cockroaches return to normal life. 135 It is supposed that this was an iridium meteor with a diameter of 10km that fell in the Yucatan area of Mexico. 136 Herbivorous dinosaurs were deprived of food and they starved. This deprived carnivorous dinosaurs of their prey, so they starved too. In addition, some mammals, rather small in size at that time, were feeding themselves with dinosaurs’ eggs. 137 It is also likely that dinosaurs came into being following an earlier meteor impact 251 million years ago. 138 Barbarians, to the Greeks, were all those who did not speak Greek. In our context, the barbarians were the non-Christians. They are taken to be ferocious foreigners who destroyed ancient Roman society, administration, law, education, culture and art. They were logistically mobile, culturally used to violence and fighting, physically tough and ethically

Theory

153

wild, as their faith did not include the Christian prohibition of taking the lives of others and respect of other tribes. 139 Because of the availability of wood, Sweden developed its huge pulp and paper industry. Strong links with suppliers contributed to a similar success for the machinery involved in the paper-production processes. 140 The first ‘modern’ corporation-like business organisation in Europe was the Church. Later came the guilds and universities. 141 R. Anderson and R. Atkins, ‘The pressure to end the 35-hour week shows that big companies are prepared to take matters into their own hands’, Financial Times, 23 July 2004. 142 L. Hockstader, ‘Enterprise belt pulls Italy into Europe’, International Herald Tribune, 6 November 1997, p. 13. 143 Consider, for instance, the applause in a theatre. Following a performance the audience will soon spontaneously start clapping at the same tempo even though nobody ‘conducts’ it. Such self-organisation occurs when many initially uncorrelated actions lock into one another’s rhythm and create a strong collective group. As Polonius said: ‘Though this be madness, yet there is method in’t’ (Shakespeare, 2000, p. 71). Or another example: imagine queues at a supermarket checkout. Which one to join? Clients judge the waiting time according to individual criteria which include: the number of persons in each queue; how full their baskets or trolleys are; how likely it is that those ahead will unload their shopping quickly; who is likely to write a cheque; who will chat with a cashier; and so on. Nobody directs these supermarket clients, but some spontaneous order is established. 144 Information in the physical world is transmitted via flows of particles such as electrons or photons and density gradients. In the social world (and to a lesser extent in the biological) there are the controls of information flows from the exercise of the principle of optimum loss of detail and use of near-decomposability within hierarchical structures – and even more so within the intricate interconnections of complicated hierarchical structures that may be embodied in an heterarchical model conceived by a human mind.

(Isard, 1996, p. 357; original emphasis)

Even though there are difficulties in applying biological analogies to social phenomena, they were widely used in the analysis of firms. Examples include the ‘life cycle’ theory of the firm, ‘viability’ of the firm and ‘homeostasis’ (habitual return to the initial equilibrium) (Penrose, 1952). 145 The relation between expectations and accomplishments is essential for satisfaction. 146 Security is a legal right given to a creditor (secured party) by a borrower. Securities include shares, bonds, stock options and transferable interests in commodities such as oil, metals or grains. Securities may be traded on a secondary market. 147 http://www.gutenberg.org/etext/12 (accessed on 8 January 2008). 148 Other hypotheses refer to geography (resources, climate), history and culture. 149 Australians were prevented from selling their wines under French names such as champagne, chablis or claret. So they simply rebranded their products as sparkling wines, cabernet sauvignon and chardonnay. The French were subsequently forced to rebrand obscure labels with various names to compete with the Australians on the British market (‘Quality counts, not geography’, Financial Times, 28 July 2003). 150 ‘IMF’s Camdessus misses the point’, Stratfor, 15 November 1999. 151 Honda, Google and Mittal Steel, for instance, were also companies that emerged from ‘nowhere’. In addition, Toyota was an unimportant maker of textile machinery in Japan. However, it grew to become one of the top worldclass car manufacturers.

Evolutionary Economic Geography

154

152 GPT (general-purpose technology) is the one that is widely used for multiple purposes and has many spillovers that enable the operation of dependent and derived technologies. New GPTs encourage and rejuvenate economic growth. 153 Acceleration in productivity (growth in output per hour) is the basis for an increase in living standards and economic progress in the long term. 154 Some even thought that economic cycles were dead. 155 The performance of computer chips at a given price doubles every eighteen months (Moore’s Law). However, this does not mean that the impact of computers on productivity follows this trend in a linear way. It depends who uses computers, how, when and for what purpose: management of stocks or computer games. 156 R. Gordon, ‘America wins with a supermarket sweep’, Financial Times,19 August 2003. 157 ‘What the Internet cannot do’, The Economist, 19 August 2000, p. 9. 158 The Wal-Mart effect (better use of information technology; hire-and-fire employment and management culture) transformed the American economy as much as it changed the landscape. The European employment and spatial planning laws and practices prevented productivity growth from taking off in retail services. This may be the result of European producers’ power at the expense of (poorer) consumers. 159 B. Mongoven, ‘The US energy debate: whether to bet on future technology’, Stratfor, 2 August 2007. 160 Western Russia is rather flat and geographically open to a frontal attack from the west. However, seasonal swamps made by the melting snow in the spring and rain in the autumn make an overland frontal attack on Russia impossible or unreasonable in March and October. 161 The first of many attempts to produce a small, economical run-around vehicle early in the twentieth century resulted in the Auto-Ped. Introduced in New York in 1915, this looked like a child’s scooter: it had no seat and there was a platform for the rider to stand on. A twohorsepower motor gave the Auto-Ped a maximum speed of 55km/h. 162 In an attempt to eliminate textile waste during the Second World War, the US government reduced by one-tenth the amount of fabric allowed for clothes. This also applied to women’s swimwear. The chain of events that resulted in the ever-decreasing amount of material necessary for the bikini was set in motion. 163 Even though a Roman Catholic country, France fought towards the end of this war (1636–48) on the side of Protestants. France feared the rising strength of the Habsburgs on its eastern frontier. 164 The worry about this exodus of manufacturing jobs to China ought to be mitigated, at least a bit and in the medium term. China is a grand processing economy. Its weakness lies in the lack of raw materials and energy. The big money does not lie in manufacturing. It lies in product design, branding and marketing. At least for now. 165 The Japanese understood this and implemented a policy of domestic rivalry among half a dozen conglomerates. This became the basis for the global competitiveness of their output such as passenger cars and consumer electronics. They also transferred it elsewhere. For example, after the failure of Leyland in the 1980s, Britain developed a relatively dynamic car production and market based to a large extent on local production by Japanese TNCs. 166 Exporters from the coastal African countries such as Kenya and Tanzania cannot ship their cargos to the US directly, as there is no direct shipping service to the US. They have first to ship their goods to Singapore, where the goods are aggregated and then shipped to the US. 167 Recall the example about Ferguson tractors. 168 Coalmining regions weakened and declined when liquid fuels were used as principal sources of energy. 169 A brand is worth more than a functionally equivalent product … People can and do pay money for access to these brands … Why would people pay more when they can get the same for less? Most often, markets will tell us, because they do not think it is the same … The most important function of brands is quality certification.

Theory

155

(Kay, 2006, pp. 102–4) So, brands may tempt the self-doubting. You are reading a book by Miroslav N. Jovanović. You might, perhaps, ignore the same text if it appeared under someone else’s name. If this is the case I have a brand with some value to it. But only then.

3 Regional policy My people perish for lack of knowledge.

(Hosea 4:6)

1 Introduction Regional policy is relevant not only for the spatial distribution of economic activity, but also because it often involves big money. It may be prone to influences, even abuse, by various interest groups. In larger countries composed of many regions and often inhabited by different nationalities, regional and social policies play, in part, the role of the state ‘glue’. Losers may accept such a negative status, but not for too long. Hence, politicians find justifications for intervention. The purpose of this chapter is to provide a brief overview of regional policy. Section 2 sets the stage for considerations to follow. Objectives and justification for regional policy are presented in Section 3. Section 4 lists policy instruments, while Section 5 considers the impact of international economic integration (market widening) on locational issues. Special attention is devoted to Europe as this is the most researched integration area. Section 6 concludes the analysis. 2 Background Given that remarkable spatial concentration is the most striking feature of the geography of production, there is clear evidence of some kind of increasing returns to scale (Krugman, 1992, p. 5). New technologies in select industries may overcome some of the obstacles to the spread of production, but not many. Hence, the pattern of regional specialisation and trade can be arbitrary and potential gains from specialisation and trade are likely to be ambiguous.1 Production and trade in goods are increasingly becoming elaborate and difficult. They depend heavily on sophisticated information and services that are included in the final good and its price. Internal trade within each TNC may account for a large part of the total trade. Hence, national and international markets become more and more incomplete and tough to manage. In addition, the listed new elements that the divergence theory considers exert a strong influence on the location of production, trade and absolute advantages (which widen regional gaps), while local comparative advantages may play a less important role. In this case, competition, trade liberalisation and economic integration generate concentration (or increase the attractiveness of certain already developed areas); they generate economic divergence and widening of regional gaps, not

Regional policy

157

economic convergence among regions and countries. This holds at least in the short and medium term. A lack of R&D and innovation, as well as acquisition and employment of new technological knowledge, may be among the principal causes of slow growth in lessdeveloped regions and countries. This weakness may be partially eased by the two-way labour mobility from developed to weaker regions and vice versa. A one-way migration of labour would only depopulate weaker regions, while it could create overcrowding in the prosperous areas.2 However, structural disparity among the regions does not necessarily mean that there is an income disparity between them. If there are two regions and if each region specialises in a different output activity, then output structure among them would change, even though the incomes in both regions may still remain comparable. In a simple model (Figure 3.1), the allocation of manufacturing between two regions determines the difference in real wages between them. Let the horizontal axis represent the share of manufacturing workers residing in Region 1 and let the vertical axis show the percentage difference in real wages between the two regions. The figure shows how various transport costs influence the curve that relates the regional manufacturing population to real wages. Transport costs are taken to epitomise all costs of doing business across geographical space. With high transport costs there is little interregional trade. Regional wages depend mainly on the local conditions in the labour market. If there is an increase in the regional labour supply, competition reduces wages. With low transport costs, there is much interregional trade. A firm can have superior market access if it is located in the region with a higher concentration of workers (consumers). It can also afford to pay higher wages. These higher wages increase purchasing power as the workers have better access to consumer goods. As the number of residents/workers increases in the region, real wages increase in the same direction. With intermediate transport costs, there is some evening out between centrifugal and centripetal forces (Krugman, 1999, p. 95). Figure 3.2 shows the (‘tomahawk-like’) situation in which the workers move to the region that offers higher real wages measured by the share of manufacturing labour force in Region 1. A set of equilibria depends on transport costs. Solid lines specify stable equilibria, while broken lines show unstable equilibria. If an economy starts with high transport costs, then there is an even distribution of manufacturing between Regions 1 and 2, respectively. Point A represents this situation. Then suppose that there is a

Evolutionary Economic Geography

158

Figure 3.1 Relation between regional manufacturing populations and real wages with varying transport costs.

Figure 3.2 Transport costs and equilibria. reduction in transport costs and that production and consumption can be separated (so the concentration of production in one of the two regions becomes possible and likely). The economy reaches Point B in this (putty– clay) situation. At this point the process of concentration of manufacturing in one region starts. Which region gets the industry often

Regional policy

159

depends on a small starting difference or on a chance. The economy would spontaneously organise itself into core–periphery geography (Krugman, 1999, p. 96). A relatively high inter-regional labour mobility is a feature of the US and Chinese economies, but not within the EU. However, agglomeration can be generated equally well by interplay between the location decisions of firms in industries that are linked through an input– output structure. Thus, even without labour mobility, there may be forces leading to agglomeration of activity at particular locations in an integrating region. (Venables, 1996b, p. 356) Evolutionary economic geography models have a number of different properties in comparison with standard neoclassical ones. Some of these properties refer to factor mobility, while others, more importantly, arise from nonlinearity. This situation produces multiple equilibria and bifurcations. Baldwin et al. (2003) considered new issues in this research area. When an industry is already clustered, agglomeration forces produce inertia that makes small policy interventions ineffective regarding spatial location of business. Agglomeration (clustering) produces rents which holds firms and factors in a certain place, even if policy action would allegedly lead to a geographical move. There is, however, a threshold. Once costs to firms and mobile factors outweigh the agglomeration rents, firms and mobile factors would move elsewhere. Here comes the principal and sharp difference with the standard neoclassical framework in which small changes in policy lead to small economic responses. Economic geography models predict dramatic non-linear effects in such circumstances. A small change at a crucial point in time (once the threshold is reached) can lead to shocking outcomes (a ‘locational catastrophe’). ‘The possibility of non-linear responses to policy makes it much more difficult to forecast the effect of a given policy change … it can be seriously misleading to base expectations of the effects of future policies on linear approximations from the past’ (p. 229). Trade liberalisation, economic integration and other policy changes may have very non-linear effects on industrialisation, location of firms and growth. Even in a next to perfectly homogeneous world, production activities will tend to cluster, not least because of a different endowment of resources and because of economies of scale. In these circumstances, government intervention in the form of various policies (industrial, regional, competition, tax, trade, technology and/or education) may find some justification. If policies mattered, then one would find an abrupt change in production activity (and material wealth) when crossing the (national) frontiers at which policies change. This was obvious between eastern and western Europe during the second half of the twentieth century or between North and South Korea. The policy is generally aimed at developing or enhancing the regional or national production potential, competences and capabilities through a (supposedly) superior allocation and use of resources from the national standpoint. The objective of such intervention is to influence the spatial and industry distribution of economic activity, as well as to create and redistribute wealth in order to ease and, eventually, solve the manysided ‘regional problem’. Governments intervene in regional matters mainly for the following three reasons:

Evolutionary Economic Geography

160

• Equity This is a strong social motive based on public pressure on the government to try to achieve and maintain a ‘proper’ balance, as well as an ‘orderly’ and just geographical distribution, of national wealth and access to public services in different regions. • Efficiency This is a desire to employ, sustain and increase national economic potentials and capabilities. • Strategic behaviour This gives public authorities a chance to shape comparative advantages and influence the output potential and capabilities of the country.3 In the 1960s, the regional problem was usually tackled with supply-side subsidies for the provision of infrastructure and the reallocation of various (public) manufacturing and service industries. Foreign competition, in particular from the newly industrialised countries, placed in jeopardy a number of industries that failed to adjust. As there had been little success with earlier approaches to the regional problem, coupled with austerity programmes in public finances there was a major change in national regional policies from the 1980s. Outright subsidies were reduced or removed and the policy was supplemented by a system intended to make the lagging regions more selfreliant (support to indigenous development). This included the development of human resources, the attraction of private investors (in particular foreign ones) and the provision of technical services. There is a great deal of uncertainty regarding the impact of international economic integration on regional matters. One thing, however, should be mentioned here. Having a ‘peripheral’ location is not an irreversible economic disadvantage. Its impact can be mitigated and even reversed, as has been successfully shown by countries such as Ireland and Finland (or Australia and Japan). The peripheral EU countries became economically more dynamic in terms of growth rates than the EU ‘core’ (France, Germany and Italy) from the 1990s (but the periphery also started from a lower base). The EU Cohesion Policy was introduced in 1989. Interesting questions for further analysis arrive by themselves. Why did strong development take place in peripheral Ireland and Finland, but not in peripheral Greece and Portugal? Why have certain Spanish regions (Catalonia or Valencia) grown faster than the national average, and why have others (Galicia or Andalusia) remained behind? What matters most for the competitiveness of a country’s goods and services on the international market is the efficient creation and employment of and compensation for the most precious economic factor: human capital (heads and hands). However, this asset may depreciate over time through indolence, ageing and vices such as drugs and heavy drinking. Therefore, it needs to be maintained through continuous education and training. Craig Barrett, the chief executive officer of Intel, referred to economic liberalisation and opening in China, India, the former socialist bloc countries and other newly emerging countries during and beyond the 1990s. He noted that the reform process touched about 3 billion people, 10 per cent of whom are highly skilled workers. ‘That’s 300 million welleducated people, able to do any job.’4 To face this and other challenges, Intel would invest heavily in R&D and hire the best and brightest employees wherever the company could find them. Some consider that the external challenges from educated, experienced and productive labour and management could be a threat to skilled workers in America and contest the prosperity of the US economy. This may be an overstatement if one looks at the short and medium terms. It takes time and resources to establish and spread first-

Regional policy

161

rate educational services and equip labour and management with the necessary sophisticated knowledge, training and experience.5 However, Barrett’s statement and Intel’s intentions highlight the crucial issue: the great importance and lasting value of human capital for the prosperity of firms and an economy. Regional policy was neglected in the EU before its first enlargement in 1973. This was followed by a period of interest in an active regional policy. However, enthusiasm for this policy waned in the 1990s as there were few obvious positive results from intervention. Public intervention in regional issues is still a highly controversial issue. Is there or is there not a selfequilibrating process among the regions in the long term? Is intervention (regional policy) necessary or not? Hence, can regional policy work at all? If yes, under what conditions? National governments introduced regional policies in order to promote growth in weak regions. The policy was based on the assumption that market failures exist. The instruments of the policy included investment aid, direct investment by the state, wage subsidies, tax allowances, licensing and the provision of infrastructure and services. However, since the 1990s, the emphasis of regional policy has moved away from the attraction of extraregional investment and general subsidies, primarily towards the development of endogenous regional growth and human resources. The national regional policy was transformed in most ‘old’ EU(15) countries into an enterprise and entrepreneurship development policy. The situation regarding regional policy at the EU level has been ambivalent. On the one hand, cohesion is becoming a leitmotif of European integration, hence EU involvement in this issue. On the other hand, market liberalisation and an active regional policy (intervention) do not go hand in hand. Since an automatic market-led adjustment mechanism operates too slowly to be politically acceptable, EU structural funds for regional development and social issues emphasise supply-side intervention. They provide assistance to infrastructure, training and structural adjustment. Intervention in the form of a regional policy, or for that matter any policy, is not a simple task as there are constant changes in technology, taste, competition, demography and available funds over time. In spite of intervention in regional affairs over the past decades, a trend appeared during the 1990s and beyond in many EU countries towards abandoning or easing old regional policies. Many regional problems remained immune to change. However, at the EU level, intervention in regional affairs remained enthusiastic, but with questionable results. 3 Objectives and justification Regional policy is intervention by the public authorities in order to influence the ‘orderly’ spatial distribution and ‘just’ direction of economic activity, as well as to reduce social and economic differences among regions. Regional policy also attempts to increase the future national GDP, but this goal may often clash with an orderly spatial distribution of economic activity in the short and medium term. It was usually a reactive (ex post) policy that primarily tried to reduce the existing regional disparities, rather than a policy that primarily prevented the creation of new regional disequilibria. A forward-looking policy attempts to boost the competitiveness and profitability of the available local resources

Evolutionary Economic Geography

162

and capabilities. It also tries to invite and keep mobile resources from elsewhere that would enhance the future operation and profitability of the locally available mobile and immobile resources and capabilities. Trade liberalisation and/or economic integration can easily provoke a country’s economic adjustment (geographical, sectoral and professional reallocation of resources). Enhanced competition may force certain regions to embark upon a painful, but potentially rewarding, alteration in the structure of the geography of production (transfer of resources from unprofitable into profitable economic activities). The politicians in the affected regions often blame integration, globalisation or trade liberalisation for adjustment ‘pains’. However, this objection may not always be justified. The most basic reason for a painful adjustment process is often earlier protection. A policy of long and excessive sheltering of the domestic economy reduces the reaction time of local enterprises to international structural pressures and opportunities and, hence, increases the cost of adjustment in the future. Adjustment costs and ‘pains’ are always present, but the potential benefits from such a change may more than compensate for the effort. Relatively low wages may attract some wage-sensitive investment into a region; however, at the same time there may be some agglomeration and concentration tendencies in other regions. If wages are set at a national level, this may act as a structural barrier that works against the lagging regions. In this case, the lack of flexibility in the level of wages among regions despite relative differences in productivity and in the labour market conditions means that the less-developed areas are unable to respond to this situation by reduced wage costs. One cannot know for sure in advance where the balance will tilt. Some developed regions may become more developed, while some lessdeveloped regions/countries may become poorer (if nothing is done to alleviate the situation), but other less-developed regions may also move up into the group of advanced regions. Hence, economic development is not necessarily a steady linear evolution, but it may be in certain cases a rather big jump from the poor to the rich club. This fits well with the historical record (just recall cases of ‘Asian tigers’). Also, China and India had a relative economic decline during large parts of the nineteenth and twentieth centuries. Such a development is totally different from the one predicted by the standard neoclassical economic theory of a relatively smooth, steady and positive convergence. The objectives of a regional policy are various, but their common denominator is that they aim to employ the regions’ unemployed or underemployed resources and potentials, to locate new ones, to attract or create if possible the missing factors, as well as to increase output and incomes. In congested or polluted areas, this policy restricts the expansion of firms and stimulates exit from the region. In the developing countries, the primary concern is economic development. This is most often coupled with regional and sectoral imbalances, but their solution is not as high on the agenda as increases in economic potentials and capabilities. In areas of economic integration, the problem of regional (in)equalities is of great concern. Countries are reluctant to lag behind their partners for any extensive length of time. This is of great concern in an economic and monetary union, where there are no balance of payments disequilibria, but rather regional disparities. Depressed regions/countries can no longer resort to devaluation as there is a single currency, while advanced regions/countries will not always be willing to finance regional disequilibria indefinitely without proof that there is structural adjustment and

Regional policy

163

improvement taking place in the regions assisted. With this in mind, a regional policy can be justified on at least nine counts. (1) In contrast to an automatic free-market adjustment, intervention is supposed to assist in the process of economic change in a desired way. The first justification for a regional policy can be found in the structural deficiencies of regions. These include market rigidities, conditions of access to the market and the structure of output. When such imperfections exist, a free-market system is ‘unable’ to achieve a satisfactory equilibrium from the social standpoint in the short and medium terms. So there may be a need for intervention in order to enhance the creation and use of the regional income-generating potentials and capabilities. (2) Regional ‘imports’ consist of goods and services. In France, for example, most of the country’s regions purchase financial, insurance, legal and other services from Paris. This is not easily quantified. A reduction in a region’s ‘exports’, for instance to Paris, may be obvious from the outset in the form of reduced orders and increased employment. Regional solutions to these disequilibria are inhibited by rules of competition and by the existence of a single currency. Herein we find the second ‘justification’ for regional intervention in an economic union. (3) The third reason may be found in the employment of factors. The neoclassical equilibrium theory of international trade usually assumes in its pure form free and full international mobility of factors, which ensures full employment everywhere at the end of the day. However, even Adam Smith noted that a man is the most difficult ‘commodity’ to transport. During recession, the employment situation is tough everywhere, therefore the potential advantages of some regions that are abundant in labour are removed. Reduced mobility of labour prevents the equalisation of discrepancies in single countries and in economic unions. Suppose that economic space consists of a core and a periphery. In the case when there are market imperfections, there is a general agreement in theory that at one point in time the core region is likely to dominate the periphery. The disagreement is about the nature, extent, evolution, effects and duration of the core–periphery relation over time.6 If there is growth in one region, this may create conditions for even higher growth in the same core region and increase inequalities among regions (polarisation or backwash effect). This may, however, also act as a ‘locomotive’ for development in other peripheral regions as there may be increased markets for food, economies of scale and innovation in other regions (spread effect). In order to adopt the idea that there are benefits from the spread effect, one must assume that there is a complementarity between the regions in question and ignore the fact that the spread effects operate in an asymmetrical way throughout space. In a situation closer to reality and which includes economies of scale, large sunk costs, externalities and other market failures, adjustment does not happen according to a relatively smooth neoclassical expectation. The regional absorption capacity of new technologies requires an industrial and general culture that incorporates continuous learning, adaptation, spread and development of human capital, as well as flexibility. Therefore, the introduction of new technologies is often uncertain and quite risky, which is why some technologies diffuse quite slowly among the regions. The reallocation of resources induced by the spread

Evolutionary Economic Geography

164

effect works at a (much) slower pace than may be politically acceptable, hence the need for certain intervention. (4) The fourth rationale can be found in the ‘need’ for compensation. Two forces created by national or international economic integration, as well as trade liberalisation affect regional development. The first one, specialisation, leads the backward regions to specialise in labour-intensive production according to their comparative advantage. The second force includes economies of scale, externalities and path dependence. This can lead to a divergence in regional incomes. As integration extends the size of the market, firms take advantage of economies of scale and externalities. The EU Single Market Programme (1985–92), for example, provided incentives for the concentration of industries with strong internal production links. In the case of other regions, the programme reduced trade costs, hence there are potentials for a spread of other industries that may benefit the periphery, where labour costs are lower. However, if this does not happen, or takes place very slowly, there is a possibility that some regions may actually be damaged by national or international integration if there are no instruments for compensation. This justifies the existence of a regional policy. Subsidies may be one of the tools of a regional policy. However, the distribution of subsidies is always subject to political pressures. It is often linked with certain mismanagement of funds, even abuse of public authority. The final disbursement of subsidies often reflects more the balance of political powers than the comparative (dis)advantage or actual need of a region. At the end of the process, regional policies implemented in such a way may do more harm than good if they diminish local incentives for adjustment and flexibility or, at best, they may have a dubious effect. Nonetheless, public policies in the backward regions, such as support for education, infrastructure and selected public goods, as well as aid for SMEs in the form of loan guarantees for the start-up of a business, have the potential to assist regional development. (5) The fifth reason is the improved allocation of resources. When the market is imperfect, ‘free’ market forces usually direct capital towards already developed regions in the short and medium terms. Certain private investors tend to maximise the speed of the safe return of their invested funds and to minimise investment in the local infrastructure. Thus, it is understandable why they direct their funds towards already advanced regions. These tendencies of agglomeration (adjoint locations of linked productions), where the developed regions tend to become more developed while the underdeveloped remain at best where they are, have significant private benefits, but also social costs and benefits. A society may reap the benefits of large-scale efficient production. However, if private entrepreneurs are not impeded in their decisionmaking by government policy, the geographical location of their business may introduce significant social costs such as pollution, congestion and traffic jams in some regions and unemployment and increased social assistance in others. (6) The sixth reason lies in the improvement of stabilisation (macroeconomic) policy. Regional differences in rates of unemployment may reduce the opportunities to control inflation and introduce a stabilisation policy. The reduction of inflation in some regions may increase unemployment in others. This may not always be the desired outcome. Diversified regions with a variety of employment opportunities will be able

Regional policy

165

to adjust in a less painful way than specialised regionswith entrenched market rigidities. (7) A regional policy may reduce public expenditure in the assisted region in the long run. Public support to firms to locate in certain regions may attract economic activity towards these regions. Unemployment may drop, reducing welfare payments and in the long term increasing tax receipts that may be spent on something else. In addition, ‘employment’ in public administration used to provide a shelter for the unemployed. Such artificial employment may be reduced or even eliminated once the private sector (with superior wages and career prospects) starts thriving in the assisted region. (8) Although regional policy is targeted at regions with some disadvantage (underdevelopment, unemployment, obsolete technologies and output structure, congestion or pollution), the benefits of a regional policy are not confined to the assisted area. Other regions, through externalities, enjoy part of the benefit. The beneficial effects of a regional policy extend beyond the assisted area itself, which is why the Germans and Swedes have an interest in helping out the Poles or the Romanians within the EU. In general, integration may be reinforced, unwanted migration of labour may be prevented, factors may be employed in a superior way and there are important non-economic gains. (9) Finally, in addition to the above arguments for regional policy, which mostly deal with ‘economic’ efficiency, there are political grounds which are at least as important as the economic ones. Solidarity, tolerance and perception of a common future are the core of any social community. The mitigation of economic disparities among the constituent regions may be the necessary reason for the unity of a state or an economic union. This is relevant as the costs and benefits of international economic integration tend to be unequally spread among the participating countries in a situation with market imperfections, large sunk costs, heterogeneous firms and multiple equilibria. Arguments of equality require the solution and/or mitigation of intolerable or growing differences in the distribution of wealth among the population which lives in different regions. The national political system does not always take fully into account the needs of backward regions. For example, the national system of setting wages may significantly reduce the wage-cost advantages of many regions, whereas welfare expenditure can contribute to a greater regional equilibrium. Complete equalisation in the standards of living in different regions is neither possible nor desirable because it may reduce incentives for change and improvement. What is needed in regional matters is a continuous adjustment of regional development within commonly agreed realistic guidelines, as well as the protection of the standard of living accepted as desirable by the group. Regional intervention may be quite costly. In addition, there may be an inefficient replication of effort at various levels of government. Therefore, one may use the coordination advantages argument in favour of national regional policy. In the case of internationally integrated countries as is the case in Europe, the EU is best placed to coordinate various national and regional actions, to improve their efficiency and to collect and transfer resources from the prosperous to the weak regions of the EU if that is the policy objective. The goals of regional policy, such as balanced growth, equal shares of social and cultural progress of the society, solidarity, regional distinctiveness and stability, are

Evolutionary Economic Geography

166

vague and may not be accurately measured. Specific objectives such as job creation, reduction in unemployment, improved housing and development of infrastructure introduce fewer problems in quantification. Care has to be given to select policy tools in order to ensure that the regional policy does not evolve into protection of the present economic structure; rather, it needs to assist in continuous evolutionary structural adjustment and necessary change of the regions. Government intervention (policy) can shift this short-term perspective towards longerterm economic considerations. However, less than perfect foresight can lead banks into bad loans, which can culminate in financial crisis, as occurred in the developing countries during the 1980s or in Japan at the end of the 1990s or the US in 2008. Japan and the newly developed countries such as Korea had been held up as examples of a successful manufacturing-led development path based on intervention. These Asian countries were delivering ‘good news’ for a long time over the 1970s and 1980s. Much of the credit for these achievements went to government planners who ‘knew what they were doing’. However, when serious regional crises emerged in the late 1990s (essentially because of lightly regulated banking, which was subject to severe moral hazard problems), the truth was revealed: they didn’t know. Sceptics argue that Japan and Korea would have had an even steeper growth curve had it not been for selective intervention. Given macroeconomic stability, equilibrium and a stable exchange rate, high and stable savings and investment rates, an enterprising spirit, a respectable level of education, relatively competitive labour markets and a relatively liberal trading system were more than enough to stimulate even faster growth. The contribution of selective intervention was negligible or harmful (Pack, 2000, p. 51). 4 Instruments Instruments of regional policy are often directed towards entrepreneurs and factor owners. They can be employed either directly (for example, support to exit from existing technologies or shift towards new technologies or business activities) or indirectly (for instance, improvement in infrastructure). Their joint effects may alter, improve and/or increase employment, investment and output in the assisted regions, at least in the short term. The dilemma of the public authorities may be whether to stimulate regional development through private investment, to invest directly in production and infrastructure for which there is no interest in the private sector and/or to stimulate public–private partnership. The available policy tools include (1) those that provide incentives and disincentives to firms and labour to move and to locate in or out of specific geographical regions and activities and (2) those that are intended to alter the level of income in the target regions. The principal policy instruments include: • tax concessions; • subsidies: regional allocation of capital, investment, infrastructure, output, social security, education and income/wage; • vocational training; • public procurement and provision of goods, services and infrastructure; • loan guarantees;

Regional policy

167

• reduction in interest rates; • protection of intellectual property; • decentralisation of government offices, education and health services; • reductions in energy and public transportation costs; • free locations; • licences for the location of business; and • trade protection. As seen from this list, the choice of policy tools is rather limited. Some of them may be hotly disputed. The secret of success in this situation is not the wide range of policy instruments, but rather their astute combination and application. If offered, cash grants may be preferred to reduced tax liabilities as regional policy instruments. Grants apply horizontally and directly to all firms, whereas reduced tax liabilities help only those that are profitable. Trade restrictions are not the wisest instruments of a regional policy in the medium and long term. The costs fall on the whole country or economic union, but they bring benefits to one or only a few regions (or social groups). Disincentives for regional expansion in congested areas appeared after the Second World War. In Britain, for example, they appeared in the form of industrial development certificates (1947–81) that had to be obtained prior to expansion or location in nonassisted regions. However, after the second oil shock (1979), as unemployment increased, these certificates were abandoned as an instrument of regional policy. In France, for instance, there were certain constraints on the expansion of manufacturing and service industries in the Paris region. They included a special tax on office floor construction and authorisations for large-scale investments. The policy of moving workers to jobs views the regional problem as being exclusively one of unemployment. It ignores the fact that other problems in a region may be made worse by this ‘forced’ mobility of labour. The productive part of the population moves out of the region while the consuming part remains in it. The shrinking local tax base may not be able to provide sufficient funds to cover the costs of local health care, schools and other social services. The overall situation may be worsened by the multiplier effect. In advanced regions that receive the migrants, new disequilibria may be created. The additional population increases congestion and rents for a certain type of housing, which may reduce the quality of life there. Education and training are often emphasised as key elements in the solution to the unemployment problem. Although they are necessary, they are not sufficient conditions to fight unemployment. Creating new jobs also requires investment and increased flexibility in the labour markets. Unemployment rates may be one of the most telling indicators of regional variations. If labour mobility is relatively low, then the movement of jobs to workers may reduce regional disparities in unemployment. One argument against this is that it may increase costs, as alternative locations may not be optimal for the efficient conduct of business. Movement of jobs in finance from the City of London or Wall Street to Norwich (England) or to Albany (New York), respectively, would be neither reasonable nor efficient for business. This business is successful because it is located where it is. However, as more industries become footloose, these costs to firms may diminish in certain cases. Improvements in infrastructure, including the training of labour and the education of management together with the spread of timely information, help in the

Evolutionary Economic Geography

168

decision-making process about the location of footloose industries which are at the beginning of their product cycle in the assisted regions. 5 Impact of international economic integration on location Basic theory Consider a case illustrated in Figure 3.3, where substitution in consumption is allowed for by smooth and convex indifference curves (Lipsey, 1960). There are three countries, A, B and C, and two goods, X and Y. In the free trade case, country A trades with country C and achieves indifference curve II. Suppose now that country A introduces a nondiscriminatory ad valorem tariff on imports. The relative price in country A is now AT. Suppose that this tariff does not give enough incentive to home entrepreneurs to embark upon production of good X. Country A achieves the indifference curve I1I1 with equilibrium at point G. If the government either returns all tariff proceeds to consumers or spends the entire amount in the same way as the consumers would have otherwise done, then the equilibrium should be on line AC (as country C is the best foreign supplier). The equilibrium point is at point K, which is the point where the line T2 (parallel to AT which illustrates compensation of consumers) intersects with terms of trade line AC. The extent of the rightward shift in the terms of trade line depends on how much people are willing to import at these prices and, hence, the volume of tariff revenue

Figure 3.3 Welfare in a trade-diverting customs union.

Regional policy

169

to be returned to them. T2 is deliberately drawn in such a way that K and L lie on the same indifference curve. The tariff has changed the structure of production and relative prices. Consumption of the home good increases, while imports and exports decrease. Suppose that country A forms a trade-diverting customs union with country B. The terms of trade line with country B is illustrated by line AB. Suppose that K and L lie on the same indifference curve I2I2. The formation of a customs union has not changed country A’s welfare, although the structure of consumption has changed. If the best situation is at point E, then the formation of a customs union for country A is a move from one sub-optimal position K, to another sub-optimal position L. Country A is indifferent. If country A obtains in a customs union terms of trade which are worse than OB/OA, then country A is worse off than in the situation with a non-discriminatory tariff. If country A, however, obtains terms of trade which are better than OB/OA, then a tradediverting customs union can be welfare improving for this country. Hence, the classical statement that trade diversion is always a bad thing is rejected. In addition, country B’s producers get more orders from country A consumers. The location of at least a part of production moves from country C to country B. Potential gains First, note that even an announcement and preliminary negotiations about a serious integration deal have a positive and significant impact on the volume and direction of trade and FDI. The period before the first eastern enlargement of the EU (2004) provides clear evidence for this observation. The 1993 Copenhagen criteria for EU entry and the publication of Agenda 2000 (in 1997) represent these announcements and turning points that influenced in part the location of production in the countries in central and eastern Europe. Elsewhere, the same happened in the case of Mexico before the start of NAFTA in 1994. Similarly, if Japan, Korea or the EU announce a trade or investment ‘war’ with the US, the location of their production business in Mexico or Canada as the export platform for the US may become quite an attractive option to consider. Economic integration may increase the average welfare of consumers in the involved countries in many different direct and indirect ways. The problem is that the benefits accrue to everyone in relatively small instalments and only in the medium and long terms. Some of these potential gains include: • secure access to the market of partner countries; • increased investment opportunities as expectations may be established with an increased degree of security; • improved efficiency in the use of resources; • elimination of trade barriers reduces cost of trade; • the market of the integrated group provides more opportunities for a wider range of goods and services that can be offered to consumers, hence there is an improvement in individuals’ utility function; • increased competition on internal market puts a downward pressure on prices; • facilitation of exchange of technical information; • competition forces firms to apply new ideas and technologies; • because of the enlarged market, producers may better exploit and benefit from economies of scale;

Evolutionary Economic Geography

170

• potential for coordination of certain economic policies; • improved bargaining position with external partners; • research and innovation may be stimulated because of tougher competition on a larger market and a possibility of sharing fixed costs in such a market environment; and • reduction in X-inefficiency which moves the production activities of firms closer to best-practice business organisation. In spite of these attractive potential gains, one must always bear in mind that international economic integration is never more than a useful supporting tool for sound domestic macro- and microeconomic policies and that it cannot act as their replacement. If these domestic policies are not healthy, integration cannot be their substitute. Reallocation of resources and adjustment costs Economic integration affects welfare through potential increases in product variety and consumption. It was assumed, along the lines of a neoclassical equilibrium model, that adjustment of production (shifts from unprofitable into profitable activities) is almost instantaneous and costless. This is a significant weakness of the model. External shocks such as changes in technology, trade liberalisation and economic integration may have an immediate impact in the form of increases in efficiency and income. The new and additional employment of local labour may be generated at a later stage. But these external shocks may have perhaps an immediate effect on the reduction of the local employment because increased competition or new technology may eliminate certain jobs. Adjustment to these shocks may require both time and government intervention. Gains from shifts in production and trade as a consequence of integration should be reduced by the cost of the adjustment needed to obtain net welfare effects of integration. Adjustment costs (the social price for change) may be quite high in the uncompetitive economies. Think of the ‘pain’ and length of time needed for the transition economies of central and eastern Europe to change to a market-type economic system. However, if these countries protect themselves from external trade for too long, such a policy can become an additional obstacle to adjustment. Adjustment costs are borne both by individuals and by society. Private adjustment costs include a reduction in wages, losses in the value of housing and a depreciation of the value of firms’ capital. Social costs include lost output from unemployed capital and labour. The common external protection may discriminate against imports from third countries in such a way that these external economies may adjust more swiftly than would otherwise be the case. In order to avoid the common external protection, the governments of those countries may, among other things, respond by shaping and evolving their own geography of production and comparative advantage in higher lines of manufacturing to gain a competitive edge in advanced products and services, and export them to the integrated group of countries. The dynamic and evolutionary economic models are not as simple, smooth and mechanically straightforward as neoclassical equilibrium models of trade and investment, but they are much closer to real life. Experiences in the EU, its enlargements, the European Free Trade Association (EFTA) and successive rounds of tariff reductions under the GATT/WTO have shown that geographical and industrial adjustment takes place relatively slowly and smoothly over a long period of time and throughout the economy.7 In the case of the EU these costs were

Regional policy

171

so much smaller than initially expected that the elimination of tariffs has advanced at a faster pace than anticipated in the Treaty of Rome. There were intraindustry adjustments, rather than inter-industry ones. The disastrous scenario of throwing home firms out of business on a large scale has not materialised at all. The buffer against such a scenario, as well as the means for the mitigation of adjustment costs, may be found in increased capital mobility and flexible rates of exchange. However, one has to add that the period of the 1960s was characterised by relatively high growth rates and near full employment, which helped the adjustment process. If freeing trade does not produce pain in the form of pressure and adjustment costs, it probably produces no gain either. The ‘compulsory’ reallocation of resources is a source of gains. The adjustment cost is a finite, one-time investment. The gains from improved resource allocation present a continuous flow over time. Therefore, there are reasons to believe that the ‘pain’ is much exaggerated (Curzon Price, 1987, p. 16). Trade liberalisation accelerates competition in participating countries. For this reason, the expectation that international economic integration is beneficial in the long run can be accepted with considerable confidence. Economic adjustment and the alteration in the geography of production is a necessary condition if countries want to maintain the growth of the economy in circumstances of high-risk choices where technology and the situation in the market changes fast. They will have to learn how to live with change in order to reap gains from such a strategy. Adjustment costs associated with shifts among economic activities include a need for a reallocation of labour. Positions will be lost in some business activities and geographical areas, while they will be created in others. Structural funds for social, training, regional and industrial issues can act as built-in stabilisers which help the initial losers to recover. An economic policy of non-interference with market forces has obvious advantages because competition and efficiency are stimulated, consumers’ tastes are satisfied and there is a reduction in the costs of government administration and intervention. On the other side of the coin, government intervention may be required because markets are imperfect, firms seldom take into account social costs of production (externalities) and adjustment, and market forces may increase inequality in the regional distribution of income and economic activity. However, intervention meant to smooth adjustment problems can develop into a deeply rooted protectionism which over time increases costs to everybody. Hence, neither a pure market system nor a paramount government intervention can take account of all private and social costs and benefits of evolutionary economic adjustment. While intervention may help to solve some of the major economic issues, market forces may be more successful in the fine-tuning of the entire economy. A country’s comparative advantage and geography of production are dynamic and evolutionary concepts. They constantly shift over time. Countries may not be sure that their current production advantages in goods and services will remain unchanged in the future. International economic integration may be a reliable way for a country to secure wide markets for home goods/services and to obtain sources of supply in the future. Geographical proximity of the partner countries ensures that gains from trade, specialisation and a wider choice of goods are not wasted on transport and other trade costs.

Evolutionary Economic Geography

172

Clusters The production geography of the US is particularly interesting to examine as it is a country without important internal borders and it is so big that it may serve as an example for economic comparison with an economically integrated Europe. Even a superficial look at the US production geography reveals a high concentration of certain industries. For example, the following widely known industrial clusters come to mind: • aviation around Seattle; • light aircraft in Wichita (Kansas); • helicopters in Phoenix (Arizona); • finance on Wall Street; • insurance in Hartford (Connecticut); • pharmaceuticals in central New Jersey; • biotechnology in San Diego; • medical equipment in Minneapolis; • medical imaging in central Massachusetts; • orthopaedic devices in Warsaw (Indiana); • weight-loss centres in Durham (North Carolina); • electronics in Silicon Valley and Boston’s Route 128; • advertising on Madison Avenue in New York City; • optics-related industries in Rochester (New York); • cars around Detroit; • pick-up trucks in San Antonio; • paints and coatings in Cleveland (Ohio); • entertainment in California; • office furniture in western Michigan; • virtual-reality software in Orlando; • firearms in Sturgis (South Dakota); • hosiery and home furnishings in North Carolina; • carpets in Dalton (Georgia); • wine in Napa Valley (California); • shoes in Massachusetts; and • gambling in Las Vegas and Atlantic City. Development and growth histories differ between Europe and the US. While the US developed, industrialised and grew as an economy under constant but integrated territorial expansion, the same process took place in Europe behind barriers of trade protection, state support and different languages and cultures. As a result, the US has a number of highly developed and specialised regional clusters throughout the country, whereas the same type of specialisation is much less obvious in Europe (Storper et al., 2002, p. 103). Widely known European clusters that come to mind include: • production of knives in Solingen; • financial services in London; • fashion garments and motorcycles in northern Italy; • carpets in Kortrijk (Belgium); • pleasure and entertainment in Paris;

Regional policy

173

• flowers in Holland; • diamonds and sex/prostitution8 in Amsterdam; • health in Montpellier; • nanotechnology in Grenoble; • electronics in Sophia Antipolis; • tomato canning in Campania; • food manufacturing in York and • watches in Geneva and the Swiss Jura. However, the European car industry has never created a cluster similar to Detroit. A list of select clusters in Italy and Germany is given in Annex III. Once the concentration of business becomes too high, there may be negative externalities for work and private life such as pollution, sewage and waste disposal problems, congestion, allergies, crime and an increase in the price of land and rents.9 This may have an impact on the spread and decentralisation of businesses and their shift to other regions as firms may wish to leave the ‘threatened’ regions. However, the EU has on average a less concentrated manufacturing geography and much more segmented markets than the US, as a long-term consequence of various NTBs which increase trade costs. The goal of the Single Market Programme (1985–92) was to eliminate all technical and administrative barriers on EU internal trade. Hence, if this takes place, the expectation was that the EU industries and its geography of production could resemble those in the US. One should not be deceived into thinking that the EU will ever have a homogeneous market like the one in the US. Most things and ways of doing things (culture) in the US are alike throughout the country (for example, food, customs, services, what towns and settlements look like). In the EU, on the other hand, citizens have distinct national and even regional consumer preferences for food and drink. This will persist. Each Italian region and town, for example, has its distinct local speciality. The local population is extremely proud of its difference from the ‘others’. In addition, the EU countries have a range of national policies regarding health, safety, social issues and worker representation that supplement the ones prescribed by the EU. Comparisons between the US and European spatial structure of industry have mixed conclusions. Industries in the US are much more spatially concentrated than in Europe (even controlling for the distribution of population and manufacturing as a whole), suggesting that regional integration in Europe could cause agglomeration at the sectoral level (for example, Germany gets engineering, the UK financial services, and so on). (Venables, 1999, p.17) (See also Aiginger and Leitner, 2002, p. 670; Brülhart, 1998a, p. 790; Krugman, 1992, p. 72; Midelfart-Knarvik et al., 2002, p. 218; Krugman and Venables, 1996, p. 960; and Puga, 1999, p. 304). Storper et al. (2002, p. 103) argued that the US had a number of highly developed and specialised regional clusters throughout the country; the same type of specialisation is much less obvious in Europe. Others, however, think differently regarding the impact of integration on the spatial location of economic activity: ‘we think

Evolutionary Economic Geography

174

that the results available so far point more towards similarities across the EU and the US than differences. These similarities occur despite underlying differences in mobility and the extent of integration’ (Combes and Overman, 2004, pp. 2899–900). One has to bear in mind that today manufacturing represents about a quarter of GDP in these developed regions. In any case, the issue may not be settled, hence this presents a splendid topic for further research. Andaluz et al. (2002) compared economic landscapes in the US and the EU. They attempted to test empirically earlier analytical statements and widely held beliefs that the European economic space (its manufacturing industry) was more varied than the American. This was based on the perceived existence of greater obstacles to free trade and factor movements in Europe compared with the EU. They found ‘no evidence in favour of the manufacturing sector being more concentrated in one of the two large economic areas’ (p. 326). In addition, they found ‘in the agricultural, forestry and fishery products sectors, as well as market services, it is the European landscape that is more concentrated’ (p. 328). Even though there are analogous micro-level similarities between the EU and the US regarding the operation of agglomeration and dispersion forces for the same kind of industries, differences between the two economies exist. They refer, but are not limited, to the lower degrees of labour mobility and product market integration in Europe in comparison with the US; personal choice, social system and taxation that influence incentives to work and to move; as well as geography and climate differences that determine that there is a higher share of GDP that is devoted in the US to heating and airconditioning than is the case in Europe. Hence, this comparison between the two regions is a general area that asks for further research and explanation. Whether the US outperforms Europe is an issue open to a wide debate. The difference in employment rate may be partially the outcome of differences in direct taxation and personal choice. The relative level in the GDP may not reveal the difference in the composition of GDP. Americans spend more of their GDP on heating and cooling because of a harsher climate in comparison to Europe. They also spend more on security and prisons than is the case in Europe. In addition, one needs also to look at productivity and, more importantly, at trends in productivity between the two areas. General happiness of the population and satisfaction with life, work and leisure also needs to be factored in such analyses and comparisons. Production geography: with trade barriers Some of the most remarkable and desired effects of economic integration may come from a switch from segmented to integrated markets among countries. These effects were neglected in the earlier analysis because of nonlinear analytical complications that result from externalities and because they deal with trade costs. Integration affects the balance between agglomeration and spread forces in the following ways: • With high transport costs and high trade barriers (or autarky), trade is impossible. Each country is likely to have manufacturing to supply its local consumers. There are no chances for agglomeration in a single location from which firms may serve the entire international market.

Regional policy

175

• With low or non-existing transport and trade costs (free trade or ‘death of distance’) location may become irrelevant. The world gets ‘smaller’ while markets get larger. Forward and backward linkages (presence of tradable intermediate goods) dominate, and there are possibilities for footloose manufacturing to agglomerate in a single country or a single location (American outcome). There is no need for producers and consumers to be in the same vicinity as trade is costless. Firms move to the lowestcost locations. The availability of local factors, in particular immobile ones that were not touched by market opening, determines what is produced where. • If the value of trade costs is intermediate, incentives for regional self-sufficiency are weakened, hence concentration becomes possible, useful and even necessary. Agglomeration takes place because of increasing returns to scale and possible migration of a part of workers. Costs of production may differ among countries depending on country size (market) and economies of scale. Firms in a smaller or peripheral country that produce tradable goods on average depend more on foreign trade than do firms in larger countries. Therefore, if trade is liberalised, firms from the small country may gain more than firms from large countries. However, firms from a large country may exploit economies of scale and have lower costs of production and prices of output than the firms in a small country. Once trade is liberalised and trade costs fall, firms from the large country may capture a large part of the small country’s market. Then comes a period of adjustment. Certain firms may relocate in search of low-cost immobile factors (land, labour) and larger markets. In small countries such as Austria, Belgium, Finland or Luxembourg, firms may successfully adjust and penetrate niche markets in a large country.10 Although the national production geography (distribution of industries) is altered, this does not mean that there will be a divergence in national per capita income. Each country may have a cluster of industries that supply the entire integrated market. However, this case brings with it a range of stable and unstable equilibria (European outcome). The locational impact of integration also depends on activity-specific backward and forward linkages (indivisible production). If the production process is functionally indivisible (strategically complementary) and strong, if it is based on a close network as is the case in chemicals or finance, then it is unlikely that integration would stimulate larger-scale relocation of production. However, if this process may be split into various bits, then it is likely that certain parts may go and locate elsewhere. Suppose that transport costs in the real world have a declining path over time. This is a relatively reasonable assumption from the time of Ricardo. One may observe a trend in the narrowing of the income gap between the advanced countries (north) and certain developing countries (south). ‘Declining trade costs first produce, then dissolve, the global inequality of nations’ (Fujita et al., 1999, p. 260). The problem is that industrialisation and economic development is not a uniform process. It takes place in a series of irregular waves and jumps. Labour-intensive industries are the first ones to leave the industrialised country because of high wages there. The less labourintensive industries that move at a later stage do so possibly more rapidly than the earlier movers. When upstream industries move, they create potentials for forward and backward production links in the target country which facilitate entry of firms from other downstream industries.

Evolutionary Economic Geography

176

Production geography: reduction in trade barriers There is, however, an initiative that may help out the lagging regions in an integrated area. Suppose that the initial reduction of barriers to the movement of goods, services and factors in a common market spurs an inflow of factors to already industrialised areas where they benefit from economies of scale and various externalities. If all barriers to internal trade and factor movements are eliminated or become insignificant, firms may benefitfrom economies of scale and externalities in other (less advanced or peripheral) regions where the variable costs of production are potentially lower than in the centre of the manufacturing or service activity (Krugman and Venables, 1990, p. 74; Venables, 1998, p. 3). In this case, the less-developed region or a country that takes part in a common market is likely to benefit on at least two counts: • it gets firms that benefit from economies of scale; and • the former regional production structure that was typified by a lack of open competition is altered. Kim confirmed this type of reasoning for the manufacturing industry in the US during the 1860–1987 period. Industry specialisation rose substantially prior to the turn of the twentieth century. At about the same time, the US was developing its transport and communication network in order to become a fully integrated economy. During the interwar years, the level of regional specialisation ‘flattened out’, but then fell substantially and continually between the 1930s and 1987. Economic integration made the US regions less specialised today than they were in 1860 (Kim, 1995, pp. 882–6). This trend of dispersion of industries and increasing regional similarity in the US continued throughout the 1990s (Midelfart-Knarvik et al., 2002, p. 253). In theory, economic integration may, but not necessarily will, bring greater benefits to the regions/countries that lag in their development behind the centre of economic activity. However, if production linkages (forward and backward) are strong and embedded (meaning that production is indivisible); and internal to an industry such as in chemicals (Basle) or banking and other financial services (London, New York); and imperfect competition prevails, economic integration will trigger agglomeration (clustering) tendencies. Regional policies that aim at spreading these industries will be counterproductive. If, however, the linkages are not limited to a relatively narrow industry group, but are strong across industries and sectors, integration will produce agglomeration tendencies in select spots. If labour is not mobile, the whole process will evolve in a way that would open up new and widen the existing regional wage differentials (Venables, 1996a). Although this may produce certain deindustrialisation tendencies in the peripheral regions (perhaps only initially), it does not mean that integration as a general policy choice is not desirable. For instance, education and regional policies increased the attractiveness of Spain as a location for various manufacturing industries. This was discovered by the EU and other foreign investors. This was particularly obvious following the Spanish entry into the EU in 1986. National education policy, together with favourable tax treatment and the English language, made peripheral Ireland an attractive location for investment, particularly by US TNCs. This transformed Ireland into one of the most high-technology countries in the world at present. The Irish problem may be, however, in relatively low links between highly specialised TNCs and the rest of the local economy. Even though

Regional policy

177

Portugal was in the EU, this country did not emulate the Irish example. One of the principal reasons for this was the relatively low level of education of labour and weaker educational policies in comparison with Ireland. Deepening of economic integration in the EU through the elimination of NTBs during the introduction and after the establishment of the Single European Market, reduction in the cost of transportation and the eurozone may diminish motives for regional and national self-sufficiency. However, integration may also stimulate agglomeration tendencies and reinforce the core– periphery problem. Production in the EU may resemble that in the US, where industrial output (both in manufacturing and in services) is concentrated in distinct geographical locations. Hence, if this takes place in the future, internal EU trade among its member countries will no longer be intraindustry, but rather inter-industry in nature. Further reduction of trade costs in the EU may lead to further concentration of production, which is subject to the economies of scale in the already existing core locations, while the periphery may specialise in manufacturing production that does not depend on scale economies. This provides support to the arguments of the new economic geography and trade theories (Brülhart and Torstensson, 1996; Amiti, 1999). Industry characteristics set the locational response to lower trade costs. Industries that have significant economies of scale and important technological intra-industry production links have a bubble- or bell-shaped relation between integration and concentration. When there are high trade costs, production is dispersed so industrial concentration is low; as these costs are lowered and reach intermediate level, there is an increase in industrial concentration; as trade costs fall to low levels, there is a decrease in concentration (spread) as industries start to respond to factor markets.11 These industries are metals, chemicals, machinery and transport equipment. Industries that are not strongly based on economies of scale (textiles, leather and food products) have a monotonously increasing path towards concentration as trade costs decrease because of integration (Forslid et al., 2002, p. 293). This kind of reasoning and regional compensation funds of the group may be used to convince the adversely affected peripheral regions to tolerate the difficult period of transition that may follow trade liberalisation and economic integration (Puga and Venables, 1997, p. 364). Similar arguments may be used in the case of the new eastern member countries of the EU. A lesson for the EU is that the Single Market Programme (integration deepening) provided opportunities for a concentration of production in certain hot spots, as well as for rationalisation of business operations. Amiti (1998, 1999) found evidence that this took place in the manufacturing industries which include industrial chemicals, petroleum, textiles, plastics, iron and steel, machinery and transport equipment. These are all industries that are subject to increasing economies of scale and that have a high proportion of intermediate inputs in final output. Hence, this provides some support to the arguments of the new theory of spatial economics. In the 1976–89 period, geographical concentration increased in thirty of sixty-five recorded industries, concentration fell in twelve industries (the biggest fall was in the manufacturing of concrete for construction), while there was no significant change in the geographical concentration in other industries (Amiti, 1999, p. 580). Midelfart-Knarvik et al. (2002) found that ‘between 1970/73 and 1994/97, the general trend towards spatial dispersion is reflected in 29 out of 36 industries’ (p. 241) and that there is an ‘impression of a spreading out of European

Evolutionary Economic Geography

178

manufacturing activity’ (p. 241). Aiginger and Pfaffermayr (2004) and Aiginger and Davies (2004) argued that geographical concentration in the EU decreased during the 1990s. This is in line with the second part of the bell-shaped relation between integration and location of production (as trade and transport costs decline, there is a spread of production). Mora et al. (2004, p. 14) referred to the 1985–95 period and found that ‘most sectors had slightly increased their degree of concentration, especially paper and printing products, chemical products, and food, beverages, and tobacco. Services sectors had also increased their concentration level.’ In addition, regions close to the EU ‘core’ and the ones with a diversified economic structure in which high-technology industries and knowledge-based services predominate benefited most from the integration process and improved their relative position in the EU. In contrast, EU regions with a low degree of diversification in production and relying on mature low-technology industries that employ low-skilled labour (therefore highly sensitive to asymmetrical shocks), away from the EU ‘core’ and dependent on farming, showed the worst relative position in terms of GDP per capita after fifteen years of integration. These regions were caught in a ‘poverty trap’ (ibid., p. 22). If this were the case, then there are important policy lessons not only for the ‘old’ EU(15), but also for the EU(27). On the other hand, a study by the European Commission (1998) observed that an ‘examination of data offers mixed evidence for the contention that the single market is leading to a geographical concentration’ (p. 67) and that there is ‘little evidence of concentration occurring in the EC’ (p. 69). Therefore, this is another topic that demands further research and analysis. Regional results If indigenous entrepreneurs are flexible, then the regional geography of production can be altered within a large integrated area without substantial external assistance or even without a separate regional currency. A clear general example is the conversion of New England’s production geography from the production of shoes and textiles in the 1960s (coupled with relatively high unemployment) to an economy based on high technology and low unemployment in the 1980s. Let us consider certain examples of the impact of economic integration on the location of production based on experiences from Europe and North America. Europe Centripetal forces may explain relatively low indices of intra-industry trade (agglomeration, clustering) in industries subject to high economies of scale and strong technological links in production. Conversely, relatively high indices of intra-industry trade (in ‘labour-intensive’ industries) may suggest the spread of industries. In the case of the EU, Brülhart (1998b, pp. 340–1) suggested that in the 1980–90 period: • there was no further concentration of already clustered industries that are subject to increasing returns in the central regions; • there was further concentration of textile-related industries at the periphery; and • there were certain indicators of a spread of ‘high-technology’ industries towards the periphery.

Regional policy

179

Others, however, argue that the outcome may be the reverse of the one just described and that an active regional policy is necessary, particularly in an Economic and Monetary Union (EMU). Emigration of people, in particular those that are of working age and experts, would discourage an entry of new businesses into such a region. That trend would further weaken the economic position of the region. Nonetheless, such a strong vicious circle has not taken place in the EU. Even though there are severe problems regarding data, particularly regarding the 1950s and 1960s, certain trends are discernible. What seems likely to be the case in the EU is that regional disparities were slowly narrowing until the early 1970s. This was followed by a decade-long period of widening the regional gaps, and then a mixture of stabilisation and widening between member countries ever since. It is not yet clear what prompted the halt in the convergence process after the 1970s. This does not mean that specific regions are irrevocably stuck in a specific position in the regional ranking order. Evidence is available that there is considerable switching from one position in the regional ranking to another (Armstrong and de Kervenoael, 1997, p. 41; MidelfartKnarvik et al., 2002). Various transfers within or between regions or countries that provide adjustment tools create a greater degree of social security and economic stability than would exist if free markets had free rein. These funds provide an adjustment tool and a (temporary) chance to bridge troubles and alter (unfavourable and presumably temporarily) the course of developments. The US provides an example of a country where interregional adjustments are eased by a system of implicit federal government transfers of real resources. The federal government spends much and taxes little in the backward or illperforming states/regions, while it does the reverse in the advanced states/ regions. A one-dollar reduction in per capita income of a US region causes a reduction in the payment of federal taxes of about 34 cents and, at the same time, an increase in transfers of about 6 cents. Between one-third and one-half of the initial one-dollar shock is absorbed by the federal government (Sala-i-Martin and Sachs, 1992, p. 216). This type of risk sharing is still ill understood in the literature. Later work turned the stylised ‘40 cents in the dollar’ in the range from 40 per cent to just 10 per cent (Melitz, 2004, p. 817). However, while considering these percentages one ought also to factor in the role of the private sector in the adjustment process, and not look only at the public sector. In any case, one has to note that a similar transfer mechanism does not (yet) exist in Europe. The EU, with its vast internal market, does not yet have economic stabilisers on the American scale that are intended to ease the adjustment troubles of both countries and regions within them. Does this mean that there is a need for a more active cohesion policy in the EU? ‘Not necessarily. We still lack evidence on whether existing cohesion policies have been effective in reducing growing regional disparities in the EU’ (Midelfart et al., 2003, p. 865). Others also share the same opinion the results of evaluation are not conclusive on the impact of Cohesion policy … even in the Cohesion countries, where EU transfers account for sizeable proportion of GDP and investment, impact estimates are considered imprecise, and the effects attributable to EU funding alone appear (according to some studies) to be limited. (Bachtler and Wren, 2006, p. 150)

Evolutionary Economic Geography

180

Does the system of spatial distribution of production have its own internal evolution and endogenous dynamics that may not be changed much by policy over longer period of time? In general, and in the case of the EU, the data on Objective 1 regions suggest that the policy (or at least the socio-economic and market context within which the policy has been activated) has favoured the convergence of less-developed regions towards the EU mean in terms of GDP per capita, rates of annual economic growth, employment levels unemployment between 1988 and 1999, and thereafter. (Leonardi, 2006, p. 164) However, this development, even trend, ought to be tested in the future. The economic history of integrated states such as the US points to the fact that integration is associated with regional convergence which predominates over economic divergence in the (very) long run. This process is rather slow, about 2 per cent a year, but it is sustained over a long period (Barro and Sala-i-Martin, 1991, p. 154). There is a ‘legendary’ annual convergence rate of about 2 per cent found in a number of studies. This led many that follow the convergence literature to believe that there is a ‘natural’ convergence constant. To verify this, one may employ meta-analysis that is based on various statistical techniques to compare and/or combine the results of various studies with similar features. Using this technique Abreu et al. (2005, p. 411) found that there is a value close to the ‘legendary constant’ of 2 per cent. Nonetheless, this is a highly complex methodological issue because of EU enlargements and variable national and regional (sub-national) growth rates. Therefore, this begs for further enquiry. In a critical survey of the new geographical turn in economics, Martin found that regional convergence is remarkably similar across the US, the EU, Canada, Japan, China and Australia and stated that ‘the observed rate of regional convergence is very slow, about 1–2% per annum, and considerably lower than predicted by the simple neoclassical growth model’ (Martin, 1999, p. 72). In any case, there are endogenous growth factors (increasing returns) at work, and regional policies based on pure transfers of funds are not working unless linked with structural changes, spillovers and factor mobility. Carrington (2003, p. 382) argued that growth in each region depends not only on its own characteristics, but also on those of the regions that comprise the neighbourhood to which it belongs (adjoining regions). Applying spatial econometrics to 110 regions of the ‘old’ EU(15), it was found that during the 1990s two opposing forces were exerting their influence across the EU(15). While regions did converge at a rate of about 2 per cent, neighbourhoods of regions diverged at an almost equal rate, leaving a net effect of convergence considerably smaller than the one previously reported in the literature. General, but very slow, regional convergence has been confirmed by the literature; however, a poor region can expect the gap between its initial level of income and the aggregate to be reduced by only 30–40 per cent at the outside (Canova and Marcet, 1995, pp. 1 and 24). Armstrong (1995, p. 149) found that the convergence rate between 1970 and 1990 was only 1 per cent a year, which is half the rate estimated by Barro and Sala-iMartin. Thus, a certain convergence process is at work, but it is ‘painfully slow’ (Armstrong and Taylor, 2003, p. 85).12

Regional policy

181

There is empirical evidence about convergence in income growth of European regions. However, there are particular ‘convergence clubs’. More precisely, agglomerations and rural peripheral regions possibly converge towards different steady state equilibria. The findings of the present study suggest a lower steady state income level for urbanised and rural areas of the EU than for highly agglomerated regions. (Bräuninger and Niebuhr, 2005, p. 14). As far as the distribution of total production in the EU is concerned, in the period between the 1970s and the 1990s, the EU countries became more specialised. However, EU regions show a mixed pattern. Approximately half of EU regions became more specialised, while the other half became less specialised. Industrial concentration also varies by industry. Those industries based on high technology and with economies of scale became more spatially concentrated than the others (Combes and Overman, 2004, p. 2857). The vast majority of EU countries ‘experienced a growing difference between their industrial structure and that of their EU partners’ (MidelfartKnarvik et al., 2002, p. 223). These growing divergences in the geography of national production may be the consequence of two factors: • The importance of history This is when countries initially have industries that grow at different rates. Hence a country with a higher proportion of high-growth industries becomes increasingly more specialised than the average or the rest of the group. • Differential change This is when countries move out of a certain type of production and evolve into another. Midelfart-Knarvik et al. (2002, p. 221) found that over 80 per cent of the change in the EU during 1980–97 was due to ‘differential change’, while the rest came from the amplification of initial differences. The most striking feature of this process was a change in the industrial structure of Ireland, Scotland and Finland, where new high-technology industries and the ones subject to increasing returns to scale were established (often with a high FDI component). The impact of integration on individual industries can also be looked at in the following way. Midelfart-Knarvik et al. (2002) divided industries into five groups and observed their concentration in the 1970/3 and 1990/7 periods. The results of the analysis were as follows: • There was a convergence in the industrial structure in most European countries during 1970s. However, this trend was reversed from the early 1980s onwards, as countries became increasingly different (more specialised) from the average of the rest of the EU. • The most remarkable change in industrial structure was the location and expansion of high-technology and high-skill industries in Ireland and Finland (countries on the EU periphery). • There was a change in industrial location. Certain industries that were initially dispersed, became more concentrated (textiles, apparel, leather products, furniture and

Evolutionary Economic Geography

182

transport equipment). Others which were previously concentrated became less concentrated (beverages, tobacco, office and computing machinery, machinery and equipment, and professional instruments). Yet others, which were concentrated and remained concentrated include motor vehicles, aircraft, electrical apparatus, chemical and petroleum products. • Econometric analysis ascertained that the principal forces that determined industrial location, were for high-technology production, the availability of high-quality human capital; that industries that have strong forward and backward linkages (high content of intermediate inputs) were sensitive to the central/peripheral location of the country; while industries that rely on a high degree of increasing returns to scale tended to locate in central regions, but this effect has noticeably diminished over time. • Services were in general more dispersed than manufacturing. • EU industries were less concentrated than those in the US. The change in the production structure data in the EU found certain support in the modification in the structure of EU trade. Trade data support the finding that there was a decrease in specialisation during the 1970s. However, the picture is mixed from the beginning of the 1980s. The growing specialisation revealed in the production data is not reflected in the change in the structure of trade. The main reason for this discrepancy is the growing volume of intra-industry trade (Midelfart-Knarvik et al., 2002, p. 225). Trade data may not be a suitable substitute for the production records. The structure of trade may change as a response to changing demand without any change in production. If there is a change in domestic tastes, then the domestic producers may sell goods on the home market rather than make extra efforts to sell them abroad. The structure and volume of trade are altered in this case, while there may be no change in the composition of the domestic output. Compared with the pre-Single Market Programme, the EU ‘problem regions’ saw an improved economic performance in the period following the completion of the Single European Market. This improvement was obvious in terms of growth in both employment rates and gross value-added (European Commission, 1997, p. 5). Certain concerns that the Single Market Programme would lead to a concentration of economic activity in the ‘core’ EU countries were not realised. There has been only a limited spread of economic activity to the ‘peripheral’ EU countries which enjoy certain cost advantages (European Commission, 1998, p. 145). This is consistent with the view that as costs of trade fall for products of certain industries, the periphery may became more attractive for investment as the returns on the capital are greater. However, this development in the EU may also be in part the consequence of the impact of regional aid (European Commission, 1997, p.34). One of the outcomes of the Single European Market was that certain clusters of firms and industries in the EU became a bit more visible. Relatively high geographical concentration of related firms in relatively small areas eased exchange of information. Following these changes, Frankfurt, London and Paris are now the areas that create jobs faster than the rest of the national economy. However, the problem is in the regions outside a large metropolitan area which still seem to remain ‘poor’. Success, like many other things, also appears to cluster. The regional policy of the EU is simply a supranational policy that is additional to, rather than a partial replacement of, various national regional policies. Its first

Regional policy

183

shortcoming is that in spite of its resources, the European Regional Development Fund (ERDF) is modest in relation to regional needs or demands, in particular following the eastern EU enlargements in 2004 and 2007. It should not be forgotten, however, that the regional policy of the EU was introduced after the British entry in 1973. The EU Cohesion Policy13 is relatively new (started in 1989)14 although it may be improving over time. Structural funds account for about 35 per cent of the entire EU budget. In these terms, the attempt to reduce disparities between rich and poor countries in the EU is impressive in circumstances of relatively slow growth and general cuts in public expenditure. A greater degree of coordination of national economic policies may avoid dissipation of scarce EU funds, which may result in help to the rich in the relatively lessdeveloped regions to the detriment of the poor in the developed ones. It could appear that there is a tendency to redirect structural spending from rural to urban areas. Regional problems are starting to concentrate in the urban areas of the problem regions. The principal beneficiaries of economic integration in the EU are usually urban areas in the central regions. This process has also created disadvantages in urban areas and in other regions of the EU. This is because the EU is a heavily urbanised group of countries. The impact of the Single European Market on regional disequilibria in the EU is ambiguous. There is an identification problem. First and foremost, one needs to answer difficult questions: what are the short- and long-term effects of the Single European Market on the regions, and what is the impact of the changes on the regional disequilibria that would have happened on their own? If the longer-term effects of the Single European Market include the liberalisation of EU trade, then output may continue to be concentrated in the already advanced regions in order to benefit from positive externalities (this implies a fall in output and wages in the less-advanced regions). However, internal trade liberalisation may reallocate some EU production activities towards the periphery in order to take advantage of lower wages and other production costs there. As the outcome is uncertain in a situation with market imperfections and multiple equilibria, the effects of the Single European Market on the regions will continue to be debated for some time until a new problem puts this issue on a back burner. Traditional forms of national regional policy started to change and to decline in importance from the early 1990s in most of the EU member countries.15 Denmark, for example, abolished all regional development grants and loans in 1991. The Dutch have restricted their regional assistance to a relatively small part of the country in the north. Owing to budgetary constraints, France had to scale down its regional aid. Germany has severely curtailed regional policy in the western part of the country, while placing priorities in the eastern Länder. Even the less-developed countries such as Greece have had to be careful with their regional expenditure because of the restrictions that originated in recession. The new spirit of national regional policies in the EU countries includes (Bachtler and Michie, 1993, pp. 721–2): • reduction of the importance of regional policy in the northern countries; • transfer of responsibilities to regional and local levels; • automatic aid replaced by discretionary assistance; and • increased involvement of the European Commission in regional matters, in particular through the rules on competition.

Evolutionary Economic Geography

184

Contemporary regional policy also includes the local development of producer services. This feature was gradually incorporated into regional policy only in the second half of the 1980s. Nonetheless, the creation of jobs in new manufacturing businesses, their extension in existing ones or relocation of businesses still account for the major part of regional intervention. The European Commission may authorise investment aid of up to 75 per cent net grant equivalent in the least-developed regions. The same limit is 30 per cent in other areas where aid is allowed. The upper limit for aid in the leastfavoured areas is intended to increase the competitiveness of these regions in attracting private investment. In practice, the country that has such regions within its confines may face severe budgetary problems and lack the necessary funds to finance projects in these areas. Hence, the impact of this concession is significantly eroded. In any case, a relatively high proportion of the allowed aid of up to 75 per cent may seriously question the degree of commercial risk borne by the private firm (Yuill et al., 1994, p. 100). Total regional outlays in the EU at national and group level are substantial. Despite this, they do not appear to deliver substantial or expected results. Evidence suggests that neither efficiency gains nor reduced regional inequalities result. If there is any possible impact at all, then it is at most a redistributional one. If transfers are mainly redistributional in nature, would policies based on non-discriminatory financing be a better route to follow? (Ulltveit-Moe, 2007, p. 1443) Many deeply embedded regional problems often remain quite immune to change. Success of various regional policies depends, among other elements, on the transfer of know-how, knowledge spillovers within and between industries, factor mobility and general trade costs. If this were the case, then the policy pursued in Europe may be a costlier option. A note of caution regarding linear and professional thinking by academic researchers: politicians may use public funds and policies for their personal political advantage of the moment with little regard for the longer-term benefit and need of the region/country they supposedly represent. North America Venables (2003) argued that integration among high-income countries (north–north integration) may produce convergence of their per capita incomes, while integration among low income countries (south–south integration) may produce divergence among their per capita incomes. Participating countries in the north–north case may successfully specialise within their market niche advantages. In the south–south case, the location of production for the group’s market may move to the richer countries at the expense of others in this group of low-income countries. Therefore, north–south integration may in these circumstances serve low-income countries better. In this arrangement, a lowincome country may expect high export demand for its output from the north. This may improve the southern country’s terms of trade which would give this country a larger share of gains from integration. Mexico initially benefited from these integration-related developments with the US and Canada within NAFTA.

Regional policy

185

Integration of Mexico with the US shifted many manufacturing industries from the central country area of around Mexico City northwards, along the border with the US. This took place not only because there was a growing demand in the US, but also because of the general trade liberalisation policy in Mexico. Once the economy turns outwards, internal production linkages weaken and firms have less incentive to stay in the congested hub. Turkey, with its well-established and fully operational customs union with the EU, may well be envied by other countries for having such a business privilege at its fingertips. Surprisingly, Turkey has failed to fulfil a role for the EU similar to the one that Mexico plays for the US. The reason is that Turkey attracts only a small fraction of FDI compared with Mexico. After the establishment of the customs union with the EU in 1995 and contrary to all expectations and predictions, FDI to Turkey fell for two consecutive years! In general, corruption, red tape and a lack of protection of intellectual property rights are the principal reasons for this weak performance in comparison with Mexico.16 An important fact concerning the location of economic activity appeared in relation to integration in North America. This is especially relevant regarding Mexico, which was a country with a relatively closed economy. NAFTA (1987), which includes the US, Canada and Mexico contributed to the contraction of economic activity around overcrowded Mexico City and its partial relocation (especially manufacturing) along the northern border with the US. However, if NAFTA evolves into the Western Hemisphere Free Trade Agreement, then this may not have a major impact on the relocation of firms in Latin America in order to improve access to the US market. Most trade between these countries and the US takes place by air or by sea. Outside Mexico, most manufacturing is located in cities such as Buenos Aires in Argentina, São Paulo in Brazil, Santiago in Chile, Lima in Peru or Caracas in Venezuela. ‘Hence, there would be no need for firms in these countries to relocate in order to improve access to the US market’ (Hanson, 1998, p. 42). Integration deal with the US in these cases could have a negligible impact on the location of production. Similarly in Britain, one observed a sharply changing hierarchy of Britain’s exporting ports. Between 1964 and 1978 (periods for which surveys were available) one could observe a dramatic rise in the hierarchy of south and east coast ports such as Southampton (from ninth to fifth), Dover (from tenth to third), Felixstowe (from fifteenth to seventh) and Ipswich (from twentieth to thirteenth) (Hoare, 1986, p. 31) to the detriment of those in the west of the country. The British entry to the EU (1973) and economic integration certainly had an obvious impact on the relocation of business activity. Economic integration in North America created high expectations that trade liberalisation would assist Mexico within NAFTA to catch up with the US and Canada in economic terms. On the occasion of the tenth anniversary of NAFTA, the World Bank published a study that attempted to assess the impact of integration on Mexico’s development. In the words of Lederman et al. (2005): the ratio of Mexican GDP per capita to the United States did increase somewhat after unilateral trade reforms were implemented in 1986 and after the implementation of NAFTA in the aftermath of the so-called

Evolutionary Economic Geography

186

Tequila crisis. However, it is noteworthy that other Latin American economies also have grown faster than the US economy since the mid1980s, especially Chile and to a lesser extent Costa Rica. Thus it is not obvious that NAFTA was particularly important in helping Mexico catch up with the United States. The experience of Puerto Rico is also interesting, given that it is an economy that started with a level of development similar to that of Mexico in the late 1950s but achieved an unprecedented level of economic and institutional integration with the United States in 1952, and subsequently experienced the fastest rates of economic growth in the developing Latin American economies. … we find that the post-NAFTA period was characterized by a substantially faster rate of productivity convergence than in previous years. However, at this time we cannot say whether the productivity convergence resulted from increased imports of intermediate goods from the United States … from competitive pressures and preferential access to the US market … or from increased Mexican innovation that might have been caused by a variety of factors, including increased domestic research and development (R&D) efforts and patenting aided by the enhanced protection of intellectual property rights contained in NAFTA … Finally, our estimates of the change in the autoregressive coefficient of the US–Mexico income per capita gap suggest that Mexico’s GDP per capita would have been about 4–5 percent lower by the end of 2002 if NAFTA had not been implemented. The cross-country evidence showed that differences in institutional features inherited from history play an important role in producing income gaps … Thus economic convergence in North America might not materialize under free trade or under any trade regime as long as fundamental differences in initial conditions persist over time. It is fortunate that some of these fundamentals should be sensitive to policy changes. (Lederman et al., 2005, pp. 27–8, 31–2, 59–60, 61) Some economists took this result at its face value as a proof of the beneficial effects of economic integration: integration, in particular integration with developed countries, assists in the development process of ‘poor’ countries. Economic convergence among countries exists. Others were sceptical and strongly questioned these results. The World Bank’s methodology in this study, in particular the GDP data set, was sharply criticised. When the same regression was used with a different set of GDP data from authoritative sources such as the IMF’s World Economic Outlook, OECD or the Penn World Tables, the result is reversed: NAFTA slowed down the rate of growth in Mexico (Weisbrot et al., 2004, p. 8). In any case all these tests may not be accepted as conclusive, but rather as an incentive for further research. Trade liberalisation and the creation of NAFTA were only two policy changes that had an impact on Mexico’s growth rate during the first NAFTA decade. Hence, they may not bear the brunt of the ‘guilt’ alone for Mexico’s slower growth.

Regional policy

187

Challenges Three factors account for most of the inequalities in regional levels of income. They deal with differences in total factor productivity, employment level and the share of agriculture in regional income. The suggested regional policy, based on the experience of Ireland from the mid-1980s, relates to the market-oriented approach as ‘the best conduit to sustained economic growth and fast convergence in per capita income’ (Boldrin and Canova, 2001, p. 211). Conceptual and measurement problems have always made the economic evaluation of the impact of regional policy initiatives a complex, difficult and uncertain task because of the problems that include: • restrictive assumptions, • simplification of reality, • causality (different from correlation), • policy objectives (not all are directly economic), • multiple beneficiary groups, • additionality, • time lags (policy initiative announcement, implementation and impact), • length of the evaluation period, • the identification problem within composite ‘hard facts’ (separation of policy provoked from autonomous and endogenous changes), • involvement of ‘soft’ qualitative indicators (regional capacity-building) and • policymeasurethatmayhaveageneraleffect and rigidly drawn specificspatial area within which the measurement of the policy impact is taking place. Certain theoretical models may show regional convergence, while others such as the new economic geography and evolutionary models based on economies of scale and imperfect competition may predict divergence (rather than convergence) in regional economic performance. In spite of substantial regional expenditure by the EU, Boldrin and Canova showed that neither convergence nor divergence was taking place among the EU(15) regions during the 1980–95 period. The adopted and implemented regional policies, as well as substantial public resources channelled to less-developed regions, do not appear to enhance the capacity of these regions. These funds only redistributed income among regions. With no prospects for change and if income distribution is the principal concern, such transfers will be necessary on a continuous basis. Therefore, there was no evidence that the adopted policies of redistribution of income were the most appropriate. As most regions were growing at a fairly uniform rate, these policies reflected internal political compromises which have little effect on fostering economic growth and convergence in the EU (Boldrin and Canova, 2001, pp. 211 and 242). Others think differently. ‘If our estimate of the cumulative effect of EU Cohesion policy is correct, then a cumulative employment effect of nearly 1 million jobs at 2002 is significant contribution’ (Martin and Tyler, 2006, p. 206). In spite of substantial expenditure on regional matters by the EU during the 1980s and 1990s, regional inequalities have not narrowed during this period, and by some measures they have even widened. ‘Income differences across states have fallen, but inequalities between regions within each state have risen. European states have developed

Evolutionary Economic Geography

188

increasingly different production structures’ (Puga, 2002, p. 400). A similar observation came from Midelfart-Knarvik and Overman (2002, p. 333) who noted that, despite the fact that the EU(15) spent about a third of its overall annual budget on regional matters, there was increasing regional inequality. If the EU regional policy and instruments were successful on the country level, they were less so on the subnational plane. In any case, if economic integration and technical progress provoke a certain concentration of economic activity and regional disequilibria, then one may find valid reasons for and justify regional policy in the EU. This is even more important in the light of the stated political EU goal to promote harmonious development in its member countries. The decision about regional policy including major spending initiatives on regional issues will always fall within the realm of politics. However, decisionmakers ought to be notified and informed (as much as possible) about the available policy tools, preconditions for success and possible effects of regional policy in order to make a choices that are as ‘intelligent’ (mistakes included) as sensibly possible and as politically workable: the near-impossibility of obtaining what evaluators in principle would seek to obtain, i.e. a reliable and credible assessment of impact, and a true legitimization for the EU interventions. This arises because of the context of multilevel governance, the ‘additionality’ issue and the heterogeneity of local actions and local areas. (Baslé, 2006, p. 227) However, the lack of widely accepted analytical tools and an absence of consensus about the evaluation of effect of the regional policy may confuse politicians and may give a good chance to the strongest lobby to push forward its narrow interest of the day to influence the evolution of the current and future course of policy action in its own favour.17 The issue of the Cohesion Policy performance and its evaluation is neither an exclusively academic nor a simple task. The policy touches big taxpayers’ money over a long period of time. Its spending ought to be legitimised. Can unsatisfactory regional (cohesion) policy impact and performance be a useful tool to convince the net and other contributors to the EU budget in the future that this spending needs to be continued at the same level? The policy evaluation process has a number of weaknesses. It deals with problems such as time lags (policy action and its observed effect in the future); how that can be measured; whether there is causality or only correlation between a policy action and its observed impact; whether EU spending is additional to local spending; and whether the multi-level governance and heterogeneity of actions just compound the problem mix. Why did the convergence and development policy succeed in Ireland and why did it fail in Greece and Portugal? Can one have a credible assessment of the impact and a reliable judgment at all? Information technologies may have the same role in India’s economic development as was the case with car manufacturing in Japan after the Second World War. While Western civilisation was initially overloaded with rationalism, measurements and geometry, and Chinese civilisation with Confucian heritage, Indians were more interested in abstract issues. They first developed the mathematical concept of zero in the sixth

Regional policy

189

century AD, followed by negative and decimal numbers, eternity and irrational numbers. Hence the distant national roots for the evolution and expansion of software industries in India provide a fertile soil there. If this is coupled with the Chinese brain- and manufacturing power, then this may present one of the greatest challenges to Western technological supremacy in the future. If costs of trade are high, industries tend to disperse. When this cost is reduced, agglomeration can take place as demand in distant places can be met by exports. When these trade costs approach zero (as is the case in writing computer software in Bangalore [Bengalooru], India or data entry in Manila and the use of the Internet), footloose production of goods and certain services18 may be fragmented, dispersed and located according to the availability of the specific resource inputs.19 Globalisation of certain industries (integrated international production) reduces the weight of physical proximity between various production units, as well as between producers and consumers. However, in industries that have strong internal links, such as those based on new knowledge (innovation activities are still highly clustered in the world), financial services or chemicals, there is a strong propensity to cluster in spite of ‘globalisation’ of other businesses. Economic development of east Asian countries, especially Japan, Taiwan and South Korea, as well as China from 1980s, was based at least to an extent on export-led growth. The consequence of this policy was a general overdevelopment of the export part of the domestic economy. This type of location of production was in part the consequence of an undervalued exchange rate over a long period of time. These countries face tough policy choices regarding revaluation of their domestic currencies vis-à-vis the US dollar. If they revalue, (1) domestic employment may suffer because of a possible drop in exports to the US and (2) the accumulated domestic reserves of the US dollars (in 2008 measured in trillions) may immediately drop in terms of the domestic currency. 6 Conclusion A country or economic union in which there are persisting differences in the levels of development and/or living standards among constituent regions that do not have at least a tendency towards certain equalisation cannot be regarded as having a well-integrated economy. Therefore, all countries and economic unions have a certain commitment to reduce regional disequilibria for a few economic and, more importantly, a variety of noneconomic reasons. The objectives of regional policy are to diminish existing and prevent new regional disparities. If a regional policy is to be effective, then the authorities at all levels of government have to coordinate their activities in order to influence decisions about the allocation of resources (spatial location of economic activity). In spite of these coercive powers, the regional policies of countries have had relatively limited positive achievements. It should therefore come as no surprise that the achievements of the regional policy of the EU, which often relies more on rules of competition, persuasion and on certain funds than on specific region directed coercion, are scant indeed. Regional policy has been based on a number of compromises, to the detriment of the purity of principles. Previous attempts to shape regional policy relied mainly on the

Evolutionary Economic Geography

190

alleviation of transport and communication costs through the expansion of infrastructure, as well as the mitigation of agglomeration disequilibria. More recently, attention and action has shifted towards a greater self-reliance for those regions that are lagging in development, as well as the enhancement of enterprise competitiveness in these regions. Statistical evidence and various surveys of past actions offer little testimony that the EU regional policy has had any significant effect. In the case of France, for example, Crozet et al. (2004, pp. 28 and 51) found that French and EU regional policy investment incentives and structural funds had no significant impact on investors. Elsewhere, the growth of clusters and regions such as northern Italy or Baden-Württemberg was not based on any public action, but rather on inside elements. Public action in regional matters such as those which took place in the Italian Mezzogiorno had very limited success and occasionally could have been harmful. Enterprise policy and the policy of macroeconomic stability are gradually being used instead of regional policy to tackle the regional (spatial) problems of a country. The new regional (development) policy ought to change the traditional intra- and inter-regional relations of dependence and hierarchy. This type of old, embedded and rigid structural organisation of firms, industries and institutions in which the state commands, taxes and shelters; big firms are protected; small and new firms are tolerated but not greatly encouraged; while the family ties and ‘old boy’ networks connect everything; ought to be replaced by structures that are open for contacts based on affinity, support and perception of common growth and positive-sum games. This new policy should include the following features: • assistance to innovation; • permanent general and specialist learning and networking; • aid to increase flexibility to face and adapt to challenges; • reduction in the traditional financial regional support; • emphasis on SMEs, business ‘incubators’ and start-ups of new firms; • backing of the producers’ services; • environmental sustainability; and • coordination with other policies. The critics argue that this policy stance lacks the firmness needed to have a direct influence on a given spatial problem. In contrast to the decentralisation and trimming down of national regional policies, regional policy in the EU has continued to widen its coverage and scope. It is, however, hard to determine the point at which regional policy distorts competition beyond the interests of the EU. Hence, uncertainty over regional policy will continue in the future. There are many arbitrary elements in the policy, as well as special cases. EU regional policy has revealed its limitations. It is still Byzantine in its complexity. A solution to the regional development problem, as well as to achieving some balance among various regions, is an urgent, difficult, but potentially highly rewarding challenge for the EU(27). Instability in regional policy at national and EU levels seems likely to continue. It is possible that the trend to decentralise the creation and implementation of national regional policy will continue in the future. However, there is a danger in the expansion of regional incentives that compete with one another.

Regional policy

191

The question arises, however, as to whether regional policy increases or decreases market imperfections. In the second-best world, all answers are possible, but not always desirable. If the costs of such a policy are less than the benefits it brings, then the policy is justified. The rationale for a regional policy that basically redistributes income (equity) must be found in solidarity among regions that constitute countries and/or the EU, as well as the fact that the area of benefit is larger than the assisted region. Demand, tastes, technology and supplies of factors often change. Regions that fail to adjust continuously to the new challenges and opportunities remain depressed and weak. One of the broad objectives of regional policy is to help the redistribution of economic activity among different regions. Its impact cannot be measured easily as it is not a simple task to construct a counterfactual world that would specify what would have happened without this policy. The difference between the actual and counterfactual situation may be attributed to the effects of regional policy. As the central and east European countries brought serious regional disequilibria into the EU in 2004 and 2007, cohesion will remain one of the major long-term issues in the EU(27). Hence, there are at least three major arguments in favour of the EU regional policy in order to preserve evolutionary unity with diversity in the enlarged EU: • In the absence of policy instruments such as tariffs, NTBs, devaluation or changes in rates of interest, regions that are not able to adjust as fast as the rest of the EU face increases in unemployment and decreases in living standards. In this situation, there is a case for the demand for short-term fiscal transfers at the EU level to ease the adjustment process. The possibility of such transfers in unforeseen cases ought to be permanent in an EMU (eurozone). Otherwise, when in need, the regions (countries) that are in trouble may not be sure that other partner countries will provide resources on a case-by-case basis. The eurozone may not be able to operate efficiently in the long term without an effective regional adjustment policy. However, the size of the structural funds for the EU(27) during the 2007–13 period fell far short of expectations. There are serious political pressures to limit and to scale down the overall EU budgetary expenditure. • Coordination of national regional policies, as well as other principal economic policies at the EU level, can avoid self-defeating divergent regional programmes that are taken in isolation. • Footloose industries, multiple equilibria, economies of scale and externalities do not guarantee that integration will bring an equitable dispersion of economic activities. Some guiding of economic adjustment and allocation of resources in the form of regional policy may be necessary. The evaluation of regional policy is a demanding and tough task. It is vital to compare policy achievements with goals at the start of the process (for example, the number and quality of created jobs). But that is not enough. An overall social cost–benefit appraisal is necessary. However, data for a thorough assessment of the success of the policy are either imperfect or insufficient or unavailable. Therefore, we are left with the second-best but useful analyses that rely on surveys, econometrics and expert opinions. There is an obvious diffusion shift in the location of higher value-added production in the EU towards the periphery (Ireland and Finland). There is a certain shift in the location of production to the new EU member countries in the east, but these countries are in

Evolutionary Economic Geography

192

strong competition with China and other Asian countries to attract footloose investments. Nonetheless, southern EU countries may suffer in terms of their attractiveness as locations for larger-scale manufacturing production. Regional policy is in good part another facet of social policy. Hence, the EU faces difficult political choices regarding regional policy. Challenges in this area include the eastern enlargement, monetary union, continuous unemployment, structural change and international competition. Obstacles to a stronger influence in this policy area come from the eurozone restrictions in the monetary and fiscal fields, lack of an automatic federal EU system of transfer of funds, slower growth and little internal migration of labour among the old EU(15) countries (while migration from the ‘new’ eastern EU countries towards the old ones is quite restricted till 2011). One has always to bear in mind that the amount of funds at the disposal of a particular policy is not always what matters most. What matters is the amount of funds that are necessary to change a particular type of behaviour in the desired direction. The amount of funds and their astute use to change certain types of behaviour will continue to be a matter for debate long into the future. Notes

1 See Brülhart (1998a) for a brief survey of theoretical strands. 2 Evidence from the EU shows that the strongest internal labour mobility took place during the 1960s and the beginning of the 1970s; it came to a halt during the 1990s. However, intra-EU capital mobility increased during the 1980s and has remained strong ever since, even though domestic labour mobility within countries such as Italy was reduced. Labour used to move northwards on a larger scale during the 1950s and 1960s. Not any more. Southerners, for instance in Italy, may easily be overqualified for simple jobs offered in the north of the country, the cost of living is higher in the north and there is also the attraction and charm in the ‘southern way of life’. 3 A successful strategic industrial (and trade) policy is based on a number of assumptions that include next to perfect information and forecast, as well as the policy of non-retaliation by foreign partners. 4 T. Sickinger, ‘Outsourcing not bad, Intel chief says’, The Oregonian, 12 May 2004. 5 While the US ‘produces’ 137,000 engineers a year with at least a bachelor’s degree, China churns out 352,000. The university requirements in China are much less demanding compared with the US. The Chinese even have a phrase for this type of student, ‘stuffed ducks’ who are good at memorising facts and passing exams but have little initiative. With its Confucian heritage, China places great emphasis on education, but there is also a heavy deference towards authority. Employers regularly complain that although they hire graduates who seem brilliant on paper, it can be very hard to get them to voice opinions. However, Not only are one-quarter of foreign PhD candidates in the US Chinese but a growing number of them are heading back home: Beijing says 170,000 Chinese who studied abroad have returned, 30,000 of them last year. (G. Dyer, ‘How China is rising through the innovation ranks’, Financial Times, 5 January 2007) 6 Is the core–periphery relation based on the purchase of goods by the core region from the periphery? Or the outflow of capital from the core region to the periphery? Or migration of labour (and capital) from the periphery region to the core? Or is there a mix of these possibilities? 7 The shipbuilding industry in Britain and Spain was in decline from the 1970s as it failed to specialise in a market niche as the Finnish industry has done (icebreakers), it was not able to

Regional policy

193

match the production costs in Korea and it lacked the potential for diversification existing in Japan. But this is the story from only one industry, and generalisations about the situation in the whole economy should never be made from a single-industry case study. 8 The location of prostitution as an ‘activity’ has always been predominantly in towns and cities. The reason is that towns and cities ‘sell’ anonymity, diversity and limited tolerance. 9 Bangalore (southern India), an important world software and outsourcing hub, may be the victim of its own success. The rapid growth of this city put a strain on its infrastructure. Population is about 6 million, roads are congested and power cuts are frequent. Local software firms are looking to expand elsewhere in India, possibly in Calcutta, Hyderabad, New Delhi and Pune. 10 Casella (1996) discussed the case of reallocation of resources and gains from an enlargement of a trade bloc in small and large countries that already belong to that bloc. She showed that smaller EU countries gained more from the entry of Spain and Portugal into the EU than the large EU countries. 11 Ireland, Spain and Sweden as peripheral countries in Europe attracted a relatively large number of foreign investors and their huge investment funds. 12 The conditional convergence between a poor and developed country may operate in select cases. If a backward country respects market forces and lets them operate on their own, if it intervenes smartly at the appropriate time and with suitable instruments, if it establishes institutions and protects property rights and if it accumulates human capital, it may converge with the developed countries. Certain south-east Asian countries provide good examples of success and failure in this matter. 13 The Cohesion Policy refers EU solidarity. It is not limited to economic development only. It deals with the accessibility of regions and cities (improving transport infrastructure), the knowledge economy (innovation and entrepreneurship) and employment (access to labour market, migrants and handicapped). The policy disposes of the Cohesion Fund (from 1993) and Structural Funds (European Regional Development Fund, European Social Fund, Guidance Section of the European Agricultural Guidance and Guarantee Fund, and Financial Instrument for Fisheries Guidance). The Cohesion Fund provides finance for transport infrastructure and environmental issues in the eastern EU member countries, as well as in Greece, Portugal and Spain. 14 Earlier, EU regional policy was primarily providing support to the national regional development policies in the member countries. The former policy was designed for the regions, not by the regions. Almost all EU regions are represented in Brussels. The EU Committee of the Regions is quite an active EU official body, but its actual impact is still quite ambiguous. 15 Britain changed its regional policy in 1980. The most important features of this alteration were the following three elements. First, the regional disequilibria started to be seen as a regional rather than a national issue, one that had to be resolved by indigenous development rather than by a transfer of resources and business activity from elsewhere. Second, direct subsidies for employment were replaced by a system of regional aid programmes based on employment creation through improved competitiveness. Third, the policy became increasingly reliant on employment cost-effectiveness (Wren, 1990, p. 62). 16 The Cotonou Agreement (2000) between the EU and developing countries still needs to prove itself in this regard. 17 The EU landscape, for example, is plagued with vanity political projects of dubious commercial importance and value. This is evident in particular in parts of southern Italy. 18 Computer programmers, financial analysts, accountants or radiologists in the developed world face competition from abroad as their services may be delivered electronically worldwide. However, hair-stylists, bakers, cooks, waiters, cleaners, construction workers or

Evolutionary Economic Geography

194

drivers may feel ‘safe’ if there is no free international mobility of labour. 19 There are, however, certain limits to this type of ‘offshoring’ and migration of such service jobs towards the developing countries. Tastes and needs change, markets are fickle, hence there will always be a need for a certain local presence, close to consumers. It may be difficult to service such consumers and markets in the EU or US from Bombay.

4 Market structure and location of production Do you not know that in a race all the runners run, but only one gets the prize? Run in such a way as to get the prize. (1 Corinthians 9:24) 1 Introduction The online Oxford English Dictionary1 defines competition as ‘The action of endeavouring to gain what another endeavours to gain at the same time’.As for commerce, competition is ‘Rivalry in the market, striving for custom between those who have the same commodities to dispose of’. As such, competition may be understood as an abstract idea. It is not something that most people would voluntarily rush to the barricades to protect. Competition permits the presentation of new ideas, it keeps prices moderate or low, and it removes business arrogance and incompetence. Geography in space (or in culture) limits competition as it creates distance and costs among producers, sellers and their clients. One of the most important functions of competition is the exchange of information. The objective is to increase the efficiency in the use of factors due to increased competition. In theory, free-market competition provides everyone with the widest opportunities for business and produces the best sectoral and spatial allocation of resources. This conclusion has been accepted by neoclassical equilibrium economic theory as a truth. It has provided the intellectual backing for competition (antitrust) policy. Competition both improves efficiency in the use of factors because of their constant reallocation and, something which is often forgotten or overlooked in neoclassical equilibrium theory, this reallocation introduces a permanent instability into the economic system. While competition may create lucrative opportunities and gains, it may also be the source of cost, problems and concerns linked with risk and uncertainty. The objective of competition policy is that markets attain and maintain the flexibility needed to promote initiative, innovation and constant improvement in the allocation of resources. The final intention and goal is, of course, to maintain and raise living standards. Hence, what matters in theory is how to play the competition game, rather than who wins or loses among the competitors. In practice, however, politicians, members of the public, firms and lobbies are quite concerned about the winners (and losers). This chapter is devoted to the issue of market structure, i.e. competition, and location of firms and industries. It is structured as follows. Section 2 presents basic issues such as monopolies and the Microsoft case. Section 3 discusses various ideas linked with

Evolutionary Economic Geography

196

innovations such as innovators’ rights, reasons for concentration of innovation, intraindustry trade and mergers and acquisitions, while dominant market position is covered in Section 4. Section 5 recaps competition issues. Specialisation and returns to scale are subject-matters of Section 6, in which comes an examination of barriers to enter and exit an industry, predatory pricing and the impact of standards on the location of firms and industries. Section 7 investigates rules of origin. Non-tariff barriers are referred to in Section 8. Conclusions are given in Section 9. 2 Background Issues Competition policy is a combination of two irreconcilable forces. On the one hand, there is an argument for the (spatial) concentration of business, which rationalises production and enables economies of scale. On the other hand, there is a case for an antitrust policy,2 which prevents monopolisation, protects individual freedom and rights and, through increased competition, increases welfare. The challenge for governments is to achieve and maintain a dynamic balance between these two tendencies. They need to keep the best parts of each of the two opposing tendencies, profit from the harmonious equilibrium between the two, avoid excessive regulation that interferes with the freedom to contract which may impair competitiveness, and employ competition policy as a tool to increase the general standard of living. As a term, ‘free market’ is almost an oxymoron. Such a system may operate ‘as expected’ even in theory only if there is a system of well-developed, accepted and enforced rules, regulations, codes and conventions.3 Even a self-declared, smug, freemarket-prone, democratic and liberal government such as the one in the US has awesome power, as it has no qualms about reminding everyone both at home and abroad. In any case, a relatively ‘free market’ has been, so far, the most efficient economic organisation model (social technology) over the longer period of time. Friedrich Hayek criticised the neoclassical competition model, which is based on perfect competition, in the following way: But I must be content with thus briefly indicating the absurdity of the usual procedure of starting the analysis with a situation in which all the facts are supposed to be known. This is a state of affairs which economic theory curiously calls ‘perfect competition’. It leaves no room whatever for the activity called competition, which is presumed to have already done its task. (Hayek, 1978, p. 182) There are, however, a number of admirable types and examples of more or less liberal markets. One size does not fit all and at all times. Apart from the American model,4 there are others which include those in the Scandinavian countries,5 France, Germany, Switzerland, Canada and Asia. Both pluralism and specific discipline operate in each of these market systems. All of them slowly evolved over centuries and shaped the form,

Market structure and location of production

197

significance and influence of national political, economic and social institutions through providing opportunities, learning, imitation, selection, reward and punishment. These cannot be mechanically and swiftly transplanted elsewhere with success like flu. Even within the liberal market economies, such as the US is supposed to be, one needs effective regulation to foster ‘free’ competition. The case in question is the access to cyberspace. The choice in the US is rather narrow: either through the local cable monopoly or through the local phone monopoly. Krugman commented on this situation in the following way: The price is high and the service is poor, but there’s nowhere else to go. Meanwhile, as a recent article in Business Week explains, the real French bureaucrats used judicious regulation to promote competition. As a result, French consumers get to choose from a variety of service providers who offer reasonably priced Internet access that’s much faster than anything I can get, and comes with free voice calls, TV and Wi-Fi. It’s too early to say how much harm the broadband lag will do to the US economy as a whole. But it’s interesting to learn that health care isn’t the only area in which the French, who can take a pragmatic approach because they aren’t prisoners of free-market ideology, simply do things better.6 Well, this example shows that regulators should not always be ridiculed. Monopoly If a monopoly exists, one should not rush to the conclusion that its presence per se leads to economic inefficiency. It is possible that in industries with high entry barriers, enormous sunk costs and economies of scale (for example, the aerospace industry) a single efficient producer may sell the good or a service at a lower price than would be charged by many inefficient producers in the same industry. In such cases it may not be a smart move to break up a concentrated industry. The authorities would be better taxing the excessive profits of firms in such industries and/or making sure that they are reinvested. If the market grows sufficiently, then the authorities need to encourage other, potentially efficient and profitable, producers to enter the same kind of business. More than a century ago, it was feared in the US that the concentration of economic influence could create possibilities for strengthening political power as well. The industries in question were steel, oil and tobacco. The Sherman Act of 1890 was a reaction to such fears. This act forbade monopolisation of any business in the US. Its intention was not only to prevent political influence by the business, but also to ensure that there was rivalry in innovation and opportunities for new firms. However, one has to note that under President George W. Bush (2001–9) the political power of big firms has never been greater in the US than it is today. Policies intended to increase competitiveness sometimes produce unexpected results. During the 1950s, the US government wanted to break up large and powerful corporations. One such corporation was AT&T. Little attention was paid during the debate to the position of Bell Labs and the fact that it had played the major part in

Evolutionary Economic Geography

198

innovation in telecommunications over the preceding century. Deregulation of the industry prohibited the US telephone-operating companies from making phones or switching equipment. As a result, US exports grew moderately, while imports exploded (Lipsey, 1992b, p.295). In addition, the fragmentation of the American telephone system provided the European and Japanese firms with the possibility of leading in mobile phone technology and production. They did not miss that chance. This had an impact on the location of production of mobile phones. The state-owned British Telecom had a monopoly for directory information for half a century. Everybody remembered the relevant phone number: 192. Then came deregulation in 2001. In theory, the more the suppliers of the service, the better and the cheaper the service is. In practice, 120 firms had entered this market by 2004. With 120 firms in the business, people bothered to remember only a few numbers. The numbers were assigned to firms by lottery. The lucky ones that got numbers such as 118 118 became new natural monopolies and seized most of the business. As people use this service sporadically, they did not bother to scan all 120 offers. Hence, the companies with new memorable numbers could charge higher prices than the old British Telecom did (Beinhocker, 2006, p. 424). The situation was similar in Switzerland. According to the Federal Office of Communication (OFCOM) directory information was fully liberalised at the start of 2007. This was supposed to benefit the consumers. However, expectations were not fulfilled at the end of the same year. The number of clients continued to decline. There were 73 million calls in 1999, but only 36 million in 2006. Clients found other technologies to obtain information: the Internet, CD-ROMs or mobile phones. The rates for directory information did not change to any noticeable degree. Firms that entered this business had both similar operating technologies and similar cost structure, hence they had little space to manoeuvre with prices. From the above examples stem arguments in favour of an astute competition policy. The intention of this policy is to improve efficiency in the use of factors with the objective of increasing welfare. However, in a real market situation that is full of imperfections (economies of scale, externalities, multiple equilibria, sunk costs, innovation, asymmetric information and so on) one may easily find arguments in favour of a certain concentration of production (mergers) and protection of intellectual property rights.7 The Microsoft case Five computer companies accused Microsoft of unfair business practices8 in the US in 1998. In 2000 the court found that Microsoft, the world’s biggest software company, was guilty of antitrust violations and of monopolistic behaviour in the software market. This is a case that has generated a lot of hot air. The judge decreed that Microsoft should be split into two companies: one to develop and produce the operating system (Windows) and the other to be involved with the application packages (Office, Internet Explorer and so on). The real outcome could be the creation of two monopolies instead of one. At the heart of the debate is the issue of capacity to innovate. There was a risk that Microsoft would limit the ability of third parties to innovate. The problem with splitting the company is that numerous operating systems and applications may emerge. However,

Market structure and location of production

199

the potential for customer confusion that could result may be exaggerated, as there is a strong commercial incentive to maintain compatibility. In addition, Microsoft was charged with exhibiting monopolistic behaviour such as operating exclusion contracts, as well as increasing sales by raising the price of older versions of software after the launch of a new version (so that the new version becomes more attractive). Microsoft duly appealed against such a judgment and argued that regulators have no right to meddle or determine which features software companies should bundle into their products in order to increase their functionality. This is particularly important in a dynamic and highly innovative industry. While bundling may bring certain gains, it also makes it harder for competitors to offer innovative services. Several questions immediately come to mind: • Is it smart to allow (or drive) an innovative firm such as Microsoft to spend as much on legal issues as it spends on R&D? • In the ‘new economy’ based on and driven by new knowledge, a firm that innovates gains a temporary monopoly power. Another firm with a superior product or service that creates a new temporary and ‘fragile’ monopoly replaces the first firm. Hence, the new economy may have more (fragile) monopolies than the ‘old economy’. If such dynamics encourage innovation and improvement, consumers can benefit. Should one encourage such market dynamics or not? When should regulation step in? How? • Where is the balance between innovation and standardisation (which may stifle innovation)? • Where is the balance between intellectual property rights and the wider public interest? • How does all this affect the long-term interests of the consumers? There are no sound data for a thorough cost–benefit analysis in such cases. Data are also missing to assess long-term interests of consumers in an industry that is at the frontier of innovation. This requires a very fine judgment and anyone who is confident of providing a clear-cut answer risks ridicule. Solving such issues may not easily fit into predetermined charts and models. Those that handle competition policy should also rely on their finer instincts. In any case, the basic principles and instincts in this area should include a preference for competition over monopoly, pluralism and spread over concentration, as well as new entrants over incumbents. The appeals court in the US confirmed the earlier ruling that Microsoft was a monopoly, but it rejected the remedy of breaking up the company. A ‘toothless’ settlement was reached with the Department of Justice in 2001. Microsoft had to: allow PC makers to replace some parts of Windows by alternatives made by other firms; reveal certain codes to ensure that software written by other firms operates smoothly; and have its operation monitored by an (enigmatic) three-man committee of experts.9 Windows became a monopoly not because of its engineering excellence (although those who had problems would dispute this statement), nor because of predatory pricing, but because of its convenience for a widespread use: the more people use it, the more it becomes handy to all of them. As a declared monopoly, Microsoft is vulnerable to legal actions in the future. The US case against Microsoft was brought because the company wanted to do away with competitors such as Netscape (does anyone still remember it?). The European Commission also investigated Microsoft. This case is linked with the bundling of Media

Evolutionary Economic Geography

200

Player in the Windows XP operating system, as well as the smooth/rough link between PCs and servers that run Windows. Microsoft’s position is that consumers want products with more features. This brings improved interoperability, and also, Microsoft may argue, lower prices. In addition, if Media Player is unbundled from Windows, then this may raise queries such as whether one should also take out printer drivers, graphics, spellcheckers, Internet browsers and so on. On the other hand, if Microsoft produces all that is required, where does that leave other firms? When the European Commission found in 2004 that Microsoft had abused its nearmonopoly market position, the company faced a fine of Є497 million (one of the biggest ever imposed on a single company) and a request to separate Windows from Media Player software.10 An additional fine of Є280 million was imposed in 2006 for failing to hand over complete documents to rivals. Some allege that this initial fine took into account Microsoft’s global operations. Microsoft’s legal battle against this ruling could last a number of years, so some Microsoft competitors may not survive to see its conclusion. In the meantime, the market situation may change, hence there may be serious doubts and little evidence that the case against Microsoft could have favourable effects on competition and on consumers. It is also a difficult task to provide evidence on what would have happened had the European Commission let Microsoft enjoy its nearmonopoly position. The American procedure in antitrust cases is carried out in courts that permit open challenge of the arguments. The same procedure (at least initially) in the EU takes place behind closed doors in the European Commission, which acts as prosecutor, judge, jury and executioner. The complex history of the Microsoft case in the US courts reveals that there is no consistency in the principles applied even without a single legal system. Uncertainty at corporate level continues. Therefore, if there are various regulators throughout the world, this should give an incentive for a harmonisation of principles and practices. Even a global competition authority may be a good option to be considered. The obvious causes of ineffectiveness of antitrust policy include the following (Crandall and Winston, 2003, p.23): • The excessive duration of court cases. The particular issue that is focused on may easily evolve into something different over time. By the time the case is resolved, it may be of little relevance. • Difficulties in formulating effective remedies. • Difficulties in ascertaining which potentially anticompetitive act or instance may jeopardise consumer welfare. • Rapid change in technology, intellectual property and dynamic competition relentlessly increases challenges of formulating and implementing effective antitrust policy. • Political influences about which antitrust cases are initiated, settled or dropped. Microsoft lost its appeal against the European Commission’s decision that it abused its dominant market position by keeping out rivals in adjacent markets for media players and server software. In 2007, the European Court of Justice upheld the European Commission’s tough decision in Microsoft’s case. Microsoft may ‘easily’ pay the fine of Є497 million (plus Є280 million) from its hefty profit. Microsoft will allow open-source software developers to access inter-operability information.

Market structure and location of production

201

This nine-year battle and legal precedent it created may turn dangerous, as they may pave the way for future scrutiny of information and other dominant technology behemoths such as Apple, Google, IBM and Intel. If the ruling cools the innovation vigour among or within firms, this may be dangerous. In spite of the ruling, high fine and long legal process, Windows Media Player still dominates in its market. The European Commission now has the precedent and the tool to use in the future. The US Federal Trade Commission and the Department of Justice, on the contrary, have been reluctant to pursue corporate giants. The ‘global’ antitrust centre of gravity moves towards Brussels. 3 Innovation Background Neoclassical equilibrium theory took the process of innovation to be an external (exogenous) and random variable in the economic function (like an earthquake or nice or bad weather). Joseph Schumpeter argued, on the contrary, that the innovation process is internal (endogenous) in an economy and that it provided energy for growth. Although innovation is an uncertain phenomenon, it has also dynamic and cumulative dimensions. The process of technological change is not only risky and costly, but also highly uncertain and, apparently, never-ending in an evolutionary framework. Failure, sometimes very great, is a part of R&D and innovation process. This uncertainty has three dimensions: • Technical: will one find something that is new, useful and will have market demand at the end of the day? • Financial: how much will it cost? Do expected profits justify such an investment? • Social: will consumers accept the new output? Perfect competition (perfectly contestable market) is based on zero entry and exit costs in an industry. This eliminates inefficient firms from a market, but at the same time rewards or keeps in business the efficient ones.11 Competition is not driven by changes in prices, but rather by innovation: introduction of new products, new methods of production or transportation, or new forms of industrial organisation. Schumpeter called this continuous process of change of economic structures from within ‘creative destruction’ (Schumpeter, 1987, p. 81). If equilibrium is ever achieved, it is at best a short-term and temporary event. If inter-country factor mobility is allowed for, the supply of factors (labour, capital, land, technology, organisation and entrepreneurship) increases. This ‘pluralism’ backs innovation in technology, goods, services, organisation and control. Competition probably operates best when a firm believes that it is in a process that is leading it towards becoming a monopolist (at best) or an oligopolist (at least). Also, the fear of the arrival of new competitors is among the strongest incentives for innovation. However, consumers may suffer in a monopolistic or oligopolistic industry structure. This can be redressed somewhat if oligopolistic firms introduce the most efficient innovations as rapidly as perfectly competitive firms would and if governments prevent such a market structure from behaving in a non-competitive way.

Evolutionary Economic Geography

202

The innovatory process is driven in general by several factors. These sources of novelty include: • The perception of existence of unexploited lucrative opportunities for the satisfaction of private and/or social needs. Spotted problems that ought to be solved are, for instance: the transformation of electricity into sound; or light into electricity; or voice or thoughts into (typing) work. These profitable opportunities ought to be perceived by potential innovators who deliberately search for new lucrative technical or organisational solutions. This search is linked with costs and risk on the one hand, and with potential and expected benefits on the other.12 • The importance of science; this was particularly relevant from the time of the First Industrial Revolution (roughly between 1750 and 1850).13 Since then, economic growth has had an unprecedented spurt that dwarfs the preceding thousands of years of economic history (at least in the western world). Compared with the current pace of innovation and economic growth, ‘not much’ happened before the 1750s. However, this refers only to material wealth. This should not be taken as a judgement about the level of happiness; harmony and closeness in family, society or with the environment; achievements in art (some may argue that pre-Industrial Revolution accomplishments in painting have not yet been surpassed, and market prices of these paintings provide a strong support to this view); or moral superiority or inferiority of past generations. • Changes in government regulation such as changes in technical and safety standards (including concerns about the environment) and trade policy. • The change in prices (relative scarcity) of raw materials, energy, labour or transport and communication. • Changes in the environment: global warming, depletion of a natural resource, epidemics … • Clients with the toughest demands may set the innovation process in motion. Searching for a challenge and rivalry, rather than staying away from them, inspires innovation. The same holds for regular contacts with research centres. • A change in consumers’ needs, tastes, sophistication and fickleness. Increases in income and decreased working time increase the demand for leisure and entertainment. At the same time, more stressful work takes its toll on workers’ health, increasing demand for medical and rehabilitation services. • Changes in the political situation: enemies with weapons of mass destruction, spreading terrorism … • Changes in institutions that introduce or ease constraints (guilds, unions, associations, state, Church … ). A state may randomly confiscate business gains or, alternatively, it may tax profits in an orderly and predictable way. Centralised state institutions may be too rigid (the former centrally planned economies). This is unfriendly with various ways and approaches to a problem that are necessary for innovation. A more pluralistic and decentralised approach (certain degree of liberty) may be a superior choice. A church may be negative regarding certain types of research efforts or business practice (genetic engineering or the charging of interest). Guilds may have tough entry and performance rules. All this may have an impact on friendliness with invention and innovation.

Market structure and location of production

203

Innovation (together with investments) is the principal activity that stimulates economic growth. It brings new ideas and enlarges the stock of knowledge that in turn enhances productivity. The benefits of past innovations are spread throughout the society. The problem is that the benefits of contemporary changes in technology are concentrated in a select few areas of both society and the business sector. Intellectual property rights link patents with innovation as they provide the means to reward someone who comes up with a worthy commercial idea. Without personal or enterprise-level expectations of rewards that may come following risky and costly innovation, in such an environment one may not expect a larger-scale innovatory effort. The process of technological change has four distinct but interrelated phases: • invention: discovery of something new which can ‘work’, something that links a private or social need with something that may satisfy it;14 • innovation: translation of invention into commercial use; • spread:diffusion in the market; and • absorption: learning from users and conversion into a public good. After these phases, according to the evolutionary path, comes maturity, stagnation and eventual replacement by something else. A production function for the whole economy (Y) may be written as: Y=f(K; L; N)=fKαLβNγ where α + β + γ =1 (Equation 4.1) where K, L and N stand for capital (including land), labour and the stock of knowledge, respectively. This economy has a constant technology and constant returns to scale. Sustained growth is, however, possible in this economy if there are exogenous changes in f. An alternative approach is offered by endogenous growth theory. This model incorporates economies of scale and argues that parameters in Equation 4.1 are α + β + γ > 1. Increasing returns to scale are assumed to be possible to all factors of production in the long run. Knowledge (Nγ) is supposed to be perhaps the most important one. Hence, growth may continue, in theory indefinitely. However, there exist resources that are nonrenewable and that may be eventually depleted. Growth may not, therefore, be sustained indefinitely according to this view. Even though certain non-renewable resources are in limited supply, new knowledge and technology may compensate for this potential loss. Nonetheless, a continuous stream of new energy is coming from the sun. If the importance of N is so relevant because of various positive externalities, then there are sound evolutionary reasons to stimulate the innovation process (R&D) beyond the point secured by free markets and or claimed by the neoclassical equilibrium economic model.15 If, because of a change in circumstances (for example, market enlargement because of integration with other countries), firms innovate (that is, realise their technological, organisational and control potentials and capabilities to develop, produce and sell goods and services) and introduce new technologies and new goods/services in order to maintain or improve their market position, then overall efficiency may increase. From a given set of resources one may expect to achieve more and/or better-quality output. This directly increases national welfare on average. In fact, for a small or medium-sized country, integration enables economic development and progress at a lower cost than

Evolutionary Economic Geography

204

does autarky. However, an introduction of the same new technology in different places (countries) and at different times should not be expected to produce identical results. The local environment, physical geography, business (suppliers) and society (institutions, law, politics) differ. There is, however, an opposing force. When there are market imperfections, such as economies of scale and externalities, firms make rents. Free competition leads to concentration and agglomeration, which may reduce competition in the future. The new theory of trade and strategic industrial policy holds that there may be fierce competition even among a few firms. Examples include the aircraft industry, the long-distance telephone call market and the Japanese market for electronic goods (largely confined to half a dozen domestic conglomerates). Technical innovations prompt legal and policy innovations. The telephone business, for example, was for a very long time considered to be a ‘natural monopoly’. Nowadays, it is a highly competitive industry. Innovation changes the mix of factors that are used in the production and/ or consumption of goods and services. Usually, an innovation brings a reduction in the quantity of factors needed to produce a good or service, as shown by an arrow in Figure 4.1. This makes output cheaper and, consequently, more competitive. Suppose that the production of a good requires two factors, f1 and f2. If production requires, per unit of output, 0A of factor f1 and 0B of factor f2, then the resource required to produce the good is represented by rectangle 0ACB. Innovation normally reduces the area of rectangle 0ACB (note the direction of the arrow). However, the market receptiveness to take up new products depends on the cost and convenience in use, as well as culture. If, however,

Figure 4.1 Effects of innovation on the use of factors.

Market structure and location of production

205

one factor, such as oil or some other raw material, suddenly becomes scarce, innovation may reduce the consumption of that factor, but disproportionally increase (at least temporarily) the requirement for the other factor. Hence, in such a special case, the area of the rectangle 0ACB may increase as the result of innovation. Economic integration opens up and secures the markets of the integrated countries to local firms. It is reasonable to expect competition to have a positive effect on innovation, but what are the effects of such a process in the long term? • If innovation spreads geographically, and if it increases competition, then competition and innovation reinforce each other. • If, however, innovation becomes centralised over time and space because it is costly, uncertain and risky so that only a few large firms can undertake it, then geographical extension of markets would have a positive effect on innovation only in the short term. It is therefore necessary to ensure that immediate positive effects continue in the long term (Geroski, 1988, pp. 377–9). In order to maintain the ‘necessary’ level of competition within the integrated area, the countries involved may decide to reduce the level of common external tariff and NTBs. The impact of competition is not restricted to prices and costs. Competition also yields other favourable effects. It stimulates technical progress, widens consumer choice, improves the quality of goods and services and rationalises the organisation of firms. It is important to remember that firms seldom lose out to competitors permanently because their product is overpriced or because their production capacity is insufficient; rather, they lose because they fail to develop and/or introduce new products and production processes as well or as quickly as their competitors (Lipsey, 1993a, p.18). If a firm does not make its own product obsolete through innovation, some other entrepreneur will. Profound economic changes most often happen by market entry (e.g. Ford, Microsoft, Google, Nokia) and exit of firms, rather than through adjustments within existing firms. Hence, the limiting aspect on a firm’s success and growth is not only the competitive and replication potential of other firms and fickleness of market demand, but also the protection and extension of the firm’s internal innovation capacities. This innovation does not limit itself on ‘hard’ production, but also ‘soft’ organisation16 and customer-relations technologies. The best talents in a firm are not enough on its own for a successful innovation investment. It is the management of that talent in a firm that matters. The ‘prospect theory’ of psychology explains that individuals are more hurt by a loss than they are encouraged by gains of equal size: that is, fear of loss is often more powerful than the expectation of gain. The courage to take hard decisions is also necessary. Evolutionary change within firms and industries is not necessary for an increase in market share or profit, it is necessary for mere survival. A number of popular products offered during the 1970s (or earlier) would attract little or no demand today (cassette recorders, videos, 35mm-film cameras, digital display watches). Hence the firms producing them have either changed their output (developed new production lines) or are out of business. A necessary ingredient is to create and maintain an evolutionary economic environment that drives firms to innovate from corner to corner of their business. The innovation process is sometimes accidental, but always uncertain and risky. It is often linked with serendipity.17 For example, in 1970, Spencer Silver, a research chemist

Evolutionary Economic Geography

206

at 3M, was working in the area of adhesive technology. His goal was to produce the strongest glue on the market and his reputation was ruined when he developed a product that was quite the opposite. He did, however, discover that his new glue could be used again and did not leave any traces on the surface to which it was applied. He tried for a decade to find an application for his glue, but there was no interest at 3M. However, Silver had a friend, Arthur Fry, who was a singer in his local church choir. Every Sunday he inserted slips of paper into his hymnbook to mark the pages of the hymns to be sung during the service. And every Sunday the slips would fall out as he opened the book in the church. Fry remembered Silver’s ‘futile’ invention and applied it to the slips of paper, which stayed in place. A year later, in 1981, 3M started producing the now omnipresent ‘Post-It’ notes (Hillman and Gibbs, 1998, p.183). Viagra, the most widespread medicine used to treat male impotence, is another example of serendipity. Dr Ian Osterloh experimented with this medicine during the early 1990s with the aim of finding a painkiller for those suffering from angina pectoris. This is a dangerous disease, characterised by sudden and severe pain in the lower part of the chest, towards the left side, and a feeling of suffocation and fear of impending death. Pfizer, the pharmaceuticals company, was ready to give up on the project. However, certain male patients who were using this medicine during the research phase reported that following its consumption their virility was boosted. The rest is generally known. A medicine sought after for four millennia was found thanks to serendipity. As a result of relentless innovation, techniques that are used in the production process change constantly. But innovation changes not only the method of production of goods and services but also our values and the way we live. For example, Hollywood films changed how we lived and worked, how we saw the world, even how young people courted (in cars, away from the eyes of parents and chaperones). Or the short-message system (SMS) on mobile phones and e-mail altered the nightlife of the young and single persons throughout the world. The impact of the Internet, e-mails or mobile phones was much more profound and widespread than was originally imagined. Technological changes can be minor or revolutionary. Lipsey (1993b, pp. 3–4) described four levels of innovation: • Incremental innovation: a series of changes and improvements to an existing technology, each small, but with a large cumulative effect. • Radical innovations: major but discontinuous changes such as the development of a new material (for example, plastic), a new source of power or new products (aircraft, computers or lasers). • Changes in the technology system: changes that affect an economic sector and industries within it, such as the changes that occurred in the chemical and related industries in the nineteenth century. • Technological revolution: innovations that change the whole technoeconomic paradigm.18 Small inventions contribute greatly to economic growth; however, big inventions rejuvenate the entire economic growth process. Without these big inventions, ‘growth would sooner or later slow down and possibly come close to a halt’ (Lipsey et al., 2005, p. 378).

Market structure and location of production

207

Innovation is an important economic driver. It will be the basis of competition in the future and will continue to affect the location of production. Production and use of electricity, as well as information technology, are probably the most important invented general-purpose technologies. Electricity is universally accepted and applied. Its effects are known. The same is not yet fully the case with information technology. The problem of innovation and new technology not resulting in an obvious and measurable increase in output is known as the ‘productivity paradox’. Some argue that there is no technological revolution and that computers are not productive: a personal computer that is a hundred times more powerful than one made ten years ago is not proportionately more productive. Others argue that decades must pass before the fruits of technological breakthroughs can be discerned. Yet others insist that there are benefits, but that standard statistical tools are inadequate to measure them. This is most obvious in the services sector. How can the output of a bank be measured? Has the economy become so complex and fast-changing19 that it has become unmeasurable? If this is so, it is increasingly difficult to regulate it, and to tax it too. There is, however, no simple answer to the question of whether international economic integration (extension of the market) stimulates or prevents innovation. There are two opposing views. First, a monopoly organisation has a secure market for its output. It can anticipate reaping normal or supernormal profits (rents) from any innovation. It is therefore easier for such a firm to innovate than one that does not have such market security. On the other hand, with no competitive pressure, a monopolist company may not feel the need to innovate. The sense of long-term stability fosters a conservative way of thinking that may restrict innovative activity. Monopolists may not wish to ‘rock the boat’ and can prevent or delay the implementation of innovations of their own production or developed by others. A quiet life is their chief objective. The huge increases in income and living standards that have been seen over the past 250 years, particularly in the western world, are the consequence of several factors. At the end of the seventeenth century there was a strong convergence between the theoretical understanding of science and the application of that knowledge. A critical mass of knowledge in the fields of mathematics and mechanics was accumulated and applied to the understanding of atmospheric pressure and the invention of the steam engine. This provided the foundation for a technological revolution that resulted in a 150year period of radical innovation in the areas of transport (rail and water), textiles, mining, tools, metallurgy and food processing. Most relatively modern inventions originated in a small group of countries, and this group has remained relatively stable over a long period of time. There is a clear geographical localisation of innovation activities. During the First Industrial Revolution, Britain led the way, joined by Germany, the US, France, Switzerland and Sweden in the second half of the nineteenth century. Membership of this select group of countries has been stable for over a century. The only major newcomer to the group was Japan just after the Second World War, although a few newly industrialised countries, such as South Korea and Taiwan, also joined this exclusive club.20 This reflects the fact that technological and innovatory activities have a strong cumulative nature in space, even though industry-specific expertise and advantages are quite different among these countries once one looks beyond the broad industry classifications.

Evolutionary Economic Geography

208

It is often argued that the public support is crucial for the creation and development of new technologies. Certain lobbies may also use the argument that the development of certain technologies or products depended on militaryrelated government expenditure. One may even argue, with certain success, that without government involvement21 space technology or the Internet may have been long delayed or even not developed at all by the private sector alone. Other technologies may be late in coming, but it is impossible to know by how much. Jet aircraft or PCs would have appeared on the market through the private sector; however, military-related government spending increased the speed of the development process. This was the case, for instance, with the Boeing 707. However, if there had been no US government procurement and assistance to the development of this aircraft and to tip market choice, the British VC-10 would have become the standard aircraft for long-range travel for perhaps a couple of decades. The public support for Airbus in Europe is primarily related to technological progress and commercial interests, rather than linked to defence-related matters and lobbies as is the case in the US. A strong counterexample of a general-purpose technology that developed and spread without public support can be found in electricity. Other examples include the internal combustion engine and laser technology. OECD countries allocated about $771 billion (about 2.25 per cent of overall GDP) to R&D in 2005. The US accounted for 42 per cent of the OECD total. This represented a little less than the combined total of the EU (30 per cent) and Japan (17 per cent). With an outlay of $115 billion in the same year, China is the leading non-OECD R&D spender and among the biggest in the world.22 The business community was the principal source of finance of R&D in OECD countries in the same year. Its contribution was 63 per cent of the total (most of the balance came from the government).23 However, this contribution varies among countries. The private sector financed 75 per cent of R&D in Japan, 65 per cent in the US and 54 per cent in the EU. Most of the R&D activity in the OECD countries, about 68 per cent, takes place in private firms (OECD, 2007, pp. 24–30). The national interest in the creation of novelty may be measured by R&D expenditure as a share of the GDP. Figure 4.2 shows trends in these outlays among major international players. Only Japan and China show a steady increasing trend. Patents may be used as a helpful proxy of innovation and technological performance. However, one needs to keep in mind the limitations and weaknesses of patents, in fact their number, as markers of innovatory activity. There are a number of inventions that are not patented; many patents do not have any commercial application; others may be ‘sleeping patents’; and there are also differences in national patent rules which impede comparisons. OECD (2007, p. 84) reports that that about 53,000 triadic patent families were filed worldwide in 2005. This represents a sharp increase from fewer than 35,000 filed in 1995. The US accounts for 31 per cent of these patent families, while Europe accounts for 28 per cent. The US lost about 3 percentage points compared with 1995, while Europe’s loss was more than 4 percentage points in the same period. Japan’s share, on the contrary, increased by 2 per cents over a decade to reach 29 per cent in 2005. If patent families are normalized using total population, Japan, Switzerland, Germany, the

Market structure and location of production

209

Figure 4.2 Trends in R&D intensity by area as a percentage of GDP, 1991–2005. Source: OECD, 2007, p. 25. Netherlands and Sweden were the most innovative countries in 2005. Japan has the highest number of patent families per million population (119), followed by Switzerland (107). Regarding the industry structure, patents are concentrated in pharmaceuticals, medicine, precision and optical instruments. The bulk of the global patent activity is highly concentrated in an ‘exclusive club’ that includes thirteen countries. This small group comprises, in descending order: the US, Japan, Germany, South Korea, France, Britain, the Netherlands, Canada, Switzerland, Italy, Sweden, China and Australia. A warning sign for the EU about a serious brain drain (spatial allocation of talent) is that ‘40 per cent of the research community in the US are European graduates’.24 The US looks after its talents well and rewards originality and creativity. A growing problem in the US is that there are significant difficulties in obtaining a US entry visa, in particular since September 2001.25 The European Council in Lisbon (2000) declared a new strategic goal for the EU to become ‘the most competitive and dynamic knowledge-based economy in the world capable of sustainable economic growth with more and better jobs and greater social cohesion’26 by 2010, which seems to be quite a challenging and ambitious task. Even the gap in the R&D expenditure between the EU on the one hand and the US and Japan on the other is not narrowing. There are already a few tough competitors (and new ones arriving) for this top position. As this deadline approaches, hard indicators show that the achievement of this ambitious goal by the EU must be taken with a lot of reservations.

Evolutionary Economic Geography

210

Innovators’ rights It is sometimes argued that it is quite costly to trade intangible technology assets at arm’s length because ‘it is a combination of skills, equipment and organisation embodied in people and institutions as much as in machinery and equipment’ (Sharp and Pavitt, 1993, p.147). If an inventor fears that his/ her patent rights27 are not sufficiently protected (enforcement, length of the patent right, level of penalties), he/she will keep the innovation secret or be disinclined to be involved in R&D in the future. Once created, certain types of knowledge can readily be passed on to others at sometimes little cost. If this innovation cannot be protected, it cannot be sold. For example, Albert Einstein (1879–1955), an ordinary clerk at the Zurich Patent Office and a sloppy student at Eidgenössische Technische Hochschule Zürich (ETHZ) (the Swiss Federal Institute of Technology, Zurich), devised the general theory of relativity with his senior fellow student Mileva Maricć (who became his first wife from 1903 to 1919). He was awarded the Nobel Prize in 1921. He kept the honour, but gave the money to Maricć. ‘This led to the university appointment that had previously eluded him, and thereafter Einstein worked in universities. Einstein was honored wherever he went. But he never became a rich man’ (Kay, 2004, p.266). A part of the innovation process may lean towards outcomes that may not be easily imitated. A conflict between static and dynamic efficiency in production, as well as between the welfare of producers and consumers, is obvious. As technology becomes older and is no longer at the core of the business activities of the innovator, it becomes more likely that the innovator will disseminate the technology, intellectual property rights, goodwill and know-how through licensing. The appropriation of returns from innovation is not a major problem if the innovator is a non-profit institution such as a research institute, university or government. Non-profitmaking innovators are most likely to make their findings public and, in fact, these findings may be disseminated immediately and widely. The problem of appropriation arises when there is a conflict between public interest in the spread of information and knowledge, and private interest in holding and employing that knowledge for lucrative purposes. If private knowledge acquired through risky investment of resources is not protected, at least for some period of time, there will be little incentive to generate innovations that drive efficiency and, hence, contribute to future growth. A sound knowledge-based economy demands strong protection of intellectual property. An analysis by Levin et al. (1987) of alternative ways of protecting the competitive (monopolistic) advantages of new and improved processes and products found that patents are the least effective means for appropriating returns. Lead time over competitors,28 a fast-track learning curve (unit costs of production fall as output increases over time) and sales/service effort were regarded by the surveyed firms as providing better returns than patents. Firms may sometimes refrain from patenting products or processes to avoid revealing the facts or details of innovation because of the possible disclosure of information to competitors and imitators. At the same time, firms have every incentive to advertise the benefits of new or improved products and disseminate them to consumers. Therefore, secrecy about innovation is both difficult and undesirable. In an evolutionary economic model, it may perhaps be better to be concerned about the creation of future business secrets and new qualities than to worry about the protection of

Market structure and location of production

211

existing ones. Being ahead of competitors has certain parallels in biology. The story concerns the evolution of dolphins. The geographical area in question is the Indian subcontinent. Certain primitive dog-type creatures, so the theory goes, were forced by predators (competitors for food and space) to move/escape into the sea. These dog-type creatures evolved over time into dolphins (a new quality). The vast majority of people have always liked to look at themselves, at least from time to time. Initially, they used curved metal sheets made of silver or a blend of copper and lead. However, these metal sheets (the first mirrors) lost their shine fast. To prevent the effect of air on the shining metal, people got the idea that a glass coating could protect the polished surface. The first technology was to cover glass with a sheet of lead, on top of which they poured mercury. Mercury dilutes lead, and the blend stuck hard to the glass surface. The process took over a month in order to obtain an equal coverage on the surface. Another technology that gave a more shining surface was to cover glass with a liquid chemical based on silver. Silver left its sediment on the surface in only about half an hour. In order to hold to the surface in a firmer way, this silver ‘film’ was coated with paint. Simple? Not quite, in around the 1650s. This mirror-producing craft was known only in Venice. It was strongly protected. Capital punishment awaited anyone who revealed this secret to foreigners. In addition, the manufacturing of fine glass and mirrors was transferred to Murano, an island just off Venice, which was closed to foreigners. Glass and mirror-making craftsmen were highly respected in the Venetian republic; however, they were forbidden to travel abroad. Capital punishment awaited not only the fugitives, but also any of their families who remained in Venice. The French were highly interested in learning the Murano-type glassmaking trade. They tried very hard and succeeded in bribing four craftsmen and others who assisted in the technology-stealing venture. Once the craftsmen were in Paris, the Venetian diplomat tried to find them in order to lure them back to Venice or, if unsuccessful, to send in the assassins. The glassmaking masters (criminals, for Venice) were not only well hidden, but they were also showered with luxuries, just to keep them in Paris. And they stayed. A new set of glass-making fugitives arrived from Venice soon after. The French royal workshops started producing glass and mirrors in Paris, to the delight of courts in Versailles, Louvre and Fontainebleau. However, in 1667, a year and a half after running away from Venice, the Venetian glassmaking masters started dying. One after another. The doctors concluded that they were poisoned. Fearing for their lives, the remaining Venetian masters asked the French to let them return to Venice. They were let free to go wherever they wanted. The French had already learned and mastered all their ‘technological secrets’. The English textile machinery firms initially made an effort to stop or restrict exports of their products, including spare parts. This new technologyprotection commercial system gave incentives to others abroad (the Swiss and others) to start the production of this machinery. Hence, one ought to bear in mind that An early ‘innovator’ may come out a winner, if it learns rapidly, and little of its learning ‘spills out’, or its competitors are sluggish in getting into the new technology themselves. On the other hand, it may come out a loser, if its learning is slow and hence the cost of operating the new technology remains high, or most of its learning ‘spills out’ and its

Evolutionary Economic Geography

212

competitors adopt it in a timely manner, taking advantage for free of the spillover. (Dosi and Nelson, 1994, p. 164). Additional profits can accrue from the production of complementary assets (cameras and memory chips that go into them [earlier there were 35mm films]; recorders and tapes/discs). Therefore, not only innovation and manufacturing (technological leadership), but also, and equally important, distribution and after-sales service (commercial leadership) are of great advantage in capturing markets and profits.29 In the fields of cameras, audio and video goods (and some segments of the car market) Japanese companies have virtually ousted most of their international competitors and changed the international geography of production in these industries30 through an uninterrupted tide of technical improvements and distribution/service networks. The benefits of increased competition will materialise only if firms compete and do not collude to avoid competition. Competition stimulates innovation. It may, in turn, bring new technologies with large sunk costs, geographical concentration of production and other entry barriers. If this is the case, then neither unfettered markets nor monopolies (oligopolies) should be ignored. Otherwise, consumer welfare would be distorted and allocation of resources may take place in a suboptimal way from a social standpoint. Hence, there is a need for a competition policy, not only in the market of a single country, but also in a much larger area. This area is generally limited by the geographical space where economic cycles are in step. The rule of law, based in part on economic theory, may modify market distortions both in single countries and in integration groups. Firms compete through product differentiation, innovation, the quality of their products and after-sale services, R&D, advertising and special close links with suppliers, clients and various institutions, as well as on price. The exceptions are, of course, raw materials and certain standardised semi-finished goods. Price competition is dominant here. In the early twentieth century Henry Ford would supply his clients with any colour of a car as long as it was black.31 His company was overtaken on the market by General Motors (GM), which would supply clients with the colour of a car that they wanted. This is an obvious clash between production for mass and niche markets. In the field of personal computers, for example, the greatest competition is no longer between hardware-assembly companies such as Compaq, IBM, Toshiba and Dell, but between companies operating in the area of added value, such as microprocessors, dominated by Intel, and operating software, where Microsoft, through its Windows program, reigns supreme. Market advantage for critical elements of the system is often held in the form of intellectual property (Borrus and Zysman, 1997; Zysman and Schwartz, 1998, p. 409). Major changes in the capacity of a firm that are linked to high sunk costs do not take place frequently. However, it is more difficult to test the impact of non-price rivalry such as competitors’ R&D, innovation, design activities and non-technical matters such as management, control and marketing (including guarantees and after-sale services) than their prices. It also takes longer to retaliate in these areas than to change prices (Schmalensee, 1988, p. 670). For example, Toyota is a corporation that has been investing heavily in continuous training of its staff. The core corporate value is

Market structure and location of production

213

permanent improvement (or kaizen). A gap between corporate objectives and reality needs to be bridged. Problems on this road do not appear suddenly on their own. They are always present and ought to be discovered and solved. Toyota thinks that its biggest enemy is complacency. Hence a continuous discontent with current achievements (no matter how positive they are) is at the centre of its corporate culture. The military wisdom of twenty-five centuries ago by Sun Tzu has full relevance across time. In a war, even though available resources often (but not always) bring success, the most victorious battles are the ones that your army does not actually fight. The example is the western victory over communism in the Cold War. Stratagem (trick) based on resources and potentials may be the key means of fight and success.32 Applied to business competition in the market, success often brings resources. The principal engagement for a firm may not always be another actual or potential competitor firm (Microsoft has hardly any of those), but rather engagement with the sometimes self-imposed objectives (this may be the creation of a new technology). The greatest threat to Microsoft in the medium and long term is perhaps not another competitor firm (such as Linux, Firefox, Flash), but rather a situation in which data processing and user needs would change and evolve beyond any recognition. Hence the need for superb Microsoft products and services may wane, notwithstanding their price and potential comfort in use. By offering free e-mail service, web-storage and online text/data treatment software, firms such as Google may render some of Microsoft’s software, stored at each PC, needless. The potentially good news for the consumers (users) is that they will not need to upgrade their PCs so often. Who today needs long-playing gramophone records to listen to music? The iPod is taking over this market, but until when? Between these music reproducing and listening technologies came audiocassettes and compact discs (just for a historical record on economic and technological evolution). Innovation, not necessarily price ‘combat’, is the principal tool for success in the long term. Concentration of innovation Why is innovation so concentrated in space? Europe was technologically backward and ‘uncivilised’ around the eleventh century, according to both Chinese standards and those in the Islamic world. What happened in these European countries in the seventeenth century and was allowed to continue undisturbed? Why did it not happen before? What prevented the Islamic world from continuing its innovative course after the thirteenth century, or what prevented China from doing so after the fifteenth century? Evidence to support theories of the evolution of these issues is still imperfect and controversial.33 Nonetheless, several overlapping factors may, in combination, provide a partial explanation. Let us consider them in turn. Values and culture Willingness and readiness to accept and live with change and adaptation (values and culture) are important social elements. Continuous change and self-reinforcing adaptation is essential to make a population and state wealthier. However, this needs to be coupled with investments and displacement of the mentality of self-sufficiency. Social institutions ought to be in place to streamline certain accepted behaviour in the desired way, even to

Evolutionary Economic Geography

214

interfere with our freedom of choice in certain cases: to tax or outlaw the consumption of things that many consider to be harmful such as spirits, tobacco, gambling and pollution; to prevent and outlaw production, trade and consumption of drugs; and to subsidise education, libraries and culture-related activities and institutions such as museums, operas and ballet. Societies of obedience, like the one in ancient Egypt, were rather static in comparison to the societies of will. Societies of will, like (north-western) Europe from the Middle Ages, could prove to be quite dynamic. However, the situation may change over time. Too generous a welfare state may reduce social dynamism in certain areas. Elsewhere, the former societies of obedience (China and India) may prove to be quite dynamic and enterprising, as has been the case from the end of the twentieth century. The Roman Catholic Church was initially quite opposed the printing press. This Church wanted to have control over the written word, its spread and interpretation. The printing press, however, liberated scientific research, and the publication and exchange of ideas that were outside the Church’s scrutiny. If a scientific paper was not published in one location, it could be published elsewhere. Galileo Galilei (1564–1642) made discoveries with his telescope in 1610 that overturned the comprehension of the world. His telescope helped him to discover Jupiter’s satellites. This discovery gave him the idea that not everything has to revolve around the earth. Galileo could not, however, persuade the inquisitors to take a look through his telescope. The inquisitors’ ideology persuaded them that what Galileo claimed to see was, according to their doctrine, not there. This was a clash between evidence and authority. Galileo was forced to renounce his heliocentric view of the world and spent the last years of his life under house arrest near Florence on orders of the Inquisition. Few reformers lead a happy social life. Galileo’s later works were published in Holland, rather than in what we know now as Italy. So it was opportunity and liberty, rather than pure ingenuity, that moved the location of scientific advances northwards in Europe. Many of the material amenities that we have today we owe to Galileo’s victory. Some 350 years after Galileo’s death, in 1992, Pope John Paul II gave an address on behalf of the Catholic Church in which he admitted that errors had been made by the theological advisors in the case of Galileo. He declared the Galileo case closed, but he did not admit that the Church was wrong to convict Galileo on a charge of heresy because of his belief that the earth rotates round the sun. There is, however, an additional and rather unconventional view about the Galileo issue (Kalajicć, 1979, pp. 293–301). Galileo is often portrayed as winning a victory for the ‘light’ of spirit over the ‘darkness’ of the Christian Church. Some may argue in this debate that it is highly questionable what was really darker: the victorious ‘light’ or the defeated ‘darkness’. Consequences of scientific exclusivity and monopoly over truth may be as fatal as the effects of Christian dogma. There is a view that the Church was a more tolerant party in the Galileo case than was Galileo himself! The church authorities did not request Galileo to give up his theories, but to give up his pretension to absolute truth. The Church was willing to accept Galileo’s views as hypotheses, while Galileo (and Giordano Bruno) not only had ambitions in science but also demanded a fundamental overhaul of church dogma. This implied the transfer of authority from the Church to science. At that time, this was a totally unacceptable request for the Church.

Market structure and location of production

215

An acceptance of the scientific views of the world could have caused a hold-up or even a degradation of the spiritual development of humans. This effect was perhaps no smaller than the impact of the acceptance of Christian dogma on the stagnation of the development of science. From the time of Galileo, science has tried to replace religion in the minds of some people with another type of ‘emptiness and superstition’–and in certain cases it has succeeded. Science tries to ‘establish’ theories and to confirm them as synonyms with truth. However, science may not always penetrate their qualitative and spiritual aspects, which are most often fundamental and point to constants. In such instances, science often refers to pure statistics, indicating only the mechanical, material and quantitative properties of samples. Science has not yet defined its principal workhorses: light and sight. The experiment by Albert Michelson and Edward Morley (1887) has shaken Newton’s views about light. What kind of credibility can we have in theories based on experiments if that approach does not know its basic experimentation tools? Wallace Kantor also seriously questioned Einstein’s constant speed of light (it is changeable in relation to the movement of its source; hence the theory of relativity needs re-examination). Perceptions are changeable and often individual. Eyes can sometimes ‘see’ illusions and fantasies, so one may not always accord absolute values to what is observed. As such, one may have certain doubts about the final use of science for the benefit of humanity. The ultimate question is: would the last drops of crude oil be used for the powering of tanks and missiles or for the movement of school buses? If there are no limits, and if some scientific doors open on to the road to hell, in one sense, then the culprit is the one that enters there. Science is not forcing anyone to step in. The humanist critique of science often limits itself to the enumeration of negative consequences of science instead of entering into its essence. In this regard, it is the weakness and failure of humanist thought that bears responsibility for the victory of science and its authority as the only replacement for lost and authentic humanist values. In this environment, science is taken to be a field in which there is certain agreement. This may lead to the situation in which there is too much confidence in its authority. Nonetheless, science may not give satisfactory replies to all desires and requests. There is therefore a tendency to obscure the frontier between scientific and humanist thought. The religious authorities in China and in the Muslim world34 were, during the same era, quite successful in their control of publishing. ‘The pluralism in Europe rendered ineffective any attempt at outright suppression or rigid censorship of the printed word’ (Lipsey et al., 2005, p. 237). This assisted in the spread of Isaac Newton’s (1642–1727) ideas related to new science (physics, mechanics, calculus, rates of change in space and time … ). The location of these advances and their spread was in Europe. Europe and Euro-Asia had a certain earlier geographical advantage over other continents. The east–west orientation of the general geographical ‘plateau’ eased and stimulated trade and migrations (nomads such as Franks, Goths, Huns, Moors, Slaves and Turks), as well as supporting the transfer of domesticated breeds throughout the area. Large animals were domesticated for food, work and communication (pigs, sheep, goats, cows, horses). Elsewhere, this was not too simple. The north–south orientation of the Americas impeded an easy transfer of breeds from one latitude to another; a similar passing-on in Africa was frustrated by sharp differences in climate and the impossibility

Evolutionary Economic Geography

216

of domesticating giraffes, hippopotamuses, zebras, antelopes or rhinoceroses (it is very hard to breed all these animals in captivity); while Australia lacked ‘useful’ animals. The convenient geographical location of Euro-Asia had other consequences. It provoked both a relatively easy transfer of diseases and a creation of immunity in the local population through natural selection. So when the Europeans got in touch with other civilisations, in the Americas for example, European diseases devastated local populations. Hence a relatively small number of Europeans, such as the Cortés group of 600 men, fifteen horses and a few light cannons, could conquer and rule enormous spaces. A relative openness of hundreds of kingdoms and fiefdoms in Europe to the continuous inflow and exchange of ideas, communication (even through marriages and/or wars) and competition, made Europe mobile, particularly about the time of the First Industrial Revolution. Elsewhere, where the selfcontained and relatively closed empires were in place, the monolithic, sometimes stagnating and non-competitive social structure was upset only after the arrival of Europeans with their commerce, ideas and guns. The Arab world now excels in a handful of technologies that include desalination, camel reproduction and falconry research. The Arab world, especially in the oil-rich countries in the Middle East, invests heavily in research centres. These countries would like to emulate the successes of research done in small countries such as Finland, Ireland and Singapore. These intentions are not without difficulties. Some of the funds are directed towards military technologies, while there is still a substantial brain drain from the Arab world.35 Politics and vested interests Politics and vested interests are often linked with resistance to change.36 All societies have tried to resist change at one time or another, but this is inevitably self-damaging and in vain in the long term. For example, ancient China looked with suspicion at new ideas introduced by foreigners. Indeed, the country’s rulers often banned such innovations and protected their own. China started silk-making some 5,000 years ago. It kept this technological secret for over 3,000 years. The punishment for revealing this secret was death. However, as the power of the empire waned over centuries, the circle of those who knew this technology widened. About 300 AD silkworms reached Korea and Japan. Silk cloth arrived in Europe in about the second century AD. The Byzantine emperor Justinian I sent several monks to China in the sixth century for the purpose of what we would now call ‘industrial espionage’. They brought silkworm eggs in hollow sticks to Greece. Later this spread to what is today known as Italy and France and elsewhere. China, once the world’s technological leader, provides a sobering lesson on how economies may slide down the international league table. In the eighteenth century it was the biggest world economy, with a GDP seven times as large as Britain’s. But it kept its doors closed to foreign goods, so it was left behind by the industrial revolution and the exposition to global trade. In 1793 Lord George Macartney was sent to Peking by King George III to establish a permanent British presence and open up trade relations with

Market structure and location of production

217

China. But the Chinese emperor Quianlong informed his visitor that ‘we have not the slightest need for your country’s manufactures’. China’s economic isolation was to last for almost another 200 years, during which real incomes fell. By 1950 China’s GDP per person had shrunk by a quarter compared with Lord Macartney’s day; Britain’s had risen fivefold.37 Nonetheless, China has changed and opened its economy, in particular since the 1990s. If there are economic historians around a few hundred years from now, what they will find remarkable may not be the relative rise of China and India in the twenty-first century, but the relative decline of these countries in the nineteenth and twentieth. Two hundred years ago, no one knew what gross domestic product was, far less how to measure it. But painstaking research by economic historians, notably Angus Maddison, gives us estimates. China and India then accounted for almost half of world production.38 China and India were the two biggest world economies up until about 1820, when the steam engine and power loom gave Britain its industrial lead. Openness to a technological change (in both mechanical and organisational technologies) and the speed of its acceptance and absorption were among the key elements in the British economic success. After two centuries, the economic powerhouse of the world is moving from the northern Atlantic to the Pacific region. Hence a possible setback in the American and/or European economies may have a much smaller impact than was the case even a few decades ago. (China and India are also among the strongest sources of greenhouse gas emission growth in absolute, but not in per capita, terms.) The West, in particular Europe, grew rich initially because ‘the societies broke loose from the norms and constraints of old institutions, and kept political process from doing too much, and let the market work’ (Nelson, 1995, p. 83). Britain also had an ample supply of raw materials both inland and from colonies, schools were open to Newtonian physics and its application to real problems (the middle class wanted to commercialise technological achievements), and its legal system constrained the capricious behaviour of the Crown. Many things in Britain would have evolved in a very different way had the absolutist king Charles I won the English Civil War of 1642–6 against Parliament (the interests of the nobility and the middle class). There are examples where this (relative) social openness in Europe was rather small. Arabic numerals were not accepted in Europe for a long time. Leonardo Fibonacci (1170–1250), considered to be the most talented mathematician of the Middle Ages, promoted the use of these numerals, but faced resistance. These numerals were even forbidden for the use in commercial transactions in the fourteenth century. The reason was that it could be easier to forge, for instance, number 1994 than the Roman MCMXCIV. And in Danzig (Gdansk) Anton Müller invented the ribbon loom in 1529. This invention, which produced six pieces of cloth at the same time, proved fatal for the inventor. The city council, fearing the unemployment and pauperisation of a large part of

Evolutionary Economic Geography

218

the local workforce, suppressed the use of the loom and ordered that Müller be secretly drowned.39 These looms (re)appeared three centuries later in England. Barthelemy Thimonnier (1793–1857) invented the first operating sewing machine in 1830. This French tailor was almost killed and his whole business was ransacked and burned down in Paris in 1831 by an angry mob of 200 tailors who feared unemployment if his invention was commercialised. Thimonnier had to run from Paris for his life, and died in poverty. These examples of novelties, sometimes fatal for their inventors, showed the initial resistance by vested interests to change and the spread of new ideas. There are quite a few innovations that appeared before their time. They were muffled and lost in their time and environment. However, they appeared after a long pause, sometimes in a different location in space. In spite of this, Europe advanced faster than other regions because it accepted them over the longer run or, better, it tolerated much more than did others the evolutionary disturbance that was offered by novelty and changes in technology. Despite the mentioned examples, Europe was in general more open to novelties, in particular since the time of the greatest visionary of the early modern period Leonardo da Vinci (1452–1519) and the titan of modern physics Isaac Newton (1643–1727). Whether Europe continues to be more open to technical and organisational novelties in the twentyfirst century, in comparison to other regions, is another issue. Institutions Institutions are taken to mean here organised interests or factors of collective behavioural order. They are ordered collections of rules, laws, regulations, property rights, habits, practices, customs and routines that are enforced and that enable, guide and normalise relations and behaviour among various parties. As already seen, these institutions may be formal (such as laws) that arrive after long considerations and informal (trust, culture, habits, group norms) that are a spontaneous product of continuous and repetitive interactions among individuals over a long period of time. Institutions shape not only relations among economic agents, but also their ability to alter and improve existing institutions (e.g. laws) needed for adjustments to newly arrived or created situations. Existing institutions have strong links with the ones prevailing yesterday and, often, with the ones from the previous century. One of the reasons why the western world ‘grew rich’ relatively fast, in particular from the eighteenth century, was that it broke ties from its old institutions. In the economically successful countries, institutions that provided a favourable environment for innovation and growth were created. In Britain, for example, the Magna Carta of 1215 gave subjects the right to their own property. They were protected from the Crown, which until then had been entitled to seize property at random. In contrast, arbitrary confiscation continued to be common in the Muslim world and in Asia. Communists were also confiscating property in eastern Europe, with negative economic and social results in the long term. In the western world, rulers quickly learned that a tax on property was more profitable than random confiscation, as tax proceeds may continue to accrue indefinitely. There also ought to be institutions that can accumulate, store, teach and extend scientific knowledge. Medieval universities in Europe, their libraries, scholars and

Market structure and location of production

219

teaching methods provided this relatively independent institutional framework. Even though China and the Muslim world had superb innovations and knowledge, they lacked such an institutional framework. The outcome of all this was a social system in Europe that promoted innovation and growth more than any previous system. Britain was the leader in the actual commercialisation of scientific achievements and the First Industrial Revolution. The reason was the understanding and application of Newtonian experimental and mechanical science (elsewhere in Europe science was more deductive and mathematics-oriented at that time). In any case, in modern times the new institutions have almost eliminated the gap between frontline science and applied technology. China lacked the institutions of higher learning that would house the collective memory of scientific discoveries and look after scholars. Everything was highly centralised in China and done in the name of the emperor. Secrecy also worked against the spread and accumulation of knowledge, mathematics in particular. The purpose of the fifteenth-century Chinese fleet of 2,100 ships was to spread the fame of the emperor and Buddhism to the ‘barbarians’. Goods, such as silk, porcelain, paper and collapsible umbrellas were carried on many large and long voyages, not for commerce but principally as presents to impress foreign people. They were undertaken more as expressions of Chinese confidence in themselves and frivolities than as economic or strategic urgency. The fleet of 1405, led by Zheng He, was far superior to anything that Europe could produce at that time or for the coming two centuries. The next Chinese emperor disdained commerce and opposed contacts with the barbarians. Hence, in 1525 all ocean-going ships were destroyed and their owners arrested (Lipsey et al., 2005, p. 285). This was also prompted by threats of possible invasion from the north.40 Gunpowder, for example, a Chinese invention, was used in China as a toy, while it was an industrial tool and a part of war technology in Europe. In addition, ignorance of Newtonian mechanics put China at a big technological disadvantage vis-à-vis Europe in the eighteenth century. R&D plays a crucial role in the innovation process as it sustains a supply of new knowledge. The great importance and lasting value of learning and institutional change may be seen in this example: The Islamic authorities resisted the use of printing because they saw it as encouraging ordinary people to interpret the holy writings for themselves rather than as dictated by the clergy, a fear shared by the Catholic church. European Protestants, on the other hand, embraced printing for the very same reason … So the decline of Islamic science can be traced to a number of developments that were different from those in the West, many of them due to historical accidents: a theocracy with no concept of degrees of jurisdiction or of a developing code of laws; religious hostility to established science and free enquiry; no major institutional innovations such as the corporation – in the form of guilds, universities, independent cities, and business organizations. In short, while the West was developing a pluralistic society, Islam was solidifying a monolithic, relatively rigid, theocratic society. (Lipsey et al., 2005, pp. 275–6)

Evolutionary Economic Geography

220

A lack of independent institutions (libraries, classrooms, scholars with full academic independence and freedom from prosecution) that could preserve the ‘collective memory’ of scientific discoveries, act as incubators of science and spread the knowledge necessary for advances and innovation was lacking both in China and in the Muslim world.41 The knowledge was ‘there’ but could be easily lost. While China was able to match Europe in many areas of manufacturing until the end of the eighteenth century, it lacked the modern science that gave Europe a lead at that time. Everything that was done in China was centralised and done to the glory of the emperor. The Muslim countries were very strong in international commerce, navigation and conquest, but internally they were basically subsistence farming economies. That is why, compared with others, Europe had a general spurt in economic growth from the nineteenth century. Japan industrialised relatively fast towards the end of the nineteenth century because it opened its economy to the introduction of European technology. Societies of obedience, like China and India, even though they were strong and prosperous, were complacent. The general social purpose was served more by continuity than by progress. External pressures were not extremely strong, so in general the society became smug. This has changed in both countries in the past decades. Both China and India are evolving into societies of will, where there may not be social peace without economic progress. Cumulation Another possible explanation for the relative constancy of the countries that innovate most is that innovation reflects the cumulative and interrelated nature of acquired knowledge. Once it exists, knowledge does not cease to exist, and discovery builds on discovery. Knowledge has a non-rivalrous property. If someone uses it, it does not prevent others from using it at the same time. Innovation also reflects a change in technological capabilities and economic incentives.42 It is related not only to the creation and absorption of new knowledge, but also to its adaptation, extension and control within an innovation-friendly environment. All in all, this provides strong grounds for the creation of dynamic comparative advantages, certain irreversibilities43 and economic growth44 of firms and nations as success breeds the potential for further success. An old saying says: ‘Nothing succeeds like success.’ The higher the levels of accumulation of knowledge and capital stock, the greater are the benefits of technological progress45 and vice versa. Linus Torvalds, a Finnish software engineer, was unsatisfied with the available proprietary computer operating systems. They jammed, crashed, resisted modification … so he embarked on building his own in a way that would be open and friendly with free modification. Torvalds was twenty-one years old when he started developing Linux. The year was 1991. A large number of volunteers were involved in the development of Linux. They were coordinating their work through the Internet. A new release was circulated each week. Many users and enthusiasts pooled their knowledge and were sending their feedback within a short time. Bugs were filtered out and Linux expanded. Bit by bit, Linux was evolving to a superior state through the system of filtering, selection, mutation, expansion and replication. This was not imposed by the proprietor, but rather by the non-pecuniary evolutionary developments. Non-rivalrous knowledge was pooled

Market structure and location of production

221

and used by many at the same time. Knowledge is not a rivalrous piece of land that can be occupied only by one person at a time. ‘The nonrivalry of knowledge meant that the crown jewels of economics, the Invisible Hand theorems, could not stand. The market simply could not get the prices of these goods “right”’ (Warsh, 2006, p. 365). The export performance of countries, as well as in differences in labour productivity, may have roots in the available knowledge and its use. Innovation is concentrated in a few firms in industries with high entry barriers such as aerospace, chemicals, automobiles, electric and electronic industries, while it is spread among many firms in machinery and the production of instruments. (Dosi et al., 1990, p. 102) Empirical studies show that monopolisation or concentration is not the main reason for innovation. Cumulative (clustered) knowledge enabled Germany to excel in the fields of chemicals and high-quality engineering, Britain in pop music and publishing, Italy in fashion and design and the US in computer software, aeronautics and the cinema. But innovation is also to be found in industries that are less concentrated and where there are no significant barriers to entry. New entrants may have greater motivation to test and develop new products and technologies than well-established firms. In many industrialised countries, the average size of firms is becoming smaller, not bigger. This reflects increased demand for more custom-made goods, produced in smaller batches, utilising production factors that can be readily switched to various alternative uses. But this is only on average. The industries with the most advanced technology are frequently the most geographically concentrated, highly profitable and often the largest. Modern technology is increasing the importance of capital, especially human capital.46 Krugman (1996a, pp. 13–14) found statistical evidence from the US economy that the ‘really high value-added’ industries (in relation to the number of employees) are cigarette manufacturing and petrol refining, whereas the so-called ‘high-technology industries’ such as aircraft and electronics were about average. However, one has to remember that high-technology industries have important externalities and linkages for the whole economy, although the number of computers a nation has is somewhat less important than how and for what purpose they are used, for example playing Tetris or Solitaire or organising inventories, production, transport and distribution. Another problem is that most computers are used in the services sector (finance, accounting and health care) where it is hard to measure output. Commercialisation of inventions It is one thing to invent or discover new or improved goods or services (product differentiation) and/or uncover a new way to produce or market already existing goods and services, and quite another to exploit that success commercially. The electric dynamo was invented in 1881, but it took firms four decades to reorganise plants to take advantage of the flexibility in production offered by electric power. The basic videocassette recorder (VCR) technology was the result of an invention by Ampex in the US in the 1950s. When Ampex started selling VCRs in 1963 the retail price was $30,000

Evolutionary Economic Geography

222

apiece (very expensive for individuals even by today’s standards). No wonder that only a few were sold over several years. Latecomers may overtake the original inventor, but they must be much better in the market race. Philips, a Dutch company, produced the first VCR aimed at the consumer market in 1971, several years before Sony introduced its Betamax model. Soon after, other Japanese manufacturers entered the market, and before long they came to dominate the international market for (home) video equipment. To avoid a repeat of earlier experience, Philips took a different tack after inventing compact discs and developed the final technology jointly with Sony. Hence, the geography of innovation and the geography of production need not coincide. The microprocessor was invented in 1971, but firms are still learning how to make best use of it. One of the reasons for this situation is Moore’slaw,47 which refers to the doubling of chip performance about every eighteen months. In a similar vein, commercial jet technology was a British invention. This early lead resulted in the production of the first jet commercial transport aircraft, the Comet. The project was ill-fated because of the then unexplained problem with metal fatigue. An early technological breakthrough was linked, as in many other cases, with huge uncertainties and shifts in the location of production. Later, the US took the lead (with Boeing and McDonnell-Douglas), which was, subsequently, seriously challenged, even overtaken in some market segments, by the European Airbus (a consortium of government-supported British, German, French and Spanish firms).48 Also, Polaroid was the first to develop specific sensor technology. After a decade of heavy investment and research, it had the prototype of a digital camera in 1992. This was converted into a product for the market in 1996. However, through a marketing and distribution mess-up, Polaroid left this business at the end of the 1990s. The Japanese entered the business and captured this market. Being first in the introduction of a new product or technology still counts. But there are many examples of firms catching up and overtaking the early entrant to a market. Nonetheless, if a firm is somehow a late entrant, it has to be a lot better than the early leader. Berkey produced the first hand-held electronic calculators, Ampex the first video recorders, Gablinger developed the first low alcohol lagers, Chux sold the first disposable nappies. But none of these companies made a sustained commercial success of their innovations. Today the calculators we use are probably made by Casio, our video recorder comes from Matsushita: our low alcohol beer is Philip Morris’s Miller Lite, the diapers we buy are made by Procter and Gamble. The pioneer was swept away. (Kay, 2006, p. 59) In addition, Apple developed the PC market, but most of the PCs in use are IBM-related. Xerox introduced fax machines, but others took this market. American Express introduced credit cards, but it is Visa and Master Card that rein in this business. The evolutionary business environment is not always ‘kind’ to those that break new technological grounds.

Market structure and location of production

223

When you are told that the key to the future business success is to see the future more quickly or more clearly than the other people, ask which established cases in business history illustrate the point. And try to remember Berkey and Ampex, Gablinger and Chux. (Kay, 2006, p. 61) Forecasting is hard even if a firm may see the business future more or less accurately. The problem is timing to get the business move ‘right’. Still there are examples from the 1980s when Sony with its Walkman and Nintendo with its Game Boy ‘got it all right’. But it was at that time and with those specific products. In a relatively integrated area such as the EU, one would expect the prices of similar goods in different countries to be similar, because of competition and trade. The stronger the competition and the larger the volume of trade, other things being equal, the smaller the price variation. Pre-tax prices of the same good would be expected to vary only as a result of differences in the cost of transportation, handling, insurance and, to some extent, marketing techniques between countries. However, this expectation is not borne out in reality. The EU was aware of the barriers to competition other than tariffs and quotas (NTBs), so it created the Single European Market in 1993 which was supposed to mitigate or even rectify this shortcoming. Various regional factors prevent full equalisation of prices. Some geographical areas may have small markets for certain goods and services (for example, parasols in Finland and anti-freeze in Greece or Egypt). So, in order to do business and make some profit there, firms may price their goods relatively higher. In this case market presence may be a much more important public policy and private business objective than requiring price homogenisation in all markets. Or, for example, differences in taste or special requirements regarding the basic ingredients of a product (for instance, chocolate) may cause the price of a good to vary. And, if there are local substitutes, foreign suppliers may modify the price of their goods in order to be competitive. Some goods (such as wine) may be regarded as luxuries in one country and taxed accordingly, but regarded as basic necessities in another country. This widens the price gap for the same good in different countries, and sometimes even in different regions of the same country. In a perfectly competitive market (with free entry and exit) for a good, free competition ensures equality of prices and drives profits to zero. In imperfectly competitive markets that are lightly regulated there is scope for price variation and, therefore, for profits or even rents (supernormal profits). Aircraft manufacturing is a deeply politicised industry. It is difficult to operate it on purely commercial basis.49 For example, Airbus has to keep a fine Franco-German spatial manufacturing and management balance. The industrial group has two chairmen, two chief executive officers, its headquarters is split between Munich and Paris, and so as long as keeping this balance is essential (e.g. potential cutting of jobs ought to be geographically balanced, not in line with operational needs), the commercial performance of the group may suffer. The effects of the weak dollar in 2007–08 were hitting Airbus industries hard. One of the effects is that Airbus pays suppliers in (strong) euros, while it sells its aircraft in (weak) dollars. A partial way out of this problem is to relocate a part of the production to

Evolutionary Economic Geography

224

the US and/or China. In fact some 500 Airbus suppliers around Toulouse are doing just that. A new agreement that would regulate trade in this duopoly market for civil aircraft is necessary. This industry and market is so important that it should not be left to the vagaries of market forces. New parties such as Japan, even Brazil, Russia and China, may need to enter into any new agreement as there is a lot of outsourcing in this business. Accounting rules ought to be transparent, in particular in the post-Enron and postParmalat world. Perhaps there ought to be a cap on allowed direct and indirect subsidies and a timetable for their reduction. There should also be a mechanism for the oversight of the business with certain and meaningful justice. Following big business accounting scandals such as Enron and Worldcom in the US, the Sarbanes–Oxley Act (2002) (commonly known as SOx) had a relatively modest goal. It was to improve the accountability of managers to shareholders and to restore confidence in American corporations. However, SOx turned out to be one of the most influential and controversial pieces of corporate legislation. Its section 404 keeps managers responsible for internal control and other procedures related to financial reporting. Harsh criminal penalties wait for transgressors. Zealous auditors imposed exorbitant compliance costs. Size of the local market The size of the local market competition and supply of skills are important ingredients in the complex links between technological opportunities and entrepreneurial decisions. Economic integration with other countries increases the size of the market. Other marketwidening tools include deregulation and privatisation. Small open economies inevitably have to rely in part on various foreign technologies. Such countries often do not have the necessary resources and need to develop basic technologies for all lines of production. If this situation is regarded as detrimental, then economic integration may increase the pool of resources (human, technological and financial) for innovation and the development of new technologies, products and inputs, which may mitigate the potential disadvantage of smallness and isolation. Such pooling started in the EU in the mid-1980s with a series of R&D programmes. A positive effect on innovation (creation of technology) in the EU was expected to come from the operation of the Single European Market. A deepened regional market would stimulate competition and provide an incentive to innovation that would further promote competition for the benefit of the consumers. If corrective measures (in support of trade, competition or industrial policy) were added, they would not necessarily violate a liberal trading system in the long term. They would simply add an adjustment mechanism to the already highly imperfect and suboptimal market situation. Intra-industry trade The ‘awkward fact’ is that trade takes place in differentiated products (Eaton and Lipsey, 1997, pp. 228–9). Strong advertising campaigns create awareness about ‘differences’ among what are basically very similar and easily substitutable goods (for instance, cars, printers, fax and photocopying machines, T-shirts, skis, soaps, perfumes, toothpastes, TV

Market structure and location of production

225

sets, painkillers, breakfast cereals, refrigerators, trainers, cigarettes, bicycles or audio-, video- or DVD-recorders and discs). This phenomenon is known in theory as intraindustry trade.50 This empirical finding had one of the strongest impacts on the way economists think about international trade and production. Intra-industry trade is the result of the desire of consumers to have a wide choice of goods and of economies of scale at the level of firms. Generally, product differentiation tends to dominate product specialisation in the internal trade of the EU. Although increased competition offers potential gains from both more efficient industrial and geographical allocation of resources in production and increased consumption, there is no guarantee that these gains will be achieved in practice. If a government takes this view and believes that domestic production will be wiped out by foreign competition, then it may pursue a policy of protection on the grounds that it is better to produce something inefficiently at home than to produce nothing at all. This disastrous scenario (locational catastrophe) has not been borne out in reality. The very existence of the EU, which has continued to expand in size as well as in scope and depth, is the best example of a positive scenario. Most firms in the EU countries have not been put out of business because of competition from firms in partner countries. Instead, many of them have constantly increased their business in the long run. They have specialised in lines of production that satisfy distinct demand segments throughout the EU. This fact cannot be explained by the neoclassical theories. Another proof can be found in North America. GM, Ford and Chrysler operated in the US, Canada and Mexico before negotiations started on the 1965 US–Canada Auto Pact (free trade in cars and components); the 1989 Canada–US Free Trade Area and the 1994 NAFTA. When the negotiations on NAFTA commenced, all three trade ministries were aware (with great relief) that the elimination of trade and investment barriers would not prompt large parts of the domestic auto industry to shut down business in one location/country and to move to another NAFTA partner country. They expected instead that local producers would just speed up the ongoing process of specialisation. Intraindustry trade was also expected to thrive, as was the case with the 1965 Auto Pact. These expectations materialised in reality (Hufbauer and Schott, 2005, pp. 365–7). There are also other examples that support the thesis of a smooth intraindustry adjustment in trade and in geography of production of the concerned countries to successive rounds of negotiations within the GATT that reduced tariffs and other trade barriers. The ensuing intra-industry adjustments in trade and specialisation among developed countries were relatively smooth. Contrary to the expectation of the factor endowment theory, intra-industry adjustment prevailed and carried fewer costs than would have been the case with inter-industry adjustment. If it is feared that foreigners will eliminate domestic firms through competition, exchange rates can act as an important safety-valve to prevent this happening and ease the process of adjustment to the new situation. Inter-industry trade between countries basically reflects differences in national factor endowment. This type of trade brings efficiency gains through resource allocation, as well as benefits to consumers that come from the supply of a new set of different products. Intra-industry trade, on the other hand, is associated with product differentiation. Consumers benefit from this type of trade through increased variety of

Evolutionary Economic Geography

226

closely substitutable goods and services, as well as through increased competition of products made with increasing returns of scale. As incomes have risen, consumers are no longer satisfied with identical or standardised goods. They demand and pay for varieties of the same basic good or service, often tailored to their individual needs and tastes. The larger the variety of demanded and available goods and services, the smaller the importance of economies of scale. Intraindustry trade refers to trade in differentiated goods. It occurs when a country simultaneously exports and imports goods (final or semi-finished) that are close substitutes in consumption. Differentiation of goods begins when various characteristics are added to the basic good or component, backed up by strong R&D and advertising campaigns. Thus, important gains from trade in differentiated goods may arise through an increase in consumer choice, not necessarily through lower prices only. The variety of goods produced in a country, the new theory suggests, is limited by the existence of scale economies in production. Thus, similar countries have an incentive to trade. Their trade may often be in goods that are produced with similar factor proportions. Such trade does not involve the major adjustment problems that are commonly found with more conventional trade patterns (Krugman, 1990a, pp. 50–1). In fact, one of the most distinctive properties of the liberalisation of trade in the EU was an increase in intra-industry trade coupled with modest adjustment costs (Sapir, 1992, pp. 1496–7). At the heart of neoclassical international trade and customs union theory is the analysis of two goods only. Therefore, it cannot satisfactorily account for preference diversity and intra-industry trade. The neoclassical theory’s ‘clean’ model of perfect competition and equilibrium is not applicable here. The potential for intra-industry trade increases with the level of economic development, similarity in preferences (tastes), openness to trade and geographical proximity, which reduces the costs of transport, marketing and after-sales service. A significant portion of trade among developed countries is intra-industry. In this case, variety may be preferred to quantity, so that some proportion of trade is attributable not only to differences in factor endowment, but also to different national preferences (tastes). This is the case in the EU. The response of successful firms to such business challenges is to find a specialist market niche and to employ economies of scope, rather than scale. In an ‘early’ example of intra-industry trade, Linder (1961, p.102) noted that ships that brought European beer to Milwaukee took American beer back to Europe. Although lacking a formal theory, the examination of international trade flows by Grubel and Lloyd (1975) noted that an important part of these flows was within the same industry classification group. Finger (1975) believed that intra-industry trade is an anomaly due to the definition of new products and processes and statistical sorting of data. However, Loertscher and Wolter (1980, p.286) demonstrated that intraindustry trade between countries is not a statistical fabrication, but a real phenomenon. This type of trade between countries is likely to be strong if: • they are both relatively highly developed; • the difference in their level of development is small; • they have large national markets; • the barriers to trade are low;

Market structure and location of production

227

• there is a high potential for product differentiation; • entry in narrow product lines is obstructed by significant barriers (sunk costs); and • transaction costs are low. Intra-industry trade is relatively high among developed countries. It refers to trade within the same trade classification group. One may, therefore, wonder whether intra-industry trade is a statistical aberration rather than an authentic phenomenon. In addition, it may be argued that two varieties of the same product are not always two distinct goods. The criteria for data aggregation in the Standard International Trade Classification (SITC) are similarity in inputs and substitutability in consumption. These criteria often contradict each other. Many of the three-digit groups in the SITC include heterogeneous goods. For example, SITC 751 (office machines) includes typewriters, word-processing machines, cash registers and photocopying machines, whereas SITC 895 (office and stationery supplies) includes filing cabinets, paper clips, fountain pens, chalk and typewriter ribbons. On these grounds one could conclude that intra-industry trade is a pure statistical fabrication. However, this is not so in reality. If one studied trade groups with more than three digits, differences could and would appear. The index of intra-industry trade (IIT) in a country is represented by the ratio of the absolute difference between exports and imports in a trade classification group to the sum of exports and imports in the same classification group: (Equation 4.2) The intra-industry trade index (Equation 4.2) is high and is equal to 1 for complete intraindustry specialisation (a country imports and exports goods in a group in the same quantity). This is a sign of a geographical spread of an industry. The index is low and equals zero for complete inter-industry specialisation. Such spatial concentration of production is usually the result of high entry barriers and economies of scale. However, the index does not distinguish between cross-hauling of final output and intra-industry trade that is the consequence of production sharing within an industry. It does not distinguish between quality-differentiated trade in goods either. In any case, intraindustry trade increases welfare because it extends the variety of available goods to the consumers. As there is no strong evidence of large adjustment costs to this type of trade and spatial reallocation of production, one may conclude that intra-industry trade makes almost everyone better off. The ex ante expectation that trade liberalisation and integration could shift the IIT index closer to 1 (suggesting a geographical spread of production) in the case of developed countries has been investigated in numerous studies.51 Research shows that incentives to intra-industry trade are similar levels of per capita income and country size, product differentiation, participation in regional integration schemes, common borders, as well as similar language and culture. Negative influences on this type of trade are exerted by standardisation (reduction in consumer choice), distance between countries (which increases the cost of information and services necessary for trade in differentiated goods) and trade barriers that reduce all trade flows (Balassa and Bauwens, 1988, p.1436).

Evolutionary Economic Geography

228

The fact that a large part of trade among developed countries is intraindustry may lead to the conclusion that the Heckscher–Ohlin (factor proportions) theory of trade (and indirectly theory of location of certain production) is not valid. Intra-industry trade is not based on differences in factor endowments among countries. Countries tend to specialise in production (locational dimension) and export goods that are demanded by the majority of domestic consumers. It is this demand (‘home market effect’) that induces production, rather than domestic factor endowment. Countries have a competitive edge in the production of these goods and thus gain an advantage in foreign markets, while they import goods demanded by a minority of the home population (Linder, 1961). The US, Japan and Germany have the greatest comparative advantage in goods for which their home market is relatively big. These are standardised goods for mass consumption. There is, however, one major exception. The major market for German dyes is the British textiles industry (and to a lesser extent that in the US). The German domestic market for dyestuffs is relatively small (Nelson, 1999, p.12). Toyota has achieved world leadership in passenger cars, even though its domestic market is smaller than that of General Motors. The size of the domestic market in this case became less significant than the technological competitive advantage. And since the 1990s there has been another big global change: the arrival of China as the global factory for a number of manufactured goods. Its domestic market and potentially huge demand is not the general explanation for Chinese success. Even though the domestic market for artificial Christmas trees in China is negligible, China is the supreme world producer of these trees. Superior performance or underperformance of firms in one country compared to their counterparts or competitors in other countries in terms of productivity (GDP per worker) is related to many factors. Regarding European comparisons with the US, these include and reflect reasons such as: • direct taxation, • personal choice (work vs leisure), • geography (more real resources are invested on average in the US in cooling or heating compared to Europe because of the harsher American climate), • security costs are higher in the US than in many other countries because of crime and fear of terrorist attacks, • there are also more resources invested on average in the US in prisons than elsewhere in the world, and • the average propensity to use private (rather than public) transport in the US is higher than elsewhere in the world (this increases US GDP more than US welfare). All these issues are open to debate.52 However, the important thing to compare is not necessarily the productivity rate, but rather general trends in productivity rates among different locations/countries. It is important to know not merely that productivity growth (or general economic growth) take place, but also and perhaps more importantly where this growth takes place. The analytical and policy interest may be in the location of this growth both in space and in industry structure, as well as the overall impact it has over time. The ‘American System of Manufacturing’ during the nineteenth century was based on abundant natural resources. Labour-saving and capital-intensive manufacturing

Market structure and location of production

229

techniques were employed in the production of standardised goods by specialised machinery. This culminated in Henry Ford’s assembly line from 1906 (the unfinished car should move, rather than the workers). Nature was less generous in Europe than in the US. Manufacturing in Europe was based on specialised craft techniques (labour-intensive and resource-saving). The general American System turned out to be capable of creating more technological improvements than the one in Europe during the twentieth century (Lipsey et al., 2005, pp. 104–5). An important analytical question is whether factor proportions (Heckscher–Ohlin) or economic geography (determined by economies of scale and market access) is more important to predict trade within industries. Limited research found that in the case of Sweden the answer was both. A large domestic market, as well as plentiful endowment of human capital, increases the quality of exports (Greenaway and Torstensson, 2000, p.277). Petrus Verdoorn suggested that the principal difference in manufacturing between US and European firms is not so much the size of the firm/plant as the length of the individual production run. The range of processes carried out in the same factory is much smaller in the US than in Europe (Hague, 1960, p.346). Compared with plants in the same industry, production runs in the US are several times larger than in Europe, even when the plants are owned by the same TNC (Pratten, 1971, pp. 195, 308–9; 1988, pp. 69–70). On the other hand, even in the most efficient developed countries, manufacturing is often carried out in factories of quite moderate size. Differences in plant productivity are best explained by (1) inappropriate labour relations, in particular where many thousands of workers need to be employed together; (2) inadequate levels of technical training; and (3) an unsatisfactory incentives structure (Prais, 1981, pp. 272 ff). In support of argument (1), it has been found that the number of strikes increases exponentially with plant size (Geroski and Jacquemin, 1985, p. 174). Because of a larger and more homogenised market, which required large production runs, labour productivity in the manufacturing industries in 1986 was some 50 per cent higher in the US than in Germany. This figure may overstate the difference in productivity between the two countries as it makes little allowance for the high quality of German manufactured products (Pratten, 1988, pp. 126–7). For example, a preoccupation with large quantities of output and economies of scale renders the taste of standard American chocolate, for a European, appalling. Intra-industry trade may be described in terms of monopolistic competition53 and product differentiation. Perfect competition is not a realistic market structure, so perfect monopolistic competition is the most perfect market structure in a situation with differentiated goods (Lancaster, 1980). Armington’s assumption states that products in the same industry, but from different countries, are imperfect substitutes (Armington, 1969, p.160). In other words, buyers’ preferences for different(iated) goods are independent. Armington’s assumption, however, overestimated the degree of market power of a particular producer. Instead of taking goods themselves as the basis for analysis, ‘address models’ of goods differentiation take characteristics that are embodied in goods as their starting point (Lipsey, 1987a). A computer is a good that can be considered as a collection of different attributes, such as memory, speed, printing, graphics and the like. Figure 4.3 illustrates two characteristics of a set of goods (computers).

Evolutionary Economic Geography

230

Each good (computer), A, B or C, has a certain combination of characteristic S (speed) and characteristic M (memory). Each good is defined by its location in the continuous space of characteristics, hence it has a certain ‘address’ in a two-dimensional space. Consumer preferences are defined by specific characteristics, not goods. Some consumers prefer memory to speed, whereas others prefer the opposite. Under the assumption that all three goods in Figure 4.3 have the same price, let a consumer have tastes embodied in

Figure 4.3 Characteristics of computers. the indifference curve II. This consumer maximises utility by purchasing good B. Each good in this model has close and distant neighbours. There are many goods and many consumers.54 Every consumer attempts to attain his or her highest indifference curve. This gives rise to intra-industry trade. Address models of localised (monopolistic) competition can be an important factor in explaining intra-industry trade and, hence, contribute to the explanation of the spatial location of production. A similar example may be found in the market for cars. Some consumers may prefer performance to comfort, while others may prefer appearance and comfort to performance. Such demand structure gives possibilities for a large variety of car manufacturers to coexist. This, however, does not work in the market for pharmaceuticals. When people are ill, they want a drug from the medical Ferrari, not someone that is perhaps the sixth runner-up. In fact, the sixth runner-up perhaps never appears on the medical market for a drug that cures a specific illness. All general explanations of trade in differentiated goods refer only to final goods traded among developed countries. In addition, intra-industry trade is affected by imperfect product markets (monopolisation) and by consumer demand for a variety of

Market structure and location of production

231

goods. Economies of scale can be another important factor in explaining trade in differentiated goods. Countries with a similar endowment and proportion of factors will still trade. Imperfect information about goods on the part of consumers might have had an impact on intraindustry trade in the past, but this aspect is rapidly diminishing as the Internet, global advertising and other methods of disseminating information worldwide gain importance. The Heckscher–Ohlin theory gives students the impression that the factor proportions theory of trade is orthodox. Linder’s theory tends to be less rigorous and thus has not made the same impression on students. Nonetheless, Leamer (1984) found evidence which supports the classical theory, while Greenaway and Torstensson (2000) found that both factor proportions and economic geography variables appear to be important in determining trade within industries (hence the spatial location of production). Linder’s approach does not reject the factor proportion theory of trade and location of production, but rather it asserts that factor proportion is not the only cause of location of production and trade. One may conclude that the factor proportions theory determines geographical location of production (specialisation) and trade among different SITC classification groups, while economies of scale and diversity in tastes determine the geography of production and trade within SITC classification groups. As most changes in demand take place within certain (statistical) clusters of goods, this is a sign that changes in technology are important driving forces of trade and (re)allocation of production in space. Mergers and acquisitions Let us start with the conceptual issues about mergers and acquisitions. In a merger two or more firms decide to pool their assets to form a new company. In this process one or more companies disappear completely. An acquisition does not constitute a merger if the acquired company does not disappear. Mergers are relatively less frequent than acquisitions. In a competitive market, mergers and acquisitions are thought to bring at least two efficiency gains: a reduction in management costs and a reduction in transaction costs. These benefits need to be weighed against the possible costs that accrue from the potential inefficiencies that may be the consequence of business concentration. If the expected efficiency gains are not realised and do not outweigh the disadvantages, the new merged enterprise may suffer as a result of differences in corporate cultures (in Germany engineers run firms, in Britain accountants, in Italy designers), inflexibility and poor coordination of business functions. A number of studies examining full legal mergers in various countries in the EU have found no evidence of substantial efficiency gains. Nor were economies of scale significant. Mergers had little or no effect on post-merger profitability. There was no significant difference in the returns per share three years after the merger.55 The costs of changes in business organisation were often greater than the benefits claimed by the promoters of takeovers. The main reasons for these disappointments include the high prices paid for target firms as managers often overestimate their own ability to run them or pursue personal reasons other than maximisation of shareholder value, overestimation of the business potential of the acquired firm and mismanagement of the integration

Evolutionary Economic Geography

232

process with the acquired firm.56 This is most obvious in the cases of mergers of firms in the production of steel or cars, as well as in airlines. McKinsey, a management consultant company, reviewed 160 mergers in the 1992–9 period and discovered that only twelve of the merged groups succeeded in lifting growth above the trends before the merger; the other 148 failed.57 Other studies, such as Dickerson et al. (1997, p.359) found no evidence that acquisition in Britain had a net beneficial effect on company performance if measured by profit criteria. In fact, this impact was detrimental and systematic. Perhaps company growth through internal investment, rather than a merger, may offer a superior profit growth rate. In a study of Fortune 500 takeovers (1981–95) in the US, Trimbath (2002) found that most of them (more than two-thirds) resulted in increased efficiency measured by cost per unit of revenue. The editorial article in the Financial Times ‘Merger mania’ (29 October 2003) stated: The prospect of investment bankers pocketing millions for ‘advising’ clients to do what turns out to be a very lousy deal is unappetising. Several academic studies published in the past couple of years have suggested that two-thirds of deals fail. The criteria include revenue growth, share price performance and meeting the targets set when transaction was announced. Daimler-Benz bought Chrysler in 1998. This was a merger of equals and thought to be ‘marriage made in heaven’. The Daimler-Benz boss was glad to ‘catch a star’ at the time when BMW and VW were fighting for ‘peanuts’. Nine years later: ‘Chrysler cost $35bn, but the private equity firm Cerberus is now snapping up four fifths of the Chrysler division for $7.4bn’.58 If one catches and later loses ‘peanuts’, the loss is relatively small. In addition, The Economist59 observed: M&A’s poor reputation stems not only from the string of spectacular failures in the 1990s, but also from studies that showed value destruction for acquiring shareholders in 80 per cent of deals. But more recent studies by economists have introduced a note of caution. Investors should look at the number of deals that succeed or fail (typically measured by the impact on the share price), rather than (as you might think) weighing them by size. Were past mergers and acquisitions judged too severely? Do all or at least some strategic business dilemmas such as mergers and acquisitions have solutions? Do good managers need to have talents and knowledge to preside over an agreeable and smooth transfer of business they manage to someone else? If the current organisation and company culture becomes dysfunctional, perhaps the first thing to do is to hire an outsider to manage this company’s business. With all this in mind, further research, analysis and evaluation are required in the area of mergers and acquisitions. In general, shareholders in acquired firms are likely to benefitfroma merger, but investors in the acquiring firms are likely to lose out. There are also problems in creating a new management culture in the merged company, whereas cost savings and new

Market structure and location of production

233

economies of scale can be negligible. This does not mean that all mergers and acquisitions have been failures, but it does mean that all ex ante claims about the splendid future of the merged company need to be taken with a pinch of salt. Business gurus, management consultants and investment bankers have reached the same conclusion: less than half of all mergers add value in the medium term.60 Thus, some conglomerates sell off parts of their business that are not at the core of their activity and competence in order to raise money for acquisitions in their main business area and to simplify their operations. Others expand and diversify their business activities and brands. Examples of ‘light’ diversification, the one that has a certain link with the core business competence and activity, include cases where chocolate bar companies enter the ice-cream market. McDonald’s is varying its normally uniform menus to meet national tastes. ‘Heavy’ diversification is exemplified by the examples of firms such as HarleyDavidson and Levi’s that entered the after-shave lotion market. EasyJet went from cheap air transport to businesses that include rent-a-car, Internet cafés, holidays, gifts and insurance. Virgin expanded from low-budget air transport to entertainment, clothing, drinks and books. Benetton, a clothing chain, took control of Autostrade, operator of highways in Italy, once they were privatised in 2000. Coca-Cola was founded as a company in 1892 and remained a one-product company for almost a century. Now it manages a portfolio of over 200 brands, most of them local.61 Smith & Wesson (the US producer of the most powerful handgun, the Magnum) started offering in its online catalogue62 items that include bicycles, grills, men’s cologne, toys, mugs and watches. Smith & Wesson targets – apologies, directs – a part of its business strategy towards middle-aged homeowners. Amazon.com, the global internet bookseller, started to offer home and garden items such as citrus juicers, bar stools, baking mats, toasters, BBQ tool sets and revolving shoe trees in 2007. Enron, an energy company, went into telecommunications and water, which just contributed to the failure of the company on a grand scale. In any case, diversification, mergers and acquisitions are not necessarily wrong business strategies, but they are very risky.63 Despite promises of reduction in costs, economies of scale and the creation of ‘champions’ to counter foreign rivals, mergers were used as defensive business policy instruments. The rationale that led to the large wave of mergers in the US (and Britain) during the 1960s and 1970s proved to be unfounded. Instead of supporting adjustment, mergers obstructed it by protecting firms from competitive pressure. This was reflected in the relatively slow response of some US firms to oil crises and to Japanese competition in certain industries during the 1980s, as well as competition from China from the start of the new millennium. Firms in the EU would be wise to avoid any repetition of the US experience. The presence in the EU of Japanese and American TNCs with their advanced technology and business organisation in some lines of manufacturing industry, as well as the ‘flight’ of certain domestic manufacturing firms towards China, may be the principal motivators to the EU domestic firms to restructure the existing business or move to something new in order to become more competitive. If your location is high-cost for the production of certain goods, then try to move upmarket, to make superior or unique products. Switzerland is a highcost location for the production of food. Nonetheless, farmers and cheesemakers found external top market segments for their Gruyere cheese. They sought and found new rich and choosy customers in Australia and, increasingly, in China. There are about 400 million middle-

Evolutionary Economic Geography

234

class consumers in the coastal provinces of China. Their income is rapidly increasing, hence they become quite choosy regarding the choice of high-quality goods. This is a chance that many European firms spotted and target as their business-location goal. A tidal wave of mergers and acquisitions in the EU was prompted by the need to restructure segmented industries and liberalise capital markets (exploiting restructuring and financial know-how from the US), as well as by the Single European Market. The volume of mergers and acquisitions relative to startup investments raises serious questions regarding competition (because of the reduction in the number of independent firms and an increase in the potential for collusion) in the Single European Market. Mergers and acquisitions are a global phenomenon. So is competition. Hence, anxieties about the effect of the location and structure of merged businesses on competition is to some extent offset by the expansion of ‘global’ over regional or national competition, as well as by the fact that there has been some restructuring of the business of the acquired firms (European Commission, 1998, pp. 144–5). One result of the numerous mergers and acquisitions is that the degree of concentration in the EU has risen compared with the pre-Single Market Programme era. This may increase price competition in the internal EU market and abroad through rationalisation of production and economies of scale. At the same time, an increased concentration (oligopolies) of business may restrict competition. Therefore, the EU introduced an important legal instrument for the ex ante control of mergers in 1990. Another important tool for the control of dominance in the EU market and for enhancing competition was the conclusion of the Uruguay Round in 1994. After ratification, this deal further liberalised international trade, increased competition and, hence, modified/limited the non-competitive behaviour of the concentrated EU industries.

4 Dominant market position States, in particular bigger ones, may dominate in the international arena by the use of three basic instruments. To put it harshly and directly, they may use: guns, money and ideas. And law, if tailored according to their wishes and needs. Firms behave somehow differently on the market: they use some of the same tools, but wrapped differently. If a firm has or achieves a dominant market position, it may significantly affect competition and the geography of production. The dominant market position may be secured by a firm in several ways, including the following five: • Firms may have innovative skills and competences not only in products (goods/services) but also in management, control, marketing and firmlevel planning. They may make risky investments in R&D, production and/or marketing that their competitors do not have the nerve for. Such first-mover advantages that are in line with the rules may result in dominance on the market and supernormal profits (Microsoft’s MS-DOS and Windows, Sony’s Playstations, as well as Nintendo’s Game Boy, are obvious examples). The life cycle of products is shortening all the time, hence the importance of innovating. In fact, most firms compete by continually assuming quasi-monopolistic positions that are based on innovation of various kinds.64

Market structure and location of production

235

In spite of continuous innovations by Microsoft, the most valuable user assets (personal files) are transferable from one Microsoft system to another. Apple, a competitor, failed to realise this in time, allowing Microsoft to achieve a nearmonopoly. The classical view that firms are only an input conversion mechanism does not reflect the contemporary world. In addition to their input conversion and valueadding functions, firms are also involved in learning-by-doing and innovation activities. Geographical concentration in industries with various entry barriers and increase in the size of firms (because of economies of scale and decreasing costs per unit of output) may occur as a direct consequence of a firm behaving efficiently. If consumers’ demand is homogeneous, then there will be a tendency towards the creation of a business and locational monopoly. • The dominant position in the market could be attained through mergers and acquisitions. This is quite typical in English-speaking countries. • A firm may achieve or protect its dominant position through anti-competitive business practices. Examples of this include exclusive dealerships and predatory pricing. However, exclusive dealerships do not always reduce welfare; for example, compared with the situation of free entry and exit among dealers/retailers, permanent and exclusive dealership (including after-sales service) might be a superior and welfareincreasing solution in the medium and longer run for certain complex goods and services. • The dominant position can be captured and maintained through a competitive and risky pricing policy. If economies of scale and learning are expected to result in a significant fall in prices as output increases over time, a risk-loving firm (for example, Texas Instruments in the 1960s and 1970s) may choose to set current prices on the basis of the expected (low) unit costs of production in the distant future. Alternatively, prices may be based on the average cost of production over the life cycle of the product. • Yet another way in which firms come to dominate a market is through the granting of a licence by public authorities. Examples can be found in ‘natural’ monopolies such as public utilities (water, gas, electricity, rail transport, postal services and the local fixed-line telephone service). In these industries, the minimum efficient scale may be so large that a single firm is necessary to serve the entire national market. From the outset one should be less concerned about the existence of a dominant firm (Microsoft) than about how the dominant firm ‘plays the game’. Many people forget that Microsoft competes with Microsoft (Vista competes with Windows XP, which competed with earlier Windows Millennium Edition, which competed with earlier Windows 98, which competed with earlier Windows 95). If the price of the new version is too high, consumers will stick to the old computer software. However, Microsoft may increase the price of the old versions when the new one is launched on the market in order to boost sales of the new version. 5 Competition: a recap Relatively small countries that are in the process of development employ industrial policies that may not always be competition-friendly. Even large and now developed countries used this type of policy at one time. Relatively large and developed countries

Evolutionary Economic Geography

236

claim that they now value the superb benefits of competition. This is what they claim when it suits them. What they continue to do now is another matter: just think of the US and EU trade policies in agricultural goods, cotton, textiles or subsidies to the aircraft industry. The objective is to keep the location of this kind of production at home. As concentration of business in certain industries increases, there is a trend to tighten antitrust policy and maintain a certain level of competition in the internal market. Increased competition would in most cases put a downward pressure on prices and costs. This would also stimulate innovation and growth. This outcome would enable economic growth with a reduced inflationary pressure. However, it is not clear how this would evolve in practice. Competition may reduce the prices of goods and services. It may increase output, but keep prices constant. In this case a reduction in prices would be offset by an increase in demand. The most likely outcome in practice would be that competition would produce a blend of benefits which accrue from increase in output of existing goods and services, creation of new ones and decrease in prices. The Single European Market Programme enhanced the dynamic process of competition through an easing of the market segmentation in the EU and, somewhat paradoxically, increasing concentration in some lines of business. This concentration permitted the employment of economies of scale and an increase in technical efficiency in production. In addition, it enhanced certain R&D through a joint sharing of high costs. A change in global competition and the relative loss of competitiveness by the EU – in comparison with the US in the computer software and aerospace industries, Japan in cars and consumer electronics, China in assembling products, and developing countries in textiles and clothing – as well as a potential loss in a number of other industries, were among the driving forces that brought about the Single European Market in 1993. Businesses reacted to this internal market-deepening by consolidations through mergers and acquisitions, as well as joint ventures. An increase in internal EU competition that came through the elimination of internal barriers gave the EU the chance of benefiting from the unexploited economies of scale that reduced, over time, costs of production and increased the global competitiveness, not only of manufacturing (aircraft or mobile telecom products), but also of services. All this provided an additional bonus to investment and the growth of the EU’s economy. Hence, not only innovating firms and those that use state-of-the-art technologies but consumers too are able to reap rewards from the opportunities provided by increased competition (telecom business is an example). The dynamic, spatially and professionally mobile segment of labour (at least) might gain from competition, openness and integration processes, as trade and competition do not determine whether there are jobs at home, but rather what kind of jobs are available. National macroeconomic policy, the business cycle, education and changes in technology are partly invisible and not easily measurable policy elements to the general public. However, these may be much more important factors for job creation and maintenance than is trade. Unfortunately for the policy-makers and the general public, trade is only one, highly visible, policy aspect that affects national jobs.

Market structure and location of production

237

6 Specialisation and returns to scale Meaning ‘Returns to scale’ refers to the relation between input requirements and output response with its impact on costs. Economies of scale comprise a number of things, from simple technical scale to phenomena such as processing complex information; direction, control and improvement of independent activities; and experience. If a firm’s output increases in the same proportion as its inputs, then that firm’s technology exhibits constant returns to scale, or, one may say, a firm has constant marginal costs. If a firm’s output increases by a greater proportion than inputs, then this firm’s technology has increasing returns to scale or it enjoys decreasing marginal costs (or increasing marginal product).65 If a firm’s output decreases by a smaller proportion than input requirements, then the firm suffers from decreasing returns to scale or increasing marginal costs. Suppose a firm uses a set of inputs X in the production of a good Y. Constant, increasing and decreasing returns to scale may be defined for homogeneous production functions (Equation 4.3). A function is homogeneous of degree k if: f(tx1; tx2)=tkf(x1; x2) (Equation 4.3) where t > 0 and k is constant. If the set of inputs X is increased by t, output is increased by tk : k = 1 implying constant returns to scale; k > 1 increasing returns to scale; and 0 < k < 1 results in decreasing returns to scale. These relations can also be presented graphically. Figure 4.4 depicts technologies with various returns to scale. A firm’s technology has constant returns to scale in case (a) where isoquants for output of quantities 1, 2 and 3, respectively, intersect the path from the origin at equal distances, making 0A = AB = BC. In the increasing returns to scale (case b), isoquants are closer together, 0A > AB > BC. The third case (c) describes decreasing returns to scale. Isoquants are further away from each other, 0A < AB < BC. Figure 4.5 illustrates the cost behaviour of different technologies. With constant returns to scale, marginal cost per unit of output Y is unchanged. With increasing returns to scale, as output increases, costs per unit of output are decreasing; while with decreasing returns to scale as output increases, costs per unit of output are also increasing. Decreasing returns to scale describe a situation in which there are some inputs (such as land) in fixed supply. With increasing returns to scale in industry Y, the ratio at which the two goods exchange in the market differs from the ratio in which they can be converted into one another through production. When factors are shifted to the increasing returns to scale industry, the gain in production in that industry is greater than the loss of output in the other, so there is a net increase in output. This increases real wages, average standards of living and GDP. The pure neoclassical equilibrium theory of international trade is concerned mostly with perfect competition. In a situation in which the minimum efficient scale is relatively large, only a few firms can survive simultaneously. Competition in such markets is not

Evolutionary Economic Geography

238

perfect. When internal markets are enlarged by international economic integration or opened up by marketdeepening projects such as the EU’s Single Market Programme which increase competition, firms might expand their production and specialise in order to achieve lower production costs per unit of output (economies of scale). The existence of internal NTBs was in many cases the cause of unexploited economies of scale in the EU. An empirical study of potential economies of scale in EU industry reported that in more than half of all manufacturing industries twenty firms of efficient size could co-exist in the EU, whereas the largest national markets can support only four each. The EU internal market offers the potential for efficiency and competition. Twenty efficient firms are more likely to ensure effective competition than only four firms (Emerson et al., 1988, p.18). This finding ignored the logic behind the role of

Figure 4.4 Technologies with constant, increasing and decreasing returns to scale.

Figure 4.5 Cost curves with constant, increasing and decreasing returns to scale. concentration in modern industries and the contribution which a few oligopolies, exploiting economies of scale and carefully monitored by appropriate regulatory authorities, can make to economic welfare. Besides, there may be fierce competition even among these four firms. Just take a look at competition in the EU and US long-distance

Market structure and location of production

239

telephone-call business or the fierce competition among a few Japanese electronic conglomerates in both the domestic and international markets. In addition, relatively high diversity of tastes in EU countries can make the achievement of output in very large production runs unnecessary and unprofitable, contrary to the US where the internal market demand is very homogenised. The approach to economies of scale does not have to be too linear or mechanical. Successful managers of firms must know where to find economies of scale so that a firm may lower its minimum efficient scale of production by buying certain components from outside. Take, for example, car production and assume that each car consists of three equally important components (X, Y and Z, respectively) plus the assembly. In addition, the annual production of the following quantities of components is necessary to achieve the minimum efficient scale: X = 800,000; Y = 500,000; Z = 250,000; and assembly 100,000. If all inputs are produced internally, then the firm must produce 800,000 cars in order to reach the minimum efficient scale in production in X. However, if the firm buys X and Y from other firms that are specialised in the production of these components and that are sold to various car manufacturers, then it only needs to produce 250,000 cars to achieve the minimum efficient scale in the production of Z. Technological change has reduced the minimum efficient scale of production, but the firms tend to increase the value of just-in-time operation and minimisation in stocks. Although the example is rather simplistic, in reality Volvo could be an example of a car producer that fits such a category. Even though the production run of Volvo is significantly smaller than, for example, Toyota, GM or Volkswagen, Volvo has often been a highly profitable car manufacturer. The US economic system is created in such a way that it is theoretically open to internal competition. Antitrust legislation is very strong. Nonetheless, the President’s Commission on Industrial Competitiveness (1985, vol. 2, pp. 192–3) recognised the potential efficiency gains which may come from the concentration of business in certain industries. Without the prospective reward of temporary monopoly, firms may not have big incentives to innovate. A nation that sacrifices a particular high-technology industry will gradually sacrifice many of the local technological capabilities nurtured by it. And the ultimate loss is likely to be even greater, since without these capabilities the nation will find it difficult to exploit the footloose technological knowledge created abroad. (Tyson, 1992, p.42) A temporary departure from free competition may be desirable, in the sense that it is better to allow the establishment of temporary monopolies as a way of inducing innovation than to seek static efficiency at the cost of technological progress (Krugman, 1990a, p.173). If the American firms fail to withstand international competition over a wide range of industries, then the US approach to competition in a situation with imperfect markets, rapidly changing technology and expanding exchange of information (both by volume and by speed) could consign the American example to the history books. In fact, the US system is subject to strong protectionist lobbies such steel, agriculture or cotton. The

Evolutionary Economic Geography

240

domestic steel lobby was successful, for instance during 2002–3 when the president introduced ‘safeguard measures’ in the form of 30 per cent tariffs on ten steel product groups, while the US cotton producers were active and quite successful regarding the protection of their business during the WTO Doha Round of negotiations on trade liberalisation. Changes in technology and demand exert a continuous pressure on plant efficiency. Hence, the minimum efficient plant size changes over time. Largescale production is attractive and profitable only if there is secure access to a wide and growing market. It is reasonable, therefore, to assume that firms which operate on a wide market are more likely to be closer to the minimum efficient plant size than those which act within a more restricted framework. However, recent developments in technology are diminishing the classic scale economies associated with mass production in large plants. In response to an increase in income and to changes in demand, modern technology increases the role of smaller, but highly specialised or flexible, plants. Economies of scale are largest in transport equipment, electronics, office machines, chemicals and other manufactured products. These are the industries in which demand has the highest growth and where technology changes rapidly. The common element in these industries is the vast investment required to produce even a small amount of output.66 Advanced technology is not necessarily based on increasing returns to scale, but such returns are frequently found in high-technology industries. In addition, these industries are under continuous pressure from international competition. Industries with relatively smaller returns to scale are those with a stagnant demand and relatively low technology content. They include the food, textile, clothing and footwear industries. The problem with the EU’s manufacturing sector is that a significant part of its export specialisation has for a long time been in industries with relatively stagnant or declining demand and relatively weak economies of scale, while the US and Japanese producers and exporters have specialised in goods produced by industries with strong economies of scale. This was especially true during the 1980s. Adam Smith pointed out that specialisation is limited by the extent of the market. A customs union increases the market area for firms in the participating countries, hence it opens up opportunities for specialisation. A country may gain from economies of scale which form an independent source for trade. If countries trade, they may gain not only from the exchange of goods and services but also from specialisation and a wider choice of goods. Trade increases the bundle of available goods and services in relation to what is available in autarky. Specialisation and alteration in the output mix in a country may take full advantage of its factor endowment. This holds not only for economic integration but also for strategies to open up and deepen the internal market, as was the case with the Single European Market Programme. It was hoped that the long-term impact of the Programme, together with additional instruments, would rectify the EU’s comparative disadvantage in comparison with the US, Japan and a few other countries in certain industries. Later on came the Lisbon (2000) declaration that intended to make the EU ‘the most competitive and dynamic knowledge-based economy in the world capable of sustainable growth with more and better jobs and greater social cohesion’ by 2010.67 This political declaration and wish turned out, in the course of time, to be just a hope that did not materialise as expected.

Market structure and location of production

241

Economies of scale may be internal to individual firms. They may also be external to these firms when the whole industry grows and when all the firms in that industry enjoy the fruits of such growth. For example, networking (phones or cyberspace exchange of data or electronic banking) is the consequence of economies of scale. The peculiarity of economies of scale is that they are not consistent with perfect competition. With imperfect market structure all welfare outcomes are possible (but not necessarily desirable). A perfectly competitive firm takes the price for its output as given. A firm with increasing returns to scale in such a market will find it profitable to produce one more unit of output, since it can do so at less than the prevailing price. Thus, it would tend to increase output until it dominates the market, while efficiency may set the price for its output. Above a certain level of output some costs, such as depreciation of equipment, can increase disproportionally in relation to output. For example, the ends of the blades on a very large turbine move at a speed that can be close to that of sound. These blades need a special cooling system, while the whole ‘architecture’ needs complex set of seals to prevent unwanted high-pressure gas leakage. At such high-speed movement, ‘metal fatigue’ increases disproportionally in relation to a turbine’s capacity. Beyond a certain level of rotation speed, the costs in all engines based on revolution increase more than the returns. Small and light aeroplanes can endure a ‘hard landing’ that would wreck a jumbo jet. Physical relations limit speed, size and weight. Similarly, it is possible to make quite small keyboards or mobile phones. However, such small keyboards are too crowded for fingers for fast typing. In addition, a strike in a firm with increasing returns to scale technology may have a profound impact on the profitability of the firm. However, the threat of competition or of removing barriers to imports may mitigate the distortion imposed by trade unions. The neoclassical equilibrium theory of international trade argues that countries should specialise in the production of those goods for which they have a comparative advantage. Modern theories are partly questioning this line of reasoning. International specialisation and trade are due to other factors as well. Economies of scale stimulate specialisation in production for a narrow market niche, but on a wide international market. This may entail only the reallocation of resources within the same industry or sometimes within a single company. Modern footloose industries are not linked to any particular region by inputs such as iron ore or power. So, a country’s comparative advantage can be created by deliberate actions of firms, banks and/or governments. Any consideration of returns to scale would be incomplete without mentioning services. The impact of returns to scale on costs in the manufacturing sector is direct and obvious indeed. This impact is less obvious and capable of significantly reducing costs in the services sector, although experience and competition can increase productivity in this sector too. Banking, insurance, advertising, transport and shipping are the industries that are most affected by international competition. However, there are serious methodological limitations in the quantification of returns to scale in the services sector (for example, how does one measure the output of an insurance company?). Returns to scale have not been thoroughly studied in economic theory for a very long time (until the 1980s) even though economists have been aware of increasing returns to scale since the times of the Pin Factory or railways or the telephone. Increasing returns have always been like underground water, copper, crude oil or gas deposits. People are

Evolutionary Economic Geography

242

aware of their existence, but they rarely see them. The reason for this neglect of increasing returns to scale in economic theory is that it is difficult to model them with rigour and clarity. Increasing returns to scale are also only a temporary phenomenon. There are limits to the scale effects of each new technology. The difficulty of modelling scale economies is compounded with the following observation: It took thirty to forty years for new electrically powered factories to replace the old steam-powered factories after the former provided superiority over the later. Once the transition is complete, however, the scale economies are exhausted and further expansion in output occurs at constant or diminishing returns. (Lipsey et al, 2005, p. 450) Increasing returns create strong tendencies towards monopolies in each industry. Large firms can produce standardised products in large batches and at lower costs than small firms can. Hence the big players put small ones out of business. For the invisible hand of the market to work and to ‘behave properly’ (to ‘get it right’) one must assume that there are many competitors in each business and that returns to scale are diminishing, not increasing. This has been the principal workhorse in economic theory for two centuries. Therefore, one should be very careful about using neoclassical equilibrium theory either as a description of what is likely to happen in, for instance, a customs union regarding the spatial location of production, or as a guide to policies that will ensure that such a union fulfils its expectations. For example: These studies do not support much of an effect of scale economies, largely because UK manufacturing during the period since EU entry has contracted sharply, especially in those areas of manufacturing where large-scale operation is prevalent; essentially UK manufacturing has become concentrated in areas known as ‘niche manufacturing’ for the most part. (Minford et al., 2005, p. 56) Another important fact is that a substantial part of production is linked with economies of scope rather than scale (lower costs per unit from making larger output). While economies of scale imply a certain level of standardisation in tastes and production, economies of scope deal with product or process diversity (cost and revenue advantages from operating in related markets). Economies of scope allow firms such as Benetton to respond swiftly to changes in the supply of inputs and shifts in demand because they come from the common control of distinct, but interrelated, production activities. For example, the same kind of fabric can be used in the production of various goods. Economies of scale are coupled with market imperfections which allow various welfare outcomes. Distortions, such as deviations from marginal cost pricing or the existence of barriers to entry and exit, mean that the creation of a customs union (or, for that matter, any other type of international economic integration) does not necessarily either improve or worsen welfare. With economies of scale and other distortions, the welfare effects of customs union creation and pattern of trade are much more affected by

Market structure and location of production

243

government policy and chance events than they are in the neoclassical theorists’ world of pure, clean and lean comparative advantage, equilibrium and factor proportions. Barriers to enter and exit an industry In perfectly contestable markets, price equals marginal cost (firms just cover production costs of the last unit and profits are zero). Economies of scale introduce imperfection into the market system, since firms set prices at average cost and make profits. Therefore, economies of scale lead to more specialisation compared with the situation with constant returns to scale. Barriers to entry and exit protect the existence of these firms with economies of scale. These barriers increase risk, costs, uncertainty and delays to a newcomer relative to the incumbent. The definition of entry barriers is important for court antitrust cases. However, this definition has a long, rich and confused history.68 Inability by lawyers and economists to reach a broad consensus about the definition of entry barriers introduced debates about the desirability of certain mergers and acquisitions. If a merger or an acquisition creates barriers to entry, it may not be socially desirable. For our purposes here, we shall take an entry barrier to mean an advantage that the incumbent firm enjoys over potential new entrants. This advantage permits the incumbent firm to increase prices of output above the equilibrium level at least for a certain period of time without attracting the entry of new firms. Entry barriers create a rent from previous investment that provided the firm with incumbency. Social welfare is reduced because of no entry or because of delayed new entry of firms into this market. Barriers to entry in an industry include: • high human and capital requirements (sunk costs necessary to achieve minimum efficient scale of output); • product differentiation, consumer preferences and brand loyalties; • indivisible production technology (a strong and close geographical presence of suppliers and consumers is necessary); • externalities (linkages); • R&D; • regulation (environment, product quality, permissions, accounting, standards, intellectual property rights); • marketing (access to distribution network, exclusive dealership); • advertising needs and costs; • reaction by market incumbents (predatory pricing and excess capacity in the existing firms to discourage any new entry); • trade and industrial policies (tariffs, quotas, domestic content requirements, NTBs); • organisational complexity; and • availability of staff. Barriers to exit include: • past sunk costs (not yet depreciated); • durable specialised assets that cannot be sold easily; • regulation (bankruptcy law; labour settlement issues; cost of dismantling facilities of nuclear power plants, for example);

Evolutionary Economic Geography

244

• public bailout (otherwise medium-sized towns may disappear); • new trade protection; • strategic business decision by a diversified firm to save the facility for the future (just in case that it may become necessary); and • emotional attachment to the facility by the managers and owners. Predatory pricing Predatory behaviour is a medium- or long-term business strategy that aims to eliminate an existing or potential future competitor from a market. A firm that employs this strategy wants to preserve its (monopoly or oligopoly) market position by three general means: • below-cost pricing of output (charging unprofitably low prices at the present time); • restriction of access to assets (refusal to supply inputs, exclusive dealings, excessive charges to enter marketing network, fidelity rebates that reward clients not to purchase from competitors or new entrants); and • building a much larger output capacity than is current and expected demand. If there is a potential entrant, the predator may slightly increase output and lower prices. This would discourage certain potential entrants from entering the market and starting production. The predatory strategy may be backed either by the home government’s support or by charging the domestic consumers high prices in a protected market. In addition, a riskloving firm may price its current output on the bases of the expected future lower costs of production and expanded demand, rather than on the current costs that may be higher (and low current demand). This was the practice of Texas Instruments during the 1960s and early 1970s in order to build volume of production, gain market share and move down the learning curve. It was Texas Instruments and not the Japanese producers of semiconductors that pioneered this strategy of ‘pre-emptive price cutting’ (Tyson, 1992, p.89). Sometimes it may be hard to distinguish predatory pricing from vigorous price competition. Predatory pricing is seldom a winning strategy in the long term. When a predator raises prices again, this tempts new competitors to consider market entry. The predator’s losses may exceed the prey’s since the prey can shut down temporarily and ‘mothball’ output potential while the predator must make substantial sales to keep prices low. If the prey goes bankrupt, the predator may need to acquire its assets in order to prevent a new rival from buying them up. It would often be less costly to merge with the prey at the outset than to drive it out of business by a predatory strategy (Schmalensee, 1988, p. 665). In addition, ‘price wars’ may have negative consequences for productivity growth in the longer term. This is because firms will have to reallocate resources from investment to cover the expenses of a ‘price war’. AKZO, a Dutch-based chemical TNC, charged different prices to buyers of comparable size. These differences in prices were not based on differences in the quality of products or changes in production costs. Therefore, the EU Court (Case C- AKZO v. Commission, 1991) concluded that this type of pricing behaviour by AKZO was not to pursue a general policy of favourable prices, but rather to adopt a strategy that could

Market structure and location of production

245

damage ECS, a small British competitor. AKZO was subsequently fined for this type of inappropriate market behaviour. In 2003, for example, the European Commission imposed a fine of Є10 million on Wanadoo, a French telecom company,69 for abuse of a dominant position in the form of predatory pricing in ADSL-based Internet access services for the general public. Wanadoo was found to have charged prices below cost in the period from March 2001 to October 2002. This practice restricted market entry and development potential for competitors. This damaged the interests of consumers on a market which is key to the development of the information society. Retaliation against predatory pricing can be achieved by means of antidumping duties. However, the existence of this kind of dumping is not that easily proved in practice. Even the GATT/WTO rules do not take into account volatile exchange rates, continuous shortening in the life cycle of high-technology goods and the complexity of modern marketing arrangements. Investments in market penetration and some of the investments in plant and R&D are not recoverable. They also often have little market value. Large sunk costs make firms reluctant to enter an industry if long-term access to wide markets is not secure. Expectations about future demand and profit play a key role in the decision-making process. This restricted or blocked entry results in a smaller number of firms in the industry than would be the case under free entry. Sunk costs provide advantages to the first entrants. Substantial rents may be earned from being at the forefront of technological change, where standards are set for future products and processes, and where, once standards are established and industries have matured, new entry becomes difficult (Tyson, 1987, p.74). A defensive strategy to counteract such developments would be to respond to the competitor’s action in order to retain market share, while an offensive strategy would be the development of improved or new goods or services in order to enter new market segments and/or to weaken the competitor. Market pre-emption and moving early may create a market position and profits that may be sustained for a very long time. The Swiss firms in pharmaceuticals and watches, the German chemical companies, Procter & Gamble in detergents since the 1930s and Microsoft in computer software since the 1990s are obvious examples. They were among the first to move, they learned ‘how to do it’, exploited economies of scale, influenced standards, formed marketing networks, created brand names and established consumer loyalty. If imitation is slow, hard, costly and risky, such developments place potential competitors at a disadvantage. The longer the time lag for imitation, the greater the possibility for the firm to reap rents as the reward for risk and the right choice of investment. This increases the possibilities for rents obtained by first movers, but also opens up a great risk of failure and of falling into a trap of rigid specialisation that prevents them from adjusting to changes or moving further or elsewhere in production. New technologies may not remain clandestine for ever. If a monopolist makes a profit, other firms will try to enter the industry. Imitators could appear so that the firm’s advantage and rents would be eroded. This is why IBM, with an ‘open’ PC system, had a greater longer-term worldwide success than Macintosh with its ‘closed’ system. Similarly, Visa or JVC (owner of VHS), who encouraged others to use their ‘technology’, had a greater market success than American Express or Sony (owner of Betamax), who wanted to control the standard and technology only for themselves.

Evolutionary Economic Geography

246

A mature technology means that factories for standardised goods are often similar in size, hence countries may differ in the number of plants because of variations in the size of the market. This is a lesson for small and medium-sized countries, as well as for the developing ones. They should pool their markets and resources in order to overcome these barriers to entry and to reap the benefits of economies of scale. Standards From the vantage point of the purely technical side of production, economies of scale might require some level of standardisation as a kind of industrial Esperanto. While this could assist in producing global products, it may also easily limit product variety and restrict consumer choice. Standardisation started with the First Industrial Revolution in western Europe (roughly between 1750 and 1850). Standards are technical regulations which specify the characteristics of goods or their production or use. They can be quite different among countries. To harmonise them in a customs union or in a wider international area for trade and competition purposes maybedifficult. Mutual recognition of standards can be an attractive strategy in the short term. This places traders in different countries on an equal footing and eliminates locational disadvantages which accrue to some of them from NTBs. As already observed, trade in general, and intra-industry trade in particular, could increase product variety in relation to autarky. Product standardisation could reduce export opportunities for small open economies such as Austria, Belgium or Switzerland. These countries have little influence on foreign tastes and tend to enjoy comparative advantages in semi-manufactured goods (Gleiser et al., 1980, p. 521). This may be in line with increasing returns to scale for standardised products, which ultimately leaves the business landscape with a domination of just a few players. However, with an increase in income, consumers want and are ready to pay for differentiated products. In this situation highly specialised firms and production in relatively short runs may have full market justification. In fact these three small countries are among the richest in the world measured by GDP per capita. The US and Japan, for example, have a market for durable consumer goods sometimes dominated by a single brand which takes advantage of economies of scale. The situation in EU countries is different. Almost every country has its domestic producer. But the American or Japanese formula cannot easily be replicated in Europe. This is because of diverse and deeply rooted national preferences. While the British want to load their washing machines from the front, the French prefer to do so from the top. The Dutch like high-powered machines which can spin most of the moisture out of the washing; the Italians prefer slower-spinning machines, leaving the southern sun to do the drying. This situation has its impact on the protectionist pressure against third-country suppliers, but also against suppliers from within the EU as they are prevented from following a panEuropean production strategy. Yet there are other (relatively new) goods where preferences among countries are identical. It is likely that French cheese producers will look for the same qualities in a photocopier, a fax machine, a PC or a global positioning device as would Italian wine exporters. Common accounting and company or banking laws, therefore, could be quite useful in the EU or in the ‘global economy’.

Market structure and location of production

247

If the EU approximation of standards takes place upwardly, the overall level of regulation and costs of production will increase, and southern and eastern European countries will find it increasingly difficult to withstand international competition and will grow at a slower pace than the rest of the EU. As growth falls, demand for protectionism increases. If, however, various types of, for instance, labour regulation are left to member states and the framework of mutual recognition is kept, then the northern and western EU countries could get rid of certain laws and southern and eastern European countries would grow rapidly. The EU would expand at a substantial rate which would permit the implementation of a liberal trade policy. The ultimate irony is that only the latter course would allow northern and western EU countries to pay high wages and sustain a heavy social expenditure (Curzon Price, 1991, p.124). Although the Single European Market removed many NTBs to internal trade in the EU, there are others that still persist, making the single internal market imperfect. Prior to the period when the single-market initiative was applied (1985–92), major obstacles to internal trade included physical border controls, technical barriers (standards and product/service regulations), public procurement, different intellectual and industrial property laws, state aid and fiscal barriers, as well as obstacles to the mobility of labour and capital. The Single European Market made a distinction between what was essential to harmonise and what could be left to mutual recognition. While EU standards are still being developed, the guiding principle should be mutual recognition of national standards. EU standards are being developed on a large scale, approximately a thousand a year. However, too many national regulations are still being produced, making it as hard as ever to achieve and maintain a truly homogeneous single market (Curzon Price, 1996a). It seems that regulation keeps on reinventing itself. In a similar vein, awaveof‘economic patriotism’ with increased national protection was raging through the EU in 2006. France, Spain, Poland and Italy, for example, impeded takeovers of domestic firms in the energy and banking business by firms from EU partner countries.70 It seems that it still matters a lot who owns certain businesses and where, no matter how close economically and politically the countries are and how deeply they are linked through law. Various attempts to harmonise diverse goods at the EU level created a public furore in many countries. Examples include fuss about prawn cocktail crisps in Britain; the permissible level of bacteria in cheese in France; undersize apples in Denmark; the use of other than durum (hard) wheat for pasta in Italy; the application of other than the Reinheitsgebot (of 1516) purity rule for brewing beer in Germany (beer may have only four ingredients: malted barley, hops, yeast and water – other countries [locations] allow various additives, including preservatives); or the removal of the tilde (˜)from computer keyboards in Spain. Each year every EU country produces thousands of new regulations on new technologies and products. Every one that relates to the smooth operation of the internal market must be submitted to the European Commission. As each one carries a possible seed of conflict, it must be treated with caution. Unless it is subject to regular monitoring, the genuine Single European Market could ultimately disappear. Thus, Decision 3052/95 (adopted in 1995 and implemented in 1997) provides an improved procedure to deal with the remaining obstacles to the free movement of goods in the single market. The decision obliges member states to notify the Commission of individual measures preventing the

Evolutionary Economic Geography

248

free movement of a model, type or category of a product that has been made or sold legally in another country. The purpose of the measure is to encourage member states to think twice before making any exceptions to the EU system of mutual recognition.71 7 Rules of origin Both free trade areas or, better, preferential trade areas and customs unions yield similar results for a participating economy and the integrated group and differ only in detail regarding trade and location of production. Free trade areas tend to result in more trade creation and less trade diversion than a tariff-averaging customs union. This detail has often been forgotten in discussions about international economic integration (Curzon Price, 1997, p.182). A tariff-averaging customs union increases the level of protection of those countries that previously did not have tariffs or whose tariffs were below the level of the common external tariff, as was the case when relatively liberal-trading EFTA countries joined the EU. As such, free trade areas place a much lower cost on third countries than tariff-averaging customs unions. If one assumes that the level of the common external tariff is equal to the lowest tariff of a member country in a customs union, then there is no theoretical difference between a free trade area and a customs union. The effects of both types of integration arrangements are identical. Of course, countries in a free trade area still take part in international negotiations about trade and tariffs on their own behalf, whereas countries in a customs union negotiate as a single unit. As there is no common commercial policy relative to third countries in a free trade area, countries (such as Canada, Iceland, Norway or Switzerland) that value their sovereignty highly have the formal chance to go it alone (as much as possible) in trade matters. These countries always respect the concerns of their principal trading partner. Free trade areas in practice usually include manufactured goods but exclude agricultural ones. This is because many countries want to preserve independence in national agricultural policies. The commercial wisdom (depending on values and priorities) of this policy choice is quite another matter. Still, in these matters, commercial wisdom and benefits play second fiddle to politics, national heritage (Catholic countries in the EU want to preserve their agriculture) and deeply rooted national culture. Rules of origin (preference) are the basis of a free trade agreement, as well as other preferential trade deals with the exception of a customs union. These rules prevent geographical trade deflection. This effect of a free trade area refers to the import of goods from third countries into the area by country A (which has a relatively lower tariff than the partner country B) in order to re-export the goods in question to country B. These speculations depend not only on the difference in the level of tariffs, but also on overall trade costs, as well as on the quality (perishability) of goods. Without rules of origin in a free trade area and with no trade costs, only the lowest tariffs would be effective. In fact, a free trade area would be equal to a customs union in which the lowest tariffs of a member country would apply. Trade deflection problems do not exist in customs unions because of the existence of the common external tariff. The proliferation of discriminatory, that is, geographically preferential, trading agreements increased the importance of the rules of origin of produced goods in trade.

Market structure and location of production

249

These rules are also becoming more complex over time because of the international (global) fragmentation in the production process. For instance, the local value added (in percentages) in heavily traded goods, as is the case with the fragmented and spatially dislocated electronics industry in Asia, is so small that very few of those goods would ever satisfy the minimum requirement for the preferential status in trade. Another reason for the diversity in the rules of origin in various preferential trade agreements stems from the fact that these rules are easily customised according to the preferences of the involved parties and various lobbies. Foreign producers and exporters seek to avoid the payment of customs duties, while the protection-seeking domestic competitors endeavour to prevent them from evading those requirements. Rules of origin are required especially in free trade areas in order to determine which goods are entitled to enjoy those trade preferences. Apart from tariff discrimination, other purposes of rules of origin include the determination of eligibility for quotas and informing consumers with a preference for goods of a certain geographical origin. Sometimes these rules of origin may be used to accommodate and facilitate trade and economic integration (partner countries want to expand the market and create an attractive location for investment), while at other times these rules may be employed to conceal protection. These protectionist intentions are strong in integration groups that were created for political purposes or in the ones that have strong industrial lobbies. There are four different methods for the determination of origin of goods: • Substantial transformation Geographical origin is determined by the country in which the good underwent the last substantial transformation (the one that gave the good a new name, a new character and a new use). Critics say that this is an imprecise and subjective method. • Change in the tariff heading Even if the good made/assembled in country A has imported components, that good can be regarded by foreign countries as a good that originated in country A if that was sufficient to change the tariff classification heading of the imported materials. Opponents of this system argue that the flaw is that the existing tariff schedules were not designed to determine the origin of the goods and that the system may be manipulated and abused by strong special interest industrial lobbies. • Value-added method A certain minimum added value (de minimis rule) must be incorporated in the good in country A in order to enable foreigners to regard the good as originating in country A. The trouble is that the method depends on controversial accounting systems, and even a slight change in the exchange rate may produce a different result regarding the origin. The minimum area content requirement can also be criticised on the grounds that it shifts the production factor mix away from the optimum, reduces rationalisation in production and can reinforce market rigidities. In any case, this method allows a certain tolerance, i.e. a specified percentage of nonoriginating inputs in the production, without affecting the origin position of the final product. • Specified technological processes A good must pass through a certain technological transformation in country A in order to be regarded by foreign partner countries as having been produced in country A. The problems here are that technology changes rapidly and that it is impossible to draft and keep updated internationally agreed

Evolutionary Economic Geography

250

classification systems and records on processes for all goods that enter into international trade. To make things more complex, there are also various exceptions to these rules and allowances. The trade-diverting effects of the rules of origin and the consequent redressing in the location of production could be mitigated by the elimination of quotas and by multilateral reductions in tariffs and NTBs. The Kyoto Customs Convention (1973) states that, unless a good is wholly produced in a country, what determines the geographical origin of a good is the country where the ‘last’ substantial process took place. Packaging and dilution with water do not change the essential features of a good, so they are not taken as important elements that change the origin of goods. Rules of origin can be restrictive or liberal. If the required value added within the area is, for example, 90 per cent, then very few goods would qualify for duty-free treatment. Liberal rules of origin require that only a minor part of the value of goods should be added within the area. Commonly, rules of origin require that 50 per cent of value added should be within the area in order that goods receive a tariff-free status (for example, EFTA or the Canada–US Free Trade Agreement). In the case of NAFTA, Mexico and Canada were in favour of relatively liberal rules of origin because of the positive impact of such rules on the location of Japanese FDI in the two countries and the potential exports of goods (cars) to the US. Initially the required local content for cars’ preferential treatment in the Canada–US Free Trade Agreement was 50 per cent, but in NAFTA it gradually increased to reach 62.5 per cent in 2002 and beyond. Quite an increase. Textile products have been a target for overt and hidden protection for a long time. The disappearance of quotas on imports on these items in the US and the EU in 2005, as a part of the Uruguay Round deal, instigated a new kind of arcane protection that is hidden behind technicalities, some of which may remain unnoticed. In NAFTA, for instance, to obtain preferential treatment in trade, a textile product must pass the triple transformation test: the fibre must be created, woven into cloth and made into clothing within NAFTA (Destler, 2005, p. 249). Chile was dismayed regarding the impact on its production and exports when the Canada–US Free Trade Agreement included Mexico in NAFTA. The case in question deals with processed tomatoes (Palmeter, 1993a, p. 51). Tomato paste (pasta de tomate) is found in Chapter 20 of the Harmonised Commodity Description and Coding System, while tomato catsup (salsa de tomate)is classified in Chapter 21 in the same System. Under NAFTA, tomato paste (Chapter 20) imported from outside NAFTA and converted inside NAFTA into tomato catsup (Chapter 21) would not grant origin of the catsup for preferential purposes. In NAFTA, therefore, the tomato paste has to be produced within NAFTA confines to receive preferential tariff treatment. Chile was the leading foreign supplier of the US imports of tomato paste in 1992. In NAFTA, created in 1994, the share of the Chilean supplies of this product fell sharply, while those of Mexico increased. This was not a benign outcome of economic integration on trade and efficient geographical location of production. Protectionist pressures increase through rules of origin. If Japan has, for example, a free trade agreement with Viet Nam and not with Cambodia, Japan may set rules of origin in its deal with Viet Nam in such a way to exclude goods that originate in part in Cambodia from the preferential treatment in trade. This may be done for various reasons, such as protection of a certain industry in Japan or

Market structure and location of production

251

because Japan wants to punish Cambodia for some (political) mischief. Conversely, a transparent, flexible (including the possibility of renegotiation) and administratively simple system of rules of origin may eliminate cost-increasing distortions and be friendly with trade and efficient spatial location of production. An important issue is the basis for the application of, for example, a 50 per cent valueadded rule. The choice is between the application of this rule to direct manufacturing costs or to invoice values that include overheads. Consider the following example. The direct cost of manufacturing good X in country A is $75, while overheads and profits are $25, making a total invoice value of $100. For good X to be exported without tariffs to the free trade area partner country B, the 50 per cent direct cost rule allows $37.50-worth of imported components, while the 50 per cent of invoice value rule allows $50-worth. The former rule offers a higher protection against the use of out-of-area inputs than does the latter. The Canada–US Free Trade Agreement uses the direct cost rule (as does the free trade deal between Australia and New Zealand). However, EFTA uses the invoice value rule (Lipsey and York, 1988, pp. 31–3). Rules of origin can be criticised on the grounds that: • they are open to much abuse (for example, a simple change of packing), • they can be avoided by unscrupulous traders (for example, fake origin statements and marks on the goods), or • the costs of monitoring the system are too high. The experience of the British Imperial preference scheme offered evidence that the operation of the system may be smooth. But in this case the parties were geographically separated, whereas it would be difficult to prevent smuggling along a reasonably open continuous land frontier (Curzon Price, 1987, pp. 22–3). Certificates of origin may be easily abused. For example, according to the European Voice, the volume of Israeli orange juice imported into the EU at preferential rates was in 1993 three times this country’s production capacity.72 As regards the EU system of preferences offered to the Lomé group of developing countries, the older and more complex the system becomes, the more administrative resources it absorbs. Many working hours were lost by firms, by customs, by forgers of origin certificates, by checkers of these forgeries, and so on. If those countries had been offered the most favoured nation (MFN) treatment, this would have been more efficient from the resource-allocation point of view, as well as cheaper to run (Curzon Price, 1996b, pp. 73–4). Free trade in manufactured goods between the EU and EFTA countries operated easily. This was due to similar and low tariff rates between the parties. Nonetheless, the rules of origin were quite costly to apply. They may require complex, troublesome and costly accounting methods both for inputs and for finished goods. The cost of formalities to determine the origin of a good was between 3 and 5 per cent of the value of shipment. Many exporters did not find it worthwhile to make use of the origin rules at all and opted for paying tariffs on their exports (Herin, 1986, p. 16). The WTO stated that the cost to collect, manage and store information needed for administration and verification of origin in Europe was about 3 per cent of product prices. ‘Such high cost could lead traders to abandon the idea of claiming preferences after having weighed the net benefit to be negligible if not negative’ (WTO, 2004, p. 51). If this is the case, then a relatively low utilisation rate of preferences (relative to total trade) does not come as a surprise. This is

Evolutionary Economic Geography

252

even more so in the countries/groups that do not have automated clearance of claims and that depend on the personal judgement of customs officers in countries that are simultaneously operating several preferential trading schemes with different partners. Hence, both administrative and production costs (in terms of suboptimal location of production) may increase for firms. Preferential trade deals are common in north–south trade relations. The EU has a ‘passion’ for and a plethora of those arrangements with the developing countries. The problem is that they are complex and incoherent. Therefore, they make compliance quite demanding and costly. This erodes the benefit of the accorded preferential trade access offered to the southern countries. In the case of the US trade preference offered to Mexico within NAFTA, Anson et al. (2005, p. 514) stated that this was 4 per cent on average in 2000. The estimated total compliance costs were 6 per cent (almost a third was attributable to administrative costs). Hence, rules of origin largely ‘undid’ preferential tariff access. NAFTA offered little real market access to the southern partner. An OECD (2005b) study noted that beneficiary countries were often eligible to ship goods to the US or EU under multiple preferential programmes (Generalised System of Preferences; Special Incentives to countries that are involved in anti-drugs campaigners; Everything But Arms concessions; Cotonou Agreement; and the like). Regarding farm exports to the EU in 2002, the ‘utilisation rate for nonreciprocal preferences as a whole exceeds 89 per cent’ (OECD, 2005b, p. 10). Considering that these numerical dissonances among different studies, as well as the fact that preference schemes are frequently revised and hence unpredictable, this area begs for more research. Additional evaluation of rules of origin and paper-based ‘hard’ certificates reveals that they do not fit into current economic reality (electronic exchange of data). These rules were designed in the past when they reflected protectionist policy objectives. Hence, such rules do not take into account integrated international ways of (global) production. They also do not make allowances for the new flexible production and distribution technologies. Their complexity increases trade costs and demands rationalisation and simplification. It is difficult to assess whether a free trade area is preferable to the formation of a customs union without knowing at least two elements: the tariff levels in the participating countries, and the limits and controls that provide rules of origin. In general, a customs union with a low common external tariff is much simpler to administer and is preferred to a free trade area with numerous complicated rules of origin that increase compliance and administrative costs (Clausing, 2000, p. 433). Hitherto, the analysis has dealt with tariffsonfinal goods. If tariffs are introduced on raw materials and semi-manufactured goods, and if one supposes that production functions are identical in all members of a free trade area, all other things being equal production will be located in the member country that has the lowest tariffs on inputs. Such a situation results in the creation of ‘tariff factories’, which distort investment decisions and introduce geographical deflection in production. Once established, the producers will resist liberalisation of trade. All this increases both the prices paid by consumers and the possibility of retaliation from abroad. A solution to the problem may be found in liberal rules of origin that encourage trade creation and reduce spatial misallocation of resources, rather than in restrictive rules which generate trade diversion.

Market structure and location of production

253

Liberalisation of trade and FDI with third countries potentially increases both trade with those countries and the share of foreign inputs in domestically produced ‘hybrid’ goods.73 Rules of origin in general, and the minimum local content requirement in particular, are used as NTBs to prevent goods with a high external content from receiving preferential treatment in the internal trade of the EU or NAFTA.74 However, it is increasingly difficult to determine the ‘nationality’ of some goods, for example cars. Countries throughout the world are free to use rules of origin as policy measures, as they are not yet regulated by the WTO. As a result, practice varies considerably, causing difficulty for both producers and traders. The present EU rules of origin are based on the Community Customs Code (CCC) established in Regulation 2913/92 and its implementing provisions established in Regulation 2454/93 . Article 23 of the CCC relates to goods produced in a single country. According to Article 24, goods whose production involved more than one country are deemed to originate in the country where they underwent their last substantial processing or working, provided that • this processing or working was economically justified, • it took place in an undertaking equipped for that purpose, and • its working or processing results in a new product or represents an important stage of manufacture. The following list presents some examples regarding ascertaining the origin: • Unrecorded diskettes (Code no. 8523.20) If neither the disc nor the upper and lower shells are manufactured in the country where assembly of the diskette takes place, the diskette shall have the origin of the country where the components representing the highest percentage of the ex-works price originated. Assembly and packing alone shall not confer origin. • Television receivers (Code no. 85.28) The 45 per cent value-added rule. When this rule is not met, the apparatus shall be treated as originating in the country of origin of parts whose ex-works price represents more than 35 per cent of the final ex-works price. When the 35 per cent rule is met in two countries, the apparatus shall be treated as originating in the country of origin of parts representing the greater percentage value. • Photocopying apparatus (Code no. 90.09) Change of tariff heading or 45 per cent value-added rule, except assembly of photocopying apparatus accompanied by the manufacture of the harness, drum, rollers, side plates, roller bearings, screws and nuts. The WTO intends to draft new standards for the determination of origin that will be universally applicable. Until it does so, the EU will continue to determine the origin of the goods according to the above legal provisions and in accordance with the Kyoto Convention (except for the provisions on which a reservation has been expressed by the EU). Yet another complication in the calculation of added value and the shares of different countries in the final origin of a good is presented by fluctuations in the rates of exchange. For example, what rate of exchange should be used to calculate added value: the rate prevailing on the date of importation of inputs or the rate on the date when the final good is exported to the EU? A case in point arose in the export of ballpoint pens

Evolutionary Economic Geography

254

(made from American components) from Switzerland to France in 1983. The European Court of Justice decided in favour of the date of importation. Controversies about origin were for a long time confined to trade in goods, but are now relevant to trade in services too. What determines the ‘nationality’ of traded services? Many are based on wide international networks. Nonetheless, certain criteria are emerging, including place of incorporation, location of headquarters, nationality of ownership and control, and the principal place of operation and intellectual input. Rules of origin are responsible for many controversies, heated debates and tensions in international trade. In a supposedly ‘globalised’ international economy where TNCs are increasingly involved in foreign production and sales, rules of origin make to some less and less sense. These rules are associated with distortions in investment decisions, spatial location of production and reduction in international production and trade, and potentially do more harm than good. 8 Non-tariff barriers Introduction The GATT was quite successful in achieving a continuous reduction in tariffs on industrial goods during the post-Second World War period. Unfortunately, at the same time NTBs have flourished and eroded the beneficial liberalising effects of tariff cuts. NTBs are all measures other than tariffs that influence international trade. They affect trade and the geography of production. Some of them are overt (quotas), while others fall into a grey area such as the application of technical standards or rules of origin. Although they are costlier in terms of resource efficiency and they do not create customs revenue, they have increased because the GATT/WTO do not permit a unilateral introduction of new tariffs, while at the same time domestic pressure groups and vested interests may be quite successful in eliciting protection. NTBs are strongest in the ‘sensitive’ groups of goods and services. In fact, the use of NTBs may determine which groups of goods are sensitive. The implementation of the Treaty of Rome eliminated tariffs and quotas on internal trade in the EU in 1968. The elimination of NTBs was the task of the Single European Market Programme (1985–92). While tariffs, like transport costs, increase the price of a good, NTBs act like import quotas but do not generate revenue for the government. Currently, NTBs present the most important and dangerous barriers to trade, fragmenting markets and production more severely than tariffshave ever done. Tariffs were, on average, reduced under the auspices of the GATT/ WTO to relatively low levels, so that they now play a relatively minor role in the protection of the economic inefficiency of a country. Nonetheless, national administrations try to obtain short-term political gains through protectionism at the expense of long-term economic benefits. Hence, NTBs are and will be high on the agenda for all future international moves to liberalise trade. The inclusion of general and social issues such as the environment, labour standards and human rights in trade liberalisation negotiations and deals may generate unwanted and harmful effects on trade and location of production in the longer term. No matter how noble the intentions, hard reality may ensure that these issues may be used in the future as

Market structure and location of production

255

handy tools for protectionism. Perhaps such important general and social issues may be better handled outside both trade deals and the WTO. Classification Consideration of NTBs has always been difficult. One reason for this is the creativity of their instigators, but another, more important, reason is the lack of data. Administrations either do not record the use of NTBs or do so only partly. It comes as no surprise that the reported impact of NTBs can lead to considerable underestimations. Our classification of NTBs is presented in Table 4.1. The greatest criticism of NTBs is their lack of transparency. Hence, they are prone to abuse and need to be monitored closely, which increases firms’ administrative costs (money that could be spent on R&D). As an economic policy instrument, tariffs are a blunt weapon against which markets can defend themselves through adjustment measures. In contrast, NTBs circumvent market forces and, hence, introduce greater distortions, risk and uncertainty. NTBs therefore prevent efficient spatial and industrial location of resources, investment and specialisation. Antidumping If a government observes that a good is or is likely to be suddenly imported into its country in excessively large quantities, it may introduce an antidumping tariff. Foreign exporters can circumvent this barrier by changing the geography of production. For example, they may move the final stage of production to the tariff-imposing country. To counter this action, the home authorities may request that goods produced domestically must have a certain local content in order to obtain preferential treatment. Once the original supplier of the good meets this requirement, the local authorities may go further, as was the case with the EU control of imports of Japanese integrated circuits (microchips). Chips that are only assembled in a country are not considered to originate in that country. The EU non-preferential rules of origin link the origin of microchips to the place where the diffusion process takes place. Assembly and testing come after the diffusion process and they add to the final value of the chip about as much as diffusion. There were two reasons for this decision by the European Commission regarding rules of origin for integrated circuits. First, the rules demand that a part of the manufacturing process is located in the EU, which requires an inflow of FDI and the creation of highquality local jobs in the EU. Second, the intention was to support EU producers of Table 4.1 Non-tariff barriers Principal group 1 Government involvement in international trade

Type Subsidies (production, exports, credit, R&D, low-cost government services) Public procurement (local, regional, central) State monopoly trading Exchange rate restrictions Embargoes Tied aid

Evolutionary Economic Geography

256

2 Customs and administrative entry procedures

Customs classification Customs valuation Monitoring measures (antidumping and countervailing duties) Rules of origin Consular formalities Trade licensing Deposits Compulsory insurance with local insurers Compulsory unloading and storage Calendar of import Administrative controls Inadequate institution to appeal against decisions by the customs

3 Standards

Technical Healtha Safety Environment and ‘green standards’ Fair trade Lack of recognition of international standards Inspection, testing and certification Abusive sampling practices Packing, volume, weight and labellingb Registration of patents and brand names (and lack of it) Pirated goods

4 Others

Quotas (tariff-free ceilings) Local content and equity rules Tax remission rules Variable levies Bilateral agreements Buy-domestic campaigns Voluntary export restriction agreements Self-limitation agreements Orderly marketing agreements Multi-fibre agreement Lack of transparency and ambiguous laws Cartel practices Precautionary principle Permission to advertise

a

Recycling at least a part of the product (e.g. paper, plastic, glass, electrical appliances or computer hardware) may be a ‘green’ requirement. However, recycling paper requires a lot of energy and chemicals (which are partly released into the environment). Would a request to plant more trees be a superior choice? b Labelling in itself is not a panacea. It ought to have an easily understandable message. But should that be a positive or a negative statement? For example: ‘meat from animals that were fed on genetically modified organisms’ or ‘meat from animals that were not fed on genetically modified organisms’?

Market structure and location of production

257

microchips, some of whom practise diffusion in the EU while assembly and testing takes place in nonEU countries, where labour costs are lower. In this case, the EU defined the origin of the good as the place where the ‘most’ (rather than the ‘last’) substantial production process took place. Origin is determined by where R&D takes place and by the location of capital equipment used in the production of such goods, rather than by the place of transformation of goods. According to these rules, Ricoh photocopiers, which are made (well, assembled) in California, originate not in the US but rather in Japan, because that is where essential parts such as drums, rollers, side plates and other working equipment originate. Having declared the origin of these machines to be Japan, the EU has an arsenal of NTBs to curtail imports of these goods from the US. This has provoked tensions in EU trade with the US.75 The EU has a so-called ‘sunset’ clause that applies to antidumping duties. Antidumping measures apply for ‘only’ five years unless there is a review in which dumping and injury are again established. There is no sunset provision in US antidumping laws, which can last almost indefinitely, as exemplified by the fact that they have been applied to Japanese colour TV sets for thirty years. In addition, the scope of the measures has been expanded to cover liquid crystal display, as well as projection television (Belderbos, 1997, p.425). A comparative study of antidumping actions in the US, the EU, Canada and Australia concluded that 90 per cent of imports that were unfairly priced according to antidumping rules would be found to be fairly priced under corresponding domestic rules of competition (UNCTAD, 1999b, p. 17). The Uruguay Round deal introduced several changes that eased the tough US antidumping policy. The US law was written and applied in favour of the petitioner (thus against the foreign producer and domestic user of its products) (Destler, 2005, p. 164). Changes introduced in the post-Uruguay era refer to new cases (not to the review of old ones) and include: a general cap (sunset) of five years for antidumping duties unless a WTO review establishes its continued justification; producers submitting a case must account for at least a quarter of domestic output of the product (the US law did not have such a minimum requirement); and an investigation must be ended if the dumping margin is less than 2 per cent or if ‘dumped’ imports from the exporting country form less than 3 per cent of total imports (US law had a 0.5 per cent minimum condition for the dumping margin). This made it harder for the petitioner in the US to obtain protection and preserve its current location and volume of inefficient production, compared with a foreigner. A firm may avoid antidumping actions by changing production geography (relocating production). A new plant may be located either in a third country or in the country that has imposed antidumping duties (tariff jumping). However, the European Commission and the US Department of Commerce have the right to include third-country exports within the scope of an antidumping measure. For example, during 1989–90, nine Japanese fax-producing plants were established in the EU even though there were no antidumping investigations. Some of these firms explicitly stated that fear of antidumping measures was their major reason for investing and location in the EU (Belderbos, 1997, pp. 434–4). Once established in the EU, these firms can sell at any price they want (below the production costs or below the prevailing prices on the local market). They are not allowed to discriminate or segment the market and they may not follow predatory pricing policies (whatever this means). The EU rules on competition may prevent this.

Evolutionary Economic Geography

258

After almost virtual elimination of tariffs and quotas through multilateral negotiations, the WTO-sanctioned antidumping trade protection devices are becoming the principal barriers to trade at the start of the twenty-first century. If a government of one country subsidises exports, then another country has the right to introduce countervailing duties. As governments of at least two countries are involved, this dispute often becomes quite visible. If a firm sells a product abroad at a price that is below that at which it sells in its home market, this constitutes dumping. The injured country may introduce an antidumping duty. Both dumping and injury are necessary conditions to impose an antidumping duty. This idea seems neat in theory, but to apply it in practice may be difficult and controversial. It is a paradise for lawyers. • Dumping What if a country produces and exports a good that is not sold at home? How should this principle be applied if, say, the Chinese produce and export each year millions of artificial Christmas trees for the EU and US markets and do not sell these products at home because there is no market there? What arbitrary method should be used to determine the domestic Chinese price for Christmas trees?76 • Injury To determine injury to the domestic industry may be even more open to random whims and abuse by the ‘injured’ industry and investigating government. What measure should be used? An increase in the volume of imports? Changes (reduction) in the domestic production? Market share? An increase in the domestic unemployment? All of these? In what proportion? Ambiguity of rules and discretion in determining injury present considerable problems. The most contentious issue relates to cumulation, whereby the investigating authority adds all imports from all countries under investigation and assesses their combined injury impact on the domestic industry. This cumulation is in use in the EU, as well as in the US, Canada and Australia. In the EU, in 45 per cent of cases in which defendants agreed to increase their prices in order to bridge the dumping (injury), no injury would have been found if the European Commission had not aggregated imports from the countries under investigation (Tharakan et al., 1998, p. 335).77 There are three types of cumulation rules (Cadot et al., 2007): • Bilateral cumulation is when producers located in one country use inputs that originate in another partner country without affecting the final status of the good in question (provided that the input originates in the partner country). This type is most common. • Diagonal cumulation allows all partners in a preferential trade agreement to use inputs from any participating country and the final status of the good for this trade group is as if the good was produced in the country where the transformation was undertaken. The EU, for example, applies this type under its Pan-European Cumulation System (PANEURO).78 • Full cumulation permits a greater geographical fragmentation of production than the other two types. Under this system, all stages of processing or transformation of a good within the participating group of countries count as qualifying content. More advanced countries with higher wages may outsource certain lower technology stages of production to low-wage countries while maintaining the preferential status in trade of the produced goods in the low-wage locations. This is the case in the ASEAN Free Trade Area where there is a robust intra-area trade in parts and components for

Market structure and location of production

259

assembly and then exports to the US and Europe. Hence the ‘easiness’ regarding rules of origin. A growing interest in east Asia for the creation of preferential trade areas and integration ought to resist the creation of complicated, protectionist and distorting rules of origin, like those applied in North America and in Europe. A simple full cumulation system ought to be preserved. The number of antidumping duties or antidumping cases (there are hundreds of them, mainly in the developed countries, although India and China also use them) is not something that matters most as an NTB. The biggest problem and discouragement to trade, investment, innovation, economies of scale and reduction in prices to consumers may result from the serious possibility of being harassed by the authority of a country that imports the good in question. A country may apply ambiguous rules and use them in an unorthodox way with the intention of sheltering inefficient domestic producers. The exposure to such risk and uncertainty cannot be fully quantified. The investigated suspect (foreign producer or exporter) has, at his/her own expense, to provide the inspectors with all operating details, including a full list of the costs of production. The result is that the ‘injured’ country’s industry obtains all important technical and accounting details about foreign competitors for free. Importers also face costs, risk and uncertainty as, once the antidumping procedure is completed, they may need to pay backdated antidumping duties. The existence of antidumping measures may endanger and reduce all expected gains that may result from multilateral free trade. This situation in international trade may prevail for quite some time. The solution may be to create, apply and enforce multilateral rules to determine dumping and injury. With such rules it may be possible to construct a counterfactual situation of a case to try to ascertain and separate effects. In addition, cumulation may be eliminated, save for the case where there is collusion among exporters. The Byrd Amendment (Continued Dumping and Subsidy Offset Act) was enacted in the US in 2000. It introduced an unpleasant irritation in trade relations. This is a law that requires the US customs administration to directly distribute antidumping and antisubsidy proceeds to the company or companies that initiated the case. This has encouraged antidumping complaints by the American firms (producers of shrimps, candles, furniture and other goods) that were seeking windfall profits. Billions of dollars were paid out to US firms. The WTO ruled in 2003 that the US policy was illegal. As this amendment was not repealed, the WTO allowed the EU, Japan and Canada to introduce trade sanctions in the form of levies or surtax on imports from the US from May 2005.79 The Amendment was repealed in 2006, but the US firms would continue to benefit from it till 2008.80 Public procurement Government at all levels (local, regional and central) is an important consumer of goods and services. It can use its procurement policies either to protect home business (SMEs in particular, as is the case in Germany) and employment, or to influence the geography of production and support young industries during their first tentative steps in order to help the creation of a country’s comparative advantage. Public procurement can also be

Evolutionary Economic Geography

260

employed as an instrument to apply regional policy and aid to the advantage of certain population groups (for instance, women and minorities). An open public procurement market in the EU is seen to be one of the major potential benefits that EU firms may exploit. The EU-wide competitive tendering directives require the publication of all tender notices above a certain (rather low) financial threshold in the Official Journal of the European Community.81 Coupled with compliance with EU technical standards and a reasonable time for the submission of the offer, together with transparent award criteria, this increases the fairness of the bidding and award procedure. The low threshold is intended to avoid understating the value of the contract in order to evade open tendering. The threshold may need to be increased. On the one hand, it may not be attractive enough for the non-national companies, while on the other, it is quite costly for local authorities to put work out to tender. The European Commission may exclude foreign firms from the public procurement market if EU enterprises do not have reciprocal access to public contracts in the countries from which these firms originate. In spite of this, there is a chance of inertia (read ‘buy domestic’). A public authority could subdivide a large public contract into a series of small ones and thus avoid the obligation to advertise and award the business locally. The public procurement market for goods and services was 16 per cent of the combined GDP of EU(25) member countries in 2004. A common procurement vocabulary, introduced in 2002, is a tool to modernise, simplify and increase the flexibility in the EU public procurement market. The share of the total number of public procurement contracts openly advertised in the Official Journal (a sign of market transparency) was 8.4 per cent in 1995. This share gradually increased to 16.6 per cent in 2002 and remained at 16.1 per cent in 2004. The value of these advertised contracts was ‘only’ 2.64 per cent of total EU(15) GDP in 2002 and 2.68 per cent of total EU(25) GDP in 2004. Even though this may seem a rather low share, competition in this market has been steadily increasing from 1995, when the same share was 1.44 per cent. Smaller countries that lack domestic industrial capacities seem prepared to source more from outside their domestic economies. Greece, for instance, advertised 36 per cent of its public purchases contracts (this represented 4.1 per cent of the GDP), while Germany did so in only 7.5 per cent of cases (this represented 1.2 per cent of the GDP) in 2004 (WTO, 2004, pp. 66–99; WTO, 2007, pp. 62–5). Technical standards Goods eligible for tariff and quota-free trade in all integration arrangements are those that are produced by countries within the group. The exceptions are the products that endanger public morality and public security and order, as well as human, animal or plant life or health. An important NTB can be found in the guise of technical standards (a set of specifications for the production and/or operation of a good). Technical standards are different from tariffs in that they increase production and operation costs, while tariffshave an additional revenue-generating effect for the government. The real intention of these standards as hidden barriers to trade is in many cases to protect the national producer despite the long-term costs in the form of higher prices that come from lower economies of scale, poorer technical/energy efficiency and reduced competition.

Market structure and location of production

261

One of the best-known examples of an NTB operating within an integrated group arose in Germany in 1979. Germany banned the importation of a French fruit-based liqueur, Cassis de Dijon, on the grounds that liqueurs consumed in Germany should have an alcohol content of at least 25 per cent, and that of Cassis de Dijon was 5 per cent less than this norm. The German administration wanted to protect its domestic consumers on the grounds that they might buy this ‘weaker’ drink thinking that it was ‘strong’. In addition, weak spirits might be better tolerated by consumers. Hence the administration wanted to reduce consumption in order to protect public health. This case was considered by the European Court of Justice, which ruled that the German ban on liqueur imports was not legitimate. The implications of this ruling are of paramount importance: any goods legally produced and traded in one member country cannot be banned from importation in partner countries on the grounds that national standards differ. The ruling opened the way to competition in all goods manufactured in the EU, with the exception of those that could jeopardise an important national interest. Where differences in national standards do exist, consumer protection is assured by the addition of a warning on the product. However, this ruling is not in itself sufficient to guarantee a uniform market throughout the EU and an associated increase in competitiveness. To achieve a uniform market requires harmonisation, with all its attendant disadvantages of excessive regulation, reduction in choice and the long time required for its implementation. The Cassis de Dijon case was cited as a precedent in 1987 when the EU forced Germany to open its market to beers from other EU countries despite the fact that some beers do not comply with the German beer purity law, Reinheitsgebot (1516), which specifies that beer must be composed only of water, barley, hops and yeast. This decision has not, as feared, endangered Germany’s domestic beer production. Many other national laws define, in particular countries, what, for example, can be sold as pasta, sausage or lemonade. In any case, since the Cassis de Dijon case in 1979, mutual recognition (and standardisation) has been the cornerstone of the Single European Market. Following a six-year legal battle, the European Court of Justice ruled in 2003 in favour of the Associazione Prosciutto di Parma in the case against British ASDA Stores Ltd (owned by the US retail giant Wal-Mart). The judges upheld the Associazione’s request to stop slicing and packing the hams in-house and selling the result as Parma ham. The ruling does not affect Parma ham sold at delicatessen counters, which is sliced in front of the client. The pre-packed and sliced ham sold on supermarket shelves must be cut in the region of origin, otherwise the reputation and authenticity of the ham may suffer. The Associazione has about 200 producers of ham in Parma. It controls the whole process of production including slicing and packaging. This guarantees the quality and genuineness of this popular and expensive product.82 The ruling also applies to Grana Padano (Parmesan) cheese, another prime gastronomical product. Italian food is highly appreciated and is becoming hugely popular throughout the world. However, this great popularity and market created a fertile ground in which many imitations may flourish and sell their forgeries cheaply. Total international ‘trade’ in counterfeit goods has, according to the OECD, a value of about $200 billion a year (this is twice the total annual budget of the EU). Business lobbies regard this as an underestimate and argue that the value ranges upwards from $600 billion.83 This trade does not include only copies of branded products, but also

Evolutionary Economic Geography

262

faked car parts and drugs. Grave dangers for transport and health safety are immediately apparent. Even though this trade is illegal, immoral and often introduces hazards to life and health, it is highly profitable for those who are involved in it. If millions are desperate to buy, there will be very many who are ‘desperate’ to produce and sell. Production and trade will take place no matter the dangers. Moral stories are one thing, actual fighting against this production, trade and consumption is another. Total victory is not feasible as there are not sufficient resources. Therefore, the concentration of effort ought to be on the most harmful activities, with a (sharp) reduction of profitability in counterfeit production and trade. An example may be a sharp increase in penalties for those including fake parts in the final product, or those who employ illegal workers.84 There are about 600 foods and 4,000 wines that enjoy the protection of EU law. In order to be protected against cheap imitations, they must pass rigorous checks to maintain quality. These goods may be luxuries, but their annual market is Є40 billion in the EU. Consumers are also interested in the protection of originality, authenticity and quality of food (e.g. feta cheese). A carefully regulated, enforced and supervised system would eliminate a host of food scares from dioxin in chicken to ‘mad cow disease’ (BSE). The European Commission drafted a plan in 2004 to introduce a ‘Made in the EU’ label on all products. This is supposed to promote European origin as a mark of distinction. In addition, the specific origin mark ignores the existence of global production chains. Brand experts criticised this plan on the grounds that consumers have few positive associations with the EU. Some governments asked the Commission to abandon the proposal. Distributors and consumers rightly want specific information about the origin of goods they sell, buy and consume. In this regard, they prefer to see, for example, a ‘Made in Germany’ stamp on tools or ‘Made in Italy’ label for clothes or shoes. Or, if they buy whisky, they think of Scotland, not of the EU.85 Many examples of NTBs arise from the application of standards, and this has a direct impact on the spatial location of production. Belgium allows margarine to be sold only in oval containers, while square ones are reserved for butter. The standard width of consumer durables in the EU countries is 60 centimetres, but in Switzerland it is a few centimetres less. A bylaw of the City of London stipulates that in taxis gentlemen must be able to sit comfortably in an upright position with their hats on. While in Britain it is illegal for lifts to have a stop button, the same button is obligatory in Belgium. In the Netherlands, beer can be sold only in returnable bottles, whereas the German standard for a beer container is non-returnable bottles. None of these technical specifications has a crucial impact on the protection of life and health. Their final effect is to reduce competition and increase prices to the consumers. Trade and industrial policy may well escape the scrutiny of voters, but small, well-organised manufacturing lobbies can wield a strong influence on government attitudes. In order to protect the integrity of the French language against an invasion of foreign words, the French government tried to insist that a French word should be used on the packaging of a particular good that is sold in France: pétales de maïs instead of ‘cornflakes’. However, according to the 1995 EU law, if a good that is legally sold on the Single European Market has a clear and unambiguous photograph together with a name in an ‘easily understood language’, it is not necessary to translate the name of the good into the language of the country in which the good is sold. In order to deal with various standards, the EU has adopted three different approaches:

Market structure and location of production

263

• Mutual recognition National regulation in other member states is recognised as equivalent. This can work only when there is a similarity among national standards. • Harmonisation of standards (old approach) Various technical standards for vehicles, food or chemicals had to be harmonised in detail and were mandatory. It is not surprising that this was a rather slow approach, not only because of the thorough technical work involved, but also because of the required unanimity in the Council of Ministers. • The new approach The EU sets only the essential characteristics of the good (minimum important requirements). The industry can have a greater degree of freedom to satisfy it and to be innovative. Harmonised standards are essential in some areas, such as safety,86 health and the environment, whereas mutual recognition (an agreement to acknowledge and tolerate diversity) may be an interim solution for traded goods until harmonisation is carried out. In any case, a long-term advantage of harmonisation is that firms will have to comply with only one set of rules instead of up to twenty-seven different sets. This may increase the gains that can be achieved from economies of scale. However, it only makes sense when the market for a good or a service extends to more than one member state. For example, although consumer preferences for relatively new goods such as DVDs, GPS devices, fax and photocopying machines, printers, CD-ROMs, computers and mobile phones are similar throughout the EU, there are widespread differences in the markets for foodstuffs and beverages, with preferences often strictly local.87 In such cases, a potential EU standard that does not embrace these distinctions may do more harm than good. The member states of the EU are, however, obliged to notify the European Commission in advance about all draft regulations and standards. If the Commission or other member states find that a new standard contains elements of a barrier to trade, they may apply remedial action as allowed under Articles 30 or 94 of the Treaty of Rome. EU member states may legislate in the non-harmonised areas, provided that they previously notify the European Commission about draft regulations and standards. This is important because of the mutual recognition principle. The Commission is quite satisfied with this system of previous notification because it avoids the creation of barriers to the free movement of goods and services. However, the European Commission estimated that the non-application of the mutual recognition principle cut intra-EU trade by Є150 billion in 2000 (WTO, 2004, p. 61). This has an important implication for the location of production, technical efficiency and costs paid by consumers. Regulatory red tape Another imaginative NTB is illustrated by the Poitiers case. In autumn 1982, the French government wanted to protect the home market from imports of Japanese VCRs. There was, however, no domestic production of VCRs. The home manufacturer, Thomson, simply imported Japanese VCRs and distributed them under its own name. Some months before Christmas 1982, the French government decreed that all Japanese VCRs should undergo a customs inspection in the town of Poitiers in order to ensure that instruction manuals in French accompanied them. Poitiers is in the middle of France, far from all main ports of entry for these goods; its customs post was staffed by only eight officers

Evolutionary Economic Geography

264

and was not well equipped. This decision increased costs (transport, insurance, interest, delays) and reduced the quantity of imported and sold VCRs to 3,000 units a month. Such a measure might make some sense in a situation where a government wanted to restrain rising consumer expenditure that was causing a shortterm drain on the balance of payments. In such instances, however, a more appropriate measure might be an excise duty or high sales tax. The moral of this tale is that the time to protect against the importation of goods whose efficient manufacturing depends on economies of scale and a steep learning curve is before external suppliers have captured most of the domestic market. The attention that this case attracted in the media was considerable, leaving the French government ‘no other choice’ than to revoke the measure. After Christmas, of course! A supplementary tax ‘replaced’ the Poitiers customs clearance procedure in January 1983. From 2000, cumbersome and costly procedures for imports of DVD players and lowmarket (cheap) television sets from China into the US and the EU and elsewhere have not been necessary for the protection of domestic jobs. Hardly anyone else produces these goods but China. There are no jobs to be protected. Harm is done to the efficient domestic retail industry. The situation is similar in imports of textiles in the developed world. Trade restrictions increase costs to domestic consumers, while they do not add to domestic jobs in the textile industry. Quotas on imports of Chinese textiles may also be futile as an instrument, as over time textile imports would come from other developing countries, not from production located in the developed world. Harm is done only to the developed world’sefficient retailers and domestic consumers. If this creates a row between trade partners, then even domestic EU exporters (and jobs) may suffer. Japanese cars in Europe Following the Second World War, Japan wanted to enter into a large-scale production of passenger cars (non-military vehicles). The Ministry of International Trade and Industry (MITI) introduced high tariffs and quotas on imports of cars from the US and Europe, hindered FDI into the car industry, and scaled down the number of home manufacturers of cars from a dozen to three. Japan also feared in 1952 that small Italian cars would make substantial inroads into its market. Therefore, Japan initiated an import quota deal. According to this arrangement, the number of Japanese cars that could be imported from Japan directly into Italy was set at 2,800 passenger cars and 800 all-terrain vehicles per year. When the situation changed and the Japanese became highly competitive in the production of cars, this NTB rebounded on Japan.88 As already noted, the bumpy evolutionary process is full of surprises and unexpected turns. Since 1987 the European Commission has refused to support Italy’s attempts to invoke Article 13489 of the Treaty of Rome to block indirect imports of Japanese cars through other EU countries. Hence, imports of Japanese cars have soared.90 Other EU countries had annual ‘caps’ on the imports of Japanese cars. For example, the ‘cap’ for Japanese cars in the French market was about 3 per cent, while in Britain it was about 10 per cent of domestic sales. A tacit limit to the market share achieved by Japanese car producers in the German market was about 15 per cent. The result is a lose–lose situation. On the one hand, consumers lose as they have to pay more for both domestic and imported cars and, on the other, the car industry located in Europe is postponing its

Market structure and location of production

265

inevitable structural adjustment and potentially losing its longer-term competitive edge relative to the Japanese. The governments of the EU countries simply bow to the shortterm interests of powerful domestic manufacturing lobbies. The EU and Japan concluded a ‘car deal’ in 1991. The EU offered the Japanese an increase in market share from 11 per cent in 1991 to 17 per cent (including Japanese cars manufactured in the EU) until 2000, when all quotas disappeared. Unfortunately, the parties placed different interpretations on the deal. The EU argued that the maximum number of Japanese cars either imported from Japan or produced in Japanese plants in the EU is based on forecasts of market growth and that, if the market grew less than forecast, it had the right to renegotiate the deal. The Japanese interpretation of the deal was that the EU had committed itself not to restrict the sale of Japanese cars in the EU. In any case the accord shows that the EU is keen to increase competition and efficiency in the EU car industry. Another accord between the EU and Japan in 1995 is intended to simplify EU exports of cars and commercial vehicles to Japan through the mutual recognition of vehicle standards, as well as allowing Japanese inspectors (who issue certificates to imported cars) to work in Europe. From 2000, the EU protection of its car industry relied mostly on tariffs, standards and market entry and distribution costs. Precautionary and prevention principles As already stated, all goods legally produced in the EU may be freely traded throughout the EU unless they endanger public morality or the life or health of humans, animals and plants. This freedom may be challenged in practice by two principles, which represent a kind of social price of prudence, that can sometimes act as NTBs: • Precautionary principle refers to a proactive regulatory method. Its origin is in Germany during the late 1960s when the country wanted to introduce new laws to protect the environment. According to this principle, there are specific cases that are linked to a certain exposure to danger and risk. They concern products, technologies or activities. Although a hazard may not yet be scientifically documented beyond reasonable doubt, it can plausibly be anticipated in the light of accumulated experience. Even though there is a lack of definitive scientific knowledge, one must react now (prohibit or regulate heavily), before the arrival of a potential hazard. Although it may have certain justification following grave hazards and dangers such as BSE (mad cow disease), this principle may be badly manipulated and abused by rent-seeking vested interests. The principle has been on the books since 1958, when the US Congress enacted an amendment to the Food, Drug and Cosmetic Act. Any new additive to food is presumed unsafe until its safety has been demonstrated through scientific procedures: The US statute places the burden of proof on the manufacturer and requires that foods containing new additives be presumed unsafe until proven safe. But in the case of GM foods, the US turned the law on its head and insists they must be presumed safe until proven not to be.91 From the 1980s, Germany was very active in introducing environmentrelated legislation based on this principle. The French court invoked the principle in 1998 when it suspended the sale of three genetically modified corn varieties. The EU may apply this principle in practice, but action should be proportional92 to the

Evolutionary Economic Geography

266

potential risk and non-discriminatory, as well as subject to review and change in the light of new scientific findings. The EU lifted this six-year-long ban in 2004 because of pressure from the US through the WTO. • Preventive (reactive) principle applies in the situation where the mischief has already occurred and where there is a need to prevent or curtail the repetition of the known type of harm and damage. France invoked the precautionary principle in 1999 to maintain its unilateral ban on the import of beef from Britain because of BSE. This was a direct challenge both to the principle of the free movement of goods and to the authority of the European Commission. After several years France lifted its ban on imports of British beef.93 The precautionary principle touches on several sensitive and emotional issues (health, children’s food safety, environment) and can provoke fierce protests from consumer groups. Although the precautionary principle is not (yet) enshrined in EU law, its political significance is that it may be applied at any point. If the application of the principle were to proliferate to any great extent, this would signify the end of the Single European Market for the goods involved. The goal of the European Food Safety Authority (established in 2002) is to provide scientific input into the EU decision-making procedure. Trade-related investment measures A trade-related investment measure (TRIM) is a new name for the ‘old’ performance requirement. UNCTC (1991a) provided a survey of TRIMs. It was found that local content requirements are more common than regulations regarding exports or employment. Although TRIMs exist in both developed and developing countries, they are more common in the developed world. They are usually concentrated in specific industries, such as carmaking, chemicals and computers. This study of 682 investment projects found that in 83 per cent of cases TRIM requirements (such as local sourcing, exporting) would have been undertaken anyway; that is, not only were TRIMs intruding into the mix of inputs in the production process, but they were also redundant. In any case, the WTO outlawed these measures in the Agreement on TRIMs. The Agreement on TRIMs, negotiated during the Uruguay Round, applies only to measures that affect trade in goods. This agreement states that no member country shall apply a measure that is prohibited by the provisions of GATT Article III (national treatment) or Article XI (quantitative restrictions). Examples of inconsistent measures, as spelled out in the Annex of the agreement, include local content or trade balancing requests. The agreement contains transitional arrangements that allowed countries to maintain notified TRIMs for a limited period of time following the entry into force of the WTO (two years in the case of developed countries, five years for developing countries and seven years for least-developed countries). The agreement also established a Committee on TRIMs to monitor the operation and implementation of these commitments.

Market structure and location of production

267

9 Conclusions Even though competition creates the environment for the creation and presentation of new ideas, ways and means, even though it may keep prices moderate or low, and even though it may remove business arrogance and incompetence, many people would not voluntarily rush to the barricades to protect it. However, those who are inventive, those who take part in the race and eventually win it, should not be drowned by the others, as was the case with Anton Müller. They are the ones that break old ties and offer new opportunities to go forward, not only in a given location but also elsewhere. In the very nature of the competitive market economy is that there is no economic gain without pain (price). After all, you have to break a few eggs if you want to make at least a bit of extra profit. If the pressure and adjustment costs that result from a relaxation of trade restrictions do not hurt someone somewhere in the economy, then they probably produce no overall benefit either. The evolutionary reallocation of resources in space, lines of business and/or professions is a source of gains. The adjustment cost is a finite, one-off investment. The gains accruing from each improved resource allocation are ongoing, at least for some time in the future. Thus, there are reasons to believe that the cost of adjustment ‘pain’ is much exaggerated (Curzon Price, 1987, p.16). Notes

1 http://dictionary.oed.com/ (accessed on 8 January 2008). 2 Competition policy does not protect individual competitors, but rather the process of competition. 3 If one had to negotiate a contract each time one bought apples, ordered a meal or purchased a train ticket, without the existence of such rules, a very long period of time would be required for each transaction. Apples would rot, the meal would arrive next month, while trains would leave almost empty (provided that they were running on time). 4 The American business model is based on four principal ingredients: material selfinterest governs our economic lives (greed is good, nice guys finish last); markets should operate freely; the size and influence of the state administration should be kept at a minimum (small government); and taxation should be low, just sufficient to cover the minimum state functions (Kay, 2004, p. 313). This model

is deficient for its naive approach to issues of human motivation, its simplistic analysis of structures of property rights, its inability to maintain efficiency in the face of imperfect information, its misleading account of markets in risk, its glossing over problems of cooperation and coordination, and its failure to describe the generation of new knowledge on which its very success depends. (ibid., p. 320)

5 The World Economic Forum states that the Scandinavian model produced and fostered highly competitive economies with exemplary R&D, education, social care and stability. This was achieved without the cruel social inequalities that are the outcome of the American model. Critics state that the Scandinavian model is based on heavy public expenditure. Even though this may be true, an overlooked fact is that EU countries such as France, Germany or Italy have a similar public expenditure, but these funds were used less efficiently than is the case

Evolutionary Economic Geography

268

in Scandinavia. Generous national welfare expenditure should not be seen as an obstacle to ‘national competitiveness’. Most workers in developed market economies will continue to work in the twenty-first century in non-traded domestic services. Their jobs do not ‘compete’ with jobs in other countries. 6 P. Krugman, ‘The French Connections’, New York Times 23 July 2007. 7 Intellectual property rights are those that come from patent, copyright and trademark rules. 8 It is conventionally argued and repeated that a country that fails to protect intellectual property rights (IPR) harms its own national interests (with an increase in crime, health hazards and reduction in taxes, among many others). Making and selling counterfeit products becomes a global industry. In the past, however, ‘It took the US almost a century to turn from a violator (ask Charles Dickens if he ever got royalties from American publishers) to an IPR champion’ (O. Shenkar, ‘Why fake goods are a natural economic motivator’, Financial Times,27 May 2005). 9 These three men are to be software experts, not lawyers. The Department of Justice and Microsoft will each select one member, and those two will select the third. Their role is to notify only the Department of Justice (not the public or the industry) about their findings. The Department may use the dossier compiled by this committee to initiate another court case. 10 Microsoft argued that Media Player was a vital part of the Windows operating system. Like other operating systems, Windows needed to play music and video clips. It was wrong and unreasonable to request by law the technical separation of the two. The Microsoft lawyer told the EU Court of First Instance that only 1,787 versions of Windows without Media Player have been ordered since the Commission forced Microsoft to market such a product – compared with the 35m units with Media Player sold in the corresponding period. The failure to offer a product that nobody wants cannot be an abuse.

(T. Buck, ‘Microsoft argues EU antitrust ruling was flawed’, Financial Times, 24 April 2006).

The European Commission ‘forced’ Microsoft to supply the market Windows without the Media Player. The problem was that the Commission dealt predominantly with the technical side of the product. The request was that the full version should not be offered more cheaply than the ‘stripped-down’ version. Hence Microsoft offered the full Windows version and the one without the Media Player at the same price. 11 Competitive markets make it quite hard to earn profits. In these markets entry into a business activity is fairly simple and quick, while exit by failing firms is relatively fast. Elsewhere, where markets are not competitive, they are often regulated. 12 Military aircraft technology rapidly advanced after the First World War. Aircraft increased their speed and flying range. Britain perceived its vulnerability to an attack from the air. How to prevent it? A social need was noticed. One of the ways to defend the country was to detect approaching aircraft. Radio waves could be reflected from flying objects. This led to the invention of radar. In addition and about the same time, it was discovered that aircraft can reach greater speed if they fly at higher altitudes where the air is thinner. However, piston-propelled aircraft could not draw in enough oxygen for proper operation at high altitudes, while propellers had less ‘bite’. This need led to the invention of a jet engine (Arthur, 2007, p. 279). 13 The First Industrial Revolution was based on the new Newtonian science and mechanistic approach to the world.

The Second Industrial Revolution of the latter part of the nineteenth century could never have happened anywhere but in Europe. That revolution was firmly science-

Market structure and location of production

269

based and could not have happened in a country that did not have access to Western science. (Lipsey et al., 2005, p. 259)

14 The actual process of invention varies from case to case. A universal formula for success does not exist. Some inventions may come from an individual working on a shoestring budget, others from a group of researchers with a huge budget. Certain results may come from long years of trial and error, others may arrive much more quickly. Nonetheless, invention does not come out of the blue. It has a certain internal logic. This logic is the link between observed need and an effect that can be exploited to satisfy it. It is most often a lengthy and ‘untidy’ recursive problem-solving exercise. This does not mean that the process is purely rational. Human agency and subconscious elements are still a mystery. But this is a known mystery (Arthur, 2007). 15 Having spent the last decades of the twentieth century mastering various types of manufacturing, China wants to climb the technological ladder. The intention is to replace the omnipresent sign ‘Made in China’ with one that says ‘Invented in China’. Therefore, China is investing heavily in R&D and improvement in education. From 2006 China ranked second in absolute world R&D expenditure, just after the US. China also successfully attracts back quite a proportion of its citizens who earn PhD degrees at US universities. 16 Money and enterprise are among the most important social-organisation technologies. In Europe, it was the Church that was the first corporation in the modern sense. Later came business associations (guilds), universities and independent cities. 17 Serendipity is the faculty of making happy and unexpected discoveries by accident. 18 ‘Schumpeterians’ point to five long waves in modern history: (1) 1780–1840, steam power drove the Industrial Revolution; (2) 1840–90, the introduction of the railways; (3) 1890– 1930, the introduction and production of electric power; (4) 1930–80, cheap oil and cars; and (5) information technology. 19 The idea that we are living in an age of dramatic technological progress is mainly hype; the reality is that we live in a time when the fundamental things are actually not changing very rapidly at all … The slightly depressing truth is that technology has been letting us down lately. (Krugman, 1998b, p.104) This may be the truth if one remembers the constant maintenance necessary for PCs. The only new good that altered life for people on a larger scale from the 1990s and beyond was the mobile phone. The Internet is the only major service that has done the same in the same period. However, there is a continuous stream of small improvements in already existing goods (and services). 20 While R&D and innovation activities in the US are mainly led by the military and double-use industries where demand is rather limited, these activities in other countries are more consumer-related: that is, they are directed towards the development of goods and services for which demand exists everywhere. In addition, R&D is chiefly mission-oriented towards big problems in countries such as the US, France and Britain. In countries such as Germany, Japan, Sweden or Switzerland, R&D is directed more, but not exclusively, towards the solution of practical problems. 21 Government involvement in R&D is usually on the state or federal level. Local authorities usually do not support R&D on a larger scale as the local tax money would be ‘wasted’ if the fruits of R&D leaked elsewhere in the country. Such concerns are not too worrying at higher levels of government. 22 China invested $136 billion in R&D in 2006. Even though this was second in the R&D expenditure ‘country league’ behind the leader, the US, which invested $330 billion, it was more than Japan, which invested $130 billion in the same year (G. Dyer, ‘China overtakes Japan on R&D’, Financial Times, 3 December 2006). 23 Foreign funding of R&D has continued to be an important source of financing. Belgium, Denmark, Hungary, Iceland and the Netherlands received more than 10 per cent of their

Evolutionary Economic Geography

270

R&D from abroad, while the same share in Austria, Britain and Greece was more than 15 per cent. 24 G. Parker, ‘Brussels points the finger at lax states’, Financial Times, 21 January 2004. 25 Annual university tuition fees for students are rather low in Switzerland (about Є1,000). This is significantly lower than is the case in Britain, Canada, Australia or the US. Switzerland is able to attract certain young and potentially productive talent to its universities. The easing of employment restrictions for EU nationals in Switzerland (and in the future those that come from outside the EU) may assist Switzerland to improve the structure of employment in the country. 26 European Council, Lisbon 2000, Presidency Conclusions, § 5. 27 A patent is granted to the inventor of something new, non-obvious, useful and potentially income-generating. Patent rights exist to protect an individual’s codifiable innovation and knowledge, whereas trademarks protect the reputation of a firm. Other knowledge that comes from learning through trial and error is largely unprotected because it may not be ‘blueprinted’. 28 Lead-time advantages over competitors was the most frequently used protection method in the EU (Eurostat, 2004, ‘Innovation output and barriers to innovation’, Statistics in Focus, Theme 9–1/2004, p. 3). 29 The importance of after-sales service varies depending on the good. This service is nonexistent for soaps and detergents, but it is essential for printing presses, online printers or photocopiers. 30 Self-satisfaction and a lack of real local competition contributed to the German cameraproducing industry being driven out by the Japanese (Porter, 1990a, p. 169). 31 Ford said this allegedly because black could dry faster than other available colours for vehicles. Hence, Ford could make more cars on the assembly line. The facts are that Ford used some thirty different black paints with different drying times. The colour was selected just because it was cheap and durable. 32 Sun Tzu (c. 544–496 BC) wrote ‘Hence to fight and conquer in all your battles is not supreme excellence; supreme excellence consists in breaking the enemy’s resistance without fighting’ (The Art of War [1910], part III. Attack by Stratagem, point 2; http://www.gutenberg.org/files/132/132.txt accessed on 8 January 2008). 33 For a discussion of this issue, see Lipsey et al. (2005). 34 The idea of innovation in the Muslim world was to restrain the pursuit of certain lines of research. The Prophet … is quoted as saying that ‘the worst things are those that are novelties, every novelty is an innovation, every innovation is an error and every error leads to Hell-fire’. In its extreme form this principle meant the rejection of every idea and amenity not known in Western Arabia in the time of Muhammad and his companions, and it has indeed been used by successive generations of ultra-conservatives to oppose tables, knives, sieves, coffee and tobacco, printing-presses and artillery, telephones, wireless, and votes for women. (Lewis, 1953, p. 52) 35 J. Wilsdon, ‘Islamic innovation is finally on a rising crescent, Financial Times,19 October 2007. 36 Suppose that labour market institutions provided opportunities to extract rents. If this were the case, and if these institutions were to be reformed (deregulated) in such a way as to reduce the size of these rents, then these institutions would become less attractive to labour. A prominent example of such a reform and the labour reaction to it occurred in France in 2003. The issue concerned performing artists. Anyone who worked for at least three months during a year could count on twelve months’ unemployment insurance coverage in the following year. Generous indeed. The consequence was a continuous and large deficit in this segment of unemployment insurance (covered by general payroll taxes). Nonetheless, the French government persisted with this system as a means of supporting culture. As always when there is a distribution of such generous ‘handouts’, certain abuses may take place. A

Market structure and location of production

271

proposal was made to modify the system slightly: basically, in order to qualify for unemployment insurance the person needed to work for four (instead of three) months. The reaction by the artists was robust: a long strike at the start of the summer, with the result that most summer festivals were cancelled. In spite of the strike and pressure, this ‘mild’ reform was accepted. In any case, this example reveals the resistance and difficulties encountered in the face of even modest reforms of labour market distortions and rigidities.

The political process has always and everywhere had a similar dynamics. Insiders always want to control the system for their own social, political and economic benefits against outsiders. A generous welfare benefits system in France creates a situation in which insiders have no reasons or incentives to voluntarily vote to give them up because of a possibility of losing them in uncertain circumstances in the future. The system creates and maintains a rigid and expensive labour market. Proposed changes in the law (First Employment Contract) had a noble theoretical aim in 2006: to make firms add staff without the obligation of a lifelong labour contract. This change in the law would affect the young under the age of twentysix. However, this provoked mass demonstrations in 2006. Outsiders to the system, such as students, the young unemployed and immigrants that had no chance or hope of becoming insiders, protested against these changes, which would allow employers to dismiss staff more easily during a two-year trial period. The young felt treated like the disposable ‘Kleenex generation’, used and then discarded. The legal change did not take effect. Alain Juppé lost his job as French prime minister in 1997 because of the reform of pensions that provoked a wave of strikes in 1995, the worst since 1968. The same thing happened earlier, to the first government of Silvio Berlusconi in Italy: pension reforms provoked strikes, so the government left office in 1994. Strikes, for example those that brought French cities to a virtual standstill in May and June 2003, illustrated trade unions’ determination to resist changes in the pension system. Similar reforms were also bitterly resisted in Germany, Italy and Austria. Nicolas Sarkozy got the electoral mandate as the French president in 2007 to reform the economy. He tried to reform the retirement system in the transport industry. The plan was to change the workers’ right to retire with full benefits after 37.5 years of work and to equate it with the 40 years that prevail in the rest of the French economy; in 2007 labour unions launched a strike that was the most severe for a decade. 37 ‘The new titans: a survey of the world economy’, The Economist, 16 September 2006, p. 33. 38 J. Kay, ‘The importance of history and geography’, Financial Times, 16 January 2006. 39 The settling of accounts by the establishment with new ideas, such as Anton Müller’s were, belong to past times. How lucky Brian Arthur is that he lives in our time and age. 40 China was nearly self-sufficient in resources over its several millennia-long history. This was a big difference between China and Europe, which was quite active in expansion of its powers. However, this is changing. Since the reforms and opening of the early 1980s, China has increased its needs and dependence on imported raw materials and oil, as well as foreign markets for its output. 41 One should not idealise Europe, however; its pass-through ratio for new ideas was initially quite low, but higher than elsewhere. This gave enough advantage to Europe to lead. 42 The era of imaginative individuals as principal sources of innovation was during the nineteenth century and earlier. To create a spacecraft that may fly to the moon and return

Evolutionary Economic Geography

272

safe and sound involves the work of a large team of experts. Nonetheless, individual creativity should not be discounted at all. 43 These irreversibilities include savings in factor (including energy) inputs per unit of output. 44 Economic growth may be propelled (among other elements) by the following three components (or by their combination): investment in capital and/or human resources, trade, and technological change. 45 The time it takes for national income per capita to double in the early stages of industrialisation has fallen dramatically. ‘When America and Britain were industrialising in the nineteenth century, they took fifty years to double their real incomes per head; today China is achieving the same feat in nine years’ (‘The new titans: a survey of the world economy’, The Economist, 16 September 2006, p. 4). The main reason is technological progress. Countries are able to purchase foreign technology to make their home factors more productive. There is also another dimension to new technology. While it was an increase in the physical capacity to influence the environment and to produce goods that brought about the Industrial Revolution in the nineteenth century, the current industrial revolution has a more qualitative dimension. Lipsey sounded a note of caution when he compared presentday England with that in the Elizabethan era:

It took 400 years for England to develop from that stage to its present one. To do the same elsewhere in half of the time of 200 years would be a tremendous achievement; to aspire to do it in 25 or 50 years may be to court disaster. (Lipsey, 1992a, p. 755)

46 The great majority of innovations did not come from formal R&D (even in organisations such as Du Pont … which had strong in-house R&D facilities). Most … came from production engineers, systems engineers, technicians, managers, maintenance personnel and of course production workers. (Freeman, 1994, p. 474, original emphasis) 47 Gordon Moore was the co-founder of Intel. 48 Government support of the Airbus provoked a sharp reaction from the US (which ignored the fact that US aircraft producers were generously subsidised through defence contracts). This led to the GATT Agreement on Trade in Civil Aircraft (1979). While ‘supporting’ the civil aircraft programmes, the signatories ‘shall seek to avoid adverse effects on trade in civil aircraft’ (Article 6.1). This is a statement that offers a number of different interpretations. (International agreements may be ambiguous. This allows diplomats to interpret them at home in their own favour. Conversely, constitutions and other domestic laws are (supposedly) transparent. This is because domestic politicians want the voters to understand them in order to win their votes.) As such, it was insufficient to calm down tensions in the aircraft trade. A Bilateral Agreement (1992) between the US and the EU was supposed to introduce a framework for all government ‘involvement’ in the development of commercial aircraft with a hundred or more seats. However, it did not take account of past damage. The deal did not eliminate, but merely constrained, subsidies (for innovation and R&D). It set quantitative limits on both direct and indirect (military) subsidies for the development of new aircraft. The permitted limit for the direct subsidy for the development cost of a new aircraft was set at 33 per cent. Identifiable benefits from indirect subsidies were limited to 4 per cent of each firm’s annual sales (Tyson, 1992, p.207). 49 The US unilaterally withdrew in 2004 from the 1992 Bilateral Agreement and initiated the WTO procedure concerning the alleged illegal ($15 billion) launch subsidies that Airbus had received from Britain, France, Germany and Spain since its creation in 1970. At the same time, the EU filed a counter-case claiming that Boeing was given $23 billion of illegal aid through R&D assistance and lucrative government contracts. While Boeing has to compete with other companies such as Lockheed-Martin for such favours, Airbus is receiving them directly from the EU member governments. An important reason behind this case and its

Market structure and location of production

273

timing is that Airbus overtook Boeing and captured 54 per cent of the global market for large aircraft in 2003. The US claims that this is due to subsidies. The WTO had to make a ruling on this case between two trading giants which both profited from subsidies. However, there was the possibility of a bilateral diplomatic solution of the issue (E. Alden and C. Daniel, ‘US files WTO case against Airbus subsidies’, Financial Times, 6 October 2004; E. Alden and R. Minder, ‘Airbus and Boeing pose huge test for EU–US relations’, Financial Times,6 October 2004; E. Alden and F. Williams, ‘US and EU ramp up dispute over aircraft subsidies’, Financial Times, 3 February 2006). 50 A large part of intra-industry trade is in parts and components. 51 See the surveys by Greenaway and Milner (1987) and Greenaway and Torstensson (1997). 52 ‘Americans tend to believe that we do everything better than anyone else. That belief makes it hard for us to learn from others’ (P. Krugman, ‘French family values’, New York Times, 29 July 2005). 53 ‘Monopolistic competition’ is an oxymoron. ‘Oligopolistic competition’ may be a more accurate term. 54 When people were undecided or uncertain about buying certain brand of computers, then the old saying ‘No one ever got fired for buying IBM’ from the mid1980s may have had a certain rationale. 55 This result was similar to the one reached in many studies on mergers in the US. 56 Jacquemin, 1990a, pp. 13–14; 1990b, p. 541; Jacquemin and Wright, 1993, p. 528; UNCTAD, 2000a, p. 138; Pautler, 2001, p. 53. 57 W. Hutton, ‘What Europe can teach Uncle Sam’, Guardian, 29 April 2002. 58 ‘Mergers, like marriages, fail without a meeting of minds’, Financial Times,14 May 2007. 59 ‘If, at first’, The Economist, 3 September 2005, pp. 13–14. [M&A = mergers and acquisitions] 60 ‘How to make mergers work’, The Economist, 9 January 1999, p. 13. 61 R. Tomkins, ‘As hostility towards America grows, will the world lose its appetite for CocaCola, McDonald’s and Nike?’, Financial Times, 27 March 2003, p. 13. 62 http://www.smith-wesson.com/ (accessed on 17 April 2007). The company’s earlier online catalogue, Crossings,offered items such as furniture, salsa bowls and bed and bath items, as well as ice buckets. 63 Europe has forty battery producers compared with five in the US, fifty tractor makers while America has four and sixteen firms building railway engines, whereas the US has two (‘Small, but perfect performing’, The Economist,23 January 1999, p. 67). Some firms think that they may prosper better if they become huge through mergers and acquisitions. 64 If competitive firms want to keep their lead, they need to follow developments not only in their own industry but also in unrelated, but potentially competing, ones. Examples of ‘learning-by-watching’ may be found in the disappearance, almost overnight, of the market for cine cameras after the appearance of video cameras, the seismic shift in the market for mechanical watches after the invention of digital ones, and the move from dot matrix to laser printers, or fibre-optics, that evolved independently of the telecommunication technology. 65 When there are significant scale and learning economies, risk-loving firms have incentives to set current prices not on the basis of current costs, but rather on the costs they expect to achieve in the future once full economies of scale are reached. 66 The first disc produced by Microsoft with Windows 2000 cost Microsoft about half a billion dollars. The second disc cost them three to four cents, just the cost of reproducing the first one. The cost of the third was three to four cents, and so on. Start-up costs are immensely high (Arthur, 2002, p. 2). 67 European Council, Lisbon 2000, Presidency Conclusions, §5. 68 See McAfee et al. (2004). 69 At the time of this case, Wanadoo was 72 per cent owned by France Telecom. 70 G. Parker, ‘EU states falling to open single market’, Financial Times, 18 July 2006.

Evolutionary Economic Geography

274

71 If a country such as Germany set higher standards than other EU countries, it could have carried on enforcing those standards with trade restrictions. The purpose of Decision is to prevent such behaviour. 72 ‘Warning of increase in VAT fraud’, European Voice, 20 November 1997, p. 7. 73 An example is cars produced by Nissan in Britain. France limited imports of Japanese cars to 3 per cent of the domestic market and claimed that the Nissan Bluebird cars produced in Britain were Japanese in origin. An earlier informal guideline stated that Japanese cars produced in the EU would receive preferential (domestic) treatment in the EU market if local EU content was at least 60 per cent. In 1988, France proposed an increase in this threshold to 80 per cent. Currently, in order to qualify for national treatment in the EU, the produced (assembled) cars must have 60 per cent of local (EU) content. 74 NAFTA offered Mexico substantial preferential treatment in exports of apparel to the partner countries, principally to the US. Mexico became a net exporter of apparel to the US (but not to the rest of the world) and a major importer of the US textiles (Lederman et al., 2005, pp. 12–13). 75 Canon invested over $100 million and located a new production facility for photocopiers in the US, rather than building it in China or Malaysia where production costs are much lower. The reason was a special NAFTA rule of origin for copiers which requires 80 per cent of local value added (UNCTAD, 1999b, p. 15). 76 The EU introduced a 16.5 per cent antidumping duty on imports of leather shoes from China in 2006. Chinese were furious as the European Commission compared prices with the ones in Brazil ‘because the EU does not recognise China as a market economy. The procedure is allowed under World Trade Organisation rules’ (A. Bounds, ‘China fights EU shoe duty’, Financial Times,20 December 2006). Lack of transparency and calculation issues tend to inflate dumping margins in the EU. A more detailed analysis of this issue may be found in Vermulst (2005). 77 Another complication arises from the practice of ‘zeroing’. Goods priced above home market prices are ignored for the calculation of average dumping margin. Therefore, the final antidumping duty is inflated. 78 The complicated maze of rules of origin was consolidated among the EU(15), ten EU candidate countries and the EFTA(4) in 1997. Different rules of origin that existed in various free trade deals among these countries were merged into the PANEURO. Value added in these countries could be cumulated in products without prejudicing their free trade status. 79 R. Minder and E. Alden, ‘EU and Canada retaliate against US imports’, Financial Times, 31 March 2005. 80 T. Kuchler, ‘EU imposes new trade sanctions on US’, EUobserver, 2 May 2006. 81 The directives give many different thresholds. For example, the threshold for public works is Є5 million; supply contracts for various goods and services all have different thresholds which are about Є0.2 million. There are, of course, loopholes. For example, certain public works may be split into a number of smaller ones that are below the limit. Another way of avoiding EU scrutiny is to underestimate the value of the procured goods and services.

The supply of military goods is excluded from rules on competition for national security reasons. This makes the military equipment market one of the most fiercely protected national markets in the EU. Its annual volume is about Є30 billion. A proposed voluntary code regarding this segmented market could reduce duplication of R&D and lower the high cost of equipment produced in relatively low series (G. Parker, ‘EU poised to increase competition in military contracts this year’, Financial Times, 29 March 2005). The defence equipment market is exempted from the Single Market rules in the EU (Article 296 of the Treaty of Rome). This market is worth about Є30 billion a year

Market structure and location of production

275

in the EU. Nonetheless, a voluntary code of conduct of 2005 for defence equipment sales is a step towards an introduction of some competition in this market. Defence purchases ought to be announced on the Internet and suppliers from the conduct contracting parties would have fair and equal treatment. Rationalisation and cost effectiveness is necessary as the total European demand for 10,000 armoured vehicles has been met by twenty-three different national projects (D. Dombey, ‘EU to open Є30bn military equipment sales’, Financial Times, 20 November 2005). 82 The Associazione exported to Britain more than 6 million packets of pre-sliced Parma ham in 2002 (P. Popham and S. Castle, ‘The “barmy Parma drama” is over’, Independent, 21 May 2003). 83 H. Williamson, ‘Forgery trade loses under $200bn’, Financial Times, 7 May 2007. 84 M. Wolf, ‘The profit motive may be universal but virtue is not’, Financial Times, 16 November 2005. 85 T. Buck, ‘Brussels plans compulsory “Made in EU” labels’, Financial Times,11 January 2004; J. Moules, ‘Minister rejects plan for “made in EU” mark’, Financial Times, 4 February 2004. 86 ‘Dutch window cleaners were outraged to discover that their standard ladders were too long to comply with EU health-and-safety regulations’ (‘Outgrowing the Union: a survey of the EU’, The Economist, 25 September 2004, p.15). 87 In 2001, non-harmonised products in trade accounted for about 21 per cent of the total tariff lines (WTO, 2004, p. 60). 88 It was the US demand to exclude agriculture from the agreement that resulted in the formation of the GATT. This resulted in a later backlash against the US in farm trade with the EU and Japan. 89 Old Article 115. 90 Between 1989 and 1991, the annual imports of Japanese cars into Italy more than doubled from 10,000 to 22,000 (D. Buchan, ‘Brussels gears up for end to import barriers’, Financial Times, 5 August 1991). 91 S. Druker, ‘America’s hypocrisy over modified produce’, Financial Times,18May 2004, p. 19. 92 Yes, proportional. This may be perfect in theory, but how to ascertain that in practice? 93 There was little that the European Commission could do against the French unilateral action. This was not because of the fear of protests in France, but because it would take up to two years to settle the dispute in the European Court of Justice. If the Commission cannot bring the ‘offenders’ into line fast and effectively, confidence in a border-free EU can be seriously damaged. Therefore, a change in the EU rules is necessary that would provide the Commission with the necessary instruments for a quick reaction (‘Crucial test case for the EU’s single market’, European Voice, 28 October 1999, p. 11).

5 International firms I pray that in all respects you may prosper and be in good health, just as your soul prospers. (3 John 2) 1 Introduction The spatial distribution of production is shaped not only by actions of national firms and governments, but also by decisions, customs and practices of foreign-owned (controlled) firms, as well as organisations that impose international rules. Some preferential trading or integration agreements such as common markets permit the free movement of factors among member countries on the condition that factors originating in partner countries are not subjected to discrimination. The promotion of geographical and sectoral factor mobility results in more efficient allocation of resources from the group’s standpoint. These improvements in the locational advantages of the group for business are due to the free internal factor flow from low-to highproductivity locations, businesses and professions and are most pronounced within the common market. Similar outcomes are also expected in a country that permits the location of TNCs. In this situation, factors respond to signals that include demand, higher productivity and higher returns within the group or a country. This chapter has five sections. It is devoted to the international mobility of capital and operations of TNCs. It considers ownership, i.e. foreign ownership and control of a company and its impact on the location of business in space. The chapter is structured as follows. General conceptual and theoretical considerations about foreign direct investment and TNCs are found in Section 2. The important issue of tax competition among various locations is also included in this section. Section 3 analyses globalisation. This is an important concept in our time and age which is often misunderstood, hence it may be easily abused. Trans-border business activities of TNCs, their determinants; linkages with the local economy, spillovers and externalities; transfer pricing; R&D; relations with host authorities; cultural differences; and corporate social responsibility are among the issues analysed in Section 4. Section 5 concludes the chapter with a comparison of costs and benefits, i.e. potential benefits and real concerns from the operation of TNCs in the host country.

International firms

277

2 Foreign direct investment and transnational corporations General considerations The theory of preferential trade explores the effect of integration on the location of production, structure of trade and changes in welfare. Little attention has been devoted to the geographical origin of ownership of firms. This gap is bridged by the theory of FDI, which studies locational advantages for investment in different countries and the competitive advantages of firms that originate in different countries, as well as the interaction between firms, local natural and government resources and capabilities in the contexts of spatial distribution of economic activities. Entrepreneurs view a country’s preferential trading and integration arrangements as long-term economic signals and hard facts, unlike changes in prices, which may reflect only a temporary situation on the market. Entrepreneurs and investors can form expectations with a higher degree of certainty. Hence, TNCs may locate a part of their production in such an expanding area and increase FDI by the creation of ‘tariff factories’. The creation of tariff factories within an integrated area is a strategy that TNCs pursue, not to take advantage of their efficiency or to employ a foreign resource (resource efficiency), but rather to benefit from (or avoid) the shield provided by the common external tariff and non-tariff barriers (NTBs). This could be one of the reasons why Japanese TNCs were eager to establish a presence in the EU prior to 1993 and the full implementation of the Single European Market. They wanted to become EU residents and thus circumvent the potential threat of a ‘Fortress Europe’. In addition to such an ‘investment creation effect’ (a strategic response of foreign firms to potential trade diversion), new prospects for improved business without tariffs and quotas on trade within the (protected) region may prompt local firms to spatially rearrange production facilities within the group. This may produce an ‘investment diversion effect’. This strategic response of firms to trade creation may have as its effect an increase in FDI in some countries in the group and a decrease in FDI in others. Capital moves among countries in the form of portfolio and direct investments. Portfolio investment is most often simply a short-term movement of claims that is speculative in nature. The main objectives include an increase in the value of assets and, perhaps, some safety. This type of capital mobility may be prompted by differences in interest rates. The recipient country will probably not wish to use these funds for investment in fixed assets and structures that must be repaid in the long term, so these movements of capital may be seen by the recipient country as hot, unstable and ‘bad’. Any firm that owns, has a lasting interest in or controls1 assets and structures in more than one country can be called a TNC. It is a wider concept than FDI since it includes non-equity business participation in another country. FDI is often the result of decisions by TNCs. Therefore, FDI may be a relatively good proxy for the investment activities of TNCs. A note of caution has to be added, however. TNCs may control trans-border business operations by non-equity involvement such as licensing. In the case of licensing, a TNC must be assured that the goods or services provided conform with the original quality standards.

Evolutionary Economic Geography

278

Asset bubbles appear when the main motive for purchase is not the good value asset, but because of rising prices. The expectation is to sell the asset soon to somebody else at a higher price. Sooner (or later) such expectations are upset because the value of an asset moves away from its fundamental worth. The ‘supply’ of new buyers dries up and bubbles ‘burst’. In the absence of bubbles, asset prices fluctuate around (uncertain) estimates about their base value. Volatility of portfolio investment complicates their analysis. The large number of portfolio investments, made in many cases by brokers, obscures who is doing what and why. Foreign direct investment reflects the goal of an entrepreneur from country A to acquire a lasting interest (including management and control) in an asset in country B. FDI is much more than plain capital. It includes technological, organisation, managerial, marketing and control knowledge and capabilities; and it also includes a network of widespread contacts that may be available only internally through TNCs. One should bear in mind that more than a third of world trade today is within the same company. In principle, FDI asks for freedom of establishment and, if possible, national treatment in foreign markets. This distinct type of international capital flow has a strong risk-taking and, often, industry-specific dimension. FDI is often the result of decisions by TNCs. Therefore, it may be a part of a proxy for the investment and location activities of TNCs (bearing in mind that TNCs may control operations abroad simply by issuing licences). International firm: theory Ownership and control of the firm was neglected in the analysis by an implicit assumption that these do not matter or by the supposition that all assets and structures are domestically owned. The presence of TNCs increases the mobility of capital, expands the availability of information and new products, widens marketing and trade networks, changes competition structure and alters substitution of labour for capital. A TNC has different locational considerations from a comparable national firm engaged in the same type and scale of activity. A tendency is that strictly national firms expand where they already are, while TNCs enter where they think they may profit from access to the largest and growing market,2 availability and favourable costs of inputs, transport and/or taxes and subsidies. TNCs have ‘organisational capital’–that is, a special culture, a common set of rules, practices, routines and values – which helps them overcome various spatial barriers through an internal network while operating in different geographical, social, cultural, legal and other environments. However, if they expand too fast and too wide, there is a chance that they may make a costly mistake somewhere along the road. There are a number of theories that explain certain aspects of why firms engage in trans-border business activities (locate abroad) and become TNCs.3 All of them are based on the existence of structural or transactional market failures. These include imperfections in goods and/or factor markets, product differentiation, economies of scale and asymmetric information. Let us consider these theories in turn. First, the motivation to control foreign firms may not come from the mere need or desire to employ assets and structures in a prudent way in foreign markets according to the market or technological efficiency criteria, but rather to remove competition from other enterprises. Hymer (1976) advocated such a market-power approach by TNCs.

International firms

279

Reuber (1973, pp. 133–4) argued in a similar vein that long-term strategic factors for FDI include the desire to eliminate competitors from foreign markets, to be within a protected foreign market, to secure a low-cost source of supply and to lock in the target country to a specific technology for a long time. Such a longer-term strategic view overshadows possible short-run variations in the profitability of FDI. The problem with this argument is that most of the TNCs (measured by their number) are small and medium-sized. There were at least 78,000 parent firms with about 780,000 foreign affiliates in 2006. Of these, about 58,000 parent TNCs were located in the developed countries, while the rest was based in the developing countries (20,000) and economies in transition (1,650) (UNCTAD, 2007, p. 12). This shows that to locate business abroad, hence to become a TNC, a firm need not be a monopolist or an oligopolist at home and try to exercise that power abroad. If there is strong competition in the market for differentiated goods and services and if there is a high degree of substitutability between products (perfumes, soaps, watches, clothing, vehicles, passenger air transport on certain lines, to mention just a few examples), then the market-power argument for the transnationalisation of business is weakened. Second, while the market-power model excludes potential rivals from competition, the internalisation theory holds that an arm’s-length relation among individual firms is in some cases less efficient (for example, trade in technology) than an intra-firm cooperative association. Intra-firm trade is trade between enterprises that belong to the same group, but are located in different countries. Profits may be maximised by means of an efficient and friendly intra-firm trade in intermediaries that eliminates sometimes excessive transaction costs (middlemen, exchange rate risk, infringement of intellectual property rights, bargaining costs) which occur when the business is conducted through the market. In these circumstances a hierarchical organisation (an enterprise) may better reward parties in the longer term, as well as curb bargaining and incentives to cheat, than markets and external contractors. Internalisation may expand until the cost of an additional transaction within a firm equals the cost of the same transaction on the open market. The problem is, however, that the larger the firm, the higher the costs of coordination and management; internal inflexibility may increase; and the labour unions may become too strong. So outsourcing certain operations or splitting them and allocation abroad may reintroduce flexibility within a firm. The share of intra-firm exports in the total exports of manufacturing affiliates under foreign control is in the range between 20 and 60 per cent in the OECD countries for which data are available. This proportion held steady at about 50 per cent throughout the 1990s and the beginning of 2000s in the US and the Netherlands. This proportion rose sharply in Sweden (from 45 to 60 per cent) and declined in Japan (from 45 to 20 per cent) during the same period (OECD, 2007, p. 182). In addition, more than a third of world trade in 2003 was within the same company.4 Payments of about 80 per cent of fees and royalties for technology ‘take place between parent firms and their foreign affiliates’ (UNCTAD, 1997, p. 20). Royalty receipts by the US TNCs were $35 billion in 2006 alone (Baily, 2007). This is an indication that TNCs play a key role in disseminating technology around the world, at least in the locations where they operate. These enormous intra-firm dealings across many countries require delicate and high-quality company governance. In certain cases this requires strong and unified lines of command;

Evolutionary Economic Geography

280

in others there is a need for decentralisation and flexibility. In any case, there is always at least a subtle overall company supervision and control. The importance and volume of intra-firm trade and various special alliances among firms indicate that markets are (becoming) incomplete. As such, it may be difficult to put these markets ‘in order’ according to free trade principles and objectives. Therefore, absolute advantages of a country or specific locations within it (clusters) may dominate and outweigh comparative advantages in their relevance for the location of production, FDI and, consequently, structure, volume and direction of trade. These absolute advantages among locations reflect in part great productivity gaps. These gaps among various locations are so huge and strongly locally embedded (in a cluster) that they become insensitive to almost any change in input prices anywhere else (e.g. London for finance services, Basle for pharmaceuticals). While Reuber and Hymer conceive TNCs as vehicles for reaping monopoly profits and for the internalisation of pecuniary externalities, the internalisation model looks at TNCs as a mode of business organisation that reduces transaction costs and internalises non-pecuniary externalities. This model of FDI may be convincing in some cases, but it may not explain the structure and location of all FDI flows. In addition to the internalisation possibilities, there ought to be ownership-specific and locational advantages for FDI. Feinberg and Kane (2005, pp. 247–8) found that American TNCs that are organised to have (large) intra-firm trade are considerably more dynamic technologically than TNCs with no such trade. TNCs with greater intra-firm trade are more dynamic technologically than the ones with no intra-firm trade. They have more intensive R&D operations in the parent country (US) and are also more likely to locate some part of R&D abroad. Affiliates that trade intra-firm are kept at the cutting edge of best technology in production, management, quality control and distribution. Affiliates that trade intra-firm experience higher real growth and have higher real wages than affiliates of firms with no intra-firm trade. This intra-TNC trade was found to persist over time in spite of both significant global reduction in tariffs and stabilisation in exchange rate fluctuations. As far as the location of a foreign affiliate is linked with export-oriented TNC strategy, this does not present a big problem for the country in which the affiliate is located. The problem may be if the affiliate is part of an import substitution strategy (protected infant industries that fail to grow up). Trade within a TNC may increase, but general trade relations and development of the host country may suffer. This may also be compounded by suboptimal technology mix if the host country requests a certain local content or has laws that request that the affiliate must have a majority domestic ownership. Excessive internalisation leads firms to diversify into unrelated technologies. This may provoke an increase in costs of production as they venture into businesses that are outside their distinctive core competence. These core capabilities or competences within a firm (unmatched and unmatchable by most competitors) are normally very few in number and individual in nature.5 Mastering new technologies that are marginal to the core business diverts time, efforts and funds from further specialisation in the core business competence. Therefore, buying from other specialised firms (outsourcing), rather than producing certain goods and services in-house, may be the preferred course of business action to achieve and maintain a greater degree of efficiency and competitiveness. Business strategists ought to ask whether a new activity extends the core competence of a

International firms

281

firm or not. If it does not, well, they and their investors and shareholders need to be leery about entering into this new business. Nonetheless, the evolutionary approach would suggest that they do not need to be negative about it from the outset. Markusen (1984) tried to integrate TNCs into general equilibrium trade models and locational patterns of firms that horizontally or vertically integrate across national borders. Intangible assets, as sources of multiplant economies, are often firm, rather than plant, specific. The intangibility of assets limits the extent to which they may be licensed out to independent firms without risk in losing control over quality. These assets include and spread over organisation, management, control, R&D, advertising, marketing and distribution. Many of these activities are centralised (finance) and present a ‘joint input’ across all production units. A TNC avoids replication of these activities that would be necessary if these units operated as independent national firms. This brings a ‘technical’ advantage to a TNC. However, if these advantages are transformed into market power, then the welfare effects of TNCs may not be clear. In any case, a firm will locate in two or more countries if trade costs are sufficiently higher relative to the disadvantage that comes from fixed costs linked to the operation of two or more plants that serve two or more local markets in different countries (Markusen and Venables, 2000, p. 221). Third, the eclectic paradigm (Dunning, 1988, pp. 42–5; 1999, pp. 1–3) explained the trans-border business activities of TNCs in terms of the following analytical framework. Trans-border business activities stem from a joint mix and interaction of three independent factors: • Ownership (O) (firm-specific) In order to locate production abroad and be commercially successful, a TNC must have or control internationally mobile incomegenerating ownership-specific advantages, assets, structures, capabilities or skills. Seen in evolutionary terms, the firm endogenously and actively generates, accumulates and perpetuates its own advantages; it does not passively respond to demand, costs, market imperfections or public incentives and restraints as assumed in neoclassical equilibrium theory. These firm-specific advantages include tangible and intangible advantages such as better technology, brand name, access to wide markets, monopoly, competence of managers to organise and control, ability to innovate, and so on, that are superior to the ones that are available to local firms (including other TNCs) in the potential target country. That is, a firm needs to operate either on a different production function from other firms or at a different point on the same function. In any case TNCs ought to have or control certain advantages over their competitors in the target location, otherwise they would be unlikely to start costly distant business with a disadvantage. • Location (L) (country-specific) Locational (non-mobile) advantages refer to the comparative or location-specific advantages of the target country. They refer both to the geographical distribution of resources and to those created by the government.6 In addition, potential positive spillover effects and benefits from an existing cluster of already settled firms might act as a locational magnet. • Internalisation (I) There must be opportunities for the internalisation of ownershipspecific advantages (management and quality control, protection of property rights, avoidance of uncertainty by buyers, and so on). It should be in the interest of the firm to transfer these advantages abroad within its own organisation, rather than sell the right to use them to other firms located in the country of intended production. Fixed

Evolutionary Economic Geography

282

exchange rates or a single international currency provide a degree of stability necessary for longer-term business planning with a high degree of confidence. The eclectic paradigm claims that the exact mix of the OLI factors facing each potential investor depends on the specific context. If a firm possesses or controls ownershipspecific advantages, then it may use licensing in order to penetrate foreign markets. If it has both ownership-specific and internalisation advantages, such an enterprise may use exports as a means of entering foreign markets. Only when a firm is able to take simultaneous return of OLI advantages does it employ FDI as a means of locating and operating in foreign markets.7 The OLI analytical framework does not apply to diversified and vertically integrated TNCs (Caves, 1996, p. 5). In addition, an emphasis is on the OLI factors: internalisation advantages and cost reduction postpone longer-term strategic decisions for FDI. Varieties of business operations abroad by TNCs (licensing, franchising or subcontracting) and certain downsizing of hierarchies through ‘outsourcing’ certain activities, question and ease some of the internalisation arguments. Fourth, the product-cycle model reasons that mature (and, perhaps, environmentally unsound) lines of production of goods (there is no explicit reference to services) are passed on to developing countries (Posner, 1961; Vernon, 1966).8 Earlier, the focus of academic research was on the impact of technological gaps on international trade, Posner and Vernon directed their attention towards production. According this model, the relation between trade and FDI has the spirit of substitution. Their proposition on spatial reallocation of production was based on the experience of Anglo-Saxon firms during the 1950s and 1960s and depended a lot on low factor costs and weak environmental standards in the target location. A snag with this model is that it is too hierarchical regarding the innovatory and technological capabilities of firms in different locations. Another problem is that this model may not be applied to products such as new software that is marketed throughout the world at the same time. The product-cycle argument as the major explanation for the location of business abroad, and with a rather vague timetable for this spread of production, cannot pass the test of recent developments. Asian firms, however, do not replicate this pattern of location of production abroad on a large scale. The Japanese auto companies invested at home and in the US, Europe and elsewhere at about the same time in a similar type of production. There is a heavy concentration of FDI in developed countries, while the majority of developing countries are relatively neglected in global FDI flows. In addition, countries start investing abroad at a much earlier stage of their development than before. The newly industrialised countries and many other developing countries are already investing abroad. In many cases these investments are in the developed world. Such developments may be prompted by the desire: • to be present in the developed countries’ markets (closer to wealthy customers); • to be near the source and cluster of the principal technological developments in manufacturing, distribution, management and already existing infrastructure (to have a foreign ‘listening and learning post’); • to participate in R&D programmes; • to avoid the dangers of protectionism in target countries; • to win public contracts; and

International firms

283

• to exploit the strength of the host country’s domestic currency. Fifth is the follow-your-leader (‘me-too’) hypothesis. Oligopolists are risk minimisers. They would like to protect their own market position and avoid destructive competition. Therefore, they typically try to minimise risk and follow each other into the new (foreign) markets (Knickerboker, 1973, p. 100). The technological gap between the home and targeted foreign location does not play a role in this model. An overview of timing of FDI by American TNCs in manufacturing seems to support this ‘snowball’ or ‘herd’ behaviour. The Japanese TNCs in automotive and consumer electronic industries were ‘following their domestic leader’ when they located their manufacturing facilities in the US and the EU during the 1980s and 1990s. This type of business location decisions was not limited to the Japanese vertical keiretsu (business group), but also included the horizontal ones. In addition, ‘French firms are attracted to regions where there are other French firms’ (Mucchielli and Puech, 2004, p. 55). Unfortunately this model of ‘bunching up’ firms does not say why the first company moves abroad. Seagate selected Singapore as the location for the transfer of production of hard disc drives in 1982. This was just three years after the founding of the firm in California. Singapore was chosen because low-cost manufacturing would more than make up for the high transport and communication costs with the US. The success of Seagate encouraged other American producers of hard drives (including Computer Memories and Tandon in 1983, Maxtor and MiniScribe in 1984, Microscien International in 1985, Micropolis in 1986, Control Data in 1987, Unysis and Western Digital in 1988) to move their production to Singapore. Seagate initially chose Singapore because the government offered tax incentives, because many local engineers spoke English, because there were local suppliers; and because American semiconductor TNCs such as Texas Instruments, Tandon and Hewlett Packard were located in Singapore from the late 1960s (Hiratsuka, 2006, p. 187). American companies were so successful in creating and producing hard discs that the US remained almost without any production of them on its own soil. The ‘herding’ foreign investment location behaviour is quite relevant during the opening phases of new markets. If a TNC has no previous experience in a certain new market, then the actions of competitors in that market may be quite informative. Examples of this herd behaviour are the actions of TNCs in China, especially after 1997 (the Asian financial crisis) and to an extent in eastern Europe from the start of the 1990s (when wages were low and when these countries had serious chances to join the EU).9 In addition, the mergers and acquisitions ‘mania’ of the 1990s in the EU and the US shows how asset-seeking by one TNC may be followed by others. The ‘catastrophic’ (total) relocation of an industry as a result of the predicted ‘snowball’ location behaviour of rival firms is one feature of the core– periphery model, one of the key economic geography models. It is postulated in theory that when trade costs lie in a certain range all firms in an industry may relocate en masse, although this would not be the optimum strategy for an individual firm were it to relocate on its own. While relatively low labour costs in China could have been the reason to start locating certain operations there at the start of the 1990s, things have changed. It was reported that China’s Pearl River delta attracted $1 billion of FDI a month. Microwave ovens are produced in Shunde, where just one of its giant factories produces 40 per cent of the global output. Shenzhen produces 70 per cent of the world’s photocopiers and 80 per cent of all artificial Christmas trees. Dongguan has 80,000 people working in a single factory

Evolutionary Economic Geography

284

making running shoes for the world’s teenagers. Flextronics is a Singapore electronics maker that produces for Microsoft, Motorola, Dell and Sony. The manager of its Chinese plant in Doumen said: ‘It is a myth that companies are coming here just for the cheap labour. It is the efficiency of the supply chain that drives them here as more and more of worldwide demand is consolidated in this area.’10 At the start of the twenty-first century, China may be an ideal location not only for many manufacturing industries that aim at global exports, but also because of its growing domestic market. Nonetheless, one ought to keep in mind that China is also moving into high(er) technology and design, particularly in its eastern and coastal region. With this in mind, many TNCs from the developed world opened Chinese innovation centres. The rise of China also has a dark side, because of domestic regional imbalances. Some 800 million peasants in the interior of the country are becoming noticeably irritated. After three decades of coastal development, the interior of the country was left behind economically. If nothing is done about it, this may provoke strong social tensions on a larger scale in the future. Sixth, the competitive international industry model for the location of business abroad refers to oligopolistic competition and rivalry within the same industry. This is basically exchange of threats (tit-for-tat strategy) regarding business moves by foreign rivals (Graham, 1978). Large firms keep an eye on the actions of their rivals: that is, they act strategically, pay attention to the likely reaction of their competitors to their own actions. What Texaco does in Europe, Shell will (try to) do in the US. Competition is not ‘cutthroat’, but rather ‘stable’ among several oligopolies. Other examples of this rivalistic trend include FDI in the manufacturing of cars and tyres or the supply of services such as hotels and advertising. SMEs such as a petrol station in the middle of nowhere may act independently in their business. However, SMEs in a cluster keep a vigilant eye on the actions of their competitors. Seventh is the diversification of portfolios model of foreign investment (Brainard and Tobin, 1992). This approach considers uncertainty. Fluctuations in the rates of return on capital invested in various countries introduce an element of risk. This inconvenience may be reduced by a diversification of portfolios. Eighth may be the strong currency argument. Firms from countries that have strong currencies are sources of FDI. These firms buy firms in countries with weak currencies (FDI hosts). This theoretical model of FDI neglects the costs of operations, available resources and labour relations in the target country. FDI may easily flow from countries with weak currencies towards the countries with strong currencies. Capital may wish to flyawayfrom instability. Firm-specific assets and exchange rates may be the ninth basic reason for FDI. Suppose that there is a target firm in the US with an innovation (a firm-specific asset) that can make the acquiring firm’s assembly line 10 per cent more productive (10 per cent more output for the same level of input). If a US firm wants to acquire the target firm, then a change in the rate of exchange makes no difference as its gains will continue to be denominated in dollars. If, however, the acquiring firm is from Japan, the gains will be denominated in yen. A depreciation of the dollar relative to the yen would increase the Japanese firm’s reservation bid, while the US firm’s bid would remain unchanged. If everything else remains the same, it is more likely that the Japanese firm will acquire the asset in this situation.

International firms

285

The actual evidence of this taking place is rather mixed. The price of US assets need not matter, only the rate of return: when the dollar depreciates, both the price of a US asset and its rate of return will go down. Blonigen (1997) found a connection between exchange rate movements (weak dollar) and higher levels of Japanese acquisitions in the US in industries which involve firm-specific assets. This refers to Japanese acquisitions in the US during 1975–92. However, in their analysis of relations between exchange rate movements and FDI flows from the US to twenty countries during the 1980–95 period, Chakrabarti and Scholnick (2002, p. 19) found that devaluation in the preceding year does not have a robust positive impact on FDI inflows. In addition to the above basic theories on why firms locate abroad, three other dimensions are relevant for coming to grips with the issue of such locational choices: cost minimisation, available technology and taxes. First, Kravis and Lipsey (1982, p. 222) argued that the location of foreign affiliates of TNCs is decided on the basis of cost minimisation. However, the intensity of this determinant varies from industry to industry. Second, Yamawaki (1993, pp. 19–20) did not dispute the importance of a relative difference in factor costs, but the availability of technology in the target country is an additional and equally important factor for the location of Japanese FDI in the EU. A Japanese TNC from a certain industry decided to locate in the EU country which has a certain advantage over other EU countries in the same industry. Britain is preferred by the Japanese TNCs for the location of production of cars and electrical/electronic equipment, Germany for precision instruments and machinery, Belgium for stone, glass and clay products, while TNCs from the chemical industry prefer Germany, the Netherlands, Spain and France. The third dimension refers to differences in taxes and fiscal incentives. This last dimension asks to be considered in a separate section. Taxes and fiscal incentives Tax authorities are aware of the possible impact that indirect taxes have both on the location of production and on trade in goods and services. Therefore, they introduced a safety device in the form of the destination and origin principle for taxation. This is of great importance to those countries that integrate. According to the principle of destination, taxes on goods are applied in the country of their consumption. This is the norm accepted in the WTO. According to the principle of origin, taxes apply in the country of their production. In equilibrium – ah, equilibrium again (its roots are deep indeed) – if taxes are harmonised among countries, they provide for the same total prices and quantities of output. There is, however, a different distribution of tax proceeds between big and small countries. Big countries prefer the origin principle, while small countries favour the destination principle. The destination or consumption principle states that consumption of all goods in one destination should be subject to the same tax, irrespective of the origin of their production. Local consumption is taxed and exports are tax exempt. This principle removes tax distortions on competition between goods on the consuming country’s market. The goods compete on equal tax conditions. This principle does not interfere with the location of production. It is widely accepted in international trade relations even though it requires the existence of fiscal frontiers among countries. However, if a country increases its tax rate, it may lose some capital to another country. The reason is that

Evolutionary Economic Geography

286

higher tax takings reduce the local demand. This makes the country in question a less attractive location for firms. The problem is that this principle may give the illusion that it stimulates exports and acts as a quasitariff on imports. The origin or production principle asserts that all goods produced in one country should be taxed in that country, despite the possibility that these goods may be exported or consumed at home. Local production is taxed and imports are tax exempt under this taxation principle. If the production tax on good X in country A is lower than the same tax on the same good in country B, then if exported at zero trade costs, good X produced in country A will have a tax advantage in country B’s market over country B’s homemade good, X. This introduces a distortion that interferes with the spatial location of production between the countries. Higher taxes equal higher local production costs. A harmonised rate of tax or zero taxation between countries is a necessary condition for spatial allocational neutrality. Even within a customs union or a common market, there may exist fiscal frontiers if the member countries accept the principle of destination. The fiscal authorities of each country should know where and when they are entitled to tax consumption of goods or services. The origin principle may have an advantage, for it does not require fiscal frontiers. This conserves scarce administrative resources. Taxes levied according to the destination and origin principles differ regarding their revenue impact. These two principles determine to which government the proceeds accrue. A full economic optimisation cannot be achieved if there are different tax rates levied on various goods. Suppose that country A levies VAT at the rate of 25 per cent on cars only, while the partner country B applies a uniform tax at the rate of 10 per cent on all goods. Suppose that both countries apply the destination principle for tax collection. In this case, the production in either country would be maximised because it would not be affected by the tax. Consumption would, however, be distorted. The relative consumer prices would be distorted because cars are dearer relative to clothes in country A than they are in country B. Consequently, country A’s consumers buy fewer cars and more clothes than they would do otherwise. The opposite tendency prevails in country B. In this case, trade between the two countries would not be optimal. Conversely, suppose now that the two countries collect taxes according to the origin principle. In this case, trade would be optimised as the relative consumer prices would be the same in each country. So, although trade is optimised, tax would still distort the maximisation of production. This is because producer prices, net of tax, would be reduced in a disproportionate way. Country A producers would be stimulated to produce clothes rather than cars, while the opposite tendency would prevail in country B. Once indirect tax is not levied at a uniform rate on all goods, the choice is between the destination principle, which maximises production but does not optimise trade, and the origin principle, which optimises trade but does not maximise production (Robson, 1987, pp. 122–3). The principle of destination offers an opportunity for tax evasion that is unavailable (if the records are not faked) with the origin principle. If taxes differ, then a consumer may be tempted to purchase a good in the state in which the relative tax burden is lower and consume it in the country where the tax burden is relatively higher. Consumers may easily purchase goods in one country and send or bring them to another one, or order these goods from abroad. This tax evasion depends on the differences in taxation, cost of

International firms

287

transport and cooperation of buyers and sellers who do not inform the tax authorities if they know that the objective of certain purchases is tax evasion. The revenue effect of a standard tax at a rate of 40 per cent in a country where tax evasion is widespread (for example, the ‘olive oil belt’ or Club Med countries of the EU) may be much smaller than the revenue impact of the same tax at a rate of 10 per cent in a country where tax evasion is not a common practice. The EU applies a hybrid system for VAT. Behrens et al. (2007, p. 272) report that taxes on firms’ transactions are levied under the origin principle, while taxes on consumers’ transactions are levied under the destination principle. As the latter method is perceived to increase administrative costs of tax collection (and increasing chances for tax evasion), the EU is proposing a full switch to the origin principle. This is a long and uncertain process, as the EU countries are at best lukewarm regarding harmonisation of their national tax systems. It always creates potential winners who push for changes and potential losers who try to prevent or slow them down. When countries are of different economic size, they usually disagree on tax principles. The reason is that big countries may raise more tax revenue under the origin principle. Being big, they have presumably more industries and diversified production, hence they can collect more tax revenue. Smaller countries may collect more tax revenue under the destination principle. They tax local consumption. Hence, slow – even glacial – moves towards tax harmonisation within a group of countries (without clear compensations) come as no surprise at all. Even if the principle is commonly agreed upon, the subsequent problem is to find accord on the tax rate. The tax system in the US relies on corporate and personal income taxes applied according to the origin principle. The tax authorities in the EU countries rely upon consumption taxes with the application of the destination principle. If the Europeans export goods to the US, they may have an advantage embodied in the difference in the tax systems. The US may contemplate the introduction of a border tax adjustment. This step may involve an addition to or reduction in the taxes already paid in Europe. The objective would be to keep competition in the US market on the same tax footing. While the origin principle does not involve visible border tax adjustment, the destination principle includes it to the full extent of the tax. The long-run effect of either principle is, however, the same. In general equilibrium, any short-run advantage to one country will be eliminated in the long run by changes in the rate of exchange and domestic prices (Johnson and Krauss, 1973, p. 241). The US is a net importer of manufactured goods from Asia and Europe. It is advantageous for the US to have these two exporting regions administer taxes on a destination, rather than origin, basis. This is correct in the short run. In the long run, general equilibrium, the operation of exchange rates and factor prices (Johnson–Krauss law), would presumably eliminate any short-run (dis)advantage to these countries. The new theory of trade and location of production, together with the evolutionary economic model, disputes the Johnson–Krauss argument. In a real situation with imperfect markets (increasing returns to scale, externalities, economies of learning and path dependence) and when the economic system is non-ergodic, once the production of, for example, aircraft or fast trains starts at a very restricted number of locations in the world, it perpetuates itself only at those few locations for a long time. There is no room in the world for even three producers of wide-body passenger aircraft. In this case, the

Evolutionary Economic Geography

288

exchange rate argument (Johnson–Krauss) is of little help for small and medium-sized countries. In addition, liberal capital mobility and relative stability of exchange rates reduce the value of the Johnson–Krauss argument. Both the destination and the origin principles are imperfect. The destination principle for taxation is able to accomplish efficiency in the location of production, but not efficiency in trade. The origin principle has an inverse effect. If taxes are applied according to this principle, then trade may be efficient, but the location of production may not be optimal. Tax competition Corporate income taxes present the price that a firm pays to make use of the host location goods and services. There are many ‘tax havens’ that attract quite a few firms (mainly from finance). There are also many countries that offer different fiscal incentives to firms in order to attract them. But, is this enough by itself to attract firms’ attention and eventually the location of their business in the given area? Are tax incentives rather more than ‘corporate welfare’? Should countries or regions enter into a ‘locations tournament’ in order to be spotted on the potential investors ‘radar screen’? Politicians usually use only three reasons to justify a local incentive programme. These are: jobs, jobs and jobs! Tax incentives to foreign investors may have quite a few forms. They include tax holidays, reduced corporate income tax rates, deduction of new investment from taxable income, reduced taxes on dividends and interest paid abroad, preferential treatment (lower customs duties) on imports, reduced social security contributions for the employees and, especially important in the developing countries, reduced taxation on hard currency earnings (export of goods or provision of services to foreign tourists). There is also a wide choice of possibilities for various types of grants and subsidies. Tax competition is not a recent policy invention at all. During the fourth century BC the eastern Mediterranean island of Rhodes was a commercial power that partially controlled certain sea routes, and it had a vibrant port. Rhodes charged a 2 per cent tax on all cargo (transit included) on all ships that entered its port. To escape this charge, Roman traders lobbied for the creation of a free port on the island of Delos. Once they got it, commerce moved away from Rhodes, which lost this harbour revenue. However, this tax competition among ports turned out to be costly to shippers. Rhodes used a part of its harbour proceeds to police the sea routes and to fight piracy. Short of revenue, Rhodes reduced policing of the sea and piracy increased significantly. Trade costs and risk increased (Schiff and Winters, 2003, p. 99). Relatively recent attention to tax competition was provoked by Britain when the government of Margaret Thatcher lowered the corporate tax rate from 52 per cent in 1982 to 35 per cent in 1986; as one of the key industrialised economies, Britain ‘forced’ other countries to follow suit. Differences in taxes on liquid fuels may influence the location of petrol stations and ‘fuel tourism’. One example may be found in the village of Martelange between Belgium and Luxembourg. One side of the main street is in Belgium, while the other is in low-tax Luxembourg. A long line of petrol stations is located on the Luxembourg side of the street, where vehicles line up to buy the cheapest fuel in the region. A similar situation is in Switzerland, where liquid fuel is cheaper than in the neighbouring countries. After a

International firms

289

customs post, there is usually a petrol station (even a line of them) as soon as one enters Switzerland by road. Switzerland has been arguing that tax competition (which springs from its deeply rooted federalism, subsidiarity and devolution) between its twentythree cantons has been one of the pillars of its economic success. Formerly backward and farming cantons (Schwytz and Zug) attracted foreign investors with local low taxes from 1950s. Other Swiss cantons had to control their spending and keep taxes low. This local tax competition attracted many TNCs to locate in Switzerland. For example, Britain, France and others in the EU were angry when American TNCs Colgate and Procter & Gamble decided to relocate their European headquarters to Geneva in 2006, and Kraft Foods to Zurich in 2007. The Swiss argued that this reallocation is not only tax-driven (8.5 per cent federal Swiss profits tax plus variable cantonal tax, which brings the grand total to 11–12.5 per cent; 12.5 per cent in Ireland and an average about 20 per cent in the EU), but also because of the general security and enviable quality of life in Switzerland.11 On the other side of the story, there are TNCs that are leaving Switzerland. Lego, for instance, relocated its manufacturing operations to the Czech Republic (EU) in 2006. This provoked no fury in the EU. Interesting. Smaller states such as Ireland or Switzerland may be able to afford to offer lower tax rates because they have fewer (large) corporations, so they want to attract them. Being furious with Switzerland and even threatening economic sanctions would not get the European Commission far. The reasons include both of the following. First, the domestic Swiss and foreign firms with operations abroad face the same tax regime. The Swiss are not favouring one firm over another. These firms (with business operations outside Switzerland) are using less of the Swiss infrastructure than firms that predominantly operate in Switzerland. Second, the EU member countries such as Ireland or those in central and eastern Europe are also playing the tax-cutting game in the globalised economy. Switzerland is a country with a small market, unfavourable terrain and a lack of natural resources. It has to offer firms something to lure them to locate and stay in the country. Relatively low taxes may be one of the advantages. The advantage in central and eastern Europe is found to an extent in relatively low wages. Italy and Spain offer southern sun for fruit and vegetable production (a lower energy bill). And for holidays. Nobody is asking for extra taxation on these ‘special’ advantages in these southern countries in order to ‘level the playing field’ with others in Europe. The evidence for the importance and influence of taxes on the location of firms is still controversial. Certain investors may avoid locations that offer financial incentives as evidence of a region’s non-competitiveness. Limited survey information reveals that fiscal variables matter little regarding business location. However, business executives often lobby hard for fiscal incentives. One can easily understand this attitude, as managers have no incentives to forgo such direct and indirect subsidies, even if they do not affect location decisions to a significant degree (Wasylenko, 1991). This was confirmation of the results of an earlier study by Carlton (1983, p. 447), which found that tax variables usually have a ‘very small and always statistically insignificant’ impact on locational choice.

Evolutionary Economic Geography

290

[T]he empirical results show that most development incentive programs have only marginally affected decisions concerning the reallocation of existing plants. Overall, the results of this study support previous findings that that the use of public funds for tax incentives to attract large industrial plants is not very effective. (Lee, 2007, p. 25) Mody and Srinivasan (1998, p. 795) found that corporate tax rates do not have a major influence over FDI either by the American or by the Japanese TNCs. Devereux and Griffith (1998) found that agglomeration effects were important for the US TNCs that invested in Europe, while the effective tax rates played a role in the choice between locations, but not in the decision of whether to locate a subsidiary in Europe. Moss Kanter (2003) argued that, contrary to common belief, the prime reason for FDI in South Carolina was not low wages and tax incentives, but rather the competence of its workforce. In another case in France, Crozet et al. (2004) found that foreign investors were not sensitive to investment incentives. In addition, French and EU regional policy investment incentives did not have any significant impact on the location of TNCs. A study of the effect of the elimination of a tax on paper for printing news (which accounted for about half of the production costs) in the Netherlands in 1869 revealed that this assisted in the creation of new newspaper firms. Two-thirds of these new entries took place in cities that already had other newspapers, while the other third occurred in provincial towns that previously did not have a newspaper. Even though a tax cut played an important role, the principal reason for this type of location of firms was the size and growth of the local market. Hence, there is no basic difference in the rationale for a firm’s location between a century and a half ago and now (Pfann and van Kranenburg, 2002). Many surveys give state and local taxes a low ranking on the list of location determinants. At the same time, many locations in various states within the US can be close substitutes. In such a situation, even a small difference in production costs can play a key role for a particular decision concerning the location of business (Bartik, 1991, p. 8). American firms became more sensitive to differences in host-country taxes towards the end of the 1984–92 period (Altshuler et al., 1998). Tax returns data from 500 US TNCs revealed that these firms were sensitive to the average effective tax rates regarding their choice of foreign location and the amount of capital invested there (Grubert and Mutti, 2000). Nonetheless, available infrastructure, stability and other factors are more important locational drivers in the target country that is the mere difference in taxation. As part of its policy to lower company taxes prevailing in the east of the EU, Austria reduced company taxes from 35 to 25 per cent in 2005. This strained relations between Austria and Germany, Bavaria in particular, because of the ‘exodus of jobs to Austria’.12 To counter the move, in March 2005 Germany proposed a cut in the company tax rate from 25 to 19 per cent.13 In addition to competition among firms, there is a specific type of competition among countries or administrative districts. Tax competition refers to the relative size of the tax burden (usually government revenue as a percentage of GDP) among different locations. Tax competition may be an effective tool to promote economic development. The twenty-three Swiss cantons traditionally have competing tax systems that ultimately

International firms

291

support economic expansion. Harmful tax competition exists, in general, when only select activities are promoted. This distorts the allocation of resources. In broad terms, tax competition lowers taxes,14 while tax harmonisation may make them rise. If left alone, tax competition may provoke a ‘race to the bottom’ in order to attract business to locate in specific places and countries. However, this scenario depends on the assumption that the firms care only about taxes when they select the location for their operations. In addition, low taxes may be too low, from both the business and the social standpoints. Business and the social infrastructure may be poor and unattractive. Hence, relatively higher tax locations may be preferred by firms to low tax areas. At the same time, authorities in these locations may safely presume that the majority of firms may not go elsewhere only because of high(er) local taxes. If there are benefits in the agglomeration of businesses and clustering of firms (strong functional intra-industry production links), then tax rates may be safely different between the centre (cluster) and the country’s periphery. Tax harmonisation (averaging) in this situation may harm both the centre and periphery. The centre would have to reduce taxes and, consequently, social and business services. The periphery would need to increase taxes and bankrupt many local firms. The tax base would shrink further. This is one of the reasons why tax harmonisation in the real world is rare, hard and lengthy. It is commonly assumed that competition between countries to attract TNCs to locate within their confines is driving down tax rates. However, this has a side-effect. Governments find it increasingly difficult to raise revenue from businesses. Hence, there is an intensified pressure to increase taxes on individuals in the same locations, particularly if these individuals are not mobile. Otherwise the social programmes may not continue. If there are agglomeration forces, then Baldwin and Krugman (2004) reversed the standard and commonly considered theoretical proposition. They argued that greater economic integration may lead to a ‘race to the top’, rather than a race to the bottom. Industrial concentration (clustering) creates ‘agglomeration rent’ that can be used to tax the mobile factor. The core region/country can charge tax at a higher rate15 than is then case in the periphery, without fear of losing capital. This is the reason why the big European countries have higher corporate taxes than smaller countries in Europe. It is a well-known fact that capital income has a high degree of international mobility. This has become more pronounced in the EU since the introduction of the euro in 1999. However, the taxation of income from savings is a sensitive issue. The European Commission would like to reduce ‘harmful’ tax competition among the EU member countries regarding income from non-resident savings. A withholding tax may find a certain rationale, as Germany16 and the Scandinavian states forgo quite a bit of tax revenue from this source since their wealthy citizens and institutions (pension funds) save in non-resident accounts, primarily in Luxembourg and Britain. However, an argument against this tax cites the logistical nightmare of administering it and distributing proceeds. In addition, it would give incentives to place holdings in non-EU financial centres such as the Channel Islands (UK), Switzerland and elsewhere. Jersey, Guernsey and the Isle of Man are not members of the EU. They have no obligation to align their fiscal system with that of the EU. Therefore, Luxembourg would never accept an accord which excludes the Channel Islands. A common tax of, for example, 20 per cent in the EU would mean an immediate capital flight from Luxembourg to the Channel Island banks.

Evolutionary Economic Geography

292

After a fourteen-year process of negotiation (horse-trading and blackmailing17) over the taxation of income from savings, the EU agreed on rules to crack down on tax fraud and tax evasion in 2003. According to this deal, twelve EU countries would start exchanging information on the savings of non-residents from mid-2005. Hence, each EU member country will be able to tax its citizens on non-resident savings in the EU. Austria, Belgium and Luxembourg were exempt from the obligation to exchange information because of the secrecy provisions in their banking systems, but they will have to levy a withholding tax instead. Three-quarters of these proceeds are to be transferred to the saver’s country of origin, while the rest will remain in the country in which the funds were saved to cover running costs. The tax rate will rise from 15 per cent in 2005 to 35 per cent in 2010. In order to prevent capital flight from the EU to ‘easier’ tax regimes such as those prevailing in Andorra, the Channel Islands, Liechtenstein, Monaco and Switzerland,18 these countries with ‘lighter’ tax regimes will have to either exchange information or impose a tax. Switzerland agreed to apply the same withholding tax, as sharing account information violates clients’ right to secrecy. However, if national authorities in the offshore countries do not apply monitoring rules with rigour, this may play into the hands of tax evaders, money launderers, arms traders, drug and human smugglers and terrorists who need to hide their money. If the elimination of tax evasion and money laundering are the principal policy objectives, it seems that cooperation, even some exchange and sharing of information on bank accounts among governments, may be the most effective policy tools. The existence of corporate taxes as they are now is quite resilient in public tax books. However, this may not last without changes. Countries that lead the way towards low taxes (as is the case in central and eastern Europe) experienced massive increases both in attracting FDI and in public revenue. But low taxes have a downside. If public revenue in a location is low, then this may have negative effects on the state of the local infrastructure, transport, education or utilities. Many firms with competitive output often look at paid local taxes as investments that enhance their competitiveness, rather than costs. At the end of the day, high taxes on firms hurt the ‘ordinary’ people they are believed to make happy. The firms may pass on these high taxes through lower wages to workers and/or higher prices to consumers. The success regarding taxation of capital in low-tax countries has been at best limited. Transfers of tax revenues to the countries in which the depositors originate were well below expectations. Capital is fleeing into Asia. Tax havens such as Singapore and Hong Kong are not covered by the EU legislation. These centres are sufficiently far away for the EU not to have a strong influence on them. They are, therefore, not very likely to cooperate with the EU lawmakers regarding exchange of information and transfers of tax proceeds. The status quo would suit them well: capital would continue to arrive and their home economies would continue to thrive. What can the EU countries do? Screening income from capital is less easy to achieve than monitoring local consumption. Hence, redressing the domestic tax system may turn towards consumption. European countries’ domestic taxes on capital may be eased, while taxation of local consumption may be increased. This is not a very winning strategy for elections. However, the alternative, of taxing global income earned from capital, looks quite bleak.

International firms

293

Conservatives argue that public spending ought to be checked. Hence tax competition introduces a degree of discipline on profligate governments. Those on the left side of the debate say that governments ought to restrain tax competition because this may undermine social programmes and the future of society. They also argue that firms that exploit a local natural resource or monopolists have to pay corporate taxes. However, nobody gives a clear answer to the question of who ultimately foots the tax bill. A popular perception is that the government has the right and means to use corporation tax in order to tax the owners of capital. However, internationally mobile shareholders may easily escape taxation or reduce a big chunk of their tax dues. In addition, they may pass on the tax burden to consumers through higher prices and to labour in the form of lower wages. If that is the case, then a more efficient solution (and an equally fair one) may be to tax both consumers and labour directly. Public authorities may tax subjects that cannot move, such as property and, to a certain extent, consumption and labour. The problem with this view is that if a government proposes such a tax strategy, it cannot count on staying in office long or winning the next election. Proponents of the flat tax argue that its simplicity is the key for efficiency and economic success. Critics claim that the flat tax (a single positive tax rate) can be unfair and that it does not have a significant impact on the economic activity. The business cycle may be better stabilised by progressive taxation, which is also perceived as a fairer income tax system. However, this conventional view was challenged from the early 1990s with the introduction of flat taxes in central and eastern Europe. The experience from Estonia, Georgia, Latvia, Lithuania, Romania, Russia, Slovakia and Ukraine does not provide evidence that the flat tax weakens automatic stabilisers or that it is inequitable. The ‘danger’ for the Europeans is not reallocation of industries within Europe, but reallocation of business elsewhere (e.g. China). The issue of the impact of taxes on the location of TNCs continues to be a controversial research area. There is a lack of consensus, as results cited in the academic literature are still mixed. High(er) taxes may be preferred to low taxes by firms if the tax proceeds are used to finance local services and infrastructure useful for the business sector. In any case, high-quality infrastructure is preferred by foreign investors to tax incentives. Transfer pricing and tax deductions in the home country provide other ways to minimise the tax burden on profit (Wheeler and Mody, 1992, pp. 71–2). The issue of tax competition and incentives for the location of FDI is open. The relevance of taxes and subsidies to the attraction of FDI depends on the objectives of TNCs. If the goal of a TNC is to serve the market of the target country, then these concessions do not matter for FDI in that country. If a TNC intends to serve the international market, all things being equal, various fiscal allowances do play a role in attracting footloose industries. However, a much superior policy to a special tax treatment has always been to ensure better domestic infrastructure and continuously improve the supply of domestic human capital through health, education and training. This could enhance the domestic absorptive capacity of novelty brought by TNCs as the principal vehicle to transfer potential spillovers to the local economy. ‘The economic desirability of locational incentives is not clear, particularly if they detract from building competitive capabilities and encourage bidding wars’ (UNCTAD, 2003, p. 126). Tax incentives for investment, in particular FDI, are conventionally not recommended. This is also the stance taken by the World Bank, the IMF and other

Evolutionary Economic Geography

294

international bodies that advise on tax matters. Tax incentives are bad in theory and bad in practice. They are bad in theory as they introduce distortions. Investment decisions by entrepreneurs are made differently from the situation where there is no special tax stimulus. They are bad in practice because of their ineffectiveness: tax considerations are rarely the principal determinant for the location of FDI. They are also inefficient, as their cost may well exceed any benefit that they may bring.19 They are difficult to administer, as there is a lack of transparency, and they are subject to abuse and corruption20 by the ‘old boy network’. Finally, they are not equitable as they benefit certain investors, but not others (Easson, 2001, p. 266). If various tax privileges are offered only to foreigners, then there is a possibility of ‘round-tripping’. Domestic capital leaves the country in order to return in the guise of FDI, so as to profitfrom better treatment in the home country. This has been the case in both Russia and China from the early 1990s. Among the leading ‘foreign investors’ in China are Hong Kong, the British Virgin Islands, Western Samoa and the Cayman Islands. In these cases, recorded FDI is not so ‘foreign’. More than a hundred countries worldwide offer tax incentives for FDI. The type of incentive that is most commonly employed is the tax holiday, which is the worst in almost every respect. Administration of these incentives is amazingly complicated; there are opportunities for abuse and avoidance; they may attract only short-term FDI to benefit from the tax holiday while it lasts; and (particularly in the developing countries) they are often beyond the capacity of tax administrations to manage and monitor. Therefore it is not surprising that tax incentives are often inefficient and ineffective (Easson, 2001, p. 375; Tanzi and Zee, 2000, p. 316). Tax holidays do not matter much for the first-time location of a TNC in a badly governed country. In spite of the above arguments, tax considerations have recently become an increasingly important factor for the spatial location of investment and business. Why is this so? Why are tax incentives becoming more and more generous? Tax considerations do not feature highly in the initial strategic decision by TNCs of whether to invest abroad or not. However, after the decision to go and locate business abroad is reached, differences in taxes between regions in the target country or differences among countries tend to play a significant role. Traditional analysis generally assumed (1) that TNCs did not base locational decisions on tax considerations and (2) that there was little competition between locations in the developed and developing world. Both of these assumptions are challenged by more recent research, which found that TNCs are becoming more responsive to locational incentives and that competition between locations in the developed and the developing countries is increasing. The responsiveness by TNCs to locational incentives is also on the increase. The developing countries may not normally match tax incentives offered in the developed countries. Instead of offering these incentives, developing countries may better use those scarce resources to improve their infrastructure, property rights, education, effectiveness of the judiciary and information-gathering and dissemination, which may benefit both TNCs and domestic firms alike (Moran et al., 2005a, p. 382). Focusing on tax incentives (corporate welfare) diverts public attention and resources away from the real problem of making serious longer-term policy reform. What makes things worse is that the lost tax revenue makes it even more difficult to pay for the profound reforms necessary for economic development.

International firms

295

Even though corporate taxation has certain impact on locational decisions, if one focuses only on tax-related matters in home and host countries and puts aside other matters and policies one may seriously overestimate and distort the idea of the effect of taxes on locational decisions. This may have an unhelpful and misleading impact and relevance on economic policy. On the whole, taxation would seem to be a relatively minor factor affecting the locational choices of multinational enterprises as compared to policies affecting the ease of entry for foreign firms, their labour costs and the functioning of product markets in the host country. (Hajkova et al., 2006, p. 6). Large local markets such as the one in China attract TNCs in spite of shabby formal institutions and regardless of local corporate taxes. Simply put, location decisions by TNCs are complex. The theoretical potential of a location to attract TNCs is one thing; its actual performance is quite another. Similarities and differences between China and India may present a good example regarding the attraction of FDI. In theory, the optimal solution to the issue of enticements of FDI may include an international agreement among countries to eliminate all tax incentives for investment (or to limit them in a uniform way). In the absence of such an agreement, and according to the prisoner’s dilemma concept, few countries would risk acting unilaterally in this subsidy-elimination way. Recap In the situation with market imperfections, there is no single theory or model that can provide a completely satisfactory answer to each issue related to the trans-border investment activities of firms. Motives for foreign investment and production are different, complex, multifaceted, varying over time and not always consistent. They alter and evolve over time in response to changes in the market, technology, needs, tastes, sophistication of players, management strategy, actions of rivals, economic policy and unpredicted events. However, if taken together, the surveyed theories may provide useful elements for an understanding of the issue. 3 Globalisation Background Globalisation is a great economic and political story of our times. There were once hopes that globalisation would benefit everyone. As time passes, globalisation’s downside becomes more and more apparent. There was a hope and expectation, fuelled by neoclassical equilibrium theory, that money would go to the developing countries, as the rate of return was higher there than is the case in the developed world. What we are witnessing is a flow in the opposite direction. Joseph Stiglitz wrote that ‘Globalisation seems to have unified so much of the world against it, perhaps because there appear to be

Evolutionary Economic Geography

296

so many losers and so few winners … Well-managed globalisation can make everyone, or at least most, better off. This has not happened.’21 Liberalisation in the national and international economy is a policy choice of governments, primarily in the developed world. It is linked with privatisation and downsizing of the activities of the public sector and the expansion of the activities of the private sector.22 Globalisation of the economy and production is a fact. It is the outcome of the behaviour of firms (TNCs); their organisation; takeovers; changing technology that allowed fragmentation in production and distribution; control and finance; and economies of scale. In part, it is also the consequence of a change in the behaviour of consumers (fickleness and declining loyalty to national producers and certain national products)23 and liberalisation of national and international economies for trade, production and finance. As a process primarily driven by technology and the actions of TNCs (power is shifted from states to firms),24 globalisation lacks two important components: transparency and accountability. Many are suspicious about corporations and their increasing power over everybody’s life. An obvious example is the influence of large pharmaceutical TNCs on governments and the WTO (trade-related intellectual property rights). The process of globalisation deals with the change in the geography of (integrated international) production and consumption as it reduces the importance of spatial proximity to inputs or markets. It widens boundaries and deepens space for the geographical location of production and consumption because of the declining costs of getting goods and services to the market. A rapid expansion of FDI is the key component of this process. Capital market liberalisation and increased capital mobility have radically reduced the influence of governments in the monetary sphere. However, governments have gained increased control in other areas. For example, computers and information technology have greatly increased potential for data collection and processing, and consequently control over firms and citizens, which is relevant for tax and other purposes. Porter (1994) succinctly explained the role of location in global competition. First he mentioned that, for the most part, earlier research showed that ‘locational choices were more of an operational detail than strategic’ (p. 35); that there is a weakening link between location and input cost minimisation; and that ‘in a world of global competition … location is no longer relevant’ (p. 35). As a reaction to a such state of affairs, Porter immediately retorted that ‘this line of reasoning about location, however, does not square with the empirical evidence’ (p. 35). A striking and persistent difference in the national economic performance or the operation of regions, even cities, indicates the importance of location. Another sign of the significance of location is spatial concentration of firms. Even though TNCs fragment and disperse their business worldwide, their most sophisticated activities are usually located and concentrated in their home country. Regionalisation is increasing in importance in the world economy. This was exemplified in the inability of 135 member countries of the WTO to agree even on the agenda for the ‘Millennium Round’ of global trade negotiations that took place in Seattle in 1999. Well-organised, vociferous and strong worldwide protests by environmentalists, farmers, enemies of genetically modified food and big business, human rights activists, labour unions, anti-capitalists and animal rights activists exacerbated the problems of that meeting and many others that followed throughout the world.

International firms

297

While international trade negotiations have taken place in the penumbra of obscurity since Second World War, this will not be the case after Seattle. There are deep divisions about the nature and extent of globalisation as shaped and conducted by the perceived instruments of global control, such as the WTO, the World Bank, the IMF and TNCs. Therefore, the creation of integrated global economic policies is likely to be a very hard task for quite some time to come. It seems that the bilateral and regional approach to economic problems will call the tune at least in the medium term. This is obvious following the indefinite suspension of negotiations under the auspices of the WTO Doha round from July 2006. Hence, in the new situation, special, strong and sometimes dangerous interests may get next to free rein. Gone are the days when liberal (free) trade on multilateral bases was the principal game in town. The existence and importance of global international institutions such as the WTO, the World Bank and the IMF is not self-evident. The liberal phase of the world economy in the period 1850–1914 operated smoothly without comparable institutions. The prevailing gold standard was not created by an international agreement. Why were these global institutions created after the Second World War? The answer to this question relies on four overlapping matters (Wolf, 2004, pp. 73–4): • Logically: struggle for dominance can lead to war. An international regime may be a solution to this problem. • Historically: as the dominant power after the Second World War, the US preferred an articulated constitutional order. The US had a passion at that time for formal international institutions. Britain preferred informal agreements during its period of hegemony prior to the First World War. • Politically: it may be hard to convince the public of the virtues of unilateral action in international trade and investment. This may look to many people like unilateral disarmament; hence an international framework may be necessary. • Economically: if an economic action is simultaneous among countries, the gains are multiplied. An international legal framework may provide greater security for all. The biggest gainers are the countries that previously had a shoddy reputation. In spite of their high profile, many global institutions are losing influence and relevance. Some of them may be losing even their purpose. Asian countries are dissatisfied with the IMF’s voting structure, which favours Europe and the US without taking into account the huge and growing reserves in Asia. The Latin American countries of Argentina and Brazil have been among the biggest clients of the Fund. They repay IMF loans as soon as possible to get rid of disastrous policies introduced by IMF’s conditionality. In addition, Horst Köhler and Rodrigo Rato, the latest two heads of the IMF, gave up their positions prematurely. That has never been a good sign. Financial markets are replacing the World Bank in project financing. The WTO is on a margin because of a long stall in the Doha Round trade liberalisation negotiations. And many have a problem remembering why the International Labour Organisation exists. At the San Francisco conference in 1945, Harry Truman stated as US policy that ‘We have all to recognise, no matter how great our strength, that we must deny ourselves the licence to do always as we please.’ By contrast, a central theme of George W. Bush’s re-election campaign in

Evolutionary Economic Geography

298

2004 was that the US would not ask for a ‘permission slip’ before taking decisions. (Ahearne et al., 2006, p. 10, emphasis in original) Certain global institutions might have exhausted their original raisons for existence, other institutions might have been supplanted by the private sector, and the motives for yet other organisations have changed, but the overall basis for the existence of the UN is more relevant than ever. Without the UN, the ideal of the world based on rules would disappear. The rapid international expansion of TNCs made them the most visible feature of globalisation. Being foreign, sometimes making visible and highly publicised mistakes and often being big in a small country, TNCs are easy targets for non-governmental organisations (NGOs), which can run successful campaigns to disgrace a mighty TNC. A number of NGOs are relatively small players who would like to ‘punch above their weight’ because of the urgency of the matter.25 Examples of their campaigns include valid targets such as land mines, HIV/AIDS or poverty and third-world debt. NGOs increased public awareness and pushed through agreements on the control of ‘greenhouse gases’ in the United Nations Conference on Environment and Development, the ‘Earth Summit’ in Rio de Janeiro in 1992. They also helped in torpedoing the Multilateral Agreement on Investment in 1998, and created much ado in Seattle in 1999 (and many other subsequent meetings), which contributed to the failure to start a new WTO round of global trade negotiations before the meeting in Doha in 2001. No matter how justified the reasons for the campaign, these ‘civil society’ protestors against global capitalism demonstrated that the tide of globalisation could be checked and even turned back. The anti-globalisation movement is moving beyond its radical, protestdriven, even anarchist legacy (which might have worked against it) towards a movement with a concrete agenda. Rock-throwers, like the Luddites (1811–16) who destroyed factory machines, were ineffectual and irrelevant in the long term. The strategy of the antiglobalisation movement is to address four weaknesses of the globalisation process: • a lack of legitimacy, • a lack of accountability,26 • a lack of organisation, and • a lack of transparency. The idea is not to be against globalisation across the board. Globalisation may contribute to the maximum viable economic activity that may create resources necessary to achieve other social goals. The idea is to put limits on it. The question is what kind of ‘globalisation’ is desirable and how to achieve it. The World Social Forum (Porto Alegre, Brazil) has made valiant attempts to put hundreds of disparate (leftist) groups under one umbrella. Nonetheless, a unified ‘global left’ is far away. These groups were generally against: poverty, laissez-faire capitalism, genetically modified food, war and the US. They all use global communication networks to push for their particular social cause. They claim that the Davos forum of the global business and financial elite has unilateral power over investment and economic policies that affect hundreds of millions of people. Another claim is that the Davos forum either ignored or made worse global social problems. Global economic and financial institutions

International firms

299

such as the World Bank, the IMF and the WTO generally operate without democratic constraints.27 In this world, big TNCs may threaten national governments with disinvestment in order to pressure them to introduce policies that favour these firms. The Porto Alegre forum recommends a range of activities, from limiting the power of TNCs to adding new responsibilities and regulatory powers to established international organisations such as the UN, the International Labour Organisation and the World Health Organisation. In this way the anti-globalisation movement may be able to gain more leverage on the national and international scene. A popular backlash against globalisation in rich countries may support these objectives. Public opinion in the developed countries is in favour of higher taxation of the rich and a general cap on the exorbitant pay of corporate executives. The anti-globalisation campaigners have shown that governments are not powerless. The authorities can as easily dismantle old trade and investment barriers as they can introduce new ones. New technology, in particular the Internet, telecommunications, computing, data processing and fragmentation of the production process can offer some of the greatest economic opportunities ever for increasing living standards in all countries. Governments and the national elite in all countries (through incompetence or indifference) have failed to explain this. However, the process needs to be coupled with balanced policies both in the rich and in the poor world. Even though global economic integration may be the best end point for the future of the world economy for the proponents of globalisation, it is more likely that other (regional) outcomes may evolve or be chosen in the future. Globalisation may be favourable for economic efficiency including standardisation, but it can be harmful for social goals and for widening of consumers’ choice. The pursuit of efficiency and more efficiency in production and conquest of markets where ‘everything is permitted’ has as its obvious consequence the destruction of balance in nature and the environment, as well as urban sprawl. Deprived of access to natural livelihood, these developments may in extreme cases even contribute to trade in human beings and/or their organs. Definition The vogue term ‘globalisation’ has not yet been well or clearly defined. Many people have an opinion about globalisation and argue about it, but without a clear idea what it actually means. It is being used to describe almost all aspects of the present capitalist era of the world economy. As such it may mean different things to different people. Hence, this fuzzy, contested and controversial, but powerful metaphor is overused, often abused and very often misleading. A relatively widely referred description of globalisation stated that: Globalisation refers to the multiplicity of linkages and interconnections between the states and societies which make up the modern world system. It describes the process by which events, decisions, and activities in one part of the world can come to have significant consequences for individuals and communities in quite distant parts of the globe. Globalisation has two distinct dimensions: scope (or stretching) and

Evolutionary Economic Geography

300

intensity (or deepening). On the one hand it defines a set of processes which embrace most of the globe or which operate world-wide; the concept therefore has a spatial connotation. Politics and other social activities are becoming stretched across the globe. On the other hand it also implies an intensification in the levels of interaction, interconnectedness or interdependence between the states and societies which constitute the world community. Accordingly, alongside the stretching goes a deepening of global processes. (McGrew, 1992, p. 23) The problem with this vision is that it does not mention TNCs – as if the business community is outside this process. Strobe Talbot, the US Deputy Secretary of State (1994–2001) during the Clinton administration, discussed the issue of globalisation in the following way: Here is one optimist’s reason for believing unity will prevail over disunity, integration over disintegration. In fact, I’ll bet that within the next hundred years (I’m giving the world time for setbacks and myself time to be out of the betting game, just in case I lose this one), nationhood as we know it will be obsolete; all states will recognize a single, global authority. A phrase briefly fashionable in the mid-twentieth century – ‘citizen of the world’–will have assumed real meaning by the end of the twenty-first century. All countries are basically social arrangements, accommodations to changing circumstances. No matter how permanent and even sacred they may seem at any one time, in fact they are all artificial and temporary. Through the ages, there has been an overall trend toward larger units claiming sovereignty and paradoxically, a gradual diminution of how much true sovereignty any one country actually has … Globalization has also contributed to the spread of terrorism, drug trafficking, AIDS and environmental degradation. But because those threats are more than any one nation can cope with on its own, they constitute an incentive for international cooperation.28 This is a long-term vision of the global American-style melting pot. At the end of the day, a single global and unified ‘grey’ standard would prevail throughout the world. One wonders if there will be any room for some differences. Talbot’s global-type claim coincided with the time when (or just after) countries such as Czechoslovakia, the Soviet Union and Yugoslavia dissolved. Whether Talbot’s prediction will be realised at the end of the twenty-first century, nobody knows. However, there is now a strong resistance to such an outcome, at least in the Middle East. Muslims, in addition, have their own and virile global expansionist agenda29 through migrations, demography, unwillingness to integrate in the host country and, some would even argue, through terrorism. Some also claim that Islam is ‘a political movement masquerading as a religion’.30 For some economists globalisation basically refers to the choices and strategies, as well as the shape, direction and significance of activities, of TNCs. ‘Globalisation has

International firms

301

been defined in business schools as the production and distribution of products and services of a homogeneous type and quality on a worldwide basis. Simply put – providing the same output to countries everywhere’ (Rugman and Hodgetts, 2001, p. 333). Others think of it as a liberal system for trade, investment and integration of the world economy (a move away from segmented national markets). For certain political scientists, globalisation means a challenge to national authority and its handover to supranational authorities. For sociologists, globalisation may mean a creation of global social interaction that provokes adjustments that challenge local social structures. This may marginalise certain social groups that may violently resist such a change. ‘Thanks’ to al-Qaida, many realised that globalisation goes well beyond links that bind TNCs, producers, traders and bankers. For these, globalisation is linked with and invigorated by new technologies in communications and information processing. It is a sum of techniques that are at the disposal of private players and states. Yet for others, globalisation is an incentive to the reform process in economic strategy in many countries, as outwardlooking economic models replace inward-looking and TNC-hostile economic policies. So globalisation is openness to trade and investment with foreign and more and more geographically distant countries. Others look at globalisation as a process that alters interactions among agents across space; still others equate globalisation with economic integration.31 Anne Krueger defined it as ‘the increasingly rapid exchange of ideas, people, and goods made possible by falling transport costs and technological advances, all leading to the closer integration of the world including – but not limited to – the economy’.32 A fall in shipping costs made globalisation possible in part. Those costs fell so much that ‘it often now costs more to ship a container by road 100 miles from a port to its final destination than it does to move the container by sea from China to Europe’.33 Apart from a partial integration of international production, globalisation brings risks and disruptions. Volatile capital flows, speculative attacks on currencies, financial crises and unpredictable reallocation of jobs are obvious examples of the increased economic and social vulnerability of many countries, in particular in the developing world. To wrap up the issue, Henry Kissinger called globalisation ‘another name for the dominant role of the United States’.34 Martin Wolf confirmed such a view: ‘Globalization is not inevitable. It depends on politics. In today’s world, it depends above all on US politics’ (Wolf, 2005, p. 8). Uniformity and homogeneity in the modern world may be a heavy price to pay for new or ‘better’ standardised things that we consume. This introduces an ever-present potential for the neo-communist risk that everyone eats and drinks the same, is dressed the same, shops in the same way, uses homogeneous (perfectly substitutable) goods and services and finally may even – or is forced to, perhaps – think ‘the same’. It colonises by annexing not territory but the thinking and behaviour of a multiplicity of policymakers at a variety of scales. So there are no formal imperial institutions, merely a shifting constellation of corporations, border-crossing networks, and territorially-defined political units representing, or at least ruling over, distinct communities. (Lovering, 2006, p. 222)

Evolutionary Economic Geography

302

If someone is not ‘in step’ with these global developments (regardless of their pseudodemocratic or legal appearances and sugar-coating), well, then … he or she may court trouble from the central globalisation ‘politburo’.

For and against globalisation Proponents of globalisation, on the one hand, point at benign effects of globalisation: increased economic opportunities and growth, and a decrease in inequality and poverty around the world. They say that globalisation as a new phenomenon cannot be compared with previous phases of capitalist development: this one has nothing to do with the past. Income disparities are acceptable if they are merited and if they support a general economic progress. Labour has to pass through continuous adjustments, as – according to this school – it is better to have a mobile than an equal society. This group announces the ‘death of distance’, the ‘death of national economic (particularly trade, industrial and monetary) policy’ and the ‘collapse of the nation-state’. This group emphasises the central role of TNCs in the globalisation process as they shape the international geography of production with their decisions to invest in certain locations or to leave them. They produce global, i.e. standardised (identical) goods and services for consumers all around the world. Production and consumption in national economies become detached as they are integrated in a global process. Sceptics, on the other hand, refer to the socioeconomic costs of globalisation: increasing income gaps and inequality around the world, uneven and unfair geographical and social distribution of gains (more than a billion people live on less than a dollar a day and many of them have never made a phone call), social tensions, environmental degradation and growing intolerance towards political diversity (many want to defend their right to be wrong). They argue that potential gains of the current phase of globalisation are largely overblown. Local forces are still strong and play a significant, useful and meaningful role. The task and influence of the state is not diminishing: it is evolving alongside globalisation and it is changing its role and authority. It is true that the behaviour and actions of TNCs shape globalisation to a large extent, but investments and sales of TNCs are mainly concentrated in specific geographical areas (principally the developed world). To operate successfully, TNCs need certain preconditions, including appropriate infrastructure and educated labour. They also need the security of property and person, which is normally provided by the state authorities. Large-scale production, efficiency and homogenisation of tastes and consumption pattern are exaggerated by the promoters of globalisation. Local and individual tastes and preferences, particularly if compounded with an increase in income, become reinforced. Consumers search for differentiated, often custom-made, goods and services, and are ready to pay for them. Despite grand talk about globalisation, even the biggest TNCs usually sell most of their output in the domestic market of the country where they produce. While there was a huge increase in absolute volume of FDI in the world, as well as liberalisation of capital markets, one important global thing did not happen. Contrary to the situation at about the

International firms

303

turn of the twentieth century, the new wave of globalisation does not have freedom of labour movements at the turn of the twenty-first.35 Critics of globalisation recall the Communist Manifesto (1848) and Marxian prophecies: unchecked global capital movements and corporations all around the world; wild capitalism and ‘globalisation’ that destroy the social and economic structure of the society.36 These critics say that global TNCs put ‘profits before people’. When these TNCs ‘start talking about how they will no longer put profits first, people (rightly) think they are lying’.37 The critics of globalisation have certain valid points. Paul Krugman wrote: The promise of export-led growth has failed in too many places. In particular, Latin America has signally failed to replicate Asia’s success: Latin nations have liberalized, privatized and deregulated, with results ranging from disappointing (Mexico) to catastrophic (Argentina).38 However, John Kay argued: ‘The essence of economic globalisation is specialisation by function and skill on an unprecedented scale. I cannot tell you how to get rich but I can tell you how to stay poor. Do not become involved in the global marketplace.’39 In any case, a large part of modern trade is intra-industry. This is the outcome of the fragmentation of the production process and specialisation for the global market. Another cost of globalisation can be found in the examples of countries such as Mexico, Thailand, Indonesia, South Korea and Russia, which suffered financial crises in the second half of the 1990s. Without the exposure to global capital markets, the crises would not have developed as they did. Critics of this view argue that these countries would not have experienced such rapid development prior to the crisis without such exposure. In any case, one ought to be fair and observe that certain ‘global tendencies’ were already present, well before the current wave of globalisation. For example, there was (for whatever reason and by whatever means) a spread of certain European languages outside Europe; a spread of Islam in Africa, Asia and Europe; and a spread of Christianity in Latin America. Child labour, begging, theft and prostitution in poor countries would certainly decline if globalisation opened advanced countries’ markets to the products made by the children’s parents. Even though globalisation spreads symbols of highly dubious value such as Coca-Cola,40 McDonald’s, MTV, Halloween and chewing gum,41 it also spreads basic values such as the rights of women and children. However, it is unfortunate that the latter takes place at a much slower pace. A still unresolved and disputed political problem is that under the current wave of globalisation, protection and spread of human rights (pretext) may sometimes be more important than state sovereignty. In earlier eras of globalisation, the pretext was the Christianisation of the barbarians or Islamisation of the infidels. The debate about globalisation is often about jobs (social dimension). Supporters argue that it is beneficial and that it creates jobs, while critics argue the opposite and say that jobs migrate to trading partners and competitors, as well as that globalisation destroys the natural balance in the environment. Certain segments of labour in all countries are suspicious of globalisation, as they no longer perceive the national government as a guaranteed protector of their concerns against external threats. To

Evolutionary Economic Geography

304

counter these fears, the best long-term policy response may be to advance the possibilities for education and training, as technical progress and globalisation have a strong bias against unskilled workers. As far as firms are concerned, efficiency-seeking enterprises, particularly some TNCs, search for seamless and wide international markets regarding trade and investment. They basically seek to break constraints on their business. On the one hand, the globalisation of economic activity is making national frontiers less divisive than ever before.42 Such worldwide economic integration and integrated international production of goods and services, whereby competitors are in one another’s backyard, are made possible by the expansion of information and telecommunication technologies.43 This process is sometimes inverted, on the other hand, by the wide spread of bilateralism and regionalism, sometimes pushed by relatively inefficient firms and governments that are driven by short-term election interests, even though the conditions for a relatively successful regional integration process, such as that in western Europe, may be largely absent. Trade and foreign investments (globalisation) may be partners, not adversaries, of social agendas, but they ought to be coupled with effective national and international institutions that ease and regulate adjustment problems (such as the polluter-pays principle). Globalisation may bring adjustment costs in the affected industries and labour markets, but this may be only transitory. The gradual opening up of markets in Japan during the 1970s, and later in southeast Asia during the 1980s, eastern Europe from the early 1990s and in China from the second half of the 1990s, demonstrates that as a country grows wealthier it ceases to be competitive in the production of labour-intensive goods. Such a country becomes at a later stage an importer of some of these goods and concentrates its production on higher value-added activities. Regional integration (a second-best solution) may be a promising form of supranational governance in areas where there is a strong case for coordination and harmonisation of national policies. Integration may resolve conflicts through positive cooperation within a cosy group but, if pushed to the limit, it may undermine multilateral (first-best) trade and investment systems and fragment the world economy into conflicting regional blocs. Regionalism and multilateralism (globalisation) need not necessarily conflict. If the regional blocs cooperate and if they adopt liberal external trade and investment policies, the outcome may be an overall welfare improvement. The pace of international trade liberalisation from the 1960s till the end of the 1990s, as well as the extension of the GATT into new areas such as services and agriculture, might have been much slower in the absence of challenges posed by the great progress in European integration. The debate should not be perhaps between regionalism and multilateralism, but rather between liberalism and controlled interventionism. Internet The ease, convenience, high speed and low cost of communication via the Internet and other means is well recognised and appreciated.44 The same holds for the participation of an increasing number of countries that take part in global production and exchange. However, in the process, companies such as Microsoft have gained an enormous power over our work and lives. This is amplified by similar possibilities by governments, thugs

International firms

305

and computer hackers. They can intrude into our databases related to work, finance, leisure, emotional relations, thoughts, ideas, privacy, habits, voting intentions, movements, conversations, addresses … and messages. They can steal our passwords, impersonate us, misuse our data and commit fraud without us being aware of it until it may be too late. They can oversee, control and even direct and manage parts of our lives. Some anti-globalisation activists appreciate the potential benefits of globalisation, but all campaigners are in favour of strict limitations and control to curb the possible excesses. The Internet symbolises the borderlessness brought about by globalisation. Physical presence in a specific location is not necessary. In the invisible continent of cyberspace, the users are ‘everywhere and nowhere’ at once. They use the Internet as a meeting place, a market, a distribution device, a library, and the like. This alters the geography of production, at least for certain goods and services. People spend more time working at home or ‘on the road’, hence certain office space (in congested cities) is vacated. However, one does not need to give an absolute value to globalisation. In spite of this process, localisation and clusters still matter. Firms that went furthest in ‘globalisation’ report that face-to-face contact is essential for the smooth organisation of business within the firm and marketing outside it. Localisation is not an attempt to turn the clock backwards. This is done by globalisation itself through a reduction in job prospects and security of employment, and also deterioration in both overall safety and in the provision and availability of basic needs. Hundreds of millions of people are affected. In addition, just try telling someone who needs a visa and wants to enter the EU or North America from outside that this is a ‘borderless world’. Borders continue to matter a great deal as people from different locations are prevented from crossing them without a cumbersome, costly, humiliating and uncertain administrative procedure. The Internet is technically based on communication and data that are located in computers that exist in the real world. Initially, this parallel universe of pure data existed everywhere freely in a ‘lawless’ world. The Internet is breaking down barriers and eliminating physical distance. However, computers exist in an identifiable geographical space. If they can be located together with their users, then they are subject to law. Now governments are increasing their control over cyberspace. For example, France prohibited the sale of Nazi memorabilia on the Internet; Iran banned access to immoral or antiIranian information; South Korea prohibited access to gambling websites; while Google decided to censor its Internet searches inside China so as not to upset the Chinese authorities. There is also increased international police cooperation in combating child pornography, or computer viruses and hackers that attack major computer networks. Hence, borders are being created in the Internet. Even though the economic impact of geographical distance is being reduced, the local economic geography is retaining its strength. Global standards The imposition of ‘global’ standards may have its justification for relatively new and standardised goods and services such as copiers, fax machines, computers, TV sets, mobile phones or better medicines, otherwise communication and exchange of information might be difficult and costly. But it may stop there. The imposition of global standards for traditional goods (for example, food) with the exception of health and the

Evolutionary Economic Geography

306

environment may not be easily justified. If needs, choices and tastes for certain types of food are strongly locally specific (even the thickness of pizzas throughout Italy differs), why should one favour or impose global Pizza Hut-type standards? For services, regionalisation is more important than globalisation. ‘Managers need to change their thinking as the end of globalization is here. Managers must “think regional and act local”’ (Rugman, 2002, p. 15). John Gray of the London School of Economics claimed that people are losing faith in globalisation, and that: [L]ed by the United States, the world’s richest states have acted on the assumption that people everywhere want to live as they do. As a result, they failed to recognise the deadly mixture of emotions – cultural resentment, the sense of injustice and a genuine rejection of western modernity – that lies behind the attacks on New York and Washington … The ideal of a universal civilisation is a recipe for unending conflict, and it is time it was given up.45 It is not that the people who live outside the western world cannot adopt a liberal attitude, rather that there are social, cultural and institutional barriers that prevent a fast transfer of western standards, values, culture and institutions elsewhere. One must also consider the choices of the ‘recipients’. Do they really want and need what is exported or ‘imposed’ on them? Wal-Mart, the world’s most powerful retailer with ‘prices that nobody could beat’, failed in and withdrew from Germany in 2006 and earlier from South Korea. German consumers already profited from local competition between Lidl and Aldi, and they did not want to spend time driving long kilometres to Wal-Mart stores and change their embedded shopping habits, while the hypermarket-shopping model was not in line with local tastes in South Korea. Local conditions may not support the one-size-fits-all approach. Globalisation may be less rewarding to firms than ambitiously predicted and expected. Consumers often demand and pay for many niche goods and services that detach them from the crowd. There was a faulty premise and a vision that all people are culturally homogeneous in that they think, act and, most of all, shop alike (or like Americans),46 that there exist institutional bases for political democracy, and that citizens are aware of their duties to one another and to the state. Checks and balances are mixed; there is mistrust vis-à-vis the government and state administration which is often deeply rooted in dictatorship, authoritarian and bureaucratic conduct, as well as in corruption. Without very long-term reform, education and trust-building, any attempt to transfer ‘global’ (social) standards quickly would fail, and fail miserably, in many parts of the world. It should not be forgotten that it took 400 years for England to develop from that stage to its present one. To do the same elsewhere in half the time of 200 years would be a tremendous achievement; to aspire to do it in 25 or 50 years may be to court disaster. (Lipsey, 1992a, p. 755)47

International firms

307

Local differences tend to be stubborn. Until the world becomes homogeneous, adaptation towards local preferences and capabilities will be necessary. Globalisation problems, which appeared in the form of concerns over progress in the WTO or in the Multilateral Agreement on Investment, illustrate signals of regional (triad: US, EU and Japan) or even local power. It is true that the greatest and matchless advantage of free markets and globalisation is that they give free choice to consumers. Economic interactions are voluntary, so consumers are free to choose, for example, between local and global goods and services. But this is only on the surface. If global products are advertised aggressively by large TNCs,48 including campaigns that are often beyond the financial capacity of local competitors,49 then the ‘free choice’ by consumers may be restricted and the local producers (and certain dimensions of the local culture and health) may be damaged. This was well understood by the Jesuits, centuries before the modern advertising campaigns by Coca-Cola and McDonald’s. The Jesuits recognised that if you can shape someone’s mind when that person is young, you may have it for life. In January 2006, for example, in order to avoid official warnings about new heavy-handed EU legislation against obesity, the Union of European Beverages Association (UNESDA) (including Coca-Cola and Pepsi) introduced self-regulation measures and voluntary restrictions about marketing soft drinks to youngsters.50 Rebottled old stuff It is often forgotten that the spirit of ‘globalisation’ does not bring anything essentially new. The quest is for international openness: more freedom for trade in goods and services and for capital mobility (FDI). The economic role of national frontiers declines as national economies merge (integrate) into a single interdependent ‘global’ unit. In a nutshell, the idea is to return to the essentials of the system that was prevailing during the first wave of globalisation (1850–1914). ‘Re-globalisation’ may be a more appropriate term for the second wave that started in and continued from the 1960s. One has to recall that, during the first wave, globalisation was imposed on the rest of the world either by ‘gunboat diplomacy’ or through colonial control: China signed a trade treaty with Britain in 1842; Britain imposed ‘free trade’ on India, while the Dutch did the same with Indonesia during the 1840s; American gunboats forced Japan to open to trade in 1858; British gunboat diplomacy opened most Latin American countries for trade somewhat earlier. ‘The West practised protection wherever necessary, but imposed free trade on the Third World’ (Nayyar, 2006, p. 139). Opening and creation of linkages between the ‘Old’ and ‘New’ World took centuries. However, the speed of the establishment of similar ties increased over time. It took only a couple of short decades to established similar linkages with China. Table 5.1 provides a list of suggestions about the basic economic features of the two globalisation waves. Both waves were supported by technological revolutions that included the fragmentation of the production process, transport and communication. The first one involved steamships, railways, telegraph and the opening of the Suez Canal. The second one includes container transport, mass air transport, cheap and reliable telephone services and the Internet. The two globalisation waves have certain differences. First, tariffs were much higher during the first than is the case during the second globalisation wave. However, NTBs are

Evolutionary Economic Geography

308

higher now than was the case during the first wave. Next, the first wave included interindustry trade (primary commodities for manufactured ones). During the second wave, an increasing share of trade is intra-industry (France both sells and buys cars from Germany). Intra-firm trade is highly important during the second wave. Third, FDI were flowing to the developing countries during the first globalisation wave, while now principal FDI flows are among the developed countries. Another difference is that the first wave of globalisation involved huge migration of people. Such migrations are strongly controlled and restricted now. Table 5.1 Similarities and differences between the two globalisation waves Feature

First wave 1850–1914

Second wave from 1960s

Technological revolution in transport and communication

Steamships, railways, telegraph and the opening of the Suez Canal (halved the distance from English to Indian’s ports). Fall in transport and communication costs.

Container transport; mass air transport; relatively widespread, cheap and reliable telephone services; and the Internet. Fall in transport and communication costs.

Production: fragmentation, unbundling, offshoring or vertical specialisation

Technology affected primarily industries and firms. This permitted spatial separation between the location of production and the place of consumption of goods. Productivity increased.

Technological changes affected industries, firms, plant, subfirm and sub-plant levels. The location of all functionally related production stages of one production process in a plant does not need to be located in the geographical proximity. Various stages of the same production process can be fragmented and located in different places. Productivity increased.

Sectors affected

Manufacturing

Manufacturing and tradable services

Trade

Inter-industry trade (primary for manufactured goods). Strong and different comparative advantages and a decline in transport costs made large-scale trade possible.

Intra-industry; intra-firm trade important. Cheap transport permits trade as soon as there is a new product or a change in tastes even if the countries are comparable.

Tariffs

High but transparent.

Low

NTBs

Low

Relatively significant, not transparent.

FDI

Mainly from Britain. Half of FDI stock was in developing countries, another half in Europe and the US.

Principally among the developed and newly industrialised countries. Enormous amounts in absolute terms. Developing countries are in general marginalised as locations for FDI. The rise of China as a destination and source of FDI from 1990s.

International firms

309

Industrialisation

Europe and North America industrialised while Asia (China and India) deindustrialised. Colonial masters prevented a spread of new technology to colonies.

Europe and North America deindustrialised while Asia industrialised. Relatively faster international spread of new technology.

Standardisation

Informal

Formal

Financial intermediaries

Banks

A variety of intermediaries: banks and institutional investors (pension funds, mutual funds).

Migration of people

Generally liberal to the Americas and Australia.

Highly controlled and restricted. Destinations: oil-rich Arab countries, the US, Canada, Australia and Europe.

Institutional arrangements

Largely informal supported and promoted by ‘gunboat diplomacy’. Based in many cases on colonial ties.

Partly formal through international economic institutions, partly informal through the economic strength and behaviour of TNCs.

Geographical spread of production by the means globalisation

Uneven

Uneven

Urbanisation

Important

Important

Who are the winners and losers from opportunities and risks brought by globalisation?

Some (elite) gained in both developing and developed world. Many lost in both developed and developing world (at least in the medium term). Firms: Gainers are large, competitive, technology leaders, risk-lovers, internationally mobile. Losers include local, small and riskaverse firms. Persons: Gainers are the educated, risk-lovers, those that work in the ‘winning’ firms, mobile and asset owners. Losers include those that do not participate in globalisation i.e. poorly and semi-educated, immobile, risk-averse and wage-earners in the uncompetitive firms and industries.

Hard to predict with certainty because of the fragmented production process often within the same firm (or plant). Certain individuals within a firm may gain, while others within the same firm may lose. Some (elite) gained in both developing and developed world. Many lost in both developed and developing world (at least in the medium term). Firms: Gainers are large, technology leaders, risk-lovers, internationally mobile. Losers include local, small and risk-averse firms. Persons: Gainers are the educated, risklovers, mobile and mobile asset owners. Losers include those that do not participate in globalisation i.e. poorly and semieducated, immobile, risk-averse, wageearners particularly in certain countries with highly regulated labour markets.

Evolutionary Economic Geography Public concern and debate

Principally within the Communist Party.

310

Social impact and issues brought by globalisation are highly controversial and fiercely debated: human rights, fair trade, labour laws, environment, NGOs …

Globalisation brings certain economic and other opportunities, but it also brings a number of risks and perils. Globalisation is a double-edged sword. Some gained from it, many others lost and suffered. Usually, the few big technology leaders, educated, internationally mobile and successful risklovers gained a lot. Others were usually on the losing side in the medium term. Globalisation may be a system that provides real opportunities to some to take the best out of the open global market economy and leave the rest to all the others. As such, this type of elite globalisation may create and accentuate various economic and social asymmetries. Nayyar (2006, p. 158) commented that: Globalisation has introduced a new dimension to the exclusion of people from development. Exclusion is no longer simply about the inability to satisfy basic human needs in terms of food, clothing, shelter, health care and education for large numbers of people. It is much more complicated. For the consumption patterns and lifestyles of the rich associated with globalisation have powerful demonstration effects. People everywhere, even the poor and the excluded, are exposed to these consumption possibility frontiers because the electronic media has spread the consumerist message far and wide. This creates both expectations and aspirations. But the simple fact of life is that those who do not have the incomes cannot buy goods and services in the market. Thus, when the paradise of consumerism is unattainable, which is the case for common people, it only creates frustration or alienation. The reaction of people who experience such exclusion differs. Some seek short cuts to the consumerist paradise through drugs, crime or violence. Some seek refuge in ethnic identities, cultural chauvinism or religious fundamentalism. Such assertion of traditional or indigenous values is often the only thing that poor people can assert, for it brings an identity and meaning to their lives. Outcomes do not always take these extreme forms. But globalisation inevitably tends to erode social stability. Thus, economic integration with the world outside may accentuate social tensions or provoke social fragmentation within countries. The spirit of globalisation might remain the same in both globalisation waves, but there are important space-related differences in their actual attainment. During the first wave the north (Europe and the US) industrialised, while the south (especially China and India) deindustrialised. There was an expansion of trade and factor movements (both labour and capital), while incomes between the two regions diverged. In the second wave of globalisation, it was the south (east Asia) that industrialised, while the north deindustrialised. Trade and capital mobility expanded (mass labour migration was small by the first-wave standards), while incomes between the two regions generally

International firms

311

converged. Urbanisation remained an important feature in both north and south during both waves of globalisation (Baldwin and Martin, 2004, pp. 2707–8). An enlarged market is an important gain for efficiency-seeking firms in a small country. In a situation without integration, foreign countries can simply threaten a small country that they will introduce protectionist measures or ‘sanctions’ against it (the US frequently makes such threats to many countries). Such a warning can seriously undermine the quality of all economic decisions in a small country. Integration enhances and secures market access for partner countries, as well as increasing the potential for long-term competitiveness of a small country’s goods and services. A common market may eliminate or harmonise national incentives to foreign TNCs to locate in partner countries (which were previously subject to countervailing duties). It also mitigates noneconomic considerations, such as political pressures on third-country investors to locate in a particular country. Other globalisation strands Transnational corporations behave like other firms: they primarily follow the opportunities for maximising profit over the long run while staying within the law. With the exception of resource-oriented FDI, the size and growth of the local market, including privileged access to international markets (instead of mere differences in the cost of labour) are the most prominent motivators for their trans-border business operations. In addition, in a situation where market liberalisation became a widely accepted policy choice, there is an increase in the importance of created assets and structures (technology and ability to create it, business culture, capability to organise and control production and marketing, communications infrastructure, marketing networks) as determinants for FDI. This is why 70 per cent of the activities of TNCs were located in developed market economies (measured by the inward stock of FDI in 2006) (UNCTAD, 2007, p. 255). In spite of the talk about ‘globalisation’, on average, a significant part of the output of affiliates is still sold on the local market.51 In this situation the developing countries and those in transition face very tough competition to attract TNCs. The introduction of the monetary union in the EU in 1999, including the euro as a subregional currency, replaced national monies for twelve countries. Expectations about the beneficial economic impact of the monetary union among the EU elite were high. However, growth rate in the eurozone countries is low, unemployment is high and countries have big trouble in sticking to the budgetary rules. The European Constitution failed in 2005. The message from the voters to the politicians may be put in the following words: ‘Do not move on other grand (global) projects unless and until you fix the current ones.’52 The relation between social spending and competitiveness of a country’s goods and services in a global economy is complex. There are no golden (global) rules and examples to follow at all times. Since 1980s, a high tax welfare state in Sweden has proven to be quite successful in the economy, as is the lower-tax and less regulated British economy. Some argue that TNCs rush to locate in countries with low social spending to save on costs. However, de Grauwe and Polan (2003) found evidence that wealthy countries that spend most on social needs on average rank highest in terms of competitiveness. Countries with highly competitive output generate extra income. They

Evolutionary Economic Geography

312

can afford generous social outlays. But there is also a reverse causality. Competitiveness depends on an absence of prolonged social conflicts, superb human capital and the quality of the government. People would be inclined to take a risk if they knew that they would not be condemned to poverty if they failed. One of the principal roles of an efficient government is to transform social security contributions into energies for social value added. Those that manage this conversion well are likely to benefit in terms of enhanced competitiveness. Foreign investors will locate their activities in a country that offers the most favourable cost mix of operation (production and marketing), provided that this factors well into the longer-term vision of potential profit. FDI can be made simpler by regionalisation of the world economy and international economic integration. However, integration/regionalisation is only a supporting tool for the tendencies that bring about international business globalisation. Modern competitive firms are usually TNCs that ‘globalise’ their fragmented business process in the search for seamless and extensive markets. Therefore, an increasing share of domestic output, even in the developed countries, is under the control of foreign TNCs. The same holds for an increasing share of foreign output of domestic TNCs. Strong FDI relations may exist even though the countries or groups of countries are not formally integrated. Just take a look at the example of two-way FDI flows between the US and the EU, or FDI into China and from it (principally in the development of resources and energy). The glue that binds transatlantic relations together is not principally trade, but FDI. Large global interpenetration of FDI reduces the possibility that regional arrangements may become closed blocs. A ‘hostage population’ of TNCs may reduce the fear of retaliatory measures. Extensive FDI links between the US and the EU have helped to reduce the strength of any potential conflict between the two partners regarding market access. The same is not yet true of Japan and China. It is hoped that Japan will mature as a foreign investor in future and that potential conflict with that country will be defused. As far as China is concerned, in the future this country may increase its FDI in the US and perhaps Europe too, in addition to the expanding investments in the development of resources in Africa, Latin America and Asia. Relatively open economies and liberal rules for investment and capital movements assisted western firms to penetrate new countries and markets for decades. This assisted the prosperity of the western world. However, western countries, in particular the US, are being (partly/slowly) replaced by China and India as engines of world economic growth. Many manufacturing and certain services jobs are going to Asia. The new Asian face of globalisation may not please many in the US and Europe. Only in 2006, the Indian entrepreneur Lakshmi Mittal took Arcelor, Europe’s largest steelmaker, and the US barred the Chinese takeover of Unocal, an oil company (China may be a source of cheap clothes, but owning an American oil company is another story). During 2006 there was also a wave of economic ‘patriotism’ in the EU. France and Spain protect their domestic energy market from takeovers, even from companies from fellow EU member countries. The same is true for banking in Italy and Poland. Will the Russians in the new global world own and control a large chunk of the EU gas distribution services? Or increase their 5 per cent share in the EADS (which not only owns Airbus, but is also active in aerospace and the defence business)?53 China owns 10 per cent in Blackstone, the world’s largest private equity firm, which has stakes in forty-

International firms

313

three industrial groups including energy, pharmaceuticals and Deutsche Telecom.54 Has the West lost control over globalisation? Can it rein in forces that it proudly unleashed and profited from for decades? Benefits of globalisation as measured by the spread of FDI, economic growth and development were much less global than the neoclassical partisans of globalisation argued. Adam Smith and David Hume taught us that in the longer term all countries grow together (not at each other’s expense). This ought to be always kept in mind. 4 Trans-border business activities Theoretical aspects There are four main theoretical types of trans-border location of business activities that are conducted by TNCs. In practice, it is hard to find examples in their pure theoretical forms. Nonetheless, in theory these types are: market-seeking, resource-based, rationalised and strategic asset-seeking (Dunning, 1999, pp. 3–4): • Market-seeking (demand-oriented) investments search for new markets, but they replace trade. They are influenced by the relative size and growth of the foreign market in which the investment is made, the relative costs of supplying that market through imports or local production, and the relative advantage of engaging in direct local production or licensing. • Resource-seeking (supply-oriented) FDI is motivated by the availability and cost of both natural resources and labour in the target location. As the products of such investments are often exported abroad, the economic climate in foreign markets, changes in technology, transport costs and barriers to trade influence the attractiveness of such investment to TNCs. In addition, evident shortage of natural resources in China made this country invest in the development of resource exploitation (crude oil) in rather volatile locations such as the Sudan and Somalia. And elsewhere in Africa, particularly from the start of the twenty-first century.55 • Rationalised investments seek efficiency. Like resource-based investments, they are complementary to trade. Their attractiveness is found in cost considerations. They are influenced by the ease with which intermediate or final products (linked to economies of scale and specialisation) can be traded on the international market. A case in question is the US loss of competitiveness as a site for labour-extensive production. The domestic US enterprises from this area of manufacturing locate abroad (Mexico, China and elsewhere in Asia). • Strategic asset-seeking FDI is aimed at protecting and augmenting the existing ownership-specific advantages of the investing firm. Alternatively, such FDI may be aimed at reducing the advantages of competitors (location of European elevator firms in the US). Determinants Large companies are in most cases also TNCs. They are not directly or completely accountable to any government, but rather have their own ethos. Chauvinism regarding

Evolutionary Economic Geography

314

the location of business (personnel matters are somehow a different issue) is alien to international firms, for their business decisions are not likely to be based on either ideological or nationalistic grounds (Rubin, 1970, p. 183). This is even more decisive for global TNCs that produce for global and low-tariff markets throughout the world. The most crucial determinants of FDI location are the relative difference in returns and profit maximisation in the long run, market presence, availability of resources, expectations of growth in demand and political stability. These determinants for the location of an affiliate may be more important than, for instance, a country’s participation in a regional economic bloc. Table 5.2 summarises principal groups of investors’ motives and determinants for FDI in the target location. It is obvious that the issue is rather complex. Therefore, forecasting FDI flows is difficult and uncertain, as determinants are many and complex, highly changeable from case to case and not always measurable. Transnational corporations behave like other market-oriented firms: primarily they look for opportunities to make a profit. Thus they compare current and expected profits at home and abroad. The principal motivators for FDI and trans-border business activity are summarised in Table 5.2, and just a few of them will be mentioned below. The size and rate of growth of a local market in the target country is approximated by income, its growth and by privileged access to regional and international markets. This was the most prominent general motivator for trans-border business operations. Depending on the industry, the availability of resources (including the cost of labour adjusted for productivity) were relevant too. Other motivators, although recognized, were less Table 5.2 Investors’ motives and determinants for FDI in the target location Investors’ motive

Determinants in the target location

1 Market

• size of private and public demand (market seeking) • its growth (current and prospective) • access to regional (integration) and global markets • local, regional and country consumer needs, tastes, sophistication, preferences and fickleness • proximity to consumers (clients) • market structure (competition, openness, protection) • objective to bypass trade and entry barriers • exercise market power (becoming a monopolist) • intentions to pre-empt (or follow) rivals • actions by competitors: follow your leader (snowball, ‘me-too’ or ‘herding behaviour’) or exchange of threats (tit-for-tat strategy among oligopolists) • demanding clients

2 Resources

• mobility of resources • raw materials (availability, prices) • skilled labour and educated management • system of education • brain drain • unskilled labour (low cost) • available technology (production functions)

International firms

315

• innovation and R&D capabilities and potentials • infrastructure • inputs including supporting services • available capital at favourable rates • organisation of labour (voluntary or mandatory labour unions) • work habits (Japanese TNCs sometimes prefer to locate in rural areas as they mistrust urban workers because of their ‘different’ work habits and mobility) 3 Efficiency

• cost of resources in the target location (listed under No. 2 above) adjusted for productivity where appropriate • raising costs of production in already operating (domestic and foreign) locations • technological production linkages • cost and easiness of operation (transactions, trade, transport, communication, culture … ) • efficiency of the supply chain • trade and marketing networks • logistics • points at the product cycle and production function curves

4 Others

• available financial resources for foreign investment (Russia invests a proportion of its ‘gas euros’ into Airbus; China invest a part of its ‘China dollars’ into the American computer industries) • existence of clusters • macroeconomic and other stability • incentives (subsidies and taxes) and concessions • rules, regulations and policies (FDI entry and protection, mergers and acquisitions, competition, labour, tax, subsidy, trade, NTBs … ) • host government privatisation and liberalisation policy • pressure by the government in the target country to locate there • movements in exchange rates • investment promotion and friendliness (helpfulness during considerations about FDI, entry, operation and exit) • dispute settlement system • establishment of a listening, learning and monitoring post • culture (way of doing business, openness, corruption, red tape, languages spoken …) • quality of life

important. This is because of the strong impact of the real market size on the minimum efficient scale for production. As the general level of protection of the national markets declines because of liberalisation (there is no need to jump over national tariff barriers) and as various types of economic integration proliferate around the world, the mere size of the national market continues to be important, but other motivators are gaining in significance and strength. The local availability of technical, managerial and organisation (networking) knowledge, capabilities and competences, including innovation, together with the existence of clusters, become prominent motives for FDI. Seeking, enhancing and protecting such strategic assets and structures becomes an important motivator for FDI. This is linked with high skills, education and experience of labour in the target location. Such labour

Evolutionary Economic Geography

316

and management is expensive everywhere. This means that low wages (adjusted for productivity) cease to be a sufficient determinant for FDI. Where market liberalisation is a widely accepted policy choice, created assets such as technology and the ability to create it, business culture, the capability to organise and control production and marketing, communications infrastructure and marketing networks internal and external to firms are increasingly important determinants for FDI. This is why most activities of TNCs are geographically located in developed market economies (measured by the stock of FDI). Neoclassical equilibrium theory cannot predict with absolute certainty the geographical location of activities resulting from capital mobility within a common market. TNCs that use complex technologies do not worry about tariffs and quotas. They are concerned, rather, with domestic regulations such as environmental standards.56 A degree of government intervention may influence the spatial location of TNCs, with important implications for the future distribution of output and trade. Some goods and services must be adapted in order to meet local needs and preferences (for example, food). Locating at least a part of the production process near the place of consumption may do this more efficiently and more cheaply. Other reasons for FDI instead of, or together with, exporting include taking advantage of a range of the host country’s incentives such as financial incentives (subsidies, reduced taxes), tariff protection, exemption from import duties, public purchases and granting monopoly rights. Other motives include market pre-emption and increase in market power, as well as the empire-building ambitions of firms. Short production runs and inefficient production technology do not mean that a subsidiary of a TNC is unprofitable and that such location for business is unappealing. During the import-substitution development phase in Mexico, Chrysler’s Mexican affiliate was the corporation’s most profitable operation in the world. Chrysler fought against the liberalisation of the Mexican automobile industry in order to safeguard what its managers described as a ‘cash cow’ (Moran, 2005, p. 285). The literature based on surveys and case studies suggests that government strategies such as tax incentives and industrial policies have little, if any, effect on the location of industry. All studies are aware of the difficulties in ascertaining the impact of public policy measures on the location of production. Public policies have mixed effects on the location of firms: on the one hand, higher taxes increase the operating costs of firms; on the other, higher taxes in some areas may be used to pay for better public services, such as education or infrastructure, that support the operation of firms (Smith and Florida, 1994, p. 31). In a survey of thirty TNCs covering seventy-four investment projects concerning cars, computers, food processing and petrochemicals, many TNCs revealed that government incentives were not an issue and, where they existed, simply made an already attractive country for the location of business operations even more attractive. Investment decisions were made on the grounds of economic and long-term strategic conditions regarding inputs, costs and markets. Overall, incentives are not an important factor in the set of elements that determine inward FDI. Only once the decision is made to invest in the target country (or region) may the incentives have an impact on the exact choice of location within the target country (UNCTAD, 1998, pp. 103–4).

International firms

317

As for geographical origin, most TNCs originate in countries with decentralised political systems such as the US, Britain and the Netherlands, which have the longest history of transnationalisation of national economic activity. In these countries, the TNCs and the state foster relations of complementarity. In centralised states, the administration does not want to share power with any enterprise; however, this has changed in past decades, particularly in France and more recently in ‘communist’ China. If producers and government in a big country A make a credible threat to close its market to exporters from a small country B, then one of the options for country B is to locate ‘fifth column’ production in country A. This would preserve country B’s market share in country A’s market. If overheads are covered in country B’s home market, then its firm in country A may sell on a marginal cost basis. Many European and Japanese TNCs have concentrated on breaking into the US market and become ‘US nationals’ in order to avoid unilateral and whimsical US economic sanctions and possible changes in the American trade regime. Making immediate profit from such an investment has played a secondary role, at least in the short term. A similar observation in the EU was the flurry of non-EU TNCs that settled in the EU prior to the full implementation of the Single Market Programme at the end of 1992. They wanted to become the domestic EU residents. The success of Japanese firms in reducing the labour content of the final output is one of the factors contributing to the expansion of their FDI into relatively high labour cost areas such as North America or the EU. High labour productivity can make up for higher wages. This is why the most significant part of all activities by TNCs takes place in a select group of developed countries. Relatively low nominal wages do not necessarily equal low labour costs, as labour productivity should be taken into consideration. In addition, low or declining wages may mean a (local) shrinking market. If this were the case, and if TNCs are interested in the local market, this may act as a deterrent to FDI. Despite the advantage of cheaper labour in Italy, Greece and Portugal, relative to the central and northern EU countries in the 1960s, labour outflow from, rather than capital inflow to, southern Europe was the equilibrating force at that time. The appearance of China as a magnet for FDI from the 1990s, with its huge domestic market and relatively low-wage and productive labour force, will have a profound long-term effect on a global reallocation of many types of manufacturing on the widest global scale in decades to come. Foreign direct investment is a tool used by enterprises that want to exploit long-term profit-making opportunities abroad. Before embarking upon FDI, an enterprise compares alternative geographical locations at home and at various places abroad. If investing abroad seems to be the more promising option, then the enterprise has to make sure that it possesses or can obtain certain mobile income-generating firm-specific advantages in production or transactions that could enable it to operate profitably in the foreign environment. These advantages include exclusive or privileged access to specific assets and structures, as well as better organising and control capabilities for both production and transactions. Local firms may have several advantages over foreign ones from the outset. For example: • they have a better knowledge of local consumer and supplier markets; • they do not have the barriers and costs of operating at a distance;

Evolutionary Economic Geography

318

• they often receive favours from the government; • they do not operate in a different, often hostile, language, tax, legal, exchange rate, social and political environment; and • TNCs may have some disadvantages in the eyes of certain local politicians in target countries as they are foreign and, sometimes, relatively large. Therefore, TNCs bear these elements in mind when deciding on the location of their affiliates. They weigh up whether they have enough advantages over other firms or whether they can create the capacity to overcome these obstacles. A difficulty for TNCs may be that they have to attract key managers and technicians from the headquarters to a foreign subsidiary. This may require both higher wages for such personnel and higher allowances for their families. TNCs that extract and export natural resources are at risk of becoming a target for nationalisation. If a TNC wants to operate in such a geographical and social environment, it must have or control special advantages over the domestic competitors in the target location. These special and mobile advantages include a superior production technology and management technique (a TNC ought to operate on a different production function or at adifferent point on the same production function than domestic firms), a well-known brand name or, especially important for services, access to markets and quality control. For these reasons, it is no surprise that TNCs are often more profitable and successful than local competitors in the same industry.57 While firm-specific income-generating and mobile advantages are a necessary condition for FDI, they are not a sufficient condition. If trade were free, firms could simply use their advantages by exporting instead of producing abroad. Various market imperfections limit the size of the market for free trade and, hence, justify FDI. These include tariffs and NTBs, differences in factor prices, sunk costs and after-sales service. Therefore, an enterprise with a specific and mobile income-generating advantage considers a number of different possibilities and restraints at various foreign locations before settling abroad. Linkages, externalities and spillovers Targeting some key productive activities that have significant linkages with the rest of the economy requires some form of government intervention. Japan’s decision to target steel, shipbuilding and toys in the 1950s and 1960s was highly profitable. This choice turned out to be questionable in the 1970s, so Japan turned to cars and machine tools. In the 1980s and 1990s, the Japanese chose electronics as their target industry (Tyson, 1987, p. 70). This target became obsolete from the 1990s as software overtook hardware in importance, and in this industry the US leads. Figure 5.1 presents in a graphical way the location of various general lowand hightechnology industries in countries at different stages of development. The straight isocost line DC for developed countries describes the situation in which capital is more abundant than labour. In general terms, such a country is the location for advanced activities such as R&D and hightechnology production. The situation in the newly industrialised countries, represented by NIC, reflects the case in which there is less capital and more labour than in the developed world. These countries are the location for the basic and standardised manufacturing and assembly. The countries with a large supply of labour in

International firms

319

relation to capital (less developed countries) are the location for primary production suggested by LDC. It is important to note that countries may both change their isocost lines and slide up and down them. The obvious examples include Argentina, Britain and China. The problem with this figure is that it does not reflect a new trend started from the Asian financial crises of 1998. Newly industrialised Asian countries, such as China, accumulated reserve funds through huge trade surpluses. The reserves of over a trillion dollars have been recycled and invested in the developed world, the US in particular. So after the ‘petro dollars’ accumulated by the oil-producing countries from the 1970s, one may consider ‘China dollars’ as important players on the international monetary stage from the start of the twenty-first century. Soon one may also start talking about the

Figure 5.1 Comparative advantage and location of economic activities. Source: Kogut (2004, p. 263). Russian ‘gas dollars’ or ‘gas euros’. The newly industrialised or ‘emerging’ economies, especially China, do not only supply the ‘global’ economy with cheap labour for manufacturing, but they also provide it (especially the US) with cheap capital. These developing countries basically prevent their domestic currency rising in relation to the dollar and they pile up dollar reserves because of trade surpluses. Their domestic financial market is weak to allocate capital, hence they find an attractive alternative (in the short and medium term) in investing in the developed countries, US government bonds in particular. By doing so, the developing

Evolutionary Economic Geography

320

countries are awkwardly (according to the standard theory) subsidising en masse the wealthiest global consumers. This outcome is something that standard neoclassical international economic theory would find totally unexpected and bizarre. Neoclassical equilibrium theory would predict that the flow of capital is towards the developing and poor countries, where capital per worker is scarce and where the rate of return is higher than in the developed world (as was the case of European export of capital to the US in the nineteenth century). As the foreigners are in part financing US public consumption, this provided a handy scapegoat for some politicians in the US to blame foreigners (east Asians, but forgetting the OPEC group) for the US trade deficit, rather than the need to settle and put US domestic accounts in order (increase savings and reduce consumption). However, as long as the foreigners are willing to finance American consumption with ‘cheap’ money, the US will have little hard incentive to reform its economy. In the past, certain select industries have had important spillover effects throughout the economy. Henry Ford’s assembly-line production technology from the 1920s spread to all manufacturing industries in the US and elsewhere. Telecommunications affect the dissemination of information, computers have an impact on data processing, while transportation equipment and logistics affect the size of the market. As a result of these linkages, private returns from these industries are smaller than social returns. The targeting and location of key industries that may have significant linkages with other industries or partner countries may have important and long-term beneficial spillover effects on the country in question, or a group of countries that integrate. Various potential linkages between the operation of TNCs and the growth process are presented in Figure 5.2. Nature offers many examples of externalities and spillovers. Bees and flowers are the most obvious ones, known to everyone. To understand and capture economic spillovers (unpaid side-effects of economic activity) in general and in irreversible time, one needs a definition of externalities. Lipsey et al. (2005, pp. 100–1) define externalities as follows:

International firms

321

Figure 5.2 Transnational corporations and the growth process. Source: UNCTAD (1992, p. 247).

Externalities are unpaid for effects conferred by the continuing and potentially variable actions of one set of agents (whom we call ‘initiating agents’) on another set of agents (whom we call ‘receiving agents’) who are not involved in the initiating agent’s activity, and for which the receiving agents would be willing to pay to receive a positive externality and to avoid a negative one. Externalities typically prevent the fulfilment of general equilibrium optimality conditions in neoclassical analysis since there will be too many of the activities that produce negative externalities and too few of the activities that produce positive externalities. [Original emphasis]

Evolutionary Economic Geography

322

The policy challenge is to increase positive and to decrease negative efficiency spillovers and externalities. If a subsidy is offered to a TNC to locate in a specific location, then this subsidy ought to be lower than are the net positive externalities generated by the project. The problem is that this is very hard to ascertain, both in technological and in pecuniary terms. Technological efficiency and other spillovers on the local economy are often, but not always, industry-specific. This is because Marshal–Arrow–Romer-type spillovers (industry-specific) may spread in an easier and more direct way than the Jacobs-type spillovers (across industries). Industry-specific spillovers occur through several horizontal and vertical channels, which include the following: • Competition If the affiliate of a TNC and local firms compete for the same local market, then this type of competition by more advanced TNCs may force the local business to increase efficiency in production and in general business (positive effect on local firms). Otherwise, local firms may lose market share. Hence, if local firms lose market share to TNCs, the fixed costs in local firms are spread on a smaller output. This increases production costs per unit of output. If nothing is done, these firms may go out of business (negative short- and medium-term effect on the locally owned economic activity). TNCs are quite sensitive about the protection of intellectual property rights in host countries. They want to have their technological advantages protected. They strongly wish to prevent unwanted technological leakages to local competitors. Hence countries that have a limited rule of law should not expect TNCs to transfer the most recent technologies there. TNCs and local firms may also compete for scarce local resources, such as qualified and experienced personnel. If a TNC affiliate attracts the finest input resource, it may leave the local firms in a competitive shadow. • Supply Dealings with local suppliers (if they exist) depend on the degree of autonomy and liberty that an affiliate has from its headquarters. If the affiliate of a TNC is fussy regarding terms of supply (high requirements regarding quality, quantity, delivery time … ), it will award contracts to the best existing local supplier. Hence, spillovers do not take place as the existing local production structure is perpetuated. However, if the ‘cherrypicking’ continues, this may serve as a strong incentive to local firms to improve and reach a technological and commercial level necessary to win orders from the TNC. This may also instigate local business start-ups (e.g. Ireland). Sometimes a TNC may, in its own interest, assist the local firms to upgrade and become its suppliers. The most frequent and appreciated forms of assistance may come in the form of advance payment for supplies; training of staff; and quality control. Other forms include providing technology; leasing machinery; organising production; and facilitating access to export markets. Improved supplies from ‘transformed’ local firms may benefit other local firms that may obtain superior inputs than before the involvement of the TNCs in the local economy. The old local production possibility frontier is improved and a new one is created. • Demonstration effect Local firms may observe, copy or even improve production, management, control and distribution technologies used by the locally established foreign TNC. This spillover channel may be also called learning by imitation. The success of this imitation depends on the complexity of the technology and processes in question, the technological gap and the absorptive capacity of the local firms.

International firms

323

• Labour mobility If local labour is employed, trained and gets experience with a TNC, and if it later moves to the local firms or creates new firms locally, then the knowledge and experience of this labour (and management) may assist the local economy to upgrade. These spillovers are strongest if this labour and management is employed in local firms that belong to the same or a related industry. The reason is industryspecific knowledge that may not be either transplanted easily or relevant elsewhere. However, if TNCs hire and keep the best local labour, local firms may experience a negative externality. The existence, quality and strength of links between a TNC and the local environment depend in part on the objectives of the TNC. If it wishes to minimise costs of production, then the relation with the local business environment may have a hierarchical form and relatively low local embeddedness. However, if the TNC pursues a policy of cooperation, rather than arm’s-length hierarchy, the linkages with the local businesses may be deeper. This may take the form of technology transfer, training of labour, sharing information about future production demands and the like. But these linkages and efficiency spillovers on the local economy also depend on the ability of the local businesses to adjust, absorb and apply knowledge from TNCs and serve as their reliable partners. In Ireland, for example, an economy with a high level of involvement by TNCs, efficiency spillovers were noted in domestic plants that operate in high-technology industries. No evidence of such spillover was found in the low-technology industries (Görg and Strobl, 2003, p. 593). Similarly, Barrios et al. (2004) found evidence of spillovers from TNCs in Ireland58 and Spain, but not in the case of Greece. The capture of positive technological spillovers depends on the size, number and absorptive (learning) capacity of the local firms. Some minimum technological threshold is necessary to start to benefit from technological spillovers. In contrast to affiliates of TNCs located in Ireland and Spain, affiliates located in Greece were predominantly in traditional and mature low-technology industries, where the possibilities of technological spillovers are less likely to occur. The absorptive capacity of a firm is in the function of this firm’s distance from the technological frontier in its industry. If a firm operates at or close to this frontier (it has ‘learned everything’ in its industry), then this firm has a high spread capability for technological externalities. If the absorptive capacity in a local firm is high, then this speeds up the technological spillovers from a TNC. Productivity gains from spillovers from TNCs on home firms are diminishing as local firms increase their absorptive capacity and approach the technological frontier. As local firms approach this ‘ceiling’, the local productivity gains are much less obvious. The problem is that local firms (domestically owned) may be or are heterogeneous. They differ in their capacity to benefit from the presence of foreign firms and their superior technology. Technological externalities are only a part of the overall externalities that come from the operation of TNCs in the host (local) economy. Another part concerns pecuniary externalities that mount up from the presence of TNCs. This part deals with an increase in local orders from TNC affiliates, or from competition for the supply of the same local market, or rivalry for local inputs (natural resources, energy, semi-finished inputs or qualified and experienced labour and management). Technological efficiency spillovers on the local economy may sometimes take the form of imitation. Bangladesh did not have a developed textile industry before the end of

Evolutionary Economic Geography

324

the 1970s, while this industry is now the country’s principal exporter. Desh Garments, for example, was a small Bangladeshi firm established in 1977. It sent 130 of its production and management staff for training in Daewoo’s state-of-the-art factories in South Korea. Following a six-month training period, they returned to Bangladesh to start a shirtmaking company. The most useful part of the training was not how to operate the technical side of the textile machinery, but rather how to transfer the Korean experience to Bangladesh with regard to exchange controls and the protectionist trading system. A year after a successful collaboration, Desh cancelled the cooperation agreement with Daewoo. Of the 130 Desh workers trained by Daewoo, 115 of them left during [the] 1980s to set up their own garment export firms. This explosion of garment companies started by ex-Desh workers gave Bangladesh its textile industry, currently responsible for three-quarters of its exports. (Margalioth, 2003, p. 175) When a firm wants to locate and produce abroad it does not necessarily need to export capital. A firm may rent capital abroad rather than purchase or build a production unit. Instead of using its own funds, a firm may borrow in its home, host country’s or a third country’s financial market. In the case of fixed exchange rates, countries may enter into an ‘interest rate war’ in order to attract capital into their economies. Integrated capital markets with harmonised rates of interest and mobility of capital may prevent this outcome. While labour markets are most often regional (in Europe), capital markets are national and, with liberalisation, they are becoming international. Neither firms nor governments depend on savings in their home markets. Interest rates and demand for funds in one country are affected by money (short-term) and capital (long-term) markets in other countries because of the links among financial markets. Of course, this holds if there is confidence in foreign borrowers. Small countries are interest rate takers, so that even the national housing (non-traded good) market feels the impact from external markets through changes in interest rates. Free international mobility of capital prevents the independent conduct of monetary policy. If a country lowers interest rates in relation to third countries, then capital will flow abroad. This also destabilises the exchange rate if all else is equal. In addition, if a country increases interest rates (in relation to foreign countries) in order to curb domestic inflation, if other things are constant, capital will flow into the country. The supply of money increases, and inflation is the consequence. Free international mobility of capital introduced during the 1990s, at least among the developed market economies, is a self-policing international market device that controls the ‘correctness’ of national monetary policy and its conformity with the ‘best practice’. In this situation, the role and the existence of the IMF became largely unnecessary and irrelevant. In the case of IMF loans that kept certain corrupt and bankrupt regimes in the developing countries afloat, this IMF role could easily have been harmful. Or in Lipsey’s words: Much of the furore against globalisation is due to a correct perception that the IMF had been a harmful force in the development of many poor countries. But the correct indictment is not against globalisation but

International firms

325

against the blind application of a simplistic theory of how to integrate an emerging economy into the globalised trading world. (Lipsey, 2004, p. 9) Financial markets may sometimes favour large companies and countries and discriminate against small ones. Large companies and countries provide apparently greater security that the funds will be returned and interest paid. A large stock of assets lends weight to this confidence. These markets may discriminate against risky investments such as seabed research, new sources of energy or materials and the like. By integrating capital markets, small countries may mitigate the effect of their relative disadvantage. In theory, TNCs may generate gains through an improved allocation of resources in the target location. These come from a transfer of local resources away from less efficient to more efficient ones. The electronics industry in Taiwan has benefited from a geographical variety of foreign investors (mainly from the US and Japan). These TNCs provided education for managers and training for workers, improved the efficiency and quality of production and created stable markets and production and marketing links with local suppliers. These initiatives provided a solid foundation for the creation of an indigenous electronics industry in Taiwan, which evolved with global standards. Mexico’s misfortune in the electronics industry was that it received FDI only from declining US component-makers and assemblers. This was coupled with a lack of integration with the domestic Mexican electronics industry. Crediting the Taiwanese government for the success of its industry and blaming the Mexican government for its failure (locational outcome) would be too simplistic (Lowe and Kenney, 1999, p. 1439). Economic and business intuition, as well as theory, accepts the existence of favourable spillover effects of TNCs on local businesses. Ex ante expectations are that TNCs would increase local employment, exports and tax revenue and that some knowledge would spill over into the local economy through movement of labour or by means of purchase orders. However, clear empirical evidence about the actual nature and magnitude of these effects is still to be ascertained. This throws certain doubt on the wisdom of offering (financial) incentives to TNCs to locate in a specific area. Görg and Strobl (2001, p. F724) found that the results of studies about the presence of spillovers were mixed. Hanson (2001, pp. 14 and 23) reported only weak and mixed evidence that FDI brings positive spillover effects for the host country. Görg and Greenaway (2004, pp. 187–90) surveyed a body of the existing literature on the benefits of FDI on the host economy. They mentioned that theory points to the reasons why spillovers may arise, but they found little robust empirical evidence on the actual existence of spillovers.59 Are TNCs protecting their assets for themselves (preventing horizontal spillovers)? Or are researchers looking for spillovers in the wrong place with incorrect research tools? Empirical studies generally look at crosssection effects on industries. An increasing availability of plant-level data makes even firm plant-level analysis feasible. However, are these spillovers and linkages vague and unmeasurable phenomena? Spillovers and links include not only hard input–output relations, but also other links such as knowledge, experience, technology, common vision and shaping of future technology, products, marketing, finance, market or competition. Certain information and hard data about these ‘other links’ may be obtained through interviews,

Evolutionary Economic Geography

326

case studies, even anecdotes. Many of these elements may often have a strong subjective slant. In a similar survey of spillovers in the host economies, Blomström and Kokko (2003, p. 14) found that the results from a number of case studies about the presence of spillovers seem to be mixed. If these spillovers have an impact on the adjoining firms, these effects depend on the adequate, even comparable, technological level of local firms to absorb them, and on hostcountry policies. Mody (2004, pp. 1209 and 1218) found little evidence that FDI speeded up productivity growth and improved corporate governance. He argued that the domestic absorption capabilities appear to be crucial for the spillover effects to take place. Rodrik (2004, p. 30) mentioned ‘very little systematic evidence of technological and other externalities from foreign direct investment’. R.E. Lipsey (2004, pp. 365 and 371) reported only mixed evidence about the existence of (positive) spillovers on the local economy. Where they existed, the domestic firms did not lag far behind the technology employed by the subsidiary of the TNC. Spillovers were largely in the form of transfer of knowledge (particularly knowledge of demand in the world market) and knowledge about the ways and means of fitting national production into the global production network. This is, however, neither immediately obvious nor easy to measure or ascertain. Similarly, the link between an inflow of FDI and economic growth in a country is neither robust nor direct. Unless there are sound domestic economic policies, an inflow of FDI in a country may not wield on its own a decisive influence on a country’s growth. As there is little or no convincing evidence that FDI brings positive spillovers for the host economy, there is little support for the idea that promoting FDI is justified on welfare grounds. What is the reason for such weak or neutral effects regarding spillovers on the host economy? Answers to this question can be found in the fact that affiliates of TNCs may outcompete the domestic firms and shift demand away from the local competitors; there are time lags in the learning process by the domestic firms and adjustment to the presence and demand of TNCs; and TNCs may guard their internal advantages and prevent leaks to local firms that may turn into future competitors. Even though there is a controversy about spillovers, there is a consensus in the literature in one respect: the national economic policy regarding investment should be general and available to all firms (notwithstanding their origin or ownership), rather than specific and targeted towards a select type of investment. So why is there a special treatment of TNCs by countries even in the absence of firm evidence about favourable spillovers on the local economy? Perhaps politicians want to do ‘something’ for certain constituencies in return for votes. They try to attract TNCs that may create new jobs. In addition, governments try to lure TNCs, and by doing so they get into competition with other potential locations that offer allowances, notwithstanding the obvious costs and uncertain and potentially illusionary future benefits. The crucial questions to the local industry regarding spillovers are these: has the growth rate of productivity in the local firm(s) increased after the location of TNCs in the same industry? Which local firms increased their rate of productivity? Why? How? What is the impact on other local industries? The impact of spillovers and externalities from the location of TNCs on the host economy depends on many factors: the size, number and learning (absorptive) capacity of local firms. It depends also on the openness and policies of TNCs towards the local economy. Policies of host countries play a role too: trade

International firms

327

protection; export orientation; domestic content rules (that meddle with the technologically optimal mix of factors); financial institutions; and ownership restrictions. Hence general policy implications and advices may be hard to find. A part of knowledge and know-how is always in part tacit. It may not be transferred mechanically from one economic agent to another. In any case, the evidence about the impact of TNCs on the local industry is incomplete and sometimes casual. Do TNCs and their affiliates create positive spillovers on the local economy in host countries? The question is of vital policy importance, especially in the developing world. ‘Survey articles have found inconclusive evidence in the literature regarding the most important effects of inward FDI, especially with respect to spillovers’ (Lipsey and Sjöholm, 2005, p. 24). Theory and evidence point attention to various market imperfections regarding potential spillovers on the local economy. This may demand a certain policy action. If, when and how to use this action and with what instruments is a hotly debated issue in theory and policy practice. Even though this debate is important for public policy, its conclusion is still without clear-cut empirical findings, as well as without convincing and robust evidence for general policy advice. The experience of the Economic Development Board of Singapore may serve as a model for others that consider to attract TNCs and to increase their potentially positive spillovers on the local economy. The Board offered to reimburse a part of the salary in each TNC affiliate of an executive who is in charge of relations with local firms, in particular the one that invites local firms to take part in the affiliate’s training programme and who searches and identifies local firms as potential suppliers (business ‘talent scouts’) (UNCTAD, 2001, p. 177). Transfer pricing Some in the general public often perceive foreign ownership and control of domestic output potentials as a burden on the domestic economy. Therefore, there are some that oppose the entry and location of TNCs into the domestic economy. A more serious argument against the operation of TNCs is that they behave in the market of the host country in an anti-competitive way through various business practices such as predatory pricing, monopolisation or transfer pricing. Manipulation in prices in trade internal to a TNC in order to reduce tax liabilities has been confirmed in studies such as Grubert and Mutti (1991), Harris (1993), Rousslang (1997) and Clausing (1998). One argument against TNCs used by host countries concerns their internal (transfer) pricing system.60 TNCs internalise intermediate product and service61 markets. Prices in trade among different sister enterprises are arrived at by non-market means.62 For example, GlaxoSmithKline develops, produces and sells drugs in a global market. It conducts R&D in seven countries, manufactures in thirty-eight and sells in 191. It may not be very meaningful to ask what part of its worldwide profits is made in Britain, as if the same profits were made in transactions among independent national firms. TNCs prosper precisely because this is not the case.63 In addition, if transfer of technology is measured by international payments of royalties and fees, then about 80 per cent of payments are undertaken on an intra-firm basis (UNCTAD, 1997, p. 20). By doing so, TNCs may shift profits out of countries with relatively high taxes to those with the lowest

Evolutionary Economic Geography

328

corporate taxes (tax avoidance)64 or they may oust competition in certain locations by global cross-subsidising product lines. In order to shift profits from relatively high-tax countries such as Britain, Spain or the US, vertical TNCs may overprice imports of inputs (including intra-group borrowing) and underprice exports.65 The world’s largest TNCs such as General Electric of the US, Toyota of Japan or Nestlé of Switzerland, have paid little or nothing to the British tax authorities in the most recently reported financial years. The same holds for British subsidiaries of foreign-owned investment banks such as Goldman Sachs, Morgan Stanley, Merrill Lynch and Credit Suisse.66 Does this mean that there is one set of rules for the big guys, and another for the small ones? It is, however, true that corporate giants contribute a lot through personal taxes and VAT, but it may be surprising to find their reported ‘low’ profitability in Britain. Even though the share of collected revenue by means of corporation tax is on the decline because of reduction in tax rates and ‘financial engineering’ (transfer pricing), the national tax authorities still get their cut by other means. They do not always squeeze firms too much, as they do not want them to go and settle elsewhere. The situation that permits transfer pricing and global tax arbitrage arises from possibilities brought by globalisation. It stimulates purely domestic firms to internationalise their business operations. A TNC locates costs, revenue, borrowing and profits in the most favourable tax jurisdiction for overall group profit. An estimate for the US found that tax loss from artificial transfer pricing was $53 billion in 2001.67 It was also reported that Microsoft ‘shaved off’ more than $500 million from its annual tax bill by putting its subsidiary in low-tax Ireland in charge of $16 billion in assets in 2005.68 Tax inspection, even academic enquiry of transfer pricing, is often difficult. Comparison with competitors is obscured by a different mix of products and corporate structures. The ownership of a huge TNC such as Procter & Gamble, for example, changed hands just four times between January and May 2003.69 In any case, the pricing system, internal to TNCs may distort both the efficient spatial location of production (resource-wise) and flows of finance and trade. One way to control operations of the TNCs in the host country could be to ask that the internal pricing system treats the parent and its subsidiary as if they were two separate companies. The enforcement and control of such a requirement may be seriously endangered if there are no substitutes for these internally traded goods and various (headquarters) services. Another solution may be to harmonise fiscal systems and their enforcement in countries where TNCs operate. Nonetheless, a note of caution needs to be added. Transfer pricing is probably used much more commonly than TNCs are willing to admit, but much less frequently than is supposed by outsiders (Plasschaert, 1994, p. 13). In any case, transfer pricing is not widespread in small and decentralised TNCs or in TNCs that operate in competitive markets. Internalisation of intermediate goods or services markets within TNCs is not always done with the primary goal of avoiding taxes on corporate profits. Another possible reason is to maintain high quality in the supply of goods and services, as local or external suppliers may not necessarily be able to maintain the high quality and timely delivery standards required by the TNC. Nonetheless, fiddling with transfer pricing is more widespread in the developing countries than in the developed world. The balance of payments position of developing countries often drives them to control flows of foreign

International firms

329

exchange. Strict controls may induce TNCs to manipulate the internal pricing in order to protect and/or increase profits. The developing countries are not well equipped either to detect or to control manipulation of the internal prices of TNCs. Research and development A potential case in the host country against TNCs is that TNCs rely heavily on the R&D of their parent companies and that their head office charges for its central services in a way that might not be supervised or controlled by the host country. This can make both subsidiaries and host countries too dependent on foreign R&D and technology. Is this really the case? When General Electric acquired Tungsram, a huge and renowned Hungarian lightbulb company in 1990, it closed down Tungsram’s R&D department. However, when it was realised that light source research was so strong in Hungary, General Electric resumed and reinforced this Hungarian research in 1994. Elsewhere, empirical research in Canada found that the subsidiaries of TNCs undertook more R&D than domestic Canadian firms (Rugman, 1985, p. 468). In another example, European lift manufacturers decided to carry out their R&D of lifts for tall buildings in their US subsidiaries because there are so many skyscrapers there. Several drugs TNCs set up their operations in China. They stress that the quality of the employees is the principal reason. Not cost, or the fact that China is not fussy about animal testing, unlike other countries, India for instance. While there is historical evidence that most R&D took place in the headquarters of TNCs, this is no longer necessarily the case on a large scale. There are many instances of foreign subsidiaries developing technologies that have benefited the parent firm, especially from the mid-1980s. For example, a US TNC such as IBM had research breakthroughs in superconductor technology in Switzerland; Hoffmann-La Roche (Switzerland) developed important new pharmaceuticals such as librium and valium in New Jersey (US); Toshiba made advances in audio technology in its British laboratory; and Matsushita’s R&D facilities for air-conditioners are in Malaysia. However, in order to arrive at a solid conclusion, more evidence at a much more disaggregated level is needed. Affiliates of foreign TNCs in the US look much like the domestic US firms: there are no particular signs of headquarters effects (Graham and Krugman, 1995, pp. 73–4, 119).70 The Japanese pharmaceutical TNCs with international R&D facilities increased their patent output in comparison with the Japanese firms in the same industry that had purely domestic R&D. Hence home and foreign R&D efforts are complements, not substitutes (Penner-Hahn and Myles Shaver, 2005, p. 137). There is a trend, based on the evolutionary view of economics, for TNCs to create and promote networks of internal and external innovation (Cantwell and Piscitello, 2005, p. 3).71 Foreign asset-seeking drives for FDI and the domestic asset base motives reinforce each other. Spillovers may operate in this case. Internationalisation, at least in certain R&D activities, is an expected businessevolutionary event. Some R&D ought to be done locally in the target market because of the necessary adaptation to the local market needs and preferences. In addition, a part of R&D may be done at the location, where it may be finished in the most efficient way. Local subsidies for R&D may also play a role.

Evolutionary Economic Geography

330

Enabling factors for the international spread of R&D in particular from the start of the 1990s are manifold. They include an international widening in production (globalisation), availability of the staff and talent in the new locations for production, strategic decisions to draw on vast new markets and a pool of new ideas (in particular in China and India), new communication technologies and lower costs of a part of R&D contributed to the international widening and deepening of R&D activities. A noted international spread of R&D may be in part a puzzling affair to some observers. It was assumed traditionally that R&D was the least mobile TNC activity. R&D is a service that is highly demanding in knowledge, experience, skill, finance and the ability to accept risk. It was also assumed that this is (and may stay) in the select clusters in the developed world because a significant part of R&D is complex, tacit in nature, face-to-face communication is necessary among the researchers, and there are difficulties and costs to develop this type of know-how. There are also risks of unwanted technological and knowledge leakages in particular in locations with weak protection of intellectual property rights. Nonetheless, Motorola established its first foreign R&D lab in China in 1993. During the 1990s, India attracted R&D activities from TNCs such as General Electric (aircraft engines, medical equipment and consumer durables), or in pharmaceuticals from AstraZeneca, GlaxoSmithKline, Novartis, Pfizer or SanofiAventis. STMicroelectronics has some of its semiconductors design in Rabat, Morocco. Boeing (US), Pratt & Whitney (US), Airbus (Britain, France, Germany and Spain) and Dassault (France) have been actively cooperating with the labs and R&D institutes of the Russian aerospace industry and the Russian Academy of Sciences since the early 1990s (UNCTAD, 2005, pp. xxiv, 149). Russia has a long and strong tradition in aerospace technologies. After all, it was Russia that sent the first satellite and the first man into space. Foreigners (foreign affiliates) hold an increasing share of general R&D. In this internationalisation of R&D and the creation of new technology, foreign ownership of domestic inventions was over 50 per cent in Russia, Luxembourg, Mexico, Hungary and the Czech Republic in the period 2001–3. This was a higher share than during the corresponding period a decade ago. Foreign ownership of domestic inventions was 40 per cent in Britain during 2001–3, up from 30 per cent in the early 1990s. This indicator in US and Germany increased from 8 and 10 to 14 and 15 per cent, respectively, over the same period. South Korea and Japan reported the lowest shares for 2001–3, 4.5 and 4 per cent, respectively. (OECD, 2007, p. 162). The geographical spread of R&D activities carried out by TNCs and the selection of the actual location for this activity is driven by several factors, including: • the available sophisticated and experienced local R&D staff; • a desire to enhance the existing and acquire new competences; • the need to adapt output to the needs, preferences and requests of the local market; • subsidies or pressure from the host government to establish local R&D facilities; • the possibility of establishing a ‘listening and learning post’ in the host country; and • the prospect of capturing spillovers. However, an international spread of locations for R&D does not mean, as seen in the discussion on competition policy, that the general innovation process has also followed this global spread on a larger scale.

International firms

331

A small country often does not have the necessary resources for large-scale basic research compared to a big and developed country. By importing technology, a small country may have access to the results of a much larger volume of R&D, wherever it is carried out. It can be both complementary and supplementary to R&D already undertaken in the domestic economy. Relying on the domestic operation of foreign-owned TNCs may be a superior economic policy choice for a risk-averse (poor and small) country than being dependent on foreign supplies of the same good produced elsewhere. Technology transfers from abroad and domestic efforts in education and endeavours to master new technology are like scissor-blades: their joint effect is greater than the impact of either one alone (Pack and Saggi, 1997, pp. 94–5). Inter-firm strategic alliances in the development of technology in the EU increased sharply during the 1980s. In addition, the European Commission became heavily involved in projects on a cost-sharing basis. Over 70 per cent of private (largely nonsubsidised) strategic technology alliances were related to joint R&D of new core technologies in the fields of informatics, new materials and biotechnology. A major field of cooperation was in information technologies, as over 40 per cent of all strategic technology alliances were in this field (Hagedoorn and Schakenraad, 1993, p. 373). A comparison between established ‘private’ cooperation in R&D and cooperation sponsored by the EU found that these two forms are very similar in the case of leading enterprises. In fact, ‘subsidised R&D networks add to already existing or emerging private networks and merely reproduce the basic structure of European large firm cooperation’ (Hagedoorn and Schakenraad, 1993, p. 387). This being the case, it is difficult to understand why leading and large firms in the EU need subsidies. If the ‘official’ (EUsponsored) spatial network largely reproduced the already existing ‘private’ one, then it may be redundant. Financial resources could have been used elsewhere (for instance, to fund programmes that are not in the field of informatics, such as biotechnology or education or infrastructure). Is such replication of R&D networks the outcome of the lobbying power of powerful firms or is it necessary to accelerate R&D in the private sector because of significant externalities?72 Perhaps such waste of scarce public money will be checked in the future by WTO rules on subsidisation. There is a tendency to reduce the risks and costs of R&D within TNCs. Over 70 per cent of international strategic technology alliances between companies from the EU and NAFTA are focused on R&D (Hagedoorn, 1998, p. 184). This may be an important incentive for mergers and acquisitions in the pharmaceuticals industry (the major reason is still the very fragmented nature of the pharmaceuticals industry compared with car manufacturing). High risk and uncertainty, as well as excessive R&D costs, drive firms to centralise some of these functions (usually in their country of origin). Another reason is the exploitation of host countries’ incentives to R&D (subsidies, tax breaks, secure contracts). In this case, the basic research remains in the headquarters, while subsidiaries undertake applied development according to local demand, regulations and incentives. It is often forgotten that many firms that create knowledge are neither TNCs nor large (for example, in the areas of computer hardware and software). An important condition for their creation in certain locations is that they operate in a competitive environment. This is often lacking in small countries.

Evolutionary Economic Geography

332

Host countries and transnational corporations The greatest power of TNCs stems from their high international mobility to enter and exit an industry or a location.73 TNCs can act as spatial capital arbitrageurs. They may borrow in countries where the rate of interest is lowest and invest in countries where they expect the highest returns. TNCs may spread overheads and risk among their subsidiaries. These enterprises also extend control over international markets. If a subsidiary is producing final goods, then other parts of a TNC may increase export components to this subsidiary. Many TNCs create sophisticated and complicated technologies. They avoid the transfer of this technology through the market in order to prevent competitors from copying it. The longer the technology gap with imitators lasts, the longer the TNC can behave like a monopolist. Thus, a TNC usually transfers technology first among its subsidiaries. It is popularly argued that TNCs invest most often in fast-growing manufacturing industries such as electronics and medicine-related industries. These industries thus ‘fall into foreign hands’, possibly giving foreigners undue influence in the host country and interfering with sovereignty.74 One example was when the US introduced a ban on the export of technology for the gas pipeline from the then USSR to western Europe in the early 1980s. Many subsidiaries of US TNCs in western Europe were affected by this decision. It is much harder for a host country to influence the parent country of a TNC through a subsidiary. Nonetheless, the gas pipeline was built. New ones are under construction. In many developing countries, mineral resources, including oil, are the property of the state. In Mexico for instance, the 1917 Constitution (Articles 27 and 28, respectively) stated that subsoil minerals are the property of the people and prohibited foreign activity in the strategic energy industries. Elsewhere, it is the government that negotiates terms of entry with TNCs. In manufacturing, the role of the government is somewhat less pronounced. In this case, the government usually sets general terms of entry, performance and exit from an industry. While the developing countries, as a rule, regulate the conditions for entry of TNCs with greater scrutiny, developed countries control their exit and the possible consequences regarding job losses and the environment. Relations between TNCs and host countries may sometimes be quite tense. Mining is an industry that requires a huge amount of investment before commercial exploitation can begin. Different countries often compete and offer incentives for FDI and effectively engage in a ‘locations tournament’ (David, 1984). Tournaments may be a costly and wasteful national policy if a country has more than one location offering various (tax) incentives and that battle against each other to try to attract the same foreign investment. At this stage a TNC has the strongest bargaining position. When TNCs locate their operations in host countries they may, for example, react to changes in the host country’s tax system or interest rates quite differently from domestic firms, which may not be able to withdraw from the home market. Thus, TNCs may become a threat to the host country’s national economic policies. For instance, Delphi is one of the largest manufacturers of automotive parts. This TNC relocated operations from Argentina to Brazil in 1998 and then back to Argentina in 2000 in response to subsidies (Albronoz and Corcos, 2007, p. 2). To counter this danger, the common market member countries should coordinate and harmonise their policies regarding competition, capital mobility and TNCs.

International firms

333

If TNCs produce final goods in host countries and if they import raw materials and components from countries which are outside the common market instead of purchasing them from the local suppliers, then they may jeopardise the process of integration within a common market. The external dependence of the area may increase instead of being reduced. Where the member countries of a common market or other type of international economic integration compete among themselves for FDI in the absence of an agreed industrial policy, it is unreasonable to expect the operation of the TNCs to result in an optimal spatial allocation of resources (Robson, 1983, p. 32). This has happened in the Caribbean region. Transnational corporations may have an adverse impact on the allocation of resources in the host country if their operation accounts for a significant proportion of production, employment, purchases and sales. Thus, the operation of TNCs may have a greater influence in certain segments of industry in developing countries than in developed countries. This state of affairs demands a coordinated approach towards TNCs by the regional groups in the developing world, as was the case with the controversial Decision 24 (1970) in the Andean Pact.75 It also requires the establishment and enforcement of a common industrial policy. This can be supported by the regional development banks and/or by joint planning. Openness to and participation in an integrated network of international production of TNCs will contribute to the growth potential of the developing countries. The selection of Costa Rica over other potential locations by Intel in 1996 offers certain interesting and relevant policy advices (Moran et al., 2005, p. 279–80; Lederman et al., 2005, pp. 378–9). Intel was considering locations in Indonesia, Thailand, Brazil, Chile and Mexico for a testing and assembly plant for its latest Pentium microprocessor. Costa Rica was not even on Intel’s shortlist of potential locations for investment. Costa Rica had a rare Latin American reputation of political stability. The country already had certain FDI in low-skilled garment factories. The newly restructured domestic FDI promotion agency had a highly trained, motivated and well-paid staff, as well as strong support from the government. It was realised that ample travel funds are necessary for the agency and that it may take several years before concrete results may appear. The agency would provide both information and feasibility studies to potential investors, which is something that the market does not supply on its own. It took the Costa Rican FDI promotion agency two years to get an appointment with senior Intel executives. In order to qualify for the Intel’s shortlist of potential locations, the Costa Rican government committed itself to construct a new cargo terminal at the national airport, to improve a power supply at the potential location of Intel’s subsidiary, to offer a labour union-free business environment and, together with Intel and the Costa Rican Institute of Technology, to create special training courses. Costa Rica addressed Intel’s concerns and requests in a faster, more effective and credible way than other potential locations (e.g. Mexico). In addition, Costa Rica ‘had to match’ tax breaks offered to Intel by other potential locations (full exemption from income taxes for the first eight years and a 50 per cent exemption for the next four). The government offered all these concessions in the belief that getting Intel into Costa Rica would incite other TNCs to come to the county (this expectation about the ‘snowball effect’ turned out to be correct later on). A

Evolutionary Economic Geography

334

potentially high profitability in itself was an insufficient motivation for the first big TNC entrant. Hence, in this and other cases, aggressive investment promotion mattered. Free economic zones (export-processing zones) were quite popular tools in attracting FDI for those manufacturing processes that could be unbundled. Many countries created these zones and offered the TNCs that located there various advantages. These do not refer only to favourable tax treatment and low-cost labour, but also to low or no tariffs on imported intermediate goods. China, for instance, started a policy of free economic zones (export-processing zones) from the mid-1980s. Foreign investors were offered superior infrastructure (telecommunications and logistics) than were available elsewhere in China, as well as tax relief. Encouraged by the FDI inflow into these zones, China removed many FDI restrictions elsewhere in the country from the early 1990s. Nonetheless, TNCs continued to locate in the zones or in the adjacent areas largely because these areas have better infrastructure than is available in most other areas of China (Moran et. al., 2005, pp. 378–9). A large amount of investment may, however, keep a TNC as a hostage of the host country. These sunk costs with limited or no ‘salvage value’ may represent a barrier to exit from the host country. In this case, the hostcountry government can show the TNC who is boss. The host governments may renegotiate the deals with the TNCs (a very bad signal for potential new investors). This kind of danger may induce the TNCs to borrow predominantly on the host country’s financial market and transfer from elsewhere only the technical and managerial expertise. A possible closure of a subsidiary in the host country would be at the cost of some local jobs in both the subsidiary and supplying firms. The affected local workers and their families may lobby in favour of the interests of the TNC, which may find allies among industries in the host country. As long as TNCs purchase from them, both subsidiaries and local firms may work together in lobbying the government for protection, subsidies, tax breaks and procurement agreements. Resource-rich Russia has strong ambitions to process them at home at a much larger scale. Paper and pulp industries are on this list. Russia is the world’s largest exporter of cut logs. It supplies 80 per cent of Finnish lumber imports, mainly to the Finnish-Swedish timber giant Stora Enso. In order to lure (or force) Stora Enso to locate at least a part of its business in Russia, Russia increased tariff on lumber export from $7 to $15 per cubic metre in 2007. This tariff would increase to $70 before 2009.76 A serious threat to the market structure of a host country may be introduced by a TNC. It may monopolise the whole domestic market and by predatory pricing prevent the entry of domestic firms to the industry. It may introduce technologies that use relatively more of the resources in short supply (capital) and relatively less of the component that is abundant (labour) in the host country. This analysis suggests that the TNCs do not pay much attention to the overall social needs of the host countries. This is true. However, it is not a duty of TNCs to meet the social needs, including the infrastructure needs, of the host country. Nobody forces host countries to accept TNCs. The fulfilment of these social demands is the role of the host country’s government. In the presence of unemployed resources, inflation, foreign and budgetary debt, famine and underdevelopment, any hard currency investment is welcome. Any attempt by the host government to restrict the entry and operation of TNCs may dry up this trickle of capital inflow. Host countries sometimes behave (or used to) as if they want FDI funds, but not the foreign investor to handle them. Unfortunately,

International firms

335

the developing countries are those that most need these investments. TNCs often ask what will happen after they invest and start production: will the utility companies increase prices? Will the employees request an immediate increase in wages? Will the consumers demand bigger loans as the TNC is a large, successful and rich company? Economic adjustment and reform in the developing world may potentially be facilitated by the presence of TNCs. These corporations have the know-how, the ability to raise funds and the global marketing channels that help export growth. They may perhaps create and maintain relevant spillover effects such as linkages with the existing local firms that may prompt economic adjustment: the creation of new local firms and the restructuring of existing ones. Culture matters The distinctiveness of national character and, hence, social value systems may have a certain impact on the attraction of FDI. There was certain anxiety about opening this topic for discussion for quite some time because of ‘political correctness’ and because this issue is out of the scope of neoclassical equilibrium theory. Nonetheless, the gloves are now off for a discussion. Economic performance is influenced by culture (how we do things here) and by social norms. For instance, Yutang Lin (1935) in his famous discussion of the Chinese character remarked (p. 58): the Chinese are a hard-boiled lot. There is no nonsense about them: they do not live in order to die, as the Christians pretend to do, nor do they seek for a Utopia on earth, as many seers of the West do. They just want to order this life on earth, which they know to be full of pain and sorrow, so that they may work peaceably, endure nobly, and live happily. Of the noble virtues of the West, of nobility, ambition, zeal for reform, public spirit, sense for adventure and heroic courage, the Chinese are devoid. They cannot be interested in climbing Mont Blanc or in exploring the North Pole. But they are tremendously interested in this commonplace world, and they have an indomitable patience, an indefatigable industry, a sense of duty, a level-headed commonsense, cheerfulness, humour, tolerance, pacificism, and that unequalled genius for finding happiness in hard environments which we call contentment – qualities that make this commonplace life enjoyable to them. And chief of these are pacificism and tolerance, which are the mark of a mellow culture, and which seem to be lacking in modern Europe. If this were true and acceptable, try to replicate it outside China. This type of hard work and cooperative relations within a closely knit and often family- or clan-type network contributed to the ‘East Asian miracle’ during the 1980s.77 However, this idealised culture-related explanation of economic development success also ought to be seen in another light. If based on culture-related contacts and networks, the success bypasses market mechanisms and open cooperation. It is also linked with an array of relations from highly qualified over strictly personal and dull alliances to corrupt behaviour. Market

Evolutionary Economic Geography

336

sharing and carving, as well as ‘settlement of accounts’ among the closed and rival Chinese clans, may be ferocious and occasionally end up in murders. This ‘discount’ ought to be presented right next to the ‘idealised’ culture-based account of the development success. The role and impact of religion on economic prosperity is often mentioned in Europe. The cases in question are certain Protestant and Roman Catholic countries in western Europe and differences in their material wealth (in favour of Protestant countries). Some argue that one of the principal causes of this difference is a ‘better’ work ethic in the Protestant countries (work and save) and tendencies for carnivals and pleasure in the Catholic countries. The style of life that favours carnivals and pleasure is quite acceptable, even desirable and preferred to other lifestyles to some (many?) in certain societies. Let us forget for the moment about Italy and let us take a look only at the area that we now know as Germany. In this country, for instance, the predominantly Protestant areas were in the past more economically prosperous than the mainly Catholic ones (although Bavaria caught up from the start of the twentieth century). The work ethic may not be the principal explanation of this difference in material wealth. Education, i.e. the creation of human capital, matters a lot. Martin Luther reasoned and argued that each person should regularly read the Bible and be in continuous touch with the word of the Lord. He translated the Bible into German in 1534 to make it more approachable to the people. So the Protestants gave a high value to literacy and continuous education. The more the literate the people are, the more they read (and write), the more educated they are and consequently the higher is the general pool of human capital. Hence the European testimony about the great importance and lasting value of continuous education and learning on material prosperity. Countries can be successful in terms of material wealth even if they have different cultures and social values. Countries in this group are as diverse as Australia and Dubai or Sweden and Japan. There is no general cultural formula for the national material (economic) success. Even though a lot needs to be done regarding research in this area, there are certain elements and norms that support economic development (Beinhocker, 2006, p. 430): • Individual behaviour Individual accountability, strong work ethic, optimism, the belief that hard work and a moral life pays in this life, not just in the next, as well as vision that one may influence life and the world around – all these help a lot. Fatalism reduces personal motivation. • Cooperative behaviour Cooperation pays, as life is a positive-sum game. Fairness and generosity ought to be recognised and rewarded, while cheaters and free riders ought to be punished. Strong values within a family, trust and internal ties in certain cultures may present a problem in building wider cooperation linkages. The belief that life is a zero-sum game puts at risk trust and cooperation among people. Corruption, low trust and fraudulent behaviour may be deeply embedded and it is linked with poverty. • Novelty Cultures that are open to experimentation, innovation and ‘heresy’, the ones that allow certain degree of freedom and are tolerant, tend to be more innovative than the others. Various types of competition, recognition and rewards support innovation. Cultures that are too egalitarian tend to reduce incentives for risk-taking.

International firms

337

• Perception of time Cultures that ‘live only for today’ have problems with the work ethic, coordination of actions and saving/investing (for the future), as tomorrow does not play any role. ‘Partnership’ with ancestors and future generations does not matter either. The cultures that have opposite behaviour on all these accounts enjoy significantly superior economic performance and wealth. No culture is equally strong on these norms. The importance of each norm may vary and change over time. As people learn and interact with other cultures, certain crossfertilisation is always possible. Many old cultures in Europe adjusted to the new competitive situation in the EU. European cultures are not on the way towards homogenisation and they have certain problems in integration of some Muslims. Communications and the media spread certain cultural values throughout the world (not always for the better, as is the case with the chewing gum, tobacco, spirits, tattoos, piercing or drugs). Organisation Efficient management of huge, complex and spatially spread companies such as railways demands a separation of certain administrative functions and their concentration. These functions often include personnel, procurement, certain repairs, finance or marketing. Head office may centralise and unify these activities (the U-form of organisation). Inefficient replication of business is avoided and costs may be reduced because of specialisation. Other companies may be efficiently organised in a multi-plant manner (the M-form of organisation). In general, each one acts as a more or less independent business centre (food production, or car production after the First World War). Transnational corporations treat their general activities in different countries as a single-business operation. Therefore, their financial service is often centralised (or confined to a very limited number of locations) as this may best meet the needs of the corporation as a whole.78 The financial service represents the backbone of the control of overall corporation operation and efficiency. This is most pronounced in the case where ownership-specific advantages of a TNC are generating high returns. Other operations such as employment, wage and labour relations are always decentralised within TNCs. Labour markets are local and often highly regulated, so decentralisation of these issues is the optimal policy choice for TNCs. Rapid international mobility and multiplant coordination and control of a growing part of business activities are among the greatest advantages of TNCs over other firms. For example, Nike (an American ‘producer’ of sports goods) keeps its R&D in Oregon, but subcontracts the actual fragmented manufacturing of clothes and shoes in more than 700 factories in fifty countries, principally in Asia.79 If wages increase in one country, Nike simply moves production to another. Market entry Modern technologies complicate the employment impact of FDI. These technologies substitute capital for labour. A greenfield investment (building a new factory) is an attractive business option when blending existing assets and structures with others is undesirable, risky and costly. Investors can tailormake the production site, and

Evolutionary Economic Geography

338

manufacturing and management techniques, to suit their particular needs. Control over the subsidiary is full. Previous management habits and labour relations are not inherited. This type of investment increases employment in the host country if it does not put domestic competitors out of business. It is also practised in locations that do not have available or suitable local firms for joint ventures or takeovers. A takeover (merger and acquisition) of an existing firm is a preferred business choice if the cost of a greenfield entry exceeds the cost of a takeover. The investor may immediately start production, but this does not add to the national output capacity in the host country at the time of entry (it is just a transfer of ownership). This is a fast way to enter rapidly changing markets, and hence, in general, a more common way to enter foreign markets than through a greenfield investment. This holds particularly for entry into the developed market economies. Takeovers do not necessarily increase employment and may reduce it if the merged firm rationalises the existing activity. A foreign-located subsidiary may begin as a unit for marketing of the final good. If it develops further, it may become a product specialist, which may increase employment (direct and indirect) of a particular type of labour. However, one has to remember that employment is determined by demand for final goods and services. Takeovers (mergers and acquisitions) provide means for the ‘external’ growth of TNCs. They are attractive to TNCs that have wide international marketing networks and are tempting to sellers who are interested in penetrating the widest possible market. This is enhanced if the acquired (or merged) enterprise is not linked to any industrial group, which helps to avoid conflicting interests and corporate cultures. This is why the AngloSaxon enterprises dominate cross-border mergers and acquisitions in Europe. TNCs may also grow ‘internally’ through subsidiaries. Mergers and acquisitions are, however, virtually unknown in Japan. Wages As far as wages are concerned, there is overwhelming evidence that TNCs in all kinds of countries pay relatively higher wages to labour of a given quality than local firms do. The reasons for such a wage policy include the following elements (R.E. Lipsey, 2004, pp. 346, 351): • TNCs may be forced to offer higher wages by host-country regulations or by pressures from their country of origin about ‘fair pay’. • Local labour may prefer domestic firms and it ought to be compensated to overcome this preference. • If TNCs bring certain proprietary technology, they may wish to reduce labour turnover. By offering higher pay they may retain labour, hence they may reduce unwanted technology leaks to domestic firms. • TNCs may wish to find and attract better workers and managers. More knowledgeable local firms may identify, get and keep these workers without this wage premium. • Certain TNCs operate in higher-wage industries of the economy. They hire highly educated managers and sophisticated workers. Differences in wages may also be due to the fact that the home firms do not sufficiently value a resource that is in short supply. If affiliates of TNCs attract better workers, then

International firms

339

vacancies in the domestic firms may be filled by less-well-trained labour, which may have an adverse effect on the growth of production in the host country. Although technologies used by the TNCs in the host countries may not be the most up-to-date, they can be superior to those which are currently in use in developing countries (environmentally sound technologies are an obvious example). Local content Certain companies, such as Nike and Wal-Mart, have sourcing structures that are more geographically spread than their sales. Nike, for example, sources 99 per cent of its products offshore (principally from China and south-east Asia). However, most of its sales (52 per cent) are in its home US market (Rugman and Verbeke, 2003, p. 14). Local content rules were often used in the past as means to promote, even force, linkages between TNCs and local firms. Local content usually means ‘national content’ (value added within a country) or value added within the group such as the EU or NAFTA. Other requirements with the same policy objective included trade (export performance) and foreign exchange balancing demands. However, over the past few decades, most countries have strategically altered their economic policies from the general protection of domestic firms to widespread openness. In addition, the WTO 1995 TRIMs Agreement (Article 2) requires the phasing out of TRIMs.80 While there might be certain initial successes in the forging of linkages with the local suppliers, forced requirements on a TNC’saffiliate may jeopardise the efficient technology mix in the production. As such, this can easily endanger the efficiency in the production process, lower competitiveness and reduce incentives for other TNCs to come and locate in that country. A far superior and more sophisticated policy that can be employed by governments to foster links and spillovers between the local industry and affiliates of TNCs is to provide information about the local business environment, to act as a matchmaker with the local firms and to upgrade the technology level of the local firms through education, training and finance. The local content requirements for locating FDI in a particular country may be criticised on the grounds that they distort investors’ input choices, stimulate suboptimal input mixes and potentially increase prices. In a survey of 682 projects, it was found that in 83 per cent of cases in which there was a requirement to accomplish the objectives of trade-related investment measures (TRIMs) such as local sourcing and exporting, the firms planned to do so anyway (UNCTC, 1991a, p. 4). That is to say, TRIMs were redundant. Intervention Suppose that two countries want to integrate in order to enlarge the market and attract the location of TNCs. If market imperfections permit rents (above-average profits), then the governments of the integrated countries may wish to intervene. The larger market of an integrated area may be better able to absorb the cost of intervention owing to the spread of such costs than would be the case for individual countries acting alone. For instance, a simplified example based on the metaphor of a prisoner’s dilemma is presented in Table 5.3.

Evolutionary Economic Geography

340

Table 5.3 Profit in the aircraft industry without intervention Britain/France Production No production Production No production

−3; −3

10; 0

0; 10

0; 0

Suppose, initially, that there are just two firms capable of producing aircraft: British Aerospace in Britain and Aérospatiale in France. Assume also that, because of sunk costs, R&D and economies of scale, only one firm can produce aircraft efficiently (profit-wise) within the EU and that public authorities prefer to purchase domestically made goods. The figures in Table 5.3 then show the profit of the two firms. If there is no intervention by the government, the firm that moves first captures the market and makes a profit. If both produce, both lose; if neither produces, there is neither gain nor loss. If only one country has the ability to produce aircraft, the government of the other may then try to persuade a foreign TNC to locate its production within the confines of its borders, thus putting the potential competitor from the partner country out of business. This possibility for a geographical location of production is a strong case for a joint treatment of TNCs by the integration groups. Now suppose that there is no domestic producer of aircraft in Britain and that the domestic government decides that it may be sensible to have aircraft production located at home and to move first to strategically and irreversibly pre-empt any other player. The reasons may include employment, export and prestige, but also, and more importantly, various externalities, including obtaining the leading edge in one of the high-technology industries and also national pride. Some early movers sustain their position for decades. For example, Procter & Gamble, Unilever and Colgate have been international leaders in washing powder production since the 1930s. With this in mind, the government decides to invite Boeing to come and locate its production in Britain. As bait, it offers various subsidies, early purchases and protection to capture the aircraft market in the integration arrangement with France. The major reason for the subsidy is not simply to increase export sales, but rather to improve the terms of trade and secure rents for the home firm, a cluster of related domestic enterprises and, finally, for the country itself. Of course, such a policy may have a significant balance of payments effect in the medium and long terms. Hence, the structure of markets and the operations of TNCs matter for the spatial distribution of production. The neoclassical equilibrium model deals with the given and perfect resources and capabilities, whereas the new theory and the evolutionary model of the economy study market imperfections, multiple equilibria, path dependence and government intervention in a dynamic set-up. This evolutionary model tries to suppress market constraints and to push economic frontiers outwards. It considers economies of scale, externalities, differentiated products, changing technology and FDI. These are all features of modern manufacturing. The evolutionary theory questions the proposition that free markets may successfully take advantage of the potential benefits in the new situation. Such an approach is different from the neoclassical one, in which TNCs were, by definition, excluded from consideration. The assumption in free markets is that there are no grounds

International firms

341

for trans-border business activities as the spatial and sectoral allocation of resources is perfect and full (first-best solution). The new theory and the evolutionary model explain why countries can trade not only when their resource endowments and production capabilities are different, as in the neoclassical situation, but also when their resource endowment and production capabilities are identical. The case in question is the intraindustry trade (such as, intra-EU trade) in passenger cars. One thing, however, ought to be clear from the outset. The evolutionary theory does not fully replace the neoclassical one. It considers only market imperfections that can be mitigated by intervention, which may introduce an adjustment instrument into an already highly imperfect situation. Table 5.4 shows what would happen to the profits of the two aircraft producers if the British government subsidised its (foreign owned/controlled) firm with monetary units that equal 5. If the firm located in Britain decides to produce, it will always have an advantage over its non-subsidised French rival. When market imperfections exist, the British government can influence the geography of production (spatial location of resources and specialisation). Nonetheless, if the choice of national champions is to be a good one, the government should be competent and well informed, otherwise the result may be costly commercial failures, such as the FrancoBritish Concorde project or computers in France. Many governments in east Asia have intervened successfully in their economies. However, governments in many east European and developing countries have intervened much more during the socialist period, and yet have singularly failed to achieve the economic successes seen in east Asia. In any case, intervention is facilitated when the number of potentially competing firms is small and production output is standardised. Engineered comparative advantages of countries and firms are becoming more important in modern footloose industries and are tending to erode inherited comparative advantages. For example, trade within the EU is to a large extent of an intra-industry Table 5.4 Profit in the aircraft industry with intervention Britain/France Production No production Production

−3; 2

10; 0

No production

0; 15

0; 0

character. It is driven much more by economies of scale than by the ‘classic’ comparative advantages of those countries. If pushed to the limit, a national reaction to a monopoly in a foreign country is the creation of a national monopoly. That is fighting fire with fire (Curzon Price, 1993, p. 394). In order to intervene/subsidise in an intelligent way, governments need a great deal of information that is quite costly to obtain: information not only about current and potential future technology and demand, but also about the strategies of other governments and TNCs. If they choose wisely, strategic policy can be a superb device. In practice, a subsidy in one country may provoke retaliation in another in the form of a subsidy or a countervailing duty. The retaliation and counter-retaliation cycle makes everyone worse off. Integration may offer some advantages to developing countries in such a situation. It may inspire these countries to negotiate the distribution of strategic industries within the area or attract TNCs in order to maximise positive externalities and reduce unnecessary

Evolutionary Economic Geography

342

subsidies. The basic argument for the government’s involvement may be that without intervention in the area, and owing to imperfections, there may be underinvestment in a strategic industry. If, however, intervention is not well managed, overinvestment may be the consequence. At the end, consumers and taxpayers would have to suffer and foot the bill. The simple model of strategic investment, industrial and trade policy is based on the expectation that the subsidised home production of a tradable good or service shifts monopoly profits (rents) to the home country and to the firms owned/controlled by it. These profits should be over and above the cost of the subsidy. If the domestic firms are affiliates of foreign TNCs, then the effect on the home country’s welfare may be uncertain. No matter what the circumstances are, the expectation that profit-shifting may enhance the home country’s welfare holds only when foreign countries do not retaliate against domestic subsidies. A cycle of retaliation and counter-retaliation would make everyone worse off. In addition, when there is a liberal treatment of FDI, bilateral trade deficits may give off misleading signals. For example, if Japan (or any other country) invests in China in order to take advantage of relatively low labour costs and to export output to the US, the bilateral deficit in trade between the US and Japan may shrink, but the overall US trade deficit may increase because of extra US imports from China. Governments need to bear in mind that general favours handed out to domestic firms (subsidies) may trickle down to foreign beneficiaries located within the confines of the jurisdiction of the government. A more effective policy may be to use subsidies to develop and upgrade the skills and health of domestic human capital, as footloose capital is increasingly attracted by, among other matters, locally created factors such as the local availability of skilled, highly trained and experienced labour and management. In advanced and integrated countries such as the US, the constituent parts may compete with one another by offering substantial grants (and grant equivalents) in order to attract world-scale manufacturing projects. The size of these grants is increasing. In 1984, Michigan offered state and local incentives worth $120 million (equivalent to $14,000 per job) to attract Mazda; in 1986, Indiana offered $110 million ($51,000 per job) to SubaruIsuzu; in 1989, Kentucky offered $325 million ($108,000 per job) to Toyota (UNCTC, 1991a, pp. 73–4); in 1983, Alabama gave $252 million ($168,000 per job) to Mercedes-Benz;81 and North Carolina handed out $130 million ($108,000 per job) to BMW. In spite of strict rules of competition, countries in the EU are sometimes allowed to dispense ‘incentives’ to TNCs to settle within their confines. Britain gave $89 million ($29,675 per job) to Samsung in 1994, while France granted $111 million ($56,923 per job) to MercedesBenz and Swatch in 1995 (UNCTAD, 1995a, p. 18). In 1996, Dow Chemicals received a subsidy of $6.8 billion for an investment in the petrochemical industry in Germany ($3.4 million per job), and in 2000, Alabama gave an incentive package worth $158 million to Honda to locate a $400 million assembly plant that would employ 1,500 workers ($105,333 per job) (UNCTAD, 2002, pp. 204–5). Similarly Germany offered hefty grants to firms to settle in the eastern part of the country throughout the 1990s after the reunification of the country. These kinds of subsidies and participation in ‘location tournaments’ are well beyond the financial capabilities of many developing countries. Developed countries were leading these location tournaments for FDI into which many developing countries were drawn. However, the price paid in terms

International firms

343

of such handouts to TNCs, additional infrastructure and environmental damage may more than offset potential gains from the location of TNCs. In 2007 the European Commission approved the Czech government regional grant package to Hyundai, totalling Є195 million, to locate a carmanufacturing plant in the Moravia region. The total investment cost is Є1.148 million. Even though the Commission does not, in general, favour grants, it was satisfied with the overall cost–benefit contribution of the project to regional development.82 Direct subsidies are not always a winning strategy for the location of TNCs. Several southern states in the US fiercely competed with financial incentives to attract a new Toyota car-assembly plant. Hundreds of millions of dollars in subsidies were at stake. In spite of these potentially highly attractive incentives, Toyota decided to locate its new plant in Ontario (Canada) in 2005. Explaining why such choice was made, Toyota stated the quality of Ontario’s workforce. The Japanese auto-manufacturing companies located in the southern part of the US were deceived by the low level of training of the local labour. Trainers in the earlier Alabama plant had to use pictures to explain to some of the local workers how to use hi-tech equipment. The Alabama governor wanted to change this situation. He intended to increase taxes on the rich in order to get funds to improve the state education system. The irony in all this is that the voters overwhelmingly turned down this proposal in 2003. Another big selling-point of Ontario is the Canadian national health insurance system. This saves investors large sums of benefit payments compared with their outlays in the US. The basic health insurance system is a right in Canada, while it is a privilege in the US. So one may be tempted to say that the Canadian taxpayers were, in effect, subsidising Toyota’s location decision. Krugman argued that the American, not the Canadian, taxpayers were hurt by the ‘northward movement of auto jobs’. These carmaking jobs pay health benefits and the US remains without them. Funny, isn’t it? Pundits tell us that the welfare state is doomed by globalization, that programs like national health insurance have become unsustainable. But Canada’s universal health insurance system is handling international competition just fine. It’s our own system, which penalizes companies that treat their workers well, that’s in trouble. … treating people decently is sometimes a competitive advantage.83 The self-reinforcing aspect of FDI begins to operate only after a certain level of development. The developing countries ‘which are already doing well in these categories do not need location tournaments. The others are not likely to profit from them’ (Wheeler and Mody, 1992, p. 72). The ability of the developing world to attract FDI depends on many factors that these countries cannot control. There are, however, certain elements that can be controlled by the developing countries. These countries need to try to stabilise their macroeconomic and political situation, improve the quality and quantity of human capital and infrastructure, liberalise trade and investment policies, actively promote their advantages to the business community and study potentials for integration or preferential trade with other (neighbouring) countries. Such a policy approach would stimulate domestic private investment and growth. It could also be argued that they might wish to pursue such policies even if they had little effect on FDI.

Evolutionary Economic Geography

344

In an examination of annual FDI flows (1960–90) towards the major integration groups in the developing world, Jovanović (1995) found no difference in FDI flows before and after the creation of the integration group. The national economic (and political) situation played a much more important role for the attraction of FDI than did participation in the integration deal. This conclusion was later reconfirmed by Blomström and Kokko (1997, p. 39). In considering the situation in Ireland, it was found that recent FDI growth has taken place at a time when the relative value of Ireland’s incentives has been eroded. This erosion stems both from domestic reductions in incentives and from the increasing use of regional incentives elsewhere in the EU. This may suggest that incentives are necessary, but not sufficient to attract internationally mobile investment. (Braunerhjelm et al., 2000, p. 85). Corporate social responsibility One of the debates instigated by the political left concerns where social responsibilities of firms begin and where they end. In order to increase local embeddedness and demonstrate corporate social responsibility (business ethics, even philanthropy), some TNCs involve themselves in local communities through sponsorship of local sporting and cultural events, education and training, health (HIV/AIDS) campaigns, even blood donations (of the local staff). Corporations such as Coca-Cola, McDonald’s, Motorola and Nike publish annual reports on social responsibility. Some may argue that this may be no more than a form of advertisement. One needs to recall the situation during the First Industrial Revolution in Britain, when industrialists who modernised factories lobbied in the early 1800s for the introduction of labour laws that curbed child labour. A noble social goal, no doubt. In essence, this was done to gain further competitive advantages over older and smaller plants that employed children. Pressing for the social responsibility agenda, the incumbent firms may turn this to be in effect a cost-increasing measure, a business entry barrier for potential new entrants into the market. The social responsibility agenda ought to be socially and business friendly, not ‘business owned’. Cemex of Mexico, for example, is involved in a wide range of community development projects around the world; in the Philippines it supports educational initiatives for children, in Costa Rica it offers scholarships, while in Egypt it assists the education of girls (UNCTAD, 2006, p. 233). Whether this is a calculated public relations ploy to avoid criticism (as in the case of some Japanese and German TNCs that operate in Britain) or a real and deep commitment towards the local communities remains unanswered (Dicken et al., 1995, p. 41). Self-interest often plays a role in business ethics. A day-care facility provided by a firm for the children of its employees may assist the education of children or the reduction of teenage crime. A good and responsible deed towards the local community, no doubt. However, the real reason for this action, which costs money, may be the firm’s wish to reduce the absence of its employees from work and to reduce costs that come from lower production.

International firms

345

In any case, TNCs endeavour to present themselves as good corporate citizens that are socially and environmentally responsible, that bring benefits to the local community and the host country. Nonetheless, rational firms, in principle, do not normally invest funds into being ethical or ‘green’ unless they expect to obtain something in return from the general business environment. They always expect such actions to bring, for instance, more clients or lighter taxation from the government or both. In addition, there is a general social dilemma regarding firms that produce cigarettes, alcohol or arms and that wish also to be socially responsible. The Global Compact, launched by the UN in 2000, is a large voluntary ‘corporate citizenship initiative’. The 3,800 participants, of whom 2,900 are from the business community, adhere to ten principles that cover issues such as human rights, labour standards, environment and business ethics. The ten principles are: Human rights 1 Protection of internationally accepted human rights. 2 No involvement in human rights abuses. Labour standards 3 Respect of freedom of association and collective bargaining. 4 Elimination of forced and compulsory labour. 5 Abolition of child labour. 6 Removal of discrimination regarding occupation and employment. Environment 7 Environmental challenges should be deal with on the basis of the precautionary principle. 8 Promotion of a greater environmental responsibility. 9 Creation and widespread employment of environmentally friendly technologies ought to be encouraged. Anti-corruption 10 Participants (and others) should work against all types and forms of corruption.84 These principles stem from the idea that corporations have to consider the impact of their operations on all stakeholders. The general inspiration has roots in the anti-capitalist movement from the 1970s. It is also linked to the Nestlé infant formula marketing case. Nestlé promoted infant formula over breast-milk in the poor developing countries. Samples of formula were given out free while the mothers and children were in maternity wards. Once they left them, the formula was no longer free. Parents had to buy the formula and to prepare it with water that, in certain cases, was contaminated. Sometimes they used more water than necessary, in order to make the formula last longer. This provoked malnutrition, disease, suffering and, in certain cases, the deaths of babies. Hence, a TNC could have used it, under the umbrella of ‘humanitarian aid’, to create markets. In 1977 this provoked fury in the US against Nestlé which also spread through Europe. The widest ‘international community’ endorses the ten Global Compact principles. The stakeholders include the business community, labour organisations and NGOs such

Evolutionary Economic Geography

346

as Transparency International and Amnesty International, as well as international organisations. The problem is still the global application of these principles on the ground, both by developed and developing country players. Hence, the anti-globalisation movement found a new area to be concerned about. National and international regulation The basis of the OECD Declaration on International Investment and Multinational Enterprises (1976) and the OECD Guidelines for Multinational Enterprises (1986) (both revised in 2000) is the principle of national treatment of foreign companies. This principle means that, provided national security is not jeopardised, TNCs have the same rights and obligations as domestic companies in similar situations. This does not put all foreign suppliers on an equal footing in the importing country market (which is what the MFN clause does); rather, it refers to the treatment of foreign suppliers in comparison with domestic ones. Resource-rich85 and prosperous countries such as Canada may exercise the greatest leverage on TNCs. Canada’s experience is one of the most interesting examples of control of the location of TNCs. This is relevant, as a significant part of the economic integration between Canada and the US has an FDI dimension. Both Canada and the US are signatories to the OECD Declaration. In 1973, a few years before this declaration was delivered, Canada established the Foreign Investment Review Agency (FIRA) to survey inward FDI. This move was a reaction to a relatively large share of foreign ownership of Canadian industry. One could argue that much of the foreign ownership of the host-country industry can be attributed to the level of the host country’s tariffs, taxes, subsidies and other incentives. TNCs overcome tariff obstacles by locating ‘tariff factories’ in the host country. FIRA’s intention was not to stop FDI in Canada, but rather to allow it only if it resulted in beneficial effects to Canada. The criterion upon which FIRA evaluated both the takeovers of existing Canadian firms and the establishment of new businesses included expanded exports, use of Canadian resources, increase in investment, employment and productivity, and compatibility with the national industrial and other economic policies. FIRA’s rejection rate of 20 per cent was an important barrier for certain investors and was high compared with rates of some 1 per cent in other countries that used a similar screening process (Lipsey, 1985, p. 101). This caused many firms to withdraw their requests in order to avoid uncertain and costly application procedures, while other firms that might have been potential investors have not even applied. The ‘FIRA effect’ on FDI was a reduction in both inward investments in Canada and in outward FDI from Canada (domestic firms took advantage of protection at the home market). At the beginning of the 1980s, FIRA’s rejection rate was reduced. The Conservative government transformed FIRA from a nationalistic authority aimed at increasing Canadian ownership in domestic industry into an organisation for the attraction of FDI, and in 1984 renamed it Investment Canada. The fear of too much foreign capital had given way to a fear of too little (Lipsey and Smith, 1986, p. 53). A strong wind of ‘new protectionism’, ‘economic patriotism’ or populism was spreading like wildfire in the EU in 2006. The French, Spanish, Italian and Polish governments showed opposition and open hostility to foreign takeovers of their domestic

International firms

347

firms.86 France and Spain safeguarded domestic energy markets, while Italy and Poland did the same with their financial markets. Hungary was defending Mol, domestic energy champion, against a bid by Austrian OMV in 2007. This ‘patriotic’ or protectionist wave in Europe was strong enough to fend off potential bidders even from partner EU countries. When Russian Gazprom eyed British Centrica in 2006, the British government said that it would scrutinise any potential deal of this kind. Britain and other countries shiver at the thought that the domestic gas industry may be purchased by the Russian firms. Certain basic principles on which the EU was established (free factor mobility and open, borderless and free internal market) were in question. The German government is preparing legislation that would scrutinise acquisitions by foreign state-controlled investors. Various wealth funds in the Middle East and Asia may acquire German companies in ‘sensitive’ businesses. In addition, soaring prices of oil and gas may endow Russia with extra funds that may be invested in Germany. The German idea is to create a light version of the American Committee on Foreign Investment in the United States (CFIUS). Its coverage would include industries directly related to security, perhaps energy, but not banks, media and consumer industries such as car-making.87 Let us now turn across the Atlantic. The Defense Production Act (1950) in the US and its amended Section 721, known as the Exxon–Florio Amendment (1988), give the US president the right to block mergers, acquisitions or takeovers of domestic firms by foreign TNCs when such action is likely to jeopardise national security. If a merger, acquisition or takeover has security implications, it has to be notified for clearance with the CFIUS, which applies the Exxon–Florio provisions. The powerful CFIUS conducted a very tough and thorough review of the controversial takeover of IBM’s PC arm by the Chinese Lenovo. National security and impact on jobs were at the heart of scrutiny. Ultimately, the deal was approved in 2005. Other important deals such as the $18.5 billion bid by CNOOC, the Chinese oil producer, for US oil producer Unocal were spoiled during the same year because of a potential threat to US national security and a risk of technology leakages. Unocal was acquired by the American Chevron company. The Chinese could not have part of the US energy supplies in their hands. Dubai Ports World (DPW) acquired British P&O (the ports and shipping company) in 2006. In this deal the DPW also acquired ports in New York, New Jersey, Baltimore, New Orleans, Miami and Philadelphia. This raised security fears and caused a big row in the US. DPW agreed to sell its US port operations to an American entity. ‘Globalisation’ and general praise and preach of free markets collided with ‘economic patriotism’ not only in Europe, but also in America. There would be serious questions if the West lost control over the evolution of the process of globalisation. Access to growing national and international markets and a stable macroeconomic environment are major incentives for the locating of FDI in some geographical areas. However, a stable, predictable and transparent legal situation also encourages FDI as it lowers risk and potentially increases profits. There are global rules that regulate international trade in goods (WTO), but as yet there are no such rules regarding FDI. A complex set of bilateral88 and some regional treaties regulate FDI, but coverage of these rules is not complete. In addition, many governments offer incentives to attract FDI. Diversity in the treatment of FDI in bilateral deals and the possibility of a sudden reversal in the liberalisation trend in times of crisis (Asia 1997–8) create a need for a multilateral treaty. This is in theory. In practice, Mody (2004, p. 1211) gave evidence that in the

Evolutionary Economic Geography

348

1982–92 period, US TNCs carried out less (not more) FDI in countries with which the US had recently concluded a bilateral investment treaty. Tax treaties reduce chances for tax evasion and transfer pricing. Where there are no such treaties, TNCs profit from loopholes in foreign tax systems and various concessions that are offered to their own advantage. The OECD drafted a Multilateral Agreement on Investment (MAI) with the aim of providing a sound legal environment, including open markets, based on the principle of non-discrimination between domestic and foreign investors. In addition, there would be instruments for dispute settlement among all involved parties (various combinations of public and private players) and enforcement of decisions. In spite of grand expectations, the MAI project got into serious trouble and was shelved in 1998. The critics argued that the exclusive OECD ‘club’ of (then) twenty-nine countries did not take into account the needs of the developing countries. Another criticism was that the MAI gave excessive power to TNCs regarding protection of the investment, transfer of funds, right of establishment and MFN treatment. These criticisms are relevant if one bears in mind that India still remembers the moment when the first ships of the East India Company arrived at the port of Surat. The company came to India as a simple trader in 1608 and ended up virtually owning it. The national legal system that regulates FDI in most OECD countries is well developed. Hence, the MAI was of little real significance for these developed countries, even though most of the FDI activity in the world is within the OECD group. The drafters’ intention was to lure the emerging players in the FDI flows to join the agreement. However, few developing countries were ready and willing to sign a deal that they did not shape (exchanging concessions such as domestic market opening for FDI in return for the opening of OECD markets for agricultural goods and migration of labour). Some critics saw the MAI even as a tool of neo-colonialism. However, the work on the MAI was not in vain, since it raised an important issue. Perhaps it will have better luck in the future – possibly within the WTO, which has a wider membership (over 150 countries), dispute settlement instruments and experience in handling difficult negotiations. 5 Benefits and concerns Estimates of the potential gains from operations of TNCs are generally based on the classical assumption that the set of traded goods is both complete and fixed. In such a case, the gains from trade and FDI appear to be quite small. Prices in the economy could be changed by government (tariff) intervention, so the quantity of produced and traded goods would change, but the list of the manufactured and traded goods would remain the same. Suppose now that the list of goods is expandable. The neo-Schumpeterian and evolutionary economic model assumes that there are no limits for the introduction of new goods and services in an economy in the long term. Let us assume, in addition, that the introduction of a product requires a large amount of fixed costs.89 This is especially true in the case of the developing countries. Perfect competition and free trade exclude (major) sunk costs from consideration. Once the fixed costs are included in the picture,

International firms

349

their presence is often used as a justification for government intervention. When important fixed costs enter the model, a substantial volume of the good needs to be sold in order to make a profit. In such a case, a tariff on the good in question may reduce demand. If this reduction in (international) demand is important, then the good may never appear on the market. Losses from such a development or gains in the reverse case may not be easily estimated, at least not for the time being. In any case, if the new goods are left out of consideration, then this kind of analysis can bring substantial underestimates of the welfare cost of trade restrictions (Romer, 1994). The production of these new goods and services is usually brought (to the developing countries) by TNCs. Integrated countries may obtain certain benefits from the trans-border business activity of firms located within the confines of their group. These gains, specific to the operation of TNCs within the area, include not only tangible resources (for example, transfer of capital on more favourable terms than could be obtained on capital markets, tax receipts, economies of scale, sourcing of inputs from local suppliers and employment) that are provided at lower cost than through the market, but also various intangible assets and structures. These intangible assets include new technologies in production, management and control of assets and structures that make the existing resources more productive; positive externalities in production through linkages; access to international marketing networks that can overcome barriers for exports into foreign markets; new ideas; clustering of related firms; training of labour; and competition. The pecuniary element of these sometimes invisible spillovers is quite difficult to measure and could easily escape the attention of a non-economist. Other issues include monopolisation, restrictive business practices, increased sourcing from the parent country with a negative impact on the balance of trade, transfer pricing, transfer of profits abroad and spatial polarisation of economic activity. All these elements affect the spatial allocation of resources in a way that is not always favourable to either the integrated country or the whole group. Suspicions about the operations of TNCs in developing countries are, for instance, linked with benefits of FDI in Africa for the local communities. A great deal of FDI activity outside extraction of natural resources has been linked with import-substitution development strategies. Foreigners were profiting from the local economic rents. Local infant industries seldom started to ‘walk on their own’ without longer-term protection. This increased local costs and extended development agony. Despite a relative academic hostility or scepticism (and ridicule of the Washington Consensus),90 public sensitivity, polemics against foreign firms that export home resources and some official anxiety about the operation of TNCs, the situation has changed in general since the 1980s. All countries in the world now welcome the location of business by TNCs. At least, so they say. Attitudes towards TNCs have their ups and downs. The wind of ‘economic patriotism’ in Europe and the US may turn into protection against FDI. Alternatively, it may turn into a selection of foreign TNCs. However, the context in which FDI occurs is relevant. If TNCs enter a certain location to serve the widest international market, the perception of such an FDI is generally favourable and encouraged. If a TNC enters in order to supply the protected local market, then the perception and results of such an entry are less favourable over time. There is a tendency towards convergence in the national rules regulating FDI. Apart from some screening, these countries provide TNCs with various incentives which include the provision of infrastructure, subsidised loans, tax exemptions, export

Evolutionary Economic Geography

350

incentives, opportunities for complete ownership and exemptions from duties. However, these countries have to bear in mind that they compete with other possible locations (and countries), hence there is a tendency towards liberal and converging national systems relating to TNCs. The growing and wide markets in the US and EU are still among the most preferred locations for TNCs. If other regions want to attract TNCs, then, ideally, they should have some ‘unique selling point’ that TNCs cannot find in Europe or North America. Or in China. Apart from these drawbacks, host countries experience significant gains and thus welcome the arrival and operation of TNCs. These corporations may bring in new technology that is superior to existing domestic technology. This has the effect, in the host country, of increasing output capacity, jobs and tax proceeds, as well as resulting in savings in unemployment benefits and perhaps also creating exports. TNCs enter into growth industries and provide technological expertise that would otherwise be missing. Some TNCs produce brand-name drinks or cigarettes, whose consumption is subject to excise duties. Governments may need or want these proceeds and therefore become an ally of the TNCs. Barriers to entry, such as huge initial capital investment in an industry, can be overcome with TNC input. When the host country’s policy is to promote exports or substitute imports, then TNCs may fill part of this role. The so-called process of globalisation has boosted profits and tilted its potential benefits towards capital and firms (and against labour) in several ways: • Production was relocated to countries with low wages and reduced manufacturing costs. • Firms eased their bargaining with labour in the developed world. In return for wage cuts and longer working hours per week, certain German firms did not move their production facilities to eastern Europe or elsewhere. • Immigration into the developed countries depressed wages there in services such as catering and construction, as well as in farming. Governments may redistribute the benefits or excessive globalisation gains through the tax and general benefits system. The stronger the distribution of these gains throughout the society, the weaker the public resistance to national economic openness and ‘globalisation’. In general, if someone loses a job in the US, the person also may lose the right to health insurance. Hence there is a stronger fuss made against ‘offshoring’ production and import barriers in the US than is the case in Europe, where the more generous public welfare takes care of citizens’ health and other social needs. If a national economy is to prosper, then the government ought to improve education of the labour force and management for jobs in the future; improve the operation of markets for products and labour; and assist in the shift of production and jobs (including incentives and compensations) from old to new and promising business activities. The globalisations gains ought to be shared throughout the economy. This, however, does not imply that they ought to be shared equally, as this may kill incentives for those who are pushing forward fastest with success. One major incentive to persuade TNCs to locate their operations in a country or an integrated region is to offer them a stable macroeconomic environment and a growing market. International economic integration provides certain conditions for this opportunity. Other carrots include tax holidays, subsidies, tariff protection and secure public purchases. In the medium and longer terms, the best TNC attraction policy,

International firms

351

together with the stable macroeconomy, is to influence the supply of educated local labour (a locally created factor) as an additional incentive to TNCs to select the country for the location of their operations. Potential benefits and concerns from the location of TNCs in a national economy may not be measured directly and in the short term. Nonetheless, many of them can be listed, as in Table 5.5. As FDI touches issues such as ownership of land and real estate, direct employment and taxes, it is more a politically charged issue than a trade issue. A popular perception, based on shoddy economics, is that the export of goods is beneficial for a country and therefore needs to be supported, whereas (some) imports are perceived to be dangerous as they might jeopardise the national output and employment potential. Through trade a country acquires useful things from abroad. Therefore, imports are a gain rather than a cost. To pay for imports, the country has to ‘send’ its useful goods and services abroad. In these terms exports are a cost, rather than a gain. However, FDI is treated in a different way. Outflows are being deterred, as there is a fear that exporting capital means exporting exports and exporting jobs (hollowing out of the parts of the domestic economy), while FDI inflows are welcomed. With protection and subscale operations by TNCs, the host country may end up worse off than if it had never received FDI. Such a situation may create a vicious circle of adverse signals and harmful incentives. The penalties in terms of lost opportunities are high, but so are the rewards for successes in attracting TNCs into well-structured projects. If TNCs operate at full scale in a reasonably competitive environment, they may provide the host economy with the ‘usual list’ of benefits, such as capital, technology and management know-how (Moran, 1998, pp. 155–6). Capital mobility has its costs and benefits. If these effects are desirable, they should be stimulated up to the point where these incentives do not become higher than the potential gains, otherwise they should be regulated, controlled and taxed. Simultaneous inflow and outflow of FDI in integrated countries is possible and quite likely. This was confirmed in the case of the EU. However, FDI flows and the consequent geography of production may be significant even without formal integration and specific trade preferences. Just take a look at transatlantic FDI flows or FDI inflows into China, or Chinese FDI throughout Africa. TNCs primarily follow the opportunities for making profits in large and growing markets in the long term. Therefore, economic integration (at least in Europe) is a, rather than the, cause of FDI. The relative loss of the international competitive position of the EU in relation to the US, Japan and the newly industrialised countries (in several lines of production) in the 1970s and 1980s has increased both interest in and the need for strengthening the competitiveness of output of EU companies. These TNCs are the key actors in improving the international competitiveness of goods made and services provided in the EU. Ambitious and well-funded public research programmes in the EU may assist the domestic firms to lead or to catch up, and later to improve their position, vis-à-vis their major international competitors if astutely linked with the production resources and potentials, as well as current and future domestic and foreign demand.

Evolutionary Economic Geography

352

Table 5.5 Potential benefits and real concerns that bring TNCs to the target country/location Potential benefits

Real concerns

Investment may add to the capital formation and production capacity of the country. This may stimulate growth of the target country.

• Flow of FDI is volatile. • Capital may be raised on the local market so there is no inflow of fresh capital. • Mergers and acquisitions relate only to the transfer of ownership. Only greenfield investment adds to the capital formation. • If higher productivity of a TNC affiliate is achieved at the expense of lower productivity of domestic firms, there may be no implications for aggregate growth. • Most of FDI goes financial services (the EU case). This does not bring a direct and large-scale improvement in the manufacturing industry in the target country. • TNCs may exert strong lobbying pressure and may request incentives such as subsidies and tax holidays. Such benefits and ‘ locations tournament’ among countries may introduce important opportunity costs, unnecessary waste of scarce public funds and serious distortions in the national economy regarding spatial and industrial location of resources. This creates disadvantages to local firms. Incentives to TNCs may be difficult to administer. • Foreigners own and control domestic assets and structures. The domestic industry falls in the hands of foreigners. Because of their high international mobility, TNCs may behave in a different way than domestic firms in identical situations. • TNCs repatriate profits abroad instead of reinvesting them in the location where they were made.

TNCs may transfer new technology, knowledge, experience, skills and ideas in production, management and control (human and organisational capital embedded in TNCs). This is also linked with the economies of scale and improved efficiency. TNCs may also develop local innovation activities and transfer certain R&D activities if there exist local knowledge, experience, skills and/or subsidies.

• TNCs may transfer out-of-date and polluting technologies. • TNC may close local and concentrate R&D activities in its headquarters. The country in which the affiliate is located may be put in a long-term dependency position regarding technology. • TNCs may introduce technologies that use more resources that are in short supply (capital) in the target country and relatively less of the locally abundant resource (labour). • TNCs may develop and employ sophisticated technologies in order to prevent easy copying. They may keep this technology internal to the firm to keep as long as possible their position of

International firms

353

monopolists in their respective businesses. Spillovers to the local industry do not take place. • TNCs pay higher wages in order to attract and keep better workers. The domestic industry is left with other workers. There is no transfer of knowledge to local firms. • Excessive internalisation of business may diversify the affiliate into unrelated technologies (outside the core competence). This may reduce the competitiveness of its output. TNCs may create, develop, expand and upgrade spillovers and links with the local economy.

• Spillovers and links depend on the technological capacity of the local economy to accept them. The closer the technological capacity of the local firms to a TNC affiliate, the greater the potential spillover. • Literature surveys such as Görg and Strobl (2001 and 2005), Hanson (2001), Görg and Greenaway (2004), Blomström and Kokko (2003), Rodrik (2004), R.E. Lipsey (2004), Girma (2005), Smarzynska Javorcik and Spatareanu (2005) and Moran et al. (2005) found only mixed, ambiguous or inconclusive evidence about the existence of positive spillovers on the local economy. This is particularly obvious in the case of developing countries.

TNCs employ local labour and contribute to an increase in the local skills.

• A takeover (merger or acquisition) of a local firm transfers only the ownership. There is no immediate change in employment. If the new owner rationalises production, employment may be reduced. • A greenfield entry may increase employment, but if the affiliate puts local competitors out of business, the level of employment may change in the target country. • TNCs often offer higher wages, hence they attract the best workers and managers. The local firms may remain with a lower quality of labour. • TNCs pay higher wages to keep the best workers and managers to prevent the leakage of knowledge to the local economy. • Local trade unions may request higher wages throughout the economy. These increases may be over and above productivity. This may jeopardise competitiveness of country s output. • TNCs may use technologies that employ factors that are in short supply locally (capital). • If labour unions demand ‘too much’ , a TNC may threaten to move its affiliate elsewhere.

TNCs may stimulate competition and force local • Restrictive business practices such as competitors and suppliers to upgrade their monopolisation of the market, predatory and business. transfer pricing may hamper the expansion of existing and development of new local firms. • TNCs may crowd out local firms from the local capital market. • Various public concessions and subsidies offered only to TNCs create distortions that damage local firms. • Unethical and criminal practices (e.g. bribery, ‘ campaign contributions , improper accounting, false billings and statements) may create huge corporate failures and scandals such as Enron

Evolutionary Economic Geography

354

(2002) in the US or Parmalat (2003) in Italy with business consequences felt throughout the world. TNCs offer access to a wide international manufacturing, marketing and trade network. They include the target country into this wellestablished network that may stimulate exports. More than a third of world trade today is within the same company. Some of these important ‘ internal

• Affiliates may produce only for the local market at sub-optimal scale of production. • Affiliates do not buy inputs in the local market, but import them from other affiliates. This may increase imports (rather than exports). Transfer pricing may jeopardise the balance of payments position.

TNCs markets may be infiltrated only through affiliates. TNCs pay taxes and contribute to the tax revenue in the host country.

• Transfer pricing may diminish local tax receipts. • Tax avoidance may do the same. • Various tax incentives may degrade the effect of taxes collected.

American investors in the EU were market-seeking in the 1950s and 1960s. The same was true of Japanese FDI in the 1970s and 1980s. As EU consumers were not able to purchase what they wanted because of NTBs, Japanese TNCs invested in the EU in order to satisfy an existing and growing local demand. TNCs from both source countries have generally looked at the EU market as a single unit. The Japanese TNCs located in the EU look at the European market as a single unit, similar to the US, and expect similar business results. While the US TNCs such as Ford had certain difficulties in coordinating operations in their ‘independent’ operations in Britain and Germany, the Japanese, as relative newcomers in Europe, had fewer difficulties of this kind. Their advantages over local EU enterprises included, in some cases, not only a superior capability to innovate products and technologies and manage multiplant production and supply, but also a willingness, experience and ability to serve consumers from a local base, rather than through exports, which was the favourite method of operation of many national firms in the EU. In addition, Ando (2005, p. 171) remarked that the non-participation of certain EU countries in the EMU reinforces ‘the asymmetrical structure of the EU’. TNCs have to adapt to this situation. Toyota, for instance, located its second plant not in Britain, but in France. There is, however, at least one difference between Japanese and American FDI in the EU. Nearly half of Japanese FDI was concentrated in banking and insurance in the early 1990s. At the same time, more than half of American FDI was in the manufacturing industry, while about a third was in financial services. Therefore, Japanese FDI had relatively less impact, at present, on EU manufacturing industry than the American TNCs (Buigues and Jacquemin, 1992, p. 22; European Commission, 1998, p. 3). This focus by Japanese investors at that time may come as a surprise, as Japan has the advantage in electronics and cars. A concentration of Japanese FDI in financial services (relying on human, rather than physical, capital) both in the EU and the US is, however, a partial reflection of Japanese balance of payments surpluses and the appreciation of the yen at that time. However, from 1994, the interest of the Japanese TNCs shifted predominantly towards the Pacific Rim countries.

International firms

355

In order to avoid weakening the competitive position of EU firms in global markets, the EU may follow two courses, although employing a mix of the two has its attractions. First, the EU may increase protection of domestic firms against foreign TNCs through various NTBs (although the use of TRIMs is outlawed by the 1995 WTO Agreement). Second, the EU may open its domestic market and encourage foreign, in particular hightechnology, TNCs to locate their manufacturing operations in the EU. As widely argued, EU firms in certain manufacturing industries are less efficient than their counterparts in the US and Japan. If the EU were to adopt and maintain a liberal economic policy and EU firms were to adjust and withstand competition from foreign TNCs, then they may, in relative terms, gain more from market liberalisation than their foreign competitors. However, China became a strong competitor and magnet that attracts the location of a wide range of manufacturing industries from 1990s. Therefore, the EU may wish to think more about the promotion of services and superspecialised manufacturing, rather than to consume its energy, resources and time on the protection of jobs and industries that would go to China and elsewhere anyway in the decades to come. The incomplete internal market in the EU was the major cause of a suboptimal production structure in the region prior to 1993. All economic agents, including TNCs, behave as welfare-maximising units in the long term, subject to the prevailing conditions. These private agents should not be criticised for actions that may be in conflict with public objectives. The start of the Single European Market, which included the removal of NTBs, assisted in a rationalisation of production and enhanced the location-specific advantages of the EU. The introduction of the eurozone in 1999 removed internal EU exchange rate risks, made prices directly comparable and consolidated earlier achievements in integration, at least for the participating countries. Coupled with eastern enlargements (2004 and 2007), this will continue to contribute to an increase in both the size and the growth of the EU market, which could, in turn, further increase investment expenditure in the region in the longer term. An extensive survey of the impact of the Single European Market on FDI found strong evidence that FDI and trade are complementary (European Commission, 1998, p. 1). Concerns that the Single European Market would lead to a concentration of general economic activity into the ‘core’ EU countries were not justified. In fact, there has been a limited spread of economic activity to the ‘peripheral’ EU countries which enjoy certain cost advantages (ibid., p. 145). As one of the major players in the international capital mobility scene, the EU would like to see multilateral rules for FDI. Even though there are multilateral agreements that deal with specific ‘technical’ FDI issues such as TRIMs or dispute settlements, or with sectoral matters such as services (GATS), there is no overall multilateral agreement on investment. However, FDI issues are discussed within the WTO. The future multilateral agreement may include the following elements: • Foreigners would have the right to invest and operate competitively in all sectors of the economy; only a few exceptions to the general rule may be allowed. • There should be no discrimination against foreign investors based on their origin. A ‘standstill’ commitment would prevent the introduction of new restrictions. • There should be a ‘roll-back’ principle to gradually eliminate national (or group) measures that run counter to the liberalisation of FDI rules.

Evolutionary Economic Geography

356

Flows of FDI and location of business abroad is sensitive and sometimes subject to unpredictable economic and political evolution. Various types of risks, including political risks, terrorist attacks, wars, populism, ‘economic patriotism’, protectionism, protestors against global capitalism or epidemics (bird flu, AIDS) may sharply influence FDI and operations of TNCs. One has to be careful while designing and implementing a policy that attracts, keeps or fends off the location of TNCs. The looming demographic crises in Europe and Japan (ageing population) as well as high prices for engery and food ought to be factored into the future policy choices and actions. Notes

1 Control may not be easy to define in all circumstances. A 51 per cent ownership of a firm is a clear sign of full control. However, in some cases 20 or even 10 per cent ownership is sufficient for the control of a firm if other shareholders are widely dispersed and unorganised. 2 Toyota makes its Tundra pick-up truck in San Antonio because roughly one in seven pick-up trucks sold in the US is in Texas (J. Reed and M. Nakamoto, ‘How Toyota went native in America’, Financial Times, 6 August 2007). 3 In an early study, Weber (1909) offered two basic reasons why firms ‘go to produce abroad’. The primary determinant is the achievement of lower labour and transport costs, while the secondary element is the benefit of large-scale production. 4 G. de Jonquières, ‘Battles among regulators could damage trade’, Financial Times, 25 May 2003. 5 Sources of irreproducible capabilities include: market structure that limits entry; firm history, i.e. it takes time to be learned and replicated; and tacitness in relations. Routines may not be transplanted easily and swiftly. No one, not even the participants themselves, fully comprehends their nature (Kay, 2006, p. 164). 6 For example, over a certain period of time governments may change the availability, quality and cost of the domestic factors. The disposable tools for this policy include training of labour and education of management, R&D, science, transport and communication infrastructure and tax policies. The host governments may also sort out labour relations. These aspects of liberal or stringent labour-relations rules and high/low quality of education and labour force may be considered among other factors that influence location of TNCs. 7 Sub-Saharan Africa may be an interesting case to study. Why have TNCs avoided it on a large scale? Apart from countries rich in oil (Nigeria and Angola) and relatively developed South Africa, TNCs have generally avoided this continent. Poor education and infrastructure; weak governance; government instability and policy uncertainty; possible civil unrest; AIDS; and vulnerability to volatile shifts in commodity prices were principal reasons stated by a number of TNCs for avoiding Sub-Saharan Africa as a location for FDI. 8 In spite of allegedly lower environmental standards in the developing countries, a massive transfer of polluting business activities to these countries has not taken place, at least not on a large scale. This is the consequence of the constant focus of TNCs on the developed countries and the adherence of TNCs to the unilateral application of environmental standards (often the same as in their home country) that are superior to those prevailing in the host developing country. TNCs fear legal challenges in their home countries for damages done abroad. 9 ‘Japanese companies follow each other,’ said a senior adviser at the Czech investment promotion agency. ‘As soon as there is a big famous investor, the others start to consider the possibilities.’ The principal reason for the location in the Czech Republic is geography (rather than relatively low wages). Another important reason for the Japanese TNCs to locate

International firms

357

in the Czech Republic is tradition in technical universities and the established electronic industry (R. Anderson, ‘Electronic giants go for Czech launchpad’, Financial Times, 10 June 2003). Car manufactures adopted a similar attitude in Slovakia, where Volkswagen established its subsidiary in 1991. After a successful and huge increase in production, Volkswagen became the biggest company in Slovakia. Peugeot decided to locate in Slovakia in 2003. This was followed by Hyundai’s decision to invest in a car-assembly plant too in 2004. 10 J. Kynge and D. Roberts, ‘The Pearl river delta is attracting $1bn of investment a month amid one of the fastest bursts of economic development in history’, Financial Times, 4 February 2003. 11 H. Simonian and R. Minder, ‘Liberal Swiss tax regime angers Brussels, Financial Times, 13 February 2006; H. Simonian, ‘Swiss canton’s tax regime irks Brussels’, Financial Times, 16 February 2007. 12 H. Simonian, ‘EU tax row set to grow over Austria investor data’, Financial Times, 17 January 2005. 13 This cut would affect only public limited companies. SMEs, often family owned, make up 85 per cent of German firms that pay income tax where rates rise progressively to 42 per cent (‘No flat tax’, Financial Times, 18 March 2005). 14 The average level of corporation tax in the thirty richest countries in the world fell from 37.5 per cent in 1996 to 30.8 per cent in 2003 (C. Swann, ‘Rivalry cuts company tax rates’, Financial Times, 1 May 2003). 15 Rates at which taxes are charged can be higher in the core region/country than in the periphery, but they should not be too high, because after a certain point capital may move en masse elsewhere. The result may then be disastrous as this may irreversibly undermine certain agglomeration economies. 16 Estimates about tax evasion are notoriously unreliable. Nonetheless, German savers alone are thought to have ‘hidden’ about Є300 billion in the ‘tax-friendly’ countries. If one adds the fourteen other ‘old’ EU countries, this number may increase eight times (F. Guerrera, ‘Curtain comes down on 14-year opera’, Financial Times, 3 June 2003). The amount of funds that the Germans took to the tax-friendly countries may even reach Є500 billion, according to other estimates. Offshore banks were able to create new financial products that were able to deal with the new situation (G. Parker and H. Simonian, ‘Brussels outfoxed as Europe’s tax havens stay one step ahead’, Financial Times, 4 August 2006). 17 In the complex overall EU bargaining in 2003, in return for agreement on this tax matter Italy demanded and got significant concessions regarding fines that its domestic farmers had to pay for breaching EU milk production quotas. The deal was that 24,000 farmers would pay these fines in yearly and interest-free instalments till 2017. This would reduce their fine by a third compared with what would be the case if they were forced to comply with EU rules and settle the fine immediately (F. Guerrera, ‘Curtain comes down on 14-year opera’, Financial Times, 3 June 2003). 18 Switzerland is estimated to handle one-third of all the money held in private accounts in offshore financial centres. The Swiss Banking Federation claims that Switzerland has applied tough legislation to monitor criminal and terrorist funds (H. Simonian, ‘Swiss approve bank secrecy initiative’, Financial Times, 3 December 2003). Switzerland taxes interest earned by EU citizens on their bank accounts in Switzerland, with 75 per cent of the proceeds going to the account holders’ national tax authorities, while the rest stays with the Swiss authorities to cover running costs. Switzerland transferred SFr 536.7 million (Є215 million) to EU countries under a system which required this country to levy a flat-rate tax on EU residents’ savings in Swiss bank accounts for the year 2006. The EU finance ministers hoped that the scheme would bring them more proceeds (H. Simonian, ‘Swiss hand over £222m tax to EU’, Financial Times, 24 April 2007).

Evolutionary Economic Geography

358

19 How should one measure these benefits? What should be measured? Should it be done in fiscal terms? Or in social terms? 20 The online Oxford English Dictionary defines corruption as ‘Perversion or destruction of integrity in the discharge of public duties by bribery or favour; the use or existence of corrupt practices, esp. in a state, public corporation, etc.’ (http://dictionary.oed.com/ accessed on 9 January 2008). It is immoral behaviour to misuse public power for private gain.

Jack Straw, foreign secretary, has let it be known that he intends to turn a blind eye to some of the requirements of two-year-old legislation banning UK businesses from bribing foreign officials … From one perspective, his stance can be viewed as pragmatic realism. So-called facilitation payments are exempted by the US Foreign Corrupt Practices Act, a model of its kind, and are not explicitly outlawed by the anti-bribery convention negotiated by members of the Organisation for Economic Co-operation and Development … the borderline between ‘facilitation’ payments and bribes aimed at winning business by underhand means is only loosely defined legally and fuzzy in practice … Nonetheless, his partial and retrospective interpretation of the law sounds curious from a politician who, as home secretary, boasted of being ‘tough on crime’. (‘De minimis non curat Straw’, Financial Times, 20 February 2004). Regarding the Saudi bribes scandal linked with the £43 billion contract to supply 120 planes and extras, Simon Jenkins wrote: If the legal position of the British government as complicit in the bribery is untenable, its moral position is laughable. It has inflated the price of an export to win a contract by corruption. It has been forced to use the dictator’s defence, that resulting embarrassment should be shrouded by ‘national security’. And it must tell African and Asian regimes that its much-trumpeted stance against corruption is meant to apply only to the poor and the weak. Such hypocrisy in Britain’s name is outrageous. (S. Jenkins, ‘How much hypocrisy can Britain get away with in this sordid deal?’, Sunday Times, 10 June 2007). Is this a modern reflection of what King Richard II said in 1381 to calm the peasant rebellion (against poll taxes, market monopolies and feudal serfdom) that threatened London? He promised the rebels, represented by Wat Tyler and Richard of Wallingford, that he would end the repression of the peasants by the nobility. As soon as the menace passed, the King said: ‘Villeins ye are, and villeins ye shall remain.’ That is to say: promises to little people do not count. And serfdom was reestablished. The IMF estimates that more than $500 billion are laundered in the global financial system each year. ‘Official estimates put the UK share at £25 billion’ (‘London’s dirty secret’, Financial Times, 29 October 2004). 21 J. Stiglitz, ‘We have become rich countries of poor people’, Financial Times,7 September 2006.

International firms

359

22 Financial scandals such as the one linked with Enron created public criticism against deregulation. 23 One should not accord an absolute value to the decline of consumer loyalty to local brands. Regional rather than ‘global’ strategies in certain food and healthcare products may offer a superior business outcome. However, there was a certain ‘global success’ by, for example, Coca-Cola, which succeeded in replacing a part of the consumption of milk and healthy natural fruit juices by an artificial and sweet liquid that has a questionable impact on health and weight. 24 It was the army contingents, then the Church, that once determined our life and order in society; then it was the state, and now it is large corporations. (The Church continues to be against laws that allow Sunday shopping and work in order to protect its attendance rolls.) Governments court firms (TNCs) to come and locate within the confines of their control. Critics say that the protection of the vulnerable in society or safeguarding of the environment is weakened, even that it is left to the mercy of big businesses. This may be the case in certain lines of production, but it all depends on specific circumstances. In some cases the regulatory framework facing TNCs was strengthened, while in others it was loosened. 25 NGOs often consider that it may be better to be ‘just about right’ and make a fuss now in order to put an issue on the policy agenda, rather than to wait for firm scientific evidence and miss the political boat to do something about the issue now. 26 The same holds for the increase in the power of special influence groups such as NGOs. Do these unelected groups of people with sometimes questionable accountability represent a risky shift of power towards special vested interests? Or do they represent a move towards the new ‘civil society’ (indeed, can anyone define this type of society)? Why does the general public sometimes believe more in these organisations, which often lack resources for rigorous analysis, than in the government? Who finances them? Why? 27 The role and influence of international organisations may sometimes be misunderstood. The World Trade Organisation does not manage world trade, the World Bank is not the world’s banker and the International Monetary Fund does not manage the world’s money. The defining characteristic of a market economy is that no individual, or corporation, or small group, determines its directions. But since this is hard to grasp, the fallacy that decisions of big import for the world economy are being made in these conference rooms is sustained by both preening politicians inside and protesting demonstrators outside.

(J. Kay, ‘There is only empty talk at the world’s top table’, Financial Times, 14 June 2005)

The same holds for the evolutionary economy, which is shaped by impersonal forces. 28 S. Talbot, ‘The birth of the global nation’, Time Magazine, 20 July 1992, p. 70. 29 Recep Tayyip Erdogan is the leader of the Justice and Development Party (Adalet ve Kalkinma Partisi, or AK) in Turkey. His pro-Islamist sympathies earned him a conviction in 1998 for inciting religious hatred. He had publicly read an Islamic poem including the lines: ‘The mosques are our barracks, the domes our helmets, the minarets our bayonets and the faithful our soldiers …’He was sentenced to ten months in jail, but was freed after four. However, because of his criminal record, he was barred from standing in elections or holding political office. (‘Turkey’s charismatic pro-Islamic leader’, BBC, 4 November 2002, http://news.bbc.co.uk/2/hi/europe/2270642.stm, accessed on 9 January 2008)

Evolutionary Economic Geography

360

Nonetheless, he won two consecutive elections and became the Turkish prime minister in 2003 and 2007. 30 Leading article: ‘An end to tolerance’, Sunday Times, 12 February 2006. 31 ‘Globalisation is no more than an (admittedly ugly) name for the process of integration across frontiers of liberalising market economies at a time of rapidly falling costs of transport and communications’ (M. Wolf, Financial Times, 3 October 2001, p. 15). 32 A. Krueger, ‘Educating globalisation’s Luddites’, Financial Times, 16 April 2004. 33 R. Wright, ‘Engines of globalisation: the story of Maersk’, Financial Times,2 October 2006. 34 H. Kissinger, ‘Globalisation: America’s role for the millennium’, Irish Independent, 13 October 1999. 35 [F]ive parallels between the United States today and the United Kingdom a century ago: overstretched, physically and financially; great power rivalry, with China now in Germany’s role; an unstable alliance system, with disintegration of the transatlantic relationship; rogue regimes (then Serbia, now Iran and North Korea); and revolutionary organizations (then Bolsheviks, now al-Qaeda).

(Wolf, 2005, p. 7)

36 The bourgeoisie has through its exploitation of the world market given a cosmopolitan character to production and consumption in every country. To the great chagrin of Reactionists, it has drawn from under the feet of industry the national ground on which it stood. All old-established national industries have been destroyed or are daily being destroyed. They are dislodged by new industries, whose introduction becomes a life and death question for all civilised nations, by industries that no longer work up indigenous raw material, but raw material drawn from the remotest zones; industries whose products are consumed, not only at home, but in every quarter of the globe. In place of the old wants, satisfied by the production of the country, we find new wants, requiring for their satisfaction the products of distant lands and climes. In place of the old local and national seclusion and self-sufficiency, we have intercourse in every direction, universal interdependence of nations. And as in material, so also in intellectual production. The intellectual creations of individual nations become common property. National one-sidedness and narrowmindedness become more and more impossible, and from the numerous national and local literatures, there arises a world literature.

(K. Marx and F. Engels [1888 – English translation] The Manifesto of the Communist Party. p. 6; http://www.marxists.org/archive/marx/works/download/manifest.pdf accessed on 9 January 2008)

37 ‘A survey of globalisation’, The Economist, 29 September 2001, p. 4. 38 P. Krugman, ‘The good news’, New York Times, 28 November 2003. 39 J. Kay, ‘Global business deserves a peaceful May Day’, Financial Times, 28 April 2004. 40 In India, a next to total monopoly of the domestic market for fizzy drinks was in the hands of two companies: Coca-Cola and Pepsi. Following a successful campaign by the local NGOs and populist politicians, in August 2006 a quarter of India’s twenty-eight states banned the sale of Coca-Cola and Pepsi on the grounds of concern for the health of consumers (‘Coke and Pepsi canned’, Financial Times, 11 August 2006). 41 One thinks of the particular consequences of chewing gum, which pollutes lifts, pavements and the underside of desks and chairs. 42 Increased international mobility of factors, increased international intra-firm transactions, expanding international cooperative arrangements between firms, the increasing importance of knowledge and a reduction in transport and communication costs all support the process of globalisation and are constituent parts of it. In these circumstances, individual actions of national governments may not increase global welfare (pollution is an example) and certain

International firms

361

supranational rules and/or coordination may be necessary in order to deliver more beneficial outcomes. 43 Trade is relatively more concentrated within regions than FDI. This suggests that trade plays a more prominent role in intra-regional integration arrangements, while FDI has a greater influence on global integration (UNCTAD, 1993, p. 7). 44 The Internet is replacing books, printed news media and visits to libraries, as well as face-toface socialisation among people. 45 ‘Is globalisation doomed?’, The Economist, 29 September 2001, p. 14. The reference here is to the attacks of 11 September 2001. 46 ‘Americans tend to believe that we do everything better than anyone else. That belief makes it hard for us to learn from others’ (P. Krugman, ‘French family values’, New York Times, 29 July 2005). 47 Institutional elements such as clear property rights are very important for foreign investors. However, the mere transfer of these solutions from the developed countries elsewhere is not a sufficient condition for successful economic performance. For example, while Russia has introduced a system of private property rights, China has retained the general socialist legal system. Nonetheless, domestic and foreign investors regarded China as a promising location for investment, and this country became one of the prime locations for investment. There was a strong credibility that investments would be protected. In contrast, this credibility was lower in Russia in spite of the property rights system, which is one reason why Russia continued to score much lower than China as a location for manufacturing FDI. 48 Apart from Russians and Poles, few people cared much about vodka prior to the early 1980s. Then a Swedish firm, Absolut Vodka, launched one of the most successful advertising campaigns in the US. Following that, vodka became one of the most popular spirits in the world. 49 ‘Soft drink manufacturers around the world have been overwhelmed by the entrance of Coca-Cola and Pepsi into their home markets. Local ice cream manufacturers find they are unable to compete with Unilever’s ice cream products’ (Stiglitz, 2002, p. 68). 50 ‘A fat lot of good’, Financial Times, 26 January 2006. 51 ‘Over 90 per cent of products produced in each of the triad regions is sold within that region. There is no global car … well over 90 per cent of MNE manufacturing is intra-regional rather than global’ (Rugman, 2002, p. 5). [MNE – multinational enterprise] 52 Nonetheless, the federalists were able to convert this draft Constitution into a Reform Treaty and push it through in 2007. The ‘old’ Constitution was stripped of state symbols such as flags and anthems, but the real old substance got a fake moustache and beard: it was distributed into various technical amendments of the previous EU treaties, incomprehensible to the non-lawyers, just to set the situation for ratification in national parliaments, rather than in uncertain national referendums.

The Irish were the only nation that had a public referendum on the Reform (Lisbon) Treaty. They rejected it in 2008. Some other EU countries rumour that the Irish should vote for a second time on the same treaty. It is interesting to note that the countries that did not have one referendum on this treaty suggest that Ireland should have two. 53 The Russian oligarchs want to spread their investment and income-generating assets outside Russia and outside their core competences. This is just in case Moscow cuts their access to domestic natural resources one rainy day. 54 ‘State foreign investments: the EU considers its response’, Bulletin Quotidien Europe, 24 July 2007, p. 3. 55 Since the start of the twenty-first century, China has been very active in making FDI in the development of resources, in particular in Africa, in order to feed its booming economy.

Evolutionary Economic Geography

362

Chinese FDI goes to places and countries which many TNCs from the developed world consider ‘inhospitable’. They are welcomed by the target countries, as China uses trade and aid. China does not meddle in internal affairs and does not attach strings to aid packages. As a developing country itself, China is regarded as a responsive supplier which may build roads, railways or dams more cheaply than most others from the developed world. However, if China’s soft power perpetuates the continuation of existence at the office of local despots that are reforms-shy, the longer-term impact on the host country may be harmful. 56 According to research carried out by the location division of Ernst & Young, firms now appear to accept higher costs of production in return for high environmental standards. ‘Regions that sell themselves on the basis of relaxed environmental legislation may soon find their approach counterproductive’ (‘Centres of attraction in the EU’, Business Europe, 12 March 1997). 57 Kneller and Pisu (2004, p. 425) revealed a significant fact: foreign firms contributed disproportionally to British exports. In 1996, foreign TNCs contributed 28 per cent to the total output, but 33 per cent of total exports; 85 per cent of TNCs export compared to 75 per cent of domestically owned firms. 58 This is also compounded with a proactive Irish policy to invite and keep TNCs. 59 Smarzynska Javorcik (2004) found data in the case of Lithuania that confirm evidence consistent with positive productivity spillovers. These were brought by links between affiliates of TNCs and their local suppliers (vertical links). At the same time, Sabrianova Peter et al. (2004) reported negative spillovers from FDI on the domestically owned Czech and Russian firms. They have not been converging to the technological frontier set by foreign-owned firms. However, foreignowned firms had positive spillovers on other foreignowned firms in both the Czech Republic and Russia. 60 Intra-company pricing refers to transactions among related units of the same firm. Not every manipulation of transfer prices increases the overall profits of the entire company, as these extra revenues stem from sales to outside customers (Plasschaert, 1994, p. 1). 61 Trade in services is expanding. Monitoring internal prices for services is a much more complex task than inspecting the same prices for goods by the employment of free-market criteria. 62 Firms are reluctant to comment and release data on their internal prices. Hence, there is a black hole in this area of research. 63 S. Bond, ‘The Kafkaesque charade of corporate taxes’, Financial Times,26 January 2004. 64 Tax avoidance, unlike tax evasion, does not involve breaking any laws, but rather making the best use of positive regulations. In practice, the difference between the two is often ambiguous. In any case the effect of either is that they reduce public revenue. This means that other citizens have to pay more for public services. It is a theft from taxpayers. 65 Transfer of profits out of the country takes place when TNCs do not find the country in question a promising location for the reinvestment of earnings. Such a transfer may send a warning signal to the local government that something is wrong in the economy and that something needs to be done about it. 66 J. Plender, ‘Top banks pay little tax to the UK exchequer’, Financial Times,20 July 2004. 67 J. Plender and M. Simons, ‘A big squeeze for governments: how transfer pricing threatens global tax revenues’, Financial Times, 22 July 2004. 68 J. Murray Brown, ‘Ireland tightens tax regime in effort to shake haven reputation’, Financial Times, 8 December 2005. 69 J. Plender and M. Simons, ‘A big squeeze for governments: how transfer pricing threatens global tax revenues’, Financial Times, 22 July 2004. 70 TNCs that come from the US and operate in the EU reported an annual R&D of almost $2,500 per employee in 1989. In contrast, Euro-affiliates of Japanese TNCs reported R&D per employee of around $725 in the same year (Gittelman et al., 1992, p. 18). The reasons for the difference include a relatively strong headquarters effect, as Japanese FDI in Europe

International firms

363

is a fairly novel phenomenon compared with FDI from the US. American TNCs have been present in Europe on a larger scale since the 1950s. In addition, Japanese TNCs may be more involved in the EU in relatively mature manufacturing industries and services where R&D expenditure is not as high as in other activities. 71 The recent emergence of internationally integrated MNC networks is best observed in Europe, where the contribution of foreign-owned MNCs to national technological capabilities is much greater than elsewhere. About one-quarter of large firm R&D carried out within Europe has been conducted under foreign ownership, while the world average is only just over one-tenth.

(Cantwell and Piscitello, 2005, p. 11)

72 Decisions taken at the EU level are easy targets for special lobbies as they are made far from the public gaze. The co-decision procedure between the Council of Ministers and the European Parliament introduced by the Maastricht Treaty tried to ameliorate this shortcoming. This is a step forward compared with the past, but there is still a potential danger that EU technology policy may become a sophisticated new form of protectionism. 73 TNCs are sometimes more economically powerful than certain countries. To support this view, fifty-one out of the world’s hundred biggest economies are corporations (Finance & Development, September 2004, p. 52). 74 Mitsubishi bought a well-known American symbol, the Rockefeller Center in Manhattan, in 1989. Many Americans felt uneasy about that ‘national loss’ and feared that the domestic ‘silver’ was being taken – apologies, purchased – by foreigners. However, this purchase turned out to be a financial disaster for the Japanese, and they left the building in 1995. 75 Decision 24 regulated entry and transfer of technology by foreign TNCs in the region. 76 ‘Global market brief: Russia tattered ties to the EU’, Stratfor, 15 November 2007. 77 The interested reader in the ‘Asian miracle’ is invited to consult World Bank (1993). 78 For example, Pirelli (Italy) coordinates and guarantees its global financial duties from a Swiss affiliate which is in charge of finance for the whole corporation. The US affiliate of Siemens (Germany) transmits daily financial data to headquarters, which is in charge of global financial management (UNCTAD, 1993, p. 124). 79 M. Skapinker, ‘Why Nike has broken into a sweat’, Financial Times, 7 March 2002, p. 8. 80 Article 4 of the same agreement allows only temporary derogations from this obligation to developing countries. 81 Mercedes-Benz might have had a hefty windfall gain in Alabama. The total investment into the car plant was $300 million while the state of Alabama spent $250 million (UNCTAD, 2000b, p. 24). 82 ‘Commission approves Є111 million aid to Hyundai’, Bulletin Quotidien Europe, 12 May 2007, p. 11. 83 P. Krugman, ‘Toyota, moving northward’, New York Times, 25 July 2005. 84 This is particularly important in the resource-rich developing countries. It is often alleged that their rulers siphon off a sizeable part of the receipts from TNCs that are in the extraction business to their private accounts. The domestic population receives very little from the exploitation of the national wealth in resources. 85 Investment in natural resources has three distinct features. First, the geographical location of non-renewable resources is not mutable; second, investments require huge amounts of capital which is linked with significant economies of scale; and, third, processing of minerals is linked with a considerable consumption of energy. 86 The French takeover decree protected the provision of private security for sensitive installations such as nuclear reactors; information security; dual-use technology; biotechnology that may be abused by terrorists; cryptology; bugging equipment; and casinos where the government is concerned about money laundering (T. Buck, ‘Brussels plans action against Paris on takeovers’, Financial Times, 30 March 2006).

Evolutionary Economic Geography

364

87 B. Benoit, T, Barber and G. Parker, ‘Germany plans for own Cfius deal watchdog’, Financial Times, 27 September 2007. 88 Among international investment treaties, there were 2,573 bilateral investment treaties and 2,651 double taxation treaties at the end of 2006 (UNCTAD, 2007, p. 16). 89 The introduction of a new good depends not only on the fixed costs and expected benefits, but also on substitution and complementarity with the goods that already exist on the market. Of course, there must be, or needs to be created, real demand for the new good or service. 90 The term ‘Washington Consensus’ was coined to express the extreme form of neoliberal policy advice given (or imposed) by the Washington-based international financial institutions (principally the World Bank and the IMF) to unlucky Latin American countries from 1989 (and even before). These policies include fiscal discipline; channelling of public expenditure priorities towards health care, primary education and infrastructure rather than investment in production (for this, loans need to come from Washington and Wall Street); tax reform (to lower marginal rates and broaden the tax base); interest rate liberalisation; a competitive exchange rate; trade and FDI liberalisation (the more, the better); privatisation; deregulation of the economy (to reduce and eliminate barriers to entry and exit); corporate governance; and protection of property rights.

Neo-liberalism and globalisation were terms often used synonymously in debates on trade and development. In any case, this type of globalisation and Washingtonstyle policy advice (1) never changed Latin American countries from being vulnerable exporters of commodities to exporters of higher-technology products and (2) led these countries into crisis and misery because they were driven to follow advice that paid little attention to distribution and fairness. In addition, loans and their reprogramming were such that the debtor countries would ‘never stop repaying them’. To this day, there is remarkably little good evidence that countries which adopted the Washington Consensus have enjoyed better growth and poverty-reduction performance more than those which adopted it less, at least if one excludes states which are barely able to do anything. Haiti over the past decade scores well by Washington Consensus criteria, Vietnam scores badly; but Vietnam has by far the better performance. … the prescriptions are couched as valid for countries at all stages of development. They ignore the ‘late development effect’, the idea that countries which begin to industrialise when other countries are already highly developed have to use different policies and institutional arrangements to those used by earlier developers, in order to compensate for the disadvantages and capture the advantages of coming late – advantages such as the potential to use more advanced technologies already used elsewhere. (K. Lee, J. Mathews and R. Wade, ‘Rethinking development policy: a new consensus’, Financial Times, 19 October 2007) Regarding FDI in the developing countries, the more FDI is not always the better. It is not in the developing countries’ long-term interest to have TNCs that operate over a long period of time under heavy protection to survive. The developed countries do not have a long-term interest in having (potential) partners in the developing world that do not increase their welfare. Poor or backward countries are

International firms

365

poor trading partners. In the long term, countries grow richer together, not at each other’s expense. The two biggest IMF debtors, Brazil and Argentina, decided in 2005 to repay their loans ahead of schedule. They wanted to get rid of loan conditions and policies that caused poverty and pain among the people. A generally low demand for IMF’s loans reflects improved national economic performance and favourable terms on the international financial market. The question is: how is the IMF going to pay in the longer term for its own annual administrative expenses of about $1 billion at a time of low demand for its loans? Perhaps the IMF needs to find a new business model for the future.

6 Conclusions A man’s steps are of the Lord; How then can a man understand his own way? (Proverbs 20:24) Economists worry for a living. Their predictions of the spatial location of firms and industries are harder than in other sciences such as meteorology or seismology. In the words of John Kay: But we can identify earthquake zones even if we cannot predict earthquakes and we can look forward to summer even if we cannot forecast the weather on June 4 next. Seismologists tell us where not to build our houses, and meteorologists help us know where to sell sun cream and when to take an umbrella. Useful economic knowledge is of a similar kind … But when someone tells you that the dollar will appreciate in the second quarter of next year, or what the level of the S & P will be at the end of 2004, they do not know what they are talking about. If they did, they would not be making these predictions. But if you stop asking economists to forecast the future, there are other interesting things they can tell you.1 It is time to recapitulate certain lessons that we have learned from the previous analysis of the economic landscape as a complex, open, evolving, selforganising and adaptive evolutionary system. Economists and analysts who argue in favour of the evolutionary model of the economy are not soothsayers. They have the (modest) right to be wrong. Niels Bohr, a Nobel Prizewinning physicist, once told a colleague who thought outside the ‘mainstream box’: ‘Herr Professor, we all agreed that your theory is crazy. However, we are not convinced that it is crazy enough to be right’ (Warsh, 2006, p. 262). Since the early 1990s, the geographical dimension of economics has gained renewed relevance. This change was partly based on the recognition that there is a relatively large concentration of production in the world in a relatively limited number of geographical areas. While mainstream economics deals with decision-making and optimisation within given structures, evolutionary economics stresses the continuing longterm process of structural change in space and time.2 The story is about how one new system replaces another. According to the evolutionary economic theory, the relations based on mutual support between economic agents and institutions might replace old relations among actors based on dependency and hierarchy. Each location might consider fostering an environment that creates, keeps and attracts ‘knowledge workers’ who may contribute to

Conclusions

367

the development of new ideas and/or new solutions to old problems. This may be a promising road towards further economic development. Where a firm or an industry should locate is not a question that has a straightforward answer. It is a highly complex and divisive issue. After two centuries, the economic powerhouse of the world is moving from the northern Atlantic back to the Pacific region. Hence a possible setback in the American and/or European economies may have a much smaller global economic impact than was the case even a few decades ago. Rich nations fear that firms and industries will move to low-wage nations; poor nations worry that the production of goods and services will migrate to developed countries; small nations are concerned that businesses will go to large countries; countries that are not integrated with others are afraid that production will be transferred to the integrated countries; while all countries are troubled because of a large-scale reallocation of manufacturing towards China, especially following the Asian financial crises of 1997. The Chinese worry about the domestic lack of primary resources, hence they locate their investment abroad in the search for these resources. The general worry about this exodus of manufacturing jobs to China ought to be mitigated, on two accounts. First, China is a grand processing economy. Its weakness lies in the lack of raw materials and energy. Second, the big money does not lie in manufacturing. It lies in the innovation, product design, branding and marketing where the developed countries still call the tune. At least for now. In a time and age of business ‘globalisation’, one might expect the importance of firm location to be diminishing. Some argue that distance is dead in terms of production and business because of the astonishing changes in communications and computers. But this flat world, weightless society or frictionless economy may be relevant only for certain manufacturing and services activities that are simple, routine, fragmentable and have codified (standardised) knowledge.3 Trade costs, i.e. all costs linked with getting the final good to the consumer (tariffs, quotas, NTBs, transport, storage, distribution, information, contract enforcement, insurance, exchange and banking) are still ‘alive and kicking’. The many components of trade costs may rise or fall in different proportion to those of distance costs. Trade costs matter not only for economic geography and economic policy but also for theory, as a number of puzzles in the international economy rest on their existence. In any case, spatial proximity (clusters) of firms that are in similar or related business is highly relevant in many complex production activities. The general competitiveness of their output hinges on the concentration of their knowledge in specific spatial locations (clusters) of firms such as chemicals in Basle or finance in London.4 Lasting competitive advantages in a global economy are often locally based. ‘The more the world economy becomes complex, knowledge based, and dynamic, the more this is true’ (Porter, 1998, p. 90). In addition, Porter (2000b) also suggested that although global sourcing mitigates disadvantages, it does not create advantages. Moreover, distant sourcing normally is a second-best solution compared to accessing a competitive local cluster in terms of productivity and innovation. Paradoxically, the most enduring competitive advantages in a global economy seem to be local. (p. 31)

Evolutionary Economic Geography

368

British long-distance trade in the twentieth century measured relative to the GDP declined (rather than increased), while trade with geographically closer Europe increased. Despite the evident decline in transportation and communication costs, international trade has become more localized rather than more globalized … a decline in transportation costs should lead to a rise in longdistance trade and a decline in the most local trade. Obviously that model isn’t good enough to fit what we actually see. (Krugman, 2007, p. 30) ‘There is very little that we economists fully understand about global trade but there is one thing that we do know – commerce declined dramatically with the distance. It’s not a small world’ (Leamer, 2007, p. 110; emphasis original). In addition, the world is not flat either: geographically, institutionally, culturally or otherwise. In the evolutionary model of the world it cannot and will not be. While economic geography covers the traditional analysis and description of economic changes, tendencies and their location in physical space, evolutionary economic geography is a relatively juvenile and unconsolidated, but growing, microeconomic academic venture. As such, the latter is subject to smallpox and other research illnesses that are becoming evident with age. The wide assortment of issues dealt with by evolutionary economic geography ought to be viewed as a liveliness, vigour and strength of the subject, rather than its weakness. This book looks at neoclassical equilibrium economics with fluent scepticism. Macroeconomic aggregates are too rough in this model, while its microeconomic part is too detached from the global picture and institutions, too individualistic and too abstract. The neoclassical ‘economic mechanics’ needs to be adjusted to allow for economic biology and evolutionary principles in which firms and entrepreneurs have behavioural capabilities to change and adapt, as is the case with living organisms. Agents do not only adapt to changes in the system, they also react to it. Relatively competitive markets5 ought to play a similar role in economics to the one played by biological evolution in the living world. Markets have been the evolutionary search mechanism for the selection of fit projects in disequilibrium. The theoretical dogma in neoclassical equilibrium economic theory is that free markets automatically balance themselves. Supply and demand end up in equilibrium. The best product is selected, while the allocation of resources is most efficient. This view of how the world operates regards the QWERTY example as an anomaly. Otherwise, the neoclassical equilibrium model cannot work. The problem is, however, that there are many QWERTY-type examples in reality. They demonstrate that small events at critical times in a complex evolutionary economy have crucial and long-term implications on business and its location in space. This is exactly the case that made sense of many otherwise unexplained or neglected problems in economics. One of biggest problems in the evolutionary model is that we know little about how it works, how this fresh uniqueness comes about6 and where it is going to be located. And why. Although the outcome of this process may be known, its origins are not. They may be the consequence of human ingenuity, entrepreneurial spirit, outcome of investment in knowledge (R&D) and serendipity. Why does it work in some locations, industries and

Conclusions

369

firms, and why doesn’t it work in others? When will/may it take place? What is the threshold (if it exists)? Is there a correct or preferred timing for the start of the new industrial and technological trajectory? What is its speed and force? What is the replication and reproduction mechanism? Does it exist? The evolutionary model of how the world operates has still a long research agenda with a plethora of new ideas. It is not a rounded-up theory. The evolutionary economic theory may never be fastened together in an elegant package, as is the case with the neoclassical equilibrium economics view of the world. Even though the evolutionary approach has taught us that there is no best strategy for an economic course of action, it has suggested that, within a constantly changing environment, there is a large set of useful strategies from which to choose. Games to be played by participants are more complex than is the case in the neoclassical equilibrium world. In this field, the greatest challenge is that the location of firms and industries is rather unpredictable at the start. This is settled over time through evolutionary mechanisms that ‘normalise’ location-related behaviour of firms into a configuration of (initial) spatial concentration. Without unpredictability (the initial chaos), there may perhaps be no progress; without order, there may perhaps be no survival. Unpredictability and order may be two sides of the same evolutionary coin, while the coin is represented by the time arrow on an imaginary graph. It is harder than ever to spot the right paths to follow. For each good path one may easily find a counterexample, or many of them. Analysis was generally focused on examples of successes in location of economic activity. The overall message may be that, with multiple equilibria, it is simpler, easier and faster to get it ‘wrong’ than to get it ‘right’. One size does not fit all and at all times in the evolutionary model. This finding may not be new; however, it fits into the strong endemically uncertain evolutionary system. A firm will always move and search for the most friendly environment and adapt to it. A ready-made recipe for the optimal location of firms and industries in time and space does not exist. Nonetheless, self-grown growth has been taking place in many locations on its own, in an endogenous way, without much advice from experts and foreign aid over time. The West developed following this path independently. Botswana, Chile, China, India, even Viet Nam offer other recent examples and hope to other countries. Evolutionary concepts consider economic changes in conditions of bounded rationality. There is a constant search for optimal solutions in conditions of scarcity, competition and endemic uncertainty. The process deals with creation, differentiation, selection (including mistakes), adaptation, innovation, retention, amplification, replication and the constant search for creative–destructive optimal solutions in a pool of many and various possibilities. Evolution is a process of unfolding, self-transformation, self-organisation after oscillations and the appearance in orderly succession of events over a long period of time. This process is irrevocable as the historical CD or card is played only once. Related to geographical and economic space, an important role is often played by unpredictable random events, chance, spatial proximity, neighbourhood effects, economies of scale, the lock-in effect and path dependency in the location of firms. While these may provide

Evolutionary Economic Geography

370

fertile soil for the development of business, they may also bring the trap of rigid specialisation. The creation and evolution of inventions, innovations and new technologies may have two opposing effects. They may ease the spread of certain economic activities in space, while they may also, at the same time, embed others (based on highly specialised knowledge) in only select locations. Such an evolution creates uncertainties and risks, but also unknown opportunities that need to be tested and exploited. Evolutionary economics has open frontiers. It is wide open for an interdisciplinary dialogue. In fact, this exchange and cross-fertilisation of various ideas is essential for its own survival and progress. The examination of the issue of location of firms and industries in space is still elastic and highly suggestive (in particular, how a historical accident may shape the production geography), rather than conclusive, convincing and general. A coherent and wide-ranging theory of the subject and general prescriptions for the policy are not yet in sight. We are faced with a choice on how to map this new approach into the research agenda and how to push it further up on the economic research totem pole. However, there are various piecemeal approaches, some driven by anecdotal evidence, that contribute to the acquired knowledge and to the raising of new questions. This leaves the topic wide open to further theoretical and empirical analysis, as well as public policy debate.7 While there is a certain intuition behind the general equilibrium theory and models (cause and the likely effect), this is not always the case with evolutionary, non-linear and complex systems. That is the reason why many scholars and policy-makers may have difficulties in reaching a decision or policy recommendation based on these nonequilibrium systems. Clearly, this is the nemesis of the evolutionary model of the economy. Even though relations between economics and evolutionary biology, for example, were known and recognised a long time ago, these interaction were unfortunately ‘put aside’ by many researchers in the field of economics. New analytical tools such as increasing returns to scale, production linkages (presence of intermediate goods and services), multiple equilibria (with centrifugal and centripetal forces), evolutionary approach and imperfect competition were introduced into the field of spatial economics in the early 1990s. These methodological tools also helped to explain why firms form clusters. Even the existence of cities may be explained, in part, by economies of scale. Even though the ever-increasing demand for quantitative rigour is making such considerations more difficult, the introduction of new analytical tools assisted spatial economics in finding its proper place in mainstream economics and becoming a hot research topic. Even though biological and economic evolution have in common such similarities as creation, differentiation, selection (including mistakes), adaptation, innovation, retention, amplification, replication and the constant search for creative–destructive optimal solutions, humans respond and create novelties intentionally and in certain cases react collectively. People can plan, react faster and are more purposive than is the case with biological evolution, but this social engineering is still evolution. This ought to be kept in mind by analysts. Evolutionary economics deals with actions, reactions, competition among various expectations, strategies and whims of economic actors. It is also concerned with the

Conclusions

371

phenomenon of how these fit into the aggregate out-of-equilibrium and always changing picture (contrary to the neoclassical path towards equilibrium). This helps us fathom, at an aggregate level, issues such as the location and evolution of clusters, cities, regions, persistence of poverty, appearance and demise of firms or the emergence of monopolies in a way that may assist us to be prepared to deal more effectively with these issues now and in the future. Economic progress in the future should not be taken for granted. Players in the evolutionary framework for business need to adopt the principles of the Red Queen race8 while searching for ways to outperform others. In doing so, they ought to seek new ways of doing things to stay in business. Nonetheless, challenges such as energy and food crises, infectious diseases, natural and political catastrophes, terrorism, mismanaged policies and other geopolitical dangers may create an even greater danger: opposition to constant change. Some flexibility in the system provides the possibility of seizing new opportunities and reducing, as much as possible, the future cost of adjustment. Even though one may learn a lot from simple equilibrium theory models, this is still confined to an illusionary situation. There are markets and fields where standard and evolutionary models simply do not work. This is the case, for example, with medical and military technologies. These two ‘markets’ involve big money and substantial employment; too big to be neglected or regarded as market imperfections or anomalies. Hence, the ‘market’ selection mechanism in these two fields is strongly mixed with and influenced by political, special group interest (lobbies) and professional pressure and authority. The dynamic and mobile segment of labour (at least) might gain from competition, openness and integration processes, as trade and competition do not determine whether there are jobs at home, but rather what kind of jobs are available. National macroeconomic policy, the business cycle, education and changes in technology are partly invisible and not easily measurable policy elements to the general public. However, these may be much more important factors for job creation and keeping them where they are now than is trade policy. Unfortunately for the policy-makers and the general public, trade is only one – though highly visible – policy aspect that affects national jobs. It can easily be used to blame others, the foreigners, for the mismanaged domestic economic policies. In the very nature of the competitive market economy there is no economic gain without pain. After all, you have to break a few eggs if you want to make at least a bit of extra profit. If the pressure and adjustment costs that result from a relaxation of trade restrictions do not hurt someone somewhere in the economy, then they probably produce no overall benefit either. The evolutionary reallocation of resources is a source of gain. The adjustment cost is a finite, one-off investment. The gains accruing from each improved resource allocation are ongoing, at least for some time in the future. There are, therefore, evolutionary reasons to believe that on the social balance and in the longer, even the medium, term, the cost of the current adjustment ‘pain’ is much exaggerated. The problem is that the short-term losers are more vociferous and politically organised than the gainers. The losers can make a strong political protest. Pluralism and ardent divergences in opinions and approaches (rather than consensus) in the evolutionary model may lead to some confusion. However, the glut of various ideas about this relevant and exciting subject ought to provoke on-going debate and be

Evolutionary Economic Geography

372

the basis and incentive for the creation and progress of new knowledge. Even though there is a certain mistiness and ambiguity on the path of research and analysis regarding the principal ideas, changes in the situation with bounded rationality and the working of models, a questionable ‘mechanical’ principle of optimality in the evolutionary process, a multiplicity of concepts may create a fertile intellectual soil for new research trajectories. A highly challenging evolutionary question is still open: would the increased speed of new knowledge-creation, and its spread, boost the possibility of more easily liberating those that are locked in? (And how?) This calls for an answer in the (near) future, otherwise the toughest judges, i.e. time and history, will have no mercy with the evolutionary strand of thought. Conversely, the evolutionary approach may easily be just one among the many open-ended issues, such as the debate about the rule of law and morality, why bad things happen to good people, what is the correct rate of inflation, what is the just system and rate of taxation, or what should the government do. The nature of these and similar concepts is that they are subject to permanent re-examination and debate. Regarding the art of policy making, the Growth Report (2008) by the high-level World Bank Commission on Growth and Development observes after several years of thorough work that It is hard to know how the economy will respond to a policy, and the right answer in the present moment may not apply in the future. Today’s bad policies are often yesterday’s good policies, applied for too long. Governing a growing economy is not a static challenge. It is more akin to a long voyage undertaken with incomplete and sometimes inaccurate charts. (p. 29) It seems to us that the correct response to uncertainty is not paralysis but experiment. Governments should not do nothing, out of a fear of failure. They should test policies, and be quick to learn from failure. If they suffer a misstep, they should try something else, not plunge ahead or retreat to the shore. (p. 31) This elite group of economists says, in essence, that we do not know how growth takes place, but it may be figured out from time to time. Quite a humble conclusion indeed. Nobody can predict in advance the time and place of the appearance from thin air of firms such as Microsoft, Google, Honda, Dell, Embraer or others and their lasting effects on growth. The neoclassical equilibrium model would argue in favour of the removal of all market barriers. The evolutionary model may, however, offer certain other policy advices. Certain freedom for a general action is most welcome, but it is not all. Many would question the available freedom, particularly political freedom, in China. This country may be in some dimensions a growth model to follow for many countries in the developing world. Economic policies were tested in China step by step from the 1980s. Responses by entrepreneurs, domestic and foreign, were positive in general. Even if some firms failed,

Conclusions

373

the society at large gained. So, careful intervention followed. This does not make China a development and growth disaster. Privatisation, for instance, was not emphasised. Nonetheless, a skilful government provided for a stable macroeconomic situation and openness to trade and investment, strongly encouraged the creation of human capital, stimulated transfer of technology and its local development, as well as motivated savings. South Korea had earlier a similar policy stance. Free markets that were supposed to ‘get prices right’ (to realise static comparative advantages), as the key ingredient of the neoclassical Washington Consensus, were not featuring high at all. Select intervention to develop dynamic local comparative advantages was the rule. These dynamic advantages in China and South Korea could, perhaps, be achieved without intervention much later (if ever). Hence this growth took place in countries with strong (not weak) and competent governments that intervened in the economy. How long this may last is another issue. The evolutionary model suggests that policies ought to be evaluated and adjusted continuously in the situation with uncertainty and endemic disequilibrium. The mere number and complexity of possibilities for the location of production in an evolutionary framework is enormous. As there is an important lacuna in our understanding of evolutionary spatial economics, it is of little wonder that there are many disagreements and only several agreements in theory. Hence, generalisations and synthesis in the form of policy recommendations are coupled with huge problems and risks. The list of conditions for the application of policy recommendation is long and their voicing is quite delicate, while final policy suggestions are few and narrow at best. Hence, policy advice on most issues (location, trade, technology, investment, taxes … ) must be based ‘on a mixture of theory, measurement and subjective judgement’ (Lipsey et al., 2005, p. 515). Nonetheless, there are certain general principles (lessons learned) that one has to keep in mind and consider during the search for a solution of a particular spatial problem. Here they are: • Starting a new activity in a given location requires building on and improving already existing accumulated expertise and specialisation of that location. One may need to stick to the production process that one understands in general terms. This bottom-up specialisation needs to be extended, but this does not at all mean that the earlier production structure has to be continued. Something (vaguely) related to it or derived from it may be the starting point which, later on, needs to be altered, sometimes fundamentally. • Using policy intervention to influence the location of firms is most effective only very early in the process. Evolutionary economics argues that very little intention is necessary to produce a successful locational result. There are few, and relatively short, moments of enhanced locational freedom for new industries that set in motion new industrial trajectories. If there is a choice of potential locations for the establishment of business, the winning location often wins by the skin of its teeth. Hence, there is only a narrow window of locational option through which a short, rare and slim beam of opportunity light can pass, during which time the policy may be effective. Market forces may be the strongest ‘socio-economic technology’ for resource allocation in the situation of disequilibrium in the medium and longer term. • Local education and training, as well as health of the local human capital, has a positive impact on the location of firms and industries.

Evolutionary Economic Geography

374

• Depending on the industry, economies of scale and functional linkages in production and its servicing (strategic complementarity) have certain impact on the location of firms and industries. • Policy should be reviewed periodically and have elements of flexibility. • Industrial success or failure of a region or a policy sometimes depends on uncontrollable factors and events that are located outside regions themselves. The rationale for policy intervention is known; what remains to be known is how to manage this situation. What exactly ought to be done; what needs to go and where; how; when; and what may work in the desired direction over time and space? However, the rationale for certain intervention does not mean that there is carte blanche for any policy: no intervention in the economy may be better than lousy intervention, as certain policy actions may have uncertain effects. Some may be so disastrous that they may require tremendous time and resources to be rectified. A small change at a crucial point in time (once the threshold is reached) can lead to shocking outcomes, which may lead to ‘locational catastrophes’. ‘The possibility of nonlinear responses to policy makes it much more difficult to forecast the effect of a given policy change … it can be seriously misleading to base expectations of the effects of future policies on linear approximations from the past’ (Baldwin et al., 2003, p. 229). Trade liberalisation, economic integration and other policy changes may have very nonlinear effects on industrialisation, location of firms and growth in an evolutionary framework. The evolutionary road is not only unpredictable, but also bumpy. Solving such issues may not easily fit into predetermined charts and models, in particular the ones that pretend to have universal application over time. Those who handle competition and location policy should also rely on their finer instincts as the evolutionary economy is shaped often by unpredictable impersonal forces.9 In any case, the basic principles and instincts in this area should include a preference for competition over monopoly, pluralism and spread over concentration, and new entrants over incumbents. The firms that will define the next twenty years of economic history and production geography are probably completely unknown to anyone today. The decoupling of value from physical production, its shift to intellectual production, is a millennial shift whose full meaning will not unfold for many generations. Uncertainty and risk are very high and the way forward is not always clear. Therefore, it is very difficult to predict with a high degree of confidence and certitude where the world’s economy is really going. Profound changes that occur little by little (such as the transformation in the economy and in our lives that was brought about by Microsoft or mobile phones) may be noticed only long afterwards. Therefore, the greatest danger and mistake is in ‘linear thinking’: the belief and expectation that what exists today will also exist tomorrow, but with a higher degree of intensity. Consider what the world looked like in 1980. No scholar ever dreamed or imagined then how the general situation would turn out in reality in 1991 or 2000. Or go back to 1900–20, or almost any other twenty-year-period, in particular during the twentieth century. Certain, even a number of claims by the evolutionary theory would inevitably turn out to be mistaken while others would need adaptation and updating. That is in the spirit of the evolutionary and complexity model of the economy that rests on non-linear, multiple equilibrium and dynamic foundations. However, a turn away from the neoclassical

Conclusions

375

equilibrium theory of the economy is clear. The future is no longer what it used to be in the neoclassical equilibrium world, while the past lasts longer. The evolutionary and complexity slant in economics is not a panacea for all enigmas that bother economics. It provides only new, wide and flexible tools to approach economic problems, old and new alike. Evolutionary aspects may be quite effective in finding solutions over medium and longer time periods. They are just like water which, in the end, always finds its way through, around, by, under or over a barrier. A vibrant, nonlinear and evolutionary economic environment is a more thrilling setting for research and analysis than is the neoclassical equilibrium world. Has a formidable problem related to the location of firms and industries in space become even more complicated after our considerations? Did we learn as much about evolutionary economic geography as we hoped to learn when we embarked on this academic voyage? No, we did not. If we did, this would mean that we equilibrated out expectations with the final outcome. This would be profoundly in contrast with the everchanging and never-ending evolutionary spirit of open-ended and relentless transformation. There is no steady state and satisfaction in any moment in time in the evolutionary model. Nonetheless, we hope that our ignorance now is at least just a bit smaller than was the case at the start of this academic voyage. The principal goal was to demonstrate and reinforce the value and importance of the evolutionary approach to economic analysis. Key theoretical issues regarding the future location of production in space are not settled. In the endemically uncertain and evolutionary world they cannot be, as there are no rules set in stone. Nonetheless, there are certain agreements and principles to follow and keep in mind, which may be of assistance in future academic voyages in research, policy debate and applications in practice. Our ignorance today is different from the ignorance prevailing over the past century. Nonetheless, economists may never discover relatively comfortable certainties such as those found by Pythagoras, Galileo, Newton or Euler within the part of the world that they studied. Contrary to some tests in the natural sciences, empirical analyses in social sciences have never given certain proofs that can survive the test of time. Even for short periods. These tests and laws in an ever-changing social life, with its varied strands, escape universal explanations and coverage. The best that economists may hope for is to set and adjust a creative frame for the study of existing and forthcoming puzzles. Whether this monograph has, in part, eased certain enigmas and challenges of evolutionary economic geography or has deepened the existing ones, it is up to the reader to judge. The modest objective was no more than to provoke thought and possibly a flurry of new thinking, discussion and research.10 Is evolutionary economic geography (1) a science, with its subject, hypothesis and method of analysis, that could conduct experiments and reproductions in order to arrive at useful predictions, statistics, logic and laws? These elements of a science should assist in understanding – or better understanding – natural phenomena (natural science) and human and social behaviour (social science). Or is it (2) art in the sense of a skill or mastery of human agency or the faculty to be creative? Or (3) a bit of both? It is up to you to decide.

Evolutionary Economic Geography

Notes

376

1 J.Kay,‘Economic forecasting will never be an exact science’, Financial Times,29 October 2003, p. 23. 2 Keynes’ often quoted dictum that ‘in the long run we’re all dead’ had a negative impact on the interest in structural and evolutionary issues, as well as in the supply side of the economy. However, one has to remember that Keynes had no children. Perhaps most people would prefer to leave a dynamic, competitive and adjustment-prone economy to their descendants than one that is spiritless. 3 ‘Finally, it seems appropriate to mention that the effect of distance on trade patterns is not diminishing over time. Contrary to popular impression, the world is not getting dramatically smaller’ (Leamer and Levinsohn, 1995, p. 1387). 4 The Japanese understood and implemented a policy of domestic rivalry of half a dozen conglomerates. This became the basis for the global competitiveness of their output such as passenger cars and consumer electronics during the 1970s and 1980s. They also transferred it elsewhere. For example, after the failure of Leyland in the 1980s, Britain had a relatively dynamic car production and market based to a large extent on the local production by Japanese TNCs. 5 The government may influence these markets in the longer run by, for example, its human resource (education) policy. An investment into a certain type of educated people now may alter the general economic comparative advantage of a nation in the future, after a time lag of some fifteen years and beyond. 6 ‘[P]aths emerge from a chaotic swirl of random events’ (Martin and Sunley, 2006, p. 424). This may be totally correct, but it does not give clear guidance to a policy-maker who is facing a real problem at a given time. Sometimes a government policy is forced by events or pressed by various lobbies to rush ahead of theoretical understanding and evidence. 7 Topics for further research include: spillovers, externalities and linkages, as they are so poorly understood; local interactions; monopolistic competition; cross-fertilisation with industrial organisation and urban economics, trade and growth; cost–benefit analysis that can include linkages; services, as research was mainly concentrated on the manufacturing sector; the speed of (exponential?) weakening of the effect of spillovers and information relative to distance, especially in developing countries; and agglomeration and spread of innovative activities. 8 Always run as fast as possible to keep staying where you are now. 9 These forces include not only wars, terrorist attacks or revolutions, but also natural disasters such as earthquakes, fires, floods, epidemics, allergies, as well as energy or food crises, misguided economic and monetary policies, increased criminal activity or a flood of clandestine immigrants. 10 The field of economics is notoriously slow to open itself to ideas that are different. It has become usual practice that most path-breaking articles in economics are rejected several times before they find a home in a journal. It took Brian Arthur six years and fourteen rewrites of his working paper ‘Competing technologies, increasing returns and lock-in by historical events’ before it was published by The Economic Journal in 1989 (‘From QWERTY to Microsoft’, Options (IIASA), winter 2007, p. 20).

Annex I Chance events The Portuguese in Latin America An interesting example of the impact of chance events can be found in the Treaty of Tordesillas. This Treaty was concluded between Spain and Portugal. It clarified the situation regarding claims on newly discovered territory across the Atlantic Ocean and was intended to pre-empt any misunderstandings following Christopher Columbus’s discovery of the New World in 1492. In fact, at first it was not fully realised that Columbus had found something ‘big’. Portugal wanted to protect its monopoly of the route to Africa. Pope Alexander IV (a Spaniard by birth) intervened to clear up a possible ‘confusion’ between the two countries. He issued a bull (decree) in 1493 that established an imaginary north–south line through the midAtlantic 100 leagues (480 km; 1 league = 4.8 km) west of the Cape Verde Islands. Spain acquired undiscovered non-Christian lands to the west of the line, while Portugal had those to the east. After further explorations, Portugal realised that the Spanish new territory was extensive. The Portuguese were discontented and initiated a renegotiation of the partition. The new partition line was established in Tordesillas, a Spanish town, in 1494. The new line ran 370 leagues (1,770 km) west of the Cape Verde Islands, which is how Portugal gained possession of Brazil. The partition line was ill defined, and over several centuries Portugal pushed the Brazilian frontier westwards. Spain raised no objection to this expansion as it was some centuries before longitude could be determined with any degree of accuracy. Cannes For long centuries Cannes was just a small sleepy fishing and agricultural village on the Côte d’Azur. The area was overgrown with cane (French canne), hence the name of the village. It stayed quiet like that until Henry Peter Brougham (1778–1868), a prominent British politician, ‘discovered’ it on the way from foggy London to his winter holiday in Nice and Italy in January 1834. His carriage had to stop in Cannes as the French authorities had sealed Nice because of cholera. He hoped that he would continue his journey to Nice in a few days’ time. In the meantime, he stayed in Cannes in a fisherman’s house, as there were no hotels. His hosts were not only friendly, but also treated their guest to the best fish and wines. Having such memories, the following year Lord Brougham returned to Cannes, a generally unknown place at that time. This sensation was noted in the London press and it gave ideas to others in London high society on where to go. Several members from this milieu followed Lord Brougham, and the local fishermen in Cannes treated them all nicely. The locals in Cannes started building hotels and restaurants that could cater to various demands of the elite visitors

Annex I

378

who began to stream into the village. So started, first, the location of the tourist industry and, then, the location of art (the film festival) in Cannes. Accordions Castelfidardo is a small commune in Ancona (Marche, Italy). It has a cluster that started producing accordions in 1864, which evolved later to electric and electronic musical instruments (keyboards). Paolo Soprani build the first barrel organ in Castelfidardo. He was nineteen when a passing pilgrim from Austria asked for hospitality at his father’s house. As a sign of gratitude and appreciation, the pilgrim left them an accordion patented by the Viennese Demian in 1829. The main activity in the Castelfidardo area at that time was agriculture. Instrument-building was an activity that supplemented farmers’ income, especially during the winters. Soprani had the time to copy and eventually improve the instrument. His workshop was opened in his house in 1864. He also involved his larger family in the business. In 1905 Soprani had 500 employees, while in 1910 the firm employed 600 (Tappi, 2005, pp. 298–9). Entrepreneurship, networking, collective learning, mobility of employees and flexibility in the business organisation were the biggest strengths in this workshop, which became a firm and, subsequently, a cluster of firms. The guitar took over the world accordion market during the 1960s. There was also a growing interest in electric and electronic musical instruments. In addition, producers of accordions from eastern Europe and China had much lower prices for their products than the Italians. At about the same time, Italian manufacturers in general started moving into transistors, electrics and electronics. This opportunity was recognised and seized by the (former) accordion producers. They moved into the production of electronic keyboards and, later, into video conferencing and video-control systems. The whole process was reinforced with relations between the business community and the Faculty of Engineering in Ancona. Castelfidardo still hosts some thirty firms that are involved in the traditional production of accordions. However, the cluster’s core business is in electronics (Tappi, 2005, p. 303). Chocolate The spread of certain products is also based on chance. For example, in 1519 the Aztec king Montezuma offered Hernán Cortés xocoatl (‘god drink’) made from roasted cocoa beans, pimento and corn flour. Subsequently, the recipe was passed on to Spain. In 1615, when the fourteen-year-old Spanish princess Ann of Austria, daughter of King Philip III, married her peer, Louis XIII of France, the Spanish hosts served chocolate to the European nobility. Ann in her turn introduced the confection to the French court, and thus chocolate crossed the Pyrenees. At about the same time, the Spanish Inquisition unwittingly instigated another chocolate crossing of the Pyrenees: some of the persecuted Jews fled to Bayonne (France), where they established cocoa-processing plants; others went to the Low Countries.

Annex I

379

Marriages Various judicious marriages (chances or historical accidents) entered into by the Hapsburgs (together with wars) created the Austro-Hungarian Empire. This altered the economic space in central and eastern Europe and elsewhere. Dukes of Burgundy also enlarged their territory by marriages and conquest. The same general strategy was used in other regions of the world too. For instance, the Khmer king Jayavarman II, who reined in the period 802–50, used marriages together with alliances and military campaigns to expand his country Kambujadesa (Cambodia).

Annex II Examples of the putty–clay locational model Relatively modest initial differences between countries or the humble beginnings of an industry or a firm in a region can have irreversible economic and other effects for a specific location or region for a long period of time. Even if two or more potential locations are initially identical, then the location case may be ‘solved’ by chance. Hence, it may be impossible to predict where a cluster will emerge. But once the industry starts developing, the process of circular interdependence or cumulative causation and acceleration (at least for some time) cuts in and may continue at the same location in space for a long time, even if the structure of the economy has changed. For example, there are interesting stories of why certain firms/industries came to be located where they are today: • For centuries Basle was an important trading and banking city because of its favourable geographical position. For instance, the Rhine is navigable from Basle. This supported first trade and later production specialisation in the textile industry. In neighbouring France, the national patent law of 1844 protected materials, not the production process. A Frenchman, Vergain, invented a red aniline dye (fuchsine) early in 1859. It was used for dyeing textiles and was patented by the dyestuff factory Renard Frères et Franc in April 1859. The patent gave the factory a virtual monopoly to produce it in the country. A few months after the granting of this patent (monopoly to produce), another Frenchman, Jean Gerber-Keller, invented a completely different method for producing the same red dye. However, the law prohibited him from producing this protected dye in France, even though the domestic demand was larger than the supply. Hence, he went abroad to Switzerland which, like many other countries, did not have patent laws. Basle was the preferred location because it was a trading and banking centre, salt was available in the vicinity, one could ‘safely’ unload arsenic into the Rhine, transport and forwarding infrastructure was readily available, and the oldest Swiss university, founded in 1460, was located there. Physician and alchemist Theophrastus Bombast von Hohenheim (1493–1541), known as Paracelsus, a precursor of homeopathy, lectured on the link between medical and chemical events, which freed medical thinking. Hence, Basle was a prime and open location to start production of something new. At the same time in France, the (monopolist) producer of the dye, mollycoddled by the system, did not invest further in R&D or improvements in production techniques, while the competitive market structure in Switzerland motivated firms to increase production efficiency. The price of the dye in Switzerland was less than half of that in France, so smuggling into France escalated, and the French monopoly subsequently collapsed (Weder, 1995). Production of pharmaceuticals emerged, almost by chance, around 1880, when the curative effects of dyes became apparent. The Swiss dye industry could not compete with its German

Annex II

381

counterpart either on scale of output or on access to raw inputs. Therefore, the Swiss specialised in high-value-added market segments of medicines, pesticides, herbicides, perfumes and flavourings. Highly sophisticated medical and health-care services and equipment followed their expansion throughout the country. Strong incentives came from an inflow of talents and the invalidation of German patents by the Allies following the two world wars, as well as a government policy in education and through the support of R&D. As a relatively small country with limited domestic market, Switzerland had to export from the outset. It became specialised and highly competitive in an industry that is not based on natural resources, which are in short supply in Switzerland. R&D and highly specialised human capital are among the biggest inputs in this business success. • Towards the end of the Middle Ages, Genevan goldsmiths and jewellers were renowned for their skills, their knowledge of precious metals and the splendour of their products. They were ‘expected’ to become involved in the new watchmaking industry that started to appear in France, Flanders, England and Germany around 1500. However, the Genevans were so prosperous that they were uninterested in the emerging industry. Nevertheless, two chance events prompted the abrupt appearance of watchmaking in Geneva. First, in 1541, John Calvin issued an edict against luxury, pleasure, elegant clothing and ‘useless jewellery’. This limited the activities of the Geneva jewellers and goldsmiths and virtually ended their craft twenty-five years later, when their activities were even more restricted. Second, at about the same time, the persecuted Protestant Huguenots from France and Flanders found sanctuary in Switzerland (Basle, Geneva and Zurich). Watchmaking knowledge and experience was brought by some of these refugees, one of whom was Charles Cusin, who settled in Geneva in 1574. As watches performed a ‘useful’ function, Geneva’s jewellers and goldsmiths learned how to make quality, stylish and durable watches in order to find a new source of income. Once the number of watchmakers increased, they instituted their own guild in 1601. Most of the production was sold outside the country as few Swiss were able and willing to afford such watches. Nonetheless, production flourished and the city became ‘congested’ with a cluster of watchmakers. It became increasingly difficult and time-consuming to obtain a master watchmaker’s licence, so many of the apprentices moved to other Swiss cantons. They mostly clustered in the Swiss Jura (Nyon, Neuchâtel, La Chaux-de-Fonds, Bienne and Basle). In contrast to Geneva, light regulations permitted the production of parts to be subcontracted to the peasants in the mountains, who had limited sources of income, especially during the long winters. These watches were, of course, less sophisticated and cheaper than the ones produced in Geneva, but they found a niche in the market and production expanded. In spite of changes in time measurement technology and hard times during the 1970s and 1980s, the Geneva watchmaking cluster (for stylish watches) and the Jura cluster (for standard watches) weathered the storm, justified their label ‘Swiss made’ and retained ‘global’ dominance (Bumbacher, 1995). • Sassuolo (near Bologna, Italy) is, perhaps, the world capital in the production of ceramic tiles. Pot-making started there in the thirteenth century. Production of ceramic tiles for street names and house numbers began in the nineteenth century. Immediately after the Second World War, there were only a few ceramic tile producers left in the area. The post-war reconstruction created a boom for all building materials, including

Annex II

382

Italian ceramic tiles. Tiles did well on the market because in the Mediterranean climate they were cool in the hot weather, because wood was scarce and expensive, and because the Italians prefer natural materials to vinyl and carpeting. The principal raw material was kaolin (white) clay. The region was abundant in red clay, hence the white had to be imported from Britain. The equipment, kilns and presses were imported (mainly from Germany) in the 1950s and 1960s. From the mid-1960s, the local entrepreneurs learned how to modify the imported equipment for the use of the regionally available red clay. In the 1970s, the Sassuolo area manufacturers became very competitive suppliers of kilns and presses, and they began to export them. A relatively saturated domestic tile market in the 1980s provided the incentive to export ceramic tiles all around the world. Producers were successful in responding quickly to the changing design and other demand conditions which gave them the competitive edge (Porter, 1990a, pp. 210–25). • Modena (Italy) is an important hub for translations from and into many languages (even though the Italians may not be, in general, widely known as strong polyglots). This area has long been well known for the production and worldwide export of agricultural machinery. As the machinery required instruction manuals in various languages, such publications were initiated in Modena; once established, the industry developed and expanded. • Provided that profitability is maintained, personal factors (that is, whims) can play a role in firm location decision. For example, car production factories were established in Detroit and Oxford as in each case the founder of one of the major manufacturers (Ford and Morris, respectively) was born or grew up there. Incidentally, Bill Gates, the founder and chairman of Microsoft, was born in Seattle. Fashion-producing firm Hugo Boss is located in Metzingen (Germany) because Hugo Boss was born there and he wanted to contribute to the development of his home town. Metzingen is also situated very close to Reutlingen (near Stuttgart) and there is quite a significant textile industry in the area. Now, not only is there Hugo Boss production and a huge outlet in Metzingen, but others have also established enterprises there, such as Escade and Bally. Elsewhere, for instance, Hilti, a toolmaker, is located in Liechtenstein because its founder Martin Hilti was born there. Bentonville, Arkansas (can you find it on a map?) is the base for Wal-Mart, the world’s leading chain of retail shops. This is where Sam Walton, father of Robson Walton, founder of Wal-Mart, grew up. Elsewhere, it has been known for golf-addicted entrepreneurs and business executives to search for locations for their businesses that are close to good golfing facilities, as in Lausanne. • The systems of paying sailors had strong effects on the formation of human capital to meet the changing location of trade patterns. The English paid their sailors by the voyage while the Italians paid by the day. Although the Italian system was probably well suited to Mediterranean conditions, it caused problems when the Italians entered the Atlantic trade. English sailors would weather tough storms to complete a voyage quickly while Italian sailors would stay in harbour awaiting the most favourable winds. Thus the English sailors endured a process of learning by doing in facing the tougher weather than did the Italians. (Lipsey et al., 2005, p. 173).

Annex II

383

[T]he Venetians did not recognize the need to alter their methods of transport and their trading organisations fast enough after the threemasted sailing ship had drastically altered trading routes and arrangements. This hastened the decline that was being brought about by their unfavourable location with respect to the new routes. (ibid., p. 415) • Policy changes such as the passage of anti-union statutes in the southern states of the US gave certain manufacturing firms the incentive to leave the northern part of the country. This process was supported by another unrelated event: the advent of airconditioning, which made the southern climate more attractive and tolerable for living and working in than that in the north of the country. • Edward Lloyd (d. 1713) ran Lloyd’sCoffee House in London, located close to the docks of the Thames. This coffee house became a popular meeting place for ship-owners, merchants and captains. As such, Lloyd was able to acquire valuable and timely information of potential commercial value: the arrival and departure of ships, their size and cargo, their crew, and the like. Lloyd meticulously collected this information and passed it on to select clients. Later, about 1688, he devoted a separate room in his coffee house and equipped with the business technology of the day (paper and ink). Out of this coffee house ‘office’ evolved one of the biggest maritime insurance companies in the world. • In New York City, as was the case earlier in London, businesses benefited from scheduled and other maritime transport services, not available elsewhere in the country. Brokers, insurers, underwriters and financiers of trade flocked to the city. Finance concentrated and expanded in Manhattan. A static locational advantage evolved into the dynamic advantages that assisted the evolution of the whole economy throughout the country. • During the Renaissance (fourteenth to seventeenth century),1 Florence was the principal centre for art in what is today known as Italy. The clients who lived there were sophisticated, they appreciated art and were able and willing to pay for it. At the same time the artists needed such clients. The bankers, the rich Medici family in Florence, wanted to be remembered after their terrestrial life. So that the good memory of their terrestrial presence would last, they chose to invest in art.2 Well, many from various walks of life would argue with gratitude and respect for centuries that their choice at that time was next to perfect. Ludovico Gonzaga in Mantua and Federico da Montefeltro in Urbino made a similar choice at about the same period. One further word about the glorious Renaissance: this was a highly elitist movement that was concentrated, touched and directly involved only 100 persons. Perhaps just a few more, but not many more. • St Martin’s Court is a short street near the Royal Opera House in London, and is full of sellers of second-hand books and prints. Potential customers go there because they expect to find a number of shops with a wide range of second-hand books, while shopowners locate there because they expect to welcome a stream of clients. A similar reasoning explains the cluster of theatres around Leicester Square and the location of restaurants in Soho. • The densest art expertise in the world can be found now along a kilometre-long stretch that connects New Bond Street, Old Bond Street and St James’s Street in London. The

Annex II

384

combined turnover of the auction houses Christie’s, Sotheby’s, Bonham’s and Phillips’ is the largest in the world. In addition, in the same area there are fifteen top art dealers. They all tend to flourish during troubled times: periods of calm and political stability have been disastrous for their business, while unpredictable crisis, death, divorce and sorrow are their best friends. Notes

1 The Renaissance got its principal boost after the fall of Constantinople to the Muslim invaders in 1453. Many Greek scholars found refuge from the Turks in what is now known as Italy. These learned persons brought their accumulated knowledge and long experience, which provided a strong force in the search for ancient beauty and wisdom. 2 Besides banking, wars, stupendous art and the luxurious life, Florence created the precursor of the social security system. This was done in the form of a Dowry Fund. The Florentines worried that young women would not marry and form families; hence, they wanted to assist financially in their family-related future. The problem was that many thriving art studios/workshops in Florence were full of young men. Homosexuality was widespread.

Annex III Select clusters in Italy and Germany Italy is an often quoted example of a country with distinct manufacturing clusters. A select group of those clusters includes the following:1 • motorcycles in Bologna [2370]; • electronics, mainly alarms for cars, in Varese [100]; • jewellery in Valenza Po [1400] (Alessandria); Vicenza [1100]; Arezzo [1300] (Florence); • spectacles in Cadore [930] (Belluno); • textiles and clothing in Sempione [3900] (Varese); • textiles around Lake Como; Prato [8481] (Florence); Olgiatese [2614] (Varese); Biella [1300] (Piedmont); Valdagno (Pisa); • clothing in Val Vibrata [1150] (Pescara); Empoli (Florence); Treviso; • female underwear in Castel Goffredo [280] (Mantova); • silk in Comasco [2600] (Como); • wool in Biella; • knitwear in Carpi [2054] (Modena); • shoes in Fermo, Montegranaro, Porto Santelpidio, Sanbenedetto, San Benedetto del Trono (Ancona); Lucca, Santa Croce Sull’Arno [1749] (Pisa); Ascoli [3100]; Riviera del Brenta [886] (Padova); • sports footwear in Montebelluna [623] (Treviso); Asolo (Treviso); • leather in Arzignano [600] (Vicenza) and Solofra (Naples); • ceramic tiles in Sassuolo [199] (Bologna); • marble in Apuo-Versiliese [1161] (Carrara); • taps and valves in Alto Cusio [300] (Novara); • furniture in Brianza Comasca Milanese [6500] (Milan); Cantu [7200] (Milan); Alto Livenza [2000] (Udine); Poggibonsi [1294] (Siena); Bovolone-Cerea [3000] (Verona); • kitchens in Pesaro [1200]; • chairs and tables in Udine [1200]; • wood machinery in Rimini [1345]; agricultural machinery in Modena [100]; • foodstuffs in Parma [215]; • saucepans and valves in Lumezzane [1008] (Brescia); • packaging machinery in Bologna; and • accordions and other musical instruments in Castelfidardo [400] (Ancona). Select clusters in Germany include: • steel in Dortmund, Essen and Düsseldorf; • cars in Hanover, Wolfsburg, Stuttgart, Munich, Ingolstadt, Neckarsulm, Bochum, Cologne and Regensburg;

Annex III

386

• trains in Berlin, Braunschweig and Nuremberg; • locksmiths’ products in Velbert; • tools in Remscheid; • machine tools in Stuttgart; • cutting tools and cutlery in Solingen; • precision tools in Enzkreis; • surgical instruments in Tuttlingen; • instruments in Hamburg, Berlin and Munich; • office machines in Hamburg, Hanover, Bremen, Frankfurt, Munich, Leipzig and Berlin; • printing machinery in Heidelberg, Würzburg, Offenbach, Augsburg, Leipzig and Frankenthal; • chemicals in Leverkusen, Frankfurt and Ludwigshafen; • pharmaceuticals in Berlin and Dresden; • jewellery in Pforzheim; • hollow glass in Holzminden, Jena, Kronach and Hildburghausen; • glass-blowing and Christmas tree decorations in Lauscha; • optics in Wetzlar and Jena; • pens and pencils in Nuremberg; • knitting in Reutlingen; and • porcelain in Wuppertal, Coesfeld and Tirschenreuth. Note

1 The list is based on data for 1996; if available, the number of related firms in the cluster is given in square brackets (although the number of firms in a cluster changes over time, and it is given here only as an indication); and a bigger city in the relative vicinity or province is given in round brackets. Source: ‘Quanti sono i distretti industriali in Italia?’ Newsletter Club dei Distretti Industriali no. 9, November 1998, Club dei distretti industriali Prato, p. 9.

Bibliography And further, my son, be admonished by these. Of making many books there is no end, and much study is wearisome to the flesh. (Ecclesiastes 12:12) Abreu, M., H. de Groot and R. Florax (2005). ‘A meta-analysis of β-convergence: the legendary 2%’, Journal of Economic Surveys, pp. 390–420. Ahearne, A., J. Pisani-Ferry, A. Sapir and N. Veron (2006). ‘The EU and the governance of globalisation’, Prime Minister’sOffice, Economic Council of Finland. Aiginger, K. and S. Davies (2004). ‘Industrial specialisation and geographic concentration: two sides of the same coin?’ Journal of Applied Economics, pp. 231–48. Aiginger, K., and W. Leitner (2002). ‘Regional concentration in the United States and Europe: who follows whom?’ Weltwirtschaftliches Archiv, pp. 652–79. Aiginger, K. and M. Pfaffermayr (2004). ‘The Single Market and geographic concentration in Europe’, Review of International Economics, pp.1–11. Albronoz, F., and G. Corcos (2007). ‘Regional integration, subsidy competition and relocation choice of MNCs’, B.E. Journal of Economic Analysis and Policy, pp.1–25. Alchian, A. (1950). ‘Uncertainty, evolution, and economic theory’, Journal of Political Economy, pp. 211–21. Altshuler, R., H. Grubert and T. Newlon (1998). ‘Has US investment abroad become more sensitive to tax rates?’, NBER Working Paper No. 6383, National Bureau of Economic Research, Cambridge, MA. Amiti, M. (1998). ‘New trade theories and industrial location in the EU: a survey of evidence’, Oxford Economic Papers, pp. 45–53. ——(1999). ‘Specialization patterns in Europe’, Weltwirtschaftliches Archiv, pp. 573–93. ——(2005). ‘Location of vertically linked industries: agglomeration versus comparative advantage’, European Economic Review, pp. 809–32. Andaluz, J., L. Lanaspa and F. Sanz (2002). ‘Geographical dynamics: a sectoral comparison between economic landscapes of the United States and Europe’, Regional Studies, pp. 321–32. Ando, K. (2005). Japanese Multinationals in Europe. Cheltenham: Edward Elgar. Anson, J., O. Cadot, A. Estevadeordal, J. de Melo, A. Suwa-Eisenmann and B. Tumurchudur (2005). ‘Rules of origin in North–South preferential trading arrangements with an application to NAFTA’, Review of International Economics, pp. 501–17. Armington, P. (1969). ‘A theory of demand for products distinguished by place of production’, International Monetary Fund Staff Papers, pp. 159–78. Armstrong, H. (1995). ‘Convergence among regions of the European Union, 1950–90’, Papers in Regional Science, pp. 143–52. ——(1996). ‘European Union regional policy: sleepwalking to a crisis’, International Regional Science Review, pp. 193–209. Armstrong, H. and R. de Kervenoael (1997). ‘Regional economic change in the European Union’, in J. Bachtler and I. Turok (eds) The Coherence of EU Regional Policy. London: Jessica Kingsley, pp. 29–47. Armstrong, H. and J. Taylor (2003). Regional Economics and Policy. Oxford: Blackwell.

Bibliography

388

Arrow, K. (1962). ‘The economic implications of learning by doing’, Review of Economic Studies, pp. 155–73. Arthur, B. (1988). ‘Competing technologies: an overview’, in G. Dosi, R. Nelson, G. Silverberg and L. Soete (eds) Technical Change and Economic Theory. London: Pinter, pp. 590–607. ——(1989). ‘Competing technologies, increasing returns, and lock-in by historical events’, Economic Journal, pp. 116–31. ——(1990a). ‘Positive feedbacks in the economy’, Scientific American, February, pp. 92–9. ——(1990b). ‘Silicon Valley locational clusters: when do increasing returns imply monopoly?’, Mathematical Social Sciences, pp. 235–351. ——(1994a). ‘Industrial location patterns and the importance of history’, in B. Arthur (ed.) Increasing Returns and Path Dependence in the Economy. Michigan: University of Michigan Press, pp. 49–67. ——(1994b). ‘Urban systems and historical path dependence’, in B. Arthur (ed.) Increasing Returns and Path Dependence in the Economy. Michigan: Michigan University Press, pp. 99–110. ——(1995). ‘Complexity in economics and financial markets’, Complexity, No.1 (April), pp. 20–5. ——(1999). ‘Complexity and the economy’, Science, Vol. 284 (2 April 1999), pp. 107–9. ——(2002). ‘How growth builds upon growth in high-technology’, Annual Sir Charles Carter Lecture. Belfast: Northern Ireland Economic Council. ——(2007). ‘The structure of invention’, Research Policy, pp. 274–87. Atkeson, A. and P. Kehoe (2007). ‘Modelling the transition to a new economy: lessons from two technological revolutions’, American Economic Review, pp. 64–88. Audretsch, D. (1998). ‘Agglomeration and the location of innovative activity’, Oxford Review of Economic Policy, pp. 18–29. Audretsch, D. and D. Dohse (2004). ‘The impact of location on firm growth’, CEPR Discussion Paper No. 4332, Centre for Economic Policy Research, London. Audretsch, D. and M. Feldman (1995). ‘Innovative clusters and the industry life cycle’, CEPR Discussion Paper No. 1161, Centre for Economic Policy Research, London. ——(1996). ‘R&D spillovers and the geography of innovation and production’, American Economic Review, pp. 630–40. ——(2004). ‘Knowledge spillovers and the geography of innovation’, in J. Henderson and J. Thisse (eds) Handbook of Regional and Urban Economics. Amsterdam: Elsevier, pp. 2713–39. Auyang, S. (1998). Foundations of Complex-System Theories in Economics, Evolutionary Biology and Statistical Physics. Cambridge: Cambridge University Press. Bachtler, J. and R. Michie (1993). ‘The restructuring of regional policy in the European Community’, Regional Studies, pp. 719–25. ——(1995). ‘A new era in EU regional policy evaluation? The appraisal of the structural funds’, Regional Studies, pp. 745–51. Bachtler, J., and C. Wren (2006). ‘Evaluation of European Union Cohesion Policy: research questions and policy challenges’, Regional Studies, pp. 143–53. Bagchi-sen, S., and H. Lawton Smith (eds) (2006). Economic Geography. London: Routledge. Baily, M. (2007). ‘The globalization of research and development and innovation’, Testimony before the Committee on Science and Technology, United States House of Representatives, 12 June 2007. Balassa, B. and L. Bauwens (1988). ‘The determinants of intra-European trade in manufactured goods’, European Economic Review, pp. 1421–37. Baldwin, R. and P. Krugman (2004). ‘Agglomeration, integration and tax harmonisation’, European Economic Review, pp. 1–23. Baldwin, R. and P. Martin (2004). ‘Agglomeration and regional growth’, in J. Henderson and J. Thisse (eds) Handbook of Regional and Urban Economics. Amsterdam: Elsevier, pp. 2671–711.

Bibliography

389

Baldwin, R., and T. Okubo (2005). ‘Heterogeneous firms, agglomeration and economic geography: spatial selection and sorting’, Geneva, Graduate Institute of International Studies, mimeo. Baldwin, R., R. Forslid, P. Martin, G. Ottaviano and F. Robert-Nicoud (2003). ‘Policy and economic geography: what’s new?’, in R. Baldwin, R. Forslid, P. Martin, G. Ottaviano and F. Robert-Nicoud, Economic Geography and Public Policy, Princeton: Princeton University Press, chapter 9, pp. 227–42. Barrios, S., S. Dimelis, H. Louri and E. Strobl (2004). ‘Efficiency spillovers from foreign direct investment in the EU periphery: a comparative study of Greece, Ireland and Spain’, Review of World Economics, pp. 688–705. Barro, R. and X. Sala-i-Martin (1991). ‘Convergence across states and regions’, Brookings Papers on Economic Activity, pp. 107–82. Bartik, T. (1991). Who Benefits from State and Local Economic Development Policies? Kalamazoo: Upjohn Institute. Baslé, M. (2006). ‘Strengths and weaknesses of European Union policy evaluation methods: expost evaluation of Objective 2, 1994–99’, Regional Studies, pp. 225–35. Beaudry, C., S. Breschi and P. Swann (2002). ‘Clusters, innovation and growth’, in J. Dunning and J. Mucchielli (eds) Multinational Firms. London: Routledge, pp. 190–213. Beckmann, M. (1999). Lectures on Location Theory. Berlin: Springer. Begg, I. and D. Mayes (1994). ‘Peripherality and Northern Ireland’, National Institute Economic Review, pp. 90–100. Behrens, K., J. Hamilton, G. Ottaviano and J. Thisse (2007). ‘Commodity tax harmonization and the location of industry’, Journal of International Economics, pp. 271–91. Beinhocker, E. (2006). The Origin of Wealth. London: Random House. Bekar, C. and R. Lipsey (2001). ‘Clusters and economic policy’, Simon Fraser University, Vancouver, mimeo. Belderbos, R. (1997). ‘Antidumping and tariff jumping: Japanese firms’ DFI in the European Union and the United States’, Weltwirtschaftliches Archiv, pp. 419–57. Bernstein, J. and D. Weinstein (2002). ‘Do endowments predict the location of production? Evidence from national and international data’, Journal of International 5Economics, pp. 55–76. Binder, S. (2003). Stalemate: Causes and Consequences of Legislative Gridlock. Washington: Brookings Institution Press. Bird, R. (1972). ‘The need for regional policy in common market’, in P. Robson (ed.) International Economic Integration. Harmondsworth: Penguin Books, pp. 257–77. Blomström, M. and A. Kokko (1997). ‘Regional integration and foreign direct investment’, Policy Research Working Paper No. 1750, World Bank, Washington, DC. ——(2003). ‘The economics of foreign direct investment incentives’, Stockholm School of Economics Working Paper No. 168, Stockholm. Blonigen, B. (1997). ‘Firm-specific assets and the link between exchange rates and foreign direct investment’, American Economic Review, pp. 447–65. Boldrin, M. and F. Canova (2001). ‘Inequality and convergence in Europe’s regions: reconsidering European regional policies’, Economic Policy, pp. 207–53. Borrus, M. and J. Zysman (1997). ‘Wintelism and the changing terms of global competition: prototype of the future?’, BRIE Working Paper No. 96B, University of California, Berkeley. Boschma, R. (2004). ‘Competitiveness of regions from an evolutionary perspective’, Regional Studies, pp. 1001–14. ——(2005). ‘Proximity and innovation: a critical assessment’, Regional Studies, pp.61–74. Boschma, R. and J. Lambooy (1999a). ‘Evolutionary economics and economic geography’, Evolutionary Economics, pp. 411–29. ——(1999b). ‘The prospects of an adjustment policy based on collective learning in old industrial regions’, GeoJournal, pp. 391–9.

Bibliography

390

Boschma, R. and R. Martin (2007). ‘Constructing an evolutionary economic geography’, Journal of Economic Geography, pp. 537–48. Bosker, M., S. Brakman, H. Garretsen and M. Schramm (2006). ‘A century of shocks: the evolution of the German city size distribution 1925–99’, CESifo Working Paper No. 1728, CESifo, Munich. Bottazzi, L. and G. Peri (2002). ‘Innovation and spillovers in regions: evidence from European patent data’, CEPR Working Paper No. 215, Centre for Economic Policy Research, London. Boulding, K. (1991). ‘What is evolutionary economics?’, Journal of Evolutionary Economics, pp.9–17. Bowles, S. (2004). Microeconomics. New York: Russell Sage Foundation. Bradford, S. and R. Lawrence (2004). Has Globalization Gone Far Enough? Washington: Institute for International Economics. Brainard, W. and J. Tobin (1992). ‘On the internationalization of portfolios’, Oxford Economic Papers, pp. 533–65. Brakman, S., H. Garretsen and M. Schramm (2002). ‘The strategic bombing of German cities during World War II and its impact on city growth’, CESifo Working Paper No. 808, CESifo, Munich. Braunerhjelm, P., R. Faini, V. Norman, F. Ruane and P. Seabright (2000). Integration and the Regions of Europe: How the Right Policies Can Prevent Polarization. London: CEPR. Bräuninger, M., and A. Niebuhr (2005). ‘Agglomeration, spatial Interaction and convergence in the EU’, HWWA Discussion Paper No. 322, HWWA, Hamburg. Breschi, S. and F. Lissoni (2001). ‘Knowledge spillovers and local innovation systems: a critical survey’, Industrial and Corporate Change, pp. 975–1005. Brülhart, M. (1998a). ‘Economic geography, industry location and trade: the evidence’, World Economy, pp. 775–801. ——(1998b). ‘Trading places: industrial specialization in the European Union’, Journal of Common Market Studies, pp. 319–46. Brülhart, M. and R. Elliott (1998). ‘Adjustment to the European single market: inferences from intra-industry trade patterns’, Journal of Economic Studies, pp. 225–47. ——(1999). ‘A survey of intra-industry trade in the European Union’, in M. Brülhart and R. Hine (eds) Intra-Industry Trade and Adjustment: The European Experience. Basingstoke: Macmillan, pp. 98–117. Brülhart, M. and R. Hine (eds) (1999). Intra-Industry Trade and Adjustment: The European Experience. Basingstoke: Macmillan. Brülhart, M. and J. Torstensson (1996). ‘Regional integration, scale economies and industry location in the European Union’, CEPR Discussion Paper No. 1435, Centre for Economic Policy Research, London. Brülhart, M. and F. Trionfetti (2004). ‘Public expenditure, international specialisation and agglomeration’, European Economic Review, pp. 851–81. Buigues, P. and A. Jacquemin (1992). ‘Foreign direct investment and exports in the Common Market’, paper presented at the Japanese Direct Investment in a Unifying Europe, conference held in INSEAD, Fontainebleau, 26–27 June 1992, mimeo. Bumbacher, U. (1995). ‘The Swiss watch industry’, in M. Enright and R. Weder (eds) Studies in Swiss Competitive Advantage. Bern: Peter Lang, pp. 113–51. Cadot, O., J. De Melo and A. Portugal-Perez (2007). ‘Rules of origin for preferential trading arrangements: implications for the ASEAN Free Trade Area of EU and US experience’, Journal of Economic Integration, pp. 288–319. Cairncross, F. (2001). The Death of Distance 2.0. London: Texere. Camagni, R. (2002). ‘On the concept of territorial competitiveness: sound or misleading?’, Urban Studies, pp. 2395–411. Canova, F. and A. Marcet (1995). ‘The poor stay poor: non-convergence across countries and regions’, CEPR Discussion Paper No. 1265, Centre for Economic Policy Research, London.

Bibliography

391

Cantwell, J. and L. Piscitello (2005). ‘Recent location of foreign-owned research and development activities by large multinational corporations in the European regions’, Regional Studies, pp. 1–16. Carlton, D. (1983). ‘The location and employment choices of new firms: an econometric model with discrete and continuous endogenous variables’, Review of Economics and Statistics, pp. 440–9. Carrington, A. (2003). ‘A divided Europe? Regional convergence and neighbourhood spillover effects’, Kyklos, pp. 381–94. Casella, A. (1996). ‘Large countries, small countries and the enlargement of trade blocs’, European Economic Review, pp. 389–415. Castellacci, F. (2007). ‘Evolutionary and new growth theories. Are they converging?’, Journal of Economic Surveys, pp. 585–627. Caves, R. (1996). Multinational Enterprise and Economic Analysis. Cambridge: Cambridge University Press. Chakrabarti, R. and B. Scholnick (2002). ‘Exchange rate expectations and foreign direct investment flows’, Weltwirtschaftliches Archiv, pp. 1–21. Christaller, W. (1933). Die Zentralen Orte in Süddeutschland. Jena: Gustav Fischer. (Translated from German by C. Baskin, Central Places in Southern Germany (1966), Englewood Cliffs: Prentice-Hall.) Clark, G., M. Gertler and M. Feldman (eds) (2000). The Oxford Handbook of Economic Geography. Oxford: Oxford University Press. Clausing, K. (1998). ‘The impact of transfer pricing on intrafirm trade’, NBER Working Paper 6688, NBER, Cambridge, MA. ——(2000). ‘Customs unions and free trade areas’, Journal of Economic Integration, pp. 418–35. Combes, P. and H. Overman (2004). ‘The spatial distribution of economic activities in the European Union’, in J. Henderson and J. Thisse (eds) Handbook of Regional and Urban Economics. Amsterdam: Elsevier, pp. 2845–909. Crafts, N. and A. Venables (2001). ‘Globalization in history: a geographical perspective’, CEPR Discussion Paper No. 3079, Centre for Economic Performance, London School of Economics, London. Crandall, R. and C. Winston (2003). ‘Does antitrust policy improve consumer welfare? Assessing the evidence’, Journal of Economic Perspectives, pp .3–26. Crozet, M., T. Mayer and J. Mucchielli (2004). ‘How do firms agglomerate? A study of FDI in France’, Regional Science and Urban Economics, pp. 27–54. Culem, C. (1988). ‘The location determinants of direct investments among industrialized countries’, European Economic Review, pp. 885–904. Curzon-Price, V. (1987). Free Trade Areas? The European Experience. Toronto: C. D. Howe Institute. ——(1991). ‘The threat of ‘Fortress Europe’ from the development of social and industrial policies at European level’, Aussenwirtschaft, pp. 119–38. ——(1993). ‘EEC’s strategic trade-cum-industrial policy: a public choice analysis’, in M. Hilf and E. Petersmann (eds) National Constitutions and International Economic Law. Deventer: Kluwer, pp. 391–405. ——(1996a). ‘Residual obstacles to trade in the Single European Market’, Euryopa, Institut européen de l’Universite de Genève. ——(1996b). ‘The role of regional trade and investment agreements’, in A. Kuyvenhoven, O. Memedovic and N. van der Windt (eds) Transition in Central and Eastern Europe. Dordrecht: Kluwer, pp. 69–88. ——(1997). ‘The European Free Trade Association’, in A. El-Agraa (ed.) Economic Integration Worldwide. London: Macmillan, pp. 175–202. David, P. (1984). ‘High technology centers and the economics of locational tournaments’, Stanford University, mimeo.

Bibliography

392

Davies, R. and D. Weinstein (2002). ‘Bones, bombs and break points: the geography of economic activity’, American Economic Review, pp. 1269–89. ——(2003). ‘Market access, economic geography and comparative advantage: an empirical test’, Journal of International Economics, pp. 1–23. de Grauwe, P. and M. Polan (2003). ‘Globalisation and social spending’, CESifo Working Paper No. 885, CESifo, Munich. Della Giusta, M., U. Kambhampati and R. Hunter Wade (eds) (2006). Critical Perspectives on Globalization. Cheltenham: Edward Elgar. Destler, I. (2005). American Trade Politics. Washington: Institute for International Economics. Devereux, M. and R. Griffith (1998). ‘Taxes and location of production: evidence from panel of US multinationals’, Journal of Public Economics, pp. 335–67. Devine, W. (1983). ‘From shafts to Wires: historical perspective on electrification’, Journal of Economic History, pp. 347–72. Dicken, P., M. Forsgren and A. Malmberg (1995). ‘The local embeddedness of transnational corporations’, in A. Amin and N. Thrift (eds) Globalization, Institutions and Regional Development in Europe. Oxford: Oxford University Press, pp. 23–45. Dickerson, A., H. Gibson and E. Tsakalotos (1997). ‘The impact of acquisitions on company performance: evidence from a large panel of UK firms’, Oxford Economic Papers, pp. 344–61. Dosi, G. and R. Nelson (1994). ‘An introduction to evolutionary theories in economics’, Journal of Evolutionary Economics, pp. 153–72. Dosi, G., K. Pavitt and L. Soete (1990). The Economics of Technical Change and International Trade. New York: Harvester Wheatsheaf. Dunning, J. (1988). Explaining International Production. London: Unwin Hyman. ——(ed.) (1993). The Theory of Transnational Corporations (The United Nations Library on Transnational Corporations). London: Routledge. ——(1994). ‘MNE activity: comparing the NAFTA and the European Community’, in L. Eden (ed.) Multinationals in North America. Calgary: University of Calgary Press, pp. 277–308. ——(1997). ‘The European Internal Market Programme and inbound foreign direct investment’, Journal of Common Market Studies (PartI)pp.1–30, (Part II) pp. 189–223. ——(1998a). ‘Location and the multinational enterprise: a neglected factor?’ Journal of International Business Studies, pp .45–66. ——(1998b). ‘Globalization, technological change and the spatial organization of economic activity’, in A. Chandler, P. Hagström and O. Sölvell (eds) The Dynamic Firm. Oxford: Oxford University Press, pp. 289–314. ——(1998c). ‘Regions, globalization and the knowledge economy: the issues stated’, Rutgers University, New Brunswick, NJ, mimeo. ——(1999). ‘The eclectic paradigm as an envelope for economic and business theories of MNE activity’, Discussion Paper No. 263, Department of Economics, University of Reading. Duranton, G. and D. Puga (2004). ‘Micro-foundations of urban agglomeration economies’ in V. Henderson and J. Thisse (eds) Handbook of Regional and Urban Economics. Amsterdam: Elsevier, pp. 2063–117. Duranton, G. and M. Storper (2006). ‘Agglomeration and growth: a dialogue between economists and geographers’, Journal of Economic Geography, pp. 1–7. Easson, A. (2001). ‘Tax incentives for foreign direct investment’, International Bureau of Fiscal Documentation Bulletin (Part I) pp. 266–74, (Part II) pp. 365–75. Eaton, C. and R. Lipsey (1975). ‘The principle of minimum differentiation reconsidered: some new developments in the theory of spatial competition’, Review of Economic Studies, pp. 27–49. ——(1976). ‘The non-uniqueness of equilibrium in the Löschian location model’, American Economic Review, pp. 77–93. ——(1979a). ‘Comparison shopping and the clustering of homogeneous firms’, Journal of Regional Science, pp. 421–35.

Bibliography

393

——(1979b). ‘The theory of market pre-emption: the persistence of excess capacity and monopoly in growing spatial markets’, Economica, pp. 149–58. ——(1982). ‘An economic theory of central places’, Economic Journal, pp. 56–72. ——(1997). ‘Introduction: beyond neoclassical competitive economics’ in C. Eaton and R. Lipsey (eds) On the Foundations of Monopolistic Competition and Economic Geography. Cheltenham: Edward Elgar, pp. ix–xxvii. Ellison, G. and E. Glaeser (1997). ‘Geographic concentration in US manufacturing industries: a dartboard approach’, Journal of Political Economy, pp. 889–927. Emerson, M., M. Auejan, M. Catinat, P. Goybet and A. Jacquemin (1988). ‘The economics of 1992’, European Economy, March. Enright, M. (1998). ‘Regional clusters and firm strategy’, in A. Chandler, P. Hagström and O. Sölvell (eds) The Dynamic Firm. Oxford: Oxford University Press, pp. 315–42. Enright, M. and R. Weder (eds) (1995). Studies in Swiss Competitive Advantage. Bern: Peter Lang. European Commission (1997). Regional Growth and Convergence. London: Kogan Page. ——(1998). Foreign Direct Investment. London: Kogan Page. Feinberg, S. and M. Keane (2005). ‘Intrafirm trade of US MNCs: findings and implications for models and policies toward trade and investment’, in T. Moran, E. Graham and M. Blomström (eds) Does Foreign Direct Investment Promote Development? Washington: Institute for International Economics, pp. 245–71. Feldman, M., J. Francis and J. Bercovitz (2005). ‘Creating a cluster while building a firm: entrepreneurs and the formation of industrial clusters’, Regional Studies,pp. 129–41. Finger, M. (1975). ‘A new view of the product cycle theory’, Weltwirtschaftliches Archiv, pp.79–99. Finger, J. and A. Yeats (1976). ‘Effective protection by transportation costs and tariffs: a comparison of magnitudes’, Quarterly Journal of Economics, pp. 169–76. Florida, R. (2002). ‘The economic geography of talent’, Annals of the Association of American Geographers, pp. 743–55. Forslid, R., J. Haaland and H. Midelfart-Knarvik (2002). ‘A U-shaped Europe? A simulation study of industrial location’, Journal of International Economics, pp.273–97. Foster, J. (2000). ‘Competitive selection, self-organisation and Joseph A. Schumpeter’, Journal of Evolutionary Economics, pp. 311–28. ——(2004). ‘Why is economics not a complex science?’, Discussion Paper No. 336, School of Economics, University of Queensland. Fratianni, M. (2007). ‘The evolutionary chain of international financial centres’, Kelly School of Business, Indiana University, Bloomington, mimeo. Freeman, C. (1994). ‘The economics of technical change’, Cambridge Journal of Economics, pp. 463–514. Frenken, K., and R. Boschma (2007). ‘A theoretical framework for evolutionary economic geography: industrial dynamics and urban growth as a branching process’, Journal of Economic Geography, pp. 635–49. Friedman, T. (2006). The World is Flat. New York: Farrar, Straus and Giroux. Fujita, M. and R. Ishii (1998). ‘Global location behavior and organizational dynamics of Japanese electronics firms and their impact on regional economies’,in A. Chandler, P. Hagström and O. Sölvell (eds) The Dynamic Firm. Oxford: Oxford University Press, pp. 343–83. Fujita, M. and T. Mori (1996). ‘The role of ports in the making of major cities: selfagglomeration and hub-effect’, Journal of Development Economics,pp.93–120. Fujita, M. and J. Thisse (1996). ‘Economics of agglomeration’, Journal of the Japanese and International Economies, pp. 339–78. Fujita, M. and J. Thisse (2003). ‘Does geographical agglomeration foster economic growth? And who gains and loses from it?’ Japanese Economic Review, pp. 121–45. Fujita, M., P. Krugman and A. Venables (1999). The Spatial Economy. Cambridge, MA: MIT Press.

Bibliography

394

Garofoli, G. (1991a). ‘Industrial districts: structure and transformation’, in G. Garofoli (ed.) Endogenous Development and Southern Europe. Aldershot: Avebury, pp. 49–60. ——(1991b). ‘Endogenous development and southern Europe: an introduction’, in G. Garofoli (ed.) Endogenous Development and Southern Europe. Aldershot: Avebury, pp. 1–13. ——(2002). ‘Local development in Europe: theoretical models and international comparisons’, European Urban and Regional Studies, pp. 225–39. Geroski, P. (1988). ‘Competition and innovation’, in Commission of the European Communities, Studies on the Economics of Integration. Brussels: European Community, pp. 339–88. Geroski, P. and A. Jacquemin (1985). ‘Industrial change, barriers to mobility, and European industrial policy’, Economic Policy, pp. 170–218. Girma, S. (2005). ‘Absorptive capacity and productivity spillovers from FDI: a threshold regression analysis’, Oxford Bulletin of Economics and Statistics, pp. 281–306. Gittelman, M., E. Graham and H. Fukukawa (1992). ‘Affiliates of Japanese firms in the European Community: performance and structure’, paper presented at the conference ‘Japanese direct investment in a unifying Europe’, INSEAD, Fontainebleau, 26–27 June, mimeo. Giuliani, E. (2005). ‘Cluster absorptive capacity: why do some clusters forge ahead and others lag behind?’, European Urban and Regional Studies, pp. 269–88. Gleiser, H., A. Jacquemin and J. Petit (1980). ‘Exports in an imperfect competition framework: an analysis of 1,446 exporters’, Quarterly Journal of Economics, pp. 507–24. Gordon, I. and P. McCann (2000). ‘Industrial clusters: complexes, agglomeration and/or social networks?’, Urban Studies, pp. 513–32. Görg, H. and D. Greenaway (2004). ‘Much ado about nothing? Do domestic firms really benefit from foreign direct investment?’, World Bank Research Observer, pp. 171–97. Görg, H. and E. Strobl (2001). ‘Multinational companies and productivity spillovers: a meta analysis’, Economic Journal, pp. F723–F739. ——(2003). ‘Multinational corporations, technology spillovers and plant survival’, Scandinavian Journal of Economics, pp. 581–95. ——(2005). ‘Spillovers from foreign firms through worker mobility: an empirical investigation’, Scandinavian Journal of Economics, pp. 693–709. 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. London: Routledge, pp. 255–77. Graham, E. (1978). ‘Transatlantic investment by multinational firms: a rivalistic phenomenon?’, Journal of Post Keynesian Economics, pp. 82–99. Graham, E. and P. Krugman (1995). Foreign Direct Investment in the United States. Washington, DC: Institute for International Economics. Greenaway, D. and C. Milner (1987). ‘Intra-industry trade: current perspectives and unresolved issues’, Weltwirtschaftliches Archiv, pp. 39–57. ——(2002). ‘Regionalism and gravity’, Scottish Journal of Political Economy, pp. 574–85. Greenaway, D. and J. Torstensson (1997). ‘Back to the future: taking stock on intraindustry trade’, Weltwirtschaftliches Archiv, pp. 248–69. ——(2000). ‘Economic geography, comparative advantage and trade within industries: evidence from the OECD’, Journal of Economic Integration, pp. 260–80. Grossman, G. (1998). ‘Comment’, in J. Frenkel (ed.) The Regionalization of the World Economy. Chicago: Chicago University Press, pp. 29–31. Grubel, H. and P. Lloyd (1975). Intra Industry Trade. London: Macmillan. Grubert, H. and J. Mutti (1991). ‘Taxes, tariffs and transfer pricing in multinational corporate decision making’, Review of Economics and Statistics, pp. 285–93. ——(2000). ‘Do taxes influence where U.S. corporations invest? National Tax Journal, pp. 825–39. Hagedoorn, J. (1998). ‘Atlantic strategic technology alliances’, in G. Boyd (ed.) The Struggle for World Markets. Cheltenham, UK and Lyme, USA: Edward Elgar, pp. 177–91.

Bibliography

395

Hagedoorn, J. and J. Schakenraad (1993). ‘A comparison of private and subsidized R&D partnerships in the European information technology industry’, Journal of Common Market Studies, pp. 373–90. Hague, D. (1960). ‘Report on the proceedings: summary record of the debate’, in E. Robinson (ed.) Economic Consequences of the Size of Nations. London: Macmillan, pp. 333–438. Hajkova, D., G. Nicoletti, L. Varia and K. Yoo (2006). ‘Taxation, business environment and FDI location in OECD countries’, OECD Economics Department Working Papers No. 502, Paris: OECD. Hanson, G. (1998). ‘North American economic integration and industry location’, Oxford Review of Economic Policy, pp. 30–44. ——(2001). ‘Should countries promote foreign direct investment?’ G-24 Discussion Paper, New York: UNCTAD. Harris, C. (1954). ‘The market as a factor in the localization of industry in the United States’, Annals of the Association of American Geographers, pp. 315–48. Harris, D. (1993). ‘The impact of US tax law revision on multinational corporation’s capital location and income shifting decisions’, Journal of Accounting Research, pp. 111–39. Hayek, F. (1978). New Studies in Philosophy, Politics, Economics and the History of Ideas. London: Routledge and Kegan Paul. Henderson, J. (1974). ‘The sizes and types of cities’, American Economic Review, pp. 640–56. Henderson, J., Z. Shalizi and A. Venables (2001). ‘Geography and development’, Journal of Economic Geography, pp. 81–106. Herin, J. (1986). ‘Rules of origin and differences between tariff levels in EFTA and in the EC’, Occasional Paper No. 16, EFTA, Geneva. Hillman, D. and D. Gibbs (1998). Century Makers. London: Weidenfeld and Nicolson. Hiratsuka, D. (2006). ‘Vertical intra-regional production networks in East Asia: a case study of the hard disc drive industry’, in D. Hiratsuka (ed.) East Asia’sde Facto Economic Integration. Houndmills: Palgrave Macmillan, pp. 181–99. Hoare, A. (1986). ‘British ports and their export hinterlands: a rapidly changing geography’, Geografiska Annaler B, pp. 29–40. Holmes, T. (1998). ‘The effects of state policies on the location of manufacturing: evidence from state borders’, Journal of Political Economy, pp. 667–705. Hospers, G. (2004). ‘Restructuring Europe’s rustbelt’, Intereconomics, pp. 147–56. Hotelling, H. (1929). ‘Stability in competition’, Economic Journal, pp. 41–57. Hufbauer, G., and J. Schott (2005). NAFTA Revisited. Washington: Institute for International Economics. Hummels, D. (1999). ‘Have international transportation costs declined?’, University of Chicago, Graduate School of Business, mimeo. http://www.mgmt.purdue.edu/faculty/hummelsd/research/decline/declined.pdf. ——(2006). ‘Transportation costs and trade over time’, in OECD, Transport and International Trade. OECD: Paris, pp. 7–26. ——(2007). ‘Transport costs and international trade in the second era of globalization’, Journal of Economic Perspectives, pp. 131–54. Hymer, S. (1976). The International Operations of National Firms: A Study of Direct Foreign Investment. Boston, MA: MIT Press. Ietto-Gillies, G. (2005). Transnational Corporations and International Production. Cheltenham: Edward Elgar. Isard, W. (1954). ‘Location theory and trade theory: short run analysis’, Quarterly Journal of Economics, pp. 305–20. ——(1956). Location and Space-Economy. Cambridge, MA: Technology Press of MIT and London: Chapman and Hall.

Bibliography

396

——(1977). ‘Location theory, agglomeration and the pattern of world trade’, i nB. Ohlin, P. Hesselborn and P. Wijkman (eds) The International Allocation of Economic Activity. London: Macmillan, pp. 159–77. ——(1996). Commonalities in Art, Science and Religion: An Evolutionary Perspective. Aldershot: Avebury. Isard, W. and M. Peck (1954). ‘Location theory and international and interregional trade theory’, Quarterly Journal of Economics, pp. 97–114. Jacobs, J. (1969). The Economy of Cities. New York: Random House. Jacquemin, A. (1990a). ‘Mergers and European policy’, in P. Admiraal (ed.) Merger and Competition Policy in the European Community. Oxford: Basil Blackwell, pp. 1–38. ——(1990b). ‘Horizontal concentration and European merger policy’, European Economic Review, pp. 539–50. ——(1991). ‘Collusive behaviour, R&D and European competition policy’, in G. Faulhaber and G. Tamburini (eds) European Economic Integration. Boston, MA: Kluwer, pp. 201–35. Jacquemin, A. and J. Marchipont (1992). ‘De nouveaux enjeux pour la politique industrielle de la Communauté’ (New challenges for the industrial policy of the Community), Revue d’Économie politique, pp. 69–97. Jacquemin, A. and L. Pench (1997). Europe Competing in the Global Economy. Cheltenham, UK and Lyme, USA: Edward Elgar. Jacquemin, A. and A. Sapir (1991). ‘The internal and external opening-up of the Single Community Market: efficiency gains, adjustment costs and new Community instruments’, International Spectator, pp. 29–48. Jacquemin, A. and D. Wright (1993). ‘Corporate strategies and European challenges post-1992’, Journal of Common Market Studies, pp. 525–37. Jacquemin, A., P. Lloyd, P. Tharakan and J. Waelbroeck (1998). ‘Competition policy in an international setting: the way ahead’, World Economy, pp. 1179–83. Johnson, H. and M. Krauss (1973). ‘Border taxes, border tax adjustment, comparative advantage and the balance of payments’, in M. Krauss (ed.) Economics of Integration. London: George Allen and Unwin, pp. 239–53. Jovanović, M. (1995). ‘Economic integration among developing countries and foreign direct investment’, Economia Internazionale, pp. 209–43. ——(1997). European Economic Integration. London: Routledge. ——(1998a). International Economic Integration: Limits and Prospects. London: Routledge. ——(ed.) (1998b). International Economic Integration: Critical Perspectives on the World Economy – Theory and Measurement (Volume I). London: Routledge. ——(ed.) (1998c). International Economic Integration: Critical Perspectives on the World Economy – Monetary, Fiscal and Factor Mobility Issues (Volume II). London: Routledge. ——(ed.) (1998d). International Economic Integration: Critical Perspectives on the World Economy – General Issues (Volume III). London: Routledge. ——(ed.) (1998e). International Economic Integration: Critical Perspectives on the World Economy – Integration Schemes (Volume IV). London: Routledge. ——(2000a). ‘Eastern enlargement of the European Union: sour grapes or sweet lemon?’, Economia Internazionale, pp. 507–36; http://papers.ssrn.com/sol3/papers.cfm?abstract_id=261059. ——(2000b). ‘Economic integration and location of industries’, in B. Schefold (ed.) Economic Interests and Cultural Determinants in European Integration. Bolzano: European Academy, pp. 169–204. ——(2001). Geography of Production and Economic Integration. London: Routledge. ——(2003a). ‘Spatial location of firms and industries: an overview of theory’, Economia Internazionale, pp .23–81; http://papers.ssrn.com/sol3/papers.cfm?abstract_id=451800. ——(2003b). ‘Local vs global location of firms and industries’, Journal of Economic Integration,pp.60–104; http://papers.ssrn.com/sol3/papers.cfm?abstract_id=394760.

Bibliography

397

——(2004a). ‘Eastern enlargement of the EU: a topsy-turvy endgame or permanent disillusionment?’ Journal of Economic Integration, pp. 830–68; http://papers.ssrn.com/sol3/papers.cfm?abstract_id=650165. ——(2004b). ‘Economic integration and spatial location of production’, Department of Economics and Finance Working Paper No. 12/2004, University of Genoa; http://papers.ssrn.com/sol3/papers.cfm?abstract_id=647382. ——(2005a). The Economics of European Integration: Limits and Prospects. Cheltenham, UK, and Northampton, MA, USA: Edward Elgar. ——(2005b). ‘Turkey in the European Union: euthanasia or the rejuvenation of Europe?’ paper presented at the conference ‘Turkey, EU and the US: high stakes, uncertain prospects’, Intercollege, Nicosia, 13–14 May. ——(2005c). ‘Spatial location of firms and industries: is local superior to global?’, in L. Yee (ed.) Contemporary Issues in Urban and Regional Economics. New York: Nova Scientific Publishers, pp. 1–53. ——(2006). The Economics of International Integration. Cheltenham, UK, and Northampton, MA, USA: Edward Elgar. ——(ed.) (2007a). Economic Integration and Spatial Location of Firms and Industries, Conceptual Issues. Cheltenham, UK, and Northampton, MA, USA: Edward Elgar. ——(ed.) (2007b). Economic Integration and Spatial Location of Firms and Industries, Agglomeration and Clusters. Cheltenham, UK, and Northampton, MA, USA: Edward Elgar. ——(ed.) (2007c). Economic Integration and Spatial Location of Firms and Industries, Transnational Corporations and Search for Evidence. Cheltenham, UK, and Northampton, MA, USA: Edward Elgar. Kalajicć, D. (1979). Smak sveta (Doomsday). Zagreb: Nakladni zavod Matice hrvatske. Kay, J. (2004). Culture and Prosperity. New York: HarperCollins. ——(2006). The Hare and the Tortoise. An Informal Guide to Business Strategy. London: Erasmus Press. Keegan, J. (1993). A History of Warfare. London: Hutchinson. Kim, S. (1995). ‘Expansion of markets and the geographic distribution of economic activities: the trends in US regional manufacturing structure, 1860–1987’, Quarterly Journal of Economics, pp. 881–908. Kim, S. and R. Margo (2004). ‘Historical perspectives on US economic geography’, in J. Henderson and J. Thisse (eds) Handbook of Regional and Urban Economics. Amsterdam: Elsevier, pp. 2981–3019. Kleinknecht, A. and J. ter Wengel (1998). ‘The myth of economic globalisation’, Cambridge Journal of Economics, pp. 637–47. Kneller, R. and M. Pisu (2004). ‘Export-oriented FDI in the UK’, Oxford Review of Economic Policy, pp. 424–39. Knickerboker, F. (1973). Oligopolistic Reaction and Multinational Enterprise. Boston, MA: Harvard University Press. Kogut, B. (2004). ‘From regions and firms to multinational highways’, in M. Kenney and R. Florida (eds) Locating Global Advantage. Stanford: Stanford University Press, pp. 261–81. Kravis, I. and R. Lipsey (1982). ‘The location of overseas production and production for export by U.S. multinational firms’, Journal of International Economics, pp. 201–23. Krugman, P. (1979). ‘Increasing returns, monopolistic competition and international trade’, Journal of International Economics, pp. 469–79. ——(1980). ‘Scale economies, product differentiation and pattern of trade’, American Economic Review, pp. 950–9. ——(1987). ‘The narrow moving band, the Dutch disease, and the competitive consequences of Mrs. Thatcher’, Journal of Development Economics, pp .41–55. ——(1990a). Rethinking International Trade. Cambridge, MA: MIT Press. ——(1990b). ‘Protectionism: try it, you’ll like it’, International Economy, June/ July, pp. 35–9.

Bibliography

398

——(1991a). ‘History versus expectations’, Quarterly Journal of Economics, pp. 651–67. ——(1991b). ‘Increasing returns and economic geography’, Journal of Political Economy, pp. 483–9. ——(1991c). ‘Is bilateralism bad?’ in E. Helpman and A. Razin (eds) International Trade and Trade Policy. Cambridge, MA: MIT Press, pp. 9–23. ——(1992). Geography and Trade. Cambridge, MA: MIT Press. ——(1993a). ‘On the number and location of cities’, European Economic Review, pp. 293–8. ——(1993b). ‘The current case for industrial policy’, in D. Salvatore (ed.) Protectionism and World Welfare. Cambridge: Cambridge University Press, pp. 160–79. ——(1993c). ‘First nature, second nature, and metropolitan location’, Journal of Regional Science, pp. 129–44. ——(1993d). ‘Lessons of Massachusetts for EMU’, in F. Torres and F. Giavazzi (eds) Adjustment and Growth in the European Monetary Union. Cambridge: Cambridge University Press, pp. 241–61. ——(1995a). ‘A reply to Professor Dunning’, International Executive, pp. 325–7. ——(1995b). ‘Themovetowardfreetradezones’, in P. King (ed.) International Economics and International Economic Policy: A Reader. New York: McGraw-Hill, pp. 163–82. ——(1996a). Pop Internationalism. Cambridge, MA: MIT Press. ——(1996b). ‘What economists can learn from evolutionary theorists’, talk given to the European Association for Evolutionary Political Economy, November; http://web.mit.edu/krugman/www/. ——(1996c). ‘The Adam Smith Address: what difference does globalization make?’, Business Economics, pp .7–10. ——(1996d). ‘Urban concentration: the role of increasing returns and transport costs’, International Regional Science Review, pp. 5–30. ——(1997). ‘What should trade negotiators negotiate about?’, Journal of Economic Literature, pp. 113–20. ——(1998a). ‘What’s new about the new economic geography’, Oxford Review of Economic Policy, pp. 7–17. ——(1998b). The Accidental Theorist and other Dispatches from the Dismal Science. New York: Norton. ——(1998c). ‘Space: the final frontier’, Journal of Economic Perspectives, pp. 161–74. ——(1999). ‘The role of geography in development’, in B. Pleskovic and J. Stiglitz (eds) Annual World Bank Conference on Development Economics 1998. Washington, DC: World Bank, pp. 89–107. ——(2000). ‘Can America stay on top?’, Journal of Economic Perspectives, pp. 169–75. ——(2003). The Great Unraveling. New York: Norton. ——(2007). ‘The “new” economic geography: where are we?’, in M. Fujita (ed.) Regional Integration in East Asia. Houndmills: Palgrave Macmillan, pp. 23-34. Krugman, P. and A. Venables (1990). ‘Integration and the competitiveness of peripheral industry’, in C. Bliss and J. Braga de Macedo (eds) Unity with Diversity in the European Economy: The Community’s Southern Frontier. Cambridge: Cambridge University Press, pp. 56–75. ——(1995). ‘Globalization and the inequality of nations’, Quarterly Journal of Economics, pp. 857–80. ——(1996). ‘Integration, specialization, and adjustment’, European Economic Review, pp. 959–67. Lambooy, J. (1997). ‘Knowledge production, organisation and agglomeration economies’, GeoJournal, pp. 293–300. Lambooy, J. and R. Boschma (2001). ‘Evolutionary economics and regional policy’, Annals of Regional Science, pp. 113–31. Lancaster, K. (1980). ‘Intra-industry trade under perfect monopolistic competition’, Journal of International Economics, pp. 151–75. Leamer, E. (1984). Sources of International Comparative Advantage. Cambridge, MA: MIT Press.

Bibliography

399

——(2007). ‘A flat world, a level playing field, a small world after all, or none of the above? A review of Thomas L. Friedman’s The World is Flat’, Journal of Economic Literature, pp.83–126. Leamer, E. and J. Levinsohn (1995). ‘International trade theory: the evidence’, in G. Grossman and K. Rogoff (eds) Handbook of International Economics. Amsterdam: Elsevier, pp. 1339–94. Lederman, D., W. Maloney and L. Servén (2005). Lessons from NAFTA for Latin America and the Caribbean. Washington: World Bank. Lee, Y. (2007). Geographic redistribution of the U.S. manufacturing and the role of state development policy’, Center of Economic Studies, Discussion Paper CES 07–06, CES, Guildford, UK.

Leijonhufvud, A. (1996). ‘Towards a not-too-rational macroeconomics’, in D. Colander (ed.) Beyond Microfoundations: Post Walrasian Macroeconomics. Cambridge: Cambridge University Press, PP. 39–55.

Leonardi, R. (2006). ‘The impact and value added of Cohesion Policy’, Regional Studies, pp. 155–66. Levin, R., A. Klevorick, R. Nelson and S. Winter (1987). ‘Appropriating the returns from industrial research and development’, Brookings Papers on Economic Activity, pp. 783–820. Lewis, B. (1953). ‘Some observations on the significance of heresy in the history of Islam’, Studia Islamica, pp. 43–63. Liebowitz, S. and S. Margolis (1999). Winners, Losers and Microsoft – Competition and Antitrust in High Technology. Oakland: Independent Institute. ——(2001). Winners, Losers and Microsoft. Oakland: Independent Institute. Lin, Y. (1935). My Country and My People. New York: John Day and Regnal and Hitchcock. Linder, S. (1961). An Essay on Trade and Transformation. Uppsala: Almquist and Wiksell. Lipsey, R.E. (2004). ‘Home- and host-country effects of foreign direct investment’, in R. Baldwin and A. Winters (eds) Challenges to Globalisation – Analyzing the Economics. Chicago: Chicago University Press, pp. 333–79. Lipsey, R.E. and F. Sjöholm (2005). ‘The impact of inward FDI on host countries: why such different answers?’, in T. Moran, E. Graham and M. Blomström (eds) Does Foreign Direct Investment Promote Development? Washington: Institute for International Economics, pp. 23–43. Lipsey, R.G. (1957). ‘The theory of customs unions: trade diversion and welfare’, Economica, pp. 40–6. ——(1960). ‘The theory of customs unions: a general survey’, Economic Journal, pp. 496–513. ——(1976). ‘Comments’, in F. Machlup (ed.) Economic Integration Worldwide, Regional, Sectoral. London: Macmillan, pp. 37–40. ——(1985). ‘Canada and the United States: the economic dimension’, in C. Doran and J. Stigler (eds) Canada and the United States: Enduring Friendship, Persistent Stress. New York: Prentice-Hall, pp. 69–108. ——(1987a). ‘Models matter when discussing competitiveness: a technical note’, in R. Lipsey and W. Dobson (eds) Shaping Comparative Advantage. Toronto: C. D. Howe Institute, pp. 155–66. ——(1987b). ‘Report on the workshop’, in R.G. Lipsey and W. Dobson (eds) Shaping Comparative Advantage. Toronto: C. D. Howe Institute, pp. 109–53. ——(1990). ‘Canada at the US–Mexico free trade dance: wallflower or partner?’, Commentary 20, C. D. Howe Institute, Toronto. ——(1991). ‘The case for trilateralism’, in S. Globerman (ed.) Continental Accord: North American Economic Integration. Vancouver: Fraser Institute, pp. 89–123. ——(1992a). An Introduction to Positive Economics. London: Weidenfeld and Nicolson. ——(1992b). ‘Global change and economic policy’, in N. Stehr and R. Ericson (eds) The Culture and the Power of Knowledge. New York: De Gruyter, pp. 279–99.

Bibliography

400

——(1993a). ‘Globalisation, technological change and economic growth’, Annual Sir Charles Carter Lecture. Northern Ireland Economic Development Office, Belfast. ——(1993b). ‘The changing technoeconomic paradigm and some implications for economic policy’, Canadian Institute for Advanced Research, Vancouver, mimeo. ——(1993c). ‘Canadian trade policy in relation to regional free trade agreements: CAFTA, NAFTA and WHFTA’, Canadian Institute for Advanced Research, Vancouver, mimeo. ——(1994). ‘Markets, technological change and economic growth’, Canadian Institute for Advanced Research, Vancouver, mimeo. ——(1997). ‘Globalization and national government policies: an economist’s view’, in J. Dunning (ed.) Governments, Globalization and International Business. Oxford: Oxford University Press, pp. 73–113. ——(2004). ‘Some plusses and minuses of globalisation’, Simon Fraser University, Department of Economics, mimeo. Lipsey, R.G. and K. Lancaster (1956–7). ‘The general theory of the second best’, Review of Economic Studies, pp .11–32. Lipsey, R.G. and M. Smith (1986). Taking the Initiative: Canada’s Trade Options in a Turbulent World. Toronto: C. D. Howe Institute. ——(1987). Global Imbalances and US Policy Responses. Toronto: C. D. Howe Institute. ——(1989). ‘The Canada–US free trade agreement: special case or wave of the future?’, in J. Schott (ed.) Free Trade Areas and US Trade Policy. Washington: Institute for International Economics, pp. 317–35. Lipsey, R.G. and R. York (1988). Evaluating the Free Trade Deal: A Guided Tour through the Canada–US Agreement. Toronto: C. D. Howe Institute. Lipsey, R.G., K. Carlaw and C. Bekar (2005). Economic Transformations: General Purpose Technologies and Long Term Economic Growth. Oxford: Oxford University Press. Loertscher, R. and F. Wolter (1980). ‘Determinants of intra-industry trade: among countries and across industries’, Weltwirtschaftliches Archiv, pp. 280–93. Lösch, A. (1938). ‘The nature of economic regions’, Southern Economic Journal, pp. 71–8. ——(1940). Die Räumliche Ordnung der Wirtschaft. Jena: Gustav Fischer. (Translated from German by W. Woglom and W. Stolper, The Economics of Location (1973), New Haven, CT: Yale University Press.) Lovering, J. (2006). ‘The new imperial geography’, in S. Bagchi-sen and H. Lawton Smith (eds) Economic Geography. London: Routledge, pp. 221–32. Lowe, N. and M. Kenney (1999). ‘Foreign investment and the global geography of production: why the Mexican consumer electronics industry failed’, World Development, pp. 1427–43. Lucas, R. (1988). ‘On the mechanics of economic development’, Journal of Monetary Economics, pp. 3–42. McAfee, P., H. Mialon and M. Williams (2004). ‘When are sunk costs barriers to entry? Entry barriers in economic and antitrust analysis’, American Economic Review, pp. 461–70. McGrew, A. (1992). ‘Conceptualizing global politics’, in A. McGrew and P. Lewis (eds) Global Politics. Cambridge: Polity Press, pp. 1–28. Machlup, F. (1946). ‘Marginal analysis and empirical research’, American Economic Review, pp. 519–54. Margalioth, Y. (2003). ‘Tax competition, foreign direct investments and growth: using the tax system to promote developing countries’, Virginia Tax Review, pp. 157–98. Markusen, A. (1996). ‘Sticky places in slippery space: a typology of industrial districts’, Economic Geography, pp. 293–313. Markusen, J. (1984). ‘Multinationals, multi-plant economies and the gains from trade’, Journal of International Economics, pp. 205–26. Markusen, J. and A. Venables (2000). ‘The theory of endowment, intra-industry and multi-national trade’, Journal of International Economics, pp. 209–34. Marshall, A. (1890). Principles of Economics. London: Macmillan.

Bibliography

401

——(1961). Principles of Economics. London: Macmillan. Martin, R. (1999). ‘The new geographical turn in economics: some critical reflections’, Cambridge Journal of Economics, pp .65–91. ——(2006). ‘Economic geography and the new discourse of regional competitiveness’, in S. Bagchi-sen and H. Lawton Smith (eds) Economic Geography. London: Routledge, pp. 159–72. Martin, R. and P. Sunley (2003). ‘Deconstructing clusters: chaotic concept or policy panacea?’, Journal of Economic Geography, pp. 5–35. ——(2006). ‘Path dependence and regional economic evolution’, Journal of Economic Geography, pp. 395–437. ——(2007). ‘Complexity thinking and evolutionary economic geography’, Journal of Economic Geography, pp. 573–601. Martin, R., and P. Tyler (2006). ‘Evaluation the impact of the structural funds on objective 1 regions: an exploratory discussion’, Regional Studies, pp. 201–10. Maskell, P. (2001). ‘Towards a knowledge-based theory of the geographical cluster’, Industrial and Corporate Change, pp. 921–43. Melitz, J. (2004). ‘Risk-sharing and EMU’, Journal of Common Market Studies, pp. 815–40. Midelfart, H., H. Overman and A. Venables (2003). ‘Monetary union and the economic geography of Europe’, Journal of Common Market Studies, pp. 847–68. Midelfart-Knarvik, K. and H. Overman (2002). ‘Delocation and European integration’, Economic Policy, pp. 323–59. Midelfart-Knarvik, K., H. Overman, S. Redding and A. Venables (2002). ‘The location of European industry’, European Economy, Special report No. 2/2002, pp. 213–69. Milanovic, B. (2003). ‘The two faces of globalization: against globalization as we know it’, World Development, pp. 667–83. Minford, P., V. Mahambare and E. Nowell (2005). Should Britain Leave the EU? Cheltenham: Edward Elgar. Mody, A. (2004). ‘Is FDI integrating the world economy?’, World Economy, pp. 1195–222. Mody, A. and K. Srinivasan (1998). ‘Japanese and U.S. firms as foreign investors: do they march to the same tune?’, Canadian Journal of Economics, pp. 778–99. Mokyr, J. (2006). ‘Economics and the biologists: a review of Geerat J. Vermeij’s Nature: An Economic History’, Journal of Economic Literature, pp. 1005–13. Mora, T., E. Vaya and J. Surinach (2004). ‘The enlargement of the European Union and the spatial distribution of economic activity’, Eastern European Economics, pp. 6–35. Moran, T. (1998). Foreign Direct Investment and Development. Washington, DC: Institute for International Economics. ——(2005). ‘How does FDI affect host country development? Using industry case studies to make reliable generalisations’, in T. Moran, E. Graham and M. Blomström (eds) Does Foreign Direct Investment Promote Development? Washington: Institute for International Economics, pp. 281–313. Moran, T., E. Graham and M. Blomström (2005a). ‘Conclusions and implications for FDI policy in developing countries, new methods of research and a future research agenda’, in T. Moran, E. Graham and M. Blomström (eds) Does Foreign Direct Investment Promote Development? Washington: Institute for International Economics, pp. 375–95. ——(eds) (2005b). Does Foreign Direct Investment Promote Development? Washington: Institute for International Economics. Moses, L. (1958). ‘Location and the theory of production’, Quarterly Journal of Economics, pp. 259–72. Moss Kanter, R. (2003). ‘Thriving locally in the global economy’, Harvard Business Review, pp. 119–28.

Bibliography

402

Mucchielli, J. and F. Puech (2004). ‘Globalization, agglomeration and FDI location: the case of French firms in Europe’, in J. Mucchielli and T. Mayer (eds) Multinational Firms’ Location and the New Economic Geography. Cheltenham: Edward Elgar, pp. 35–58. Murphy. K., A. Shleifer and R. Vishny (1989). ‘Industrialization and the Big Push’, Journal of Political Economy, pp. 1003–26. Nadvi, K. (1998). ‘International competitiveness and small firm clusters – evidence from Pakistan’, Small Enterprise Development, pp. 12–24. Nayyar, D. (2006). ‘Globalisation, history and development: a tale of two centuries’, Cambridge Journal of Economics, pp. 137–59. Nelson, R. (1995). ‘Recent evolutionary theorizing about economic change’, Journal of Economic Literature, pp. 48–90. ——(1999). ‘The sources of industrial leadership: a perspective on industrial policy’, De Economist, pp. 1–18. Nelson, R. and S. Winter (1982). An Evolutionary Theory of Economic Change. Cambridge: Harvard University Press. ——(2002). ‘Evolutionary Theorizing in Economics’, Journal of Economic Perspectives, pp. 23–46. North, D. (1955). ‘Location theory and regional economic growth’, Journal of Political Economy, pp. 243–58. OECD (2005a). Business Clusters. Paris: OECD. ——(2005b). Preferential Trading Arrangements in Agricultural and Food Markets. Paris: OECD. ——(2007). OECD Science, Technology and Industry Scoreboard. Paris: OECD. Ohlin, B. (1933). Interregional and International Trade. Cambridge, MA: Harvard University Press. Pack, H. (2000). ‘Industrial policy: growth elixir or poison?’ World Bank Research Observer, pp. 47–67. Pack, H. and K. Saggi (1997). ‘Inflows of foreign technology and indigenous technological development’, Review of Development economics, pp. 81–98. Paelinck, J. and M. Polèse (1999). ‘Modelling the regional impact of continental Economic Integration: lessons from the European Union for NAFTA’, Regional Studies, pp. 727–38. Palmeter, D. (1993a). ‘Pacific regional trade liberalization and rules of origin’, Journal of World Trade, pp. 49–62. ——(1993b). ‘Rules of origin in customs unions and free trade areas’, in K. Anderson and R. Blackhurst (eds) Regional Integration and the Global Trading System. New York: Harvester Wheatsheaf, pp. 326–43. Papageorgiou, G. (1979). ‘Agglomeration’, Regional Science and Urban Economics, pp. 41–59. Pautler, P. (2001). ‘Evidence on mergers and acquisitions’, Bureau of Economics, Federal Trade Commission, Washington, DC, mimeo. Penner-Hahn, J. and J. Myles Shaver (2005). ‘Does international research and development increase patent output? An analysis of Japanese pharmaceutical firms’, Strategic Management Journal, pp. 121–40. Penrose, E. (1952). ‘Biological analogies in the theory of the firm’, American Economic Review, pp. 804–19. Perroux, F. (1950). ‘Economic space: theory and applications’, Quarterly Journal of Economics, pp. 89–104. ——(1955). ‘Note sur la notion de, < pôle de croissance >’ (A note on the notion of a pole of growth), Économie Appliqué, pp. 307–20. ——(1961). L’Économie du XXe siècle (Economics of the Twentieth Century). Paris: Presses Universitaires de France. Perry, M. (2006). Business Clusters. London: Routledge. Pfann, G. and H. van Kranenburg (2002). ‘Tax policy, location choices and market structure’, IZA Discussion Paper No. 499, Institute for the Study of Labour, Bonn.

Bibliography

403

Pinch, S. and N. Henry (1999). ‘Paul Krugman’s geographical economics, industrial clustering and the British motor sport industry’, Regional Studies, pp. 815–27. Plasschaert, S. (1994). ‘Introduction: transfer pricing and taxation’, in S. Plasschaert (ed.) Transnational Corporations: Transfer Pricing and Taxation. London: Routledge, pp. 1–21. Plato [360 BC] (1980). The Laws. Harmondsworth: Penguin Books. Porter, M. (1990a). The Competitive Advantage of Nations. New York: Free Press. ——(1990b). ‘The competitive advantage of nations’, Harvard Business Review, pp. 73–93. ——(1994). ‘The role of location in competition’, Journal of the Economics of Business, pp. 35–39. ——(1996). ‘Competitive advantage, agglomeration economies, and regional policy’, International Regional Science Review, pp. 85–94. ——(1998a). ‘Clusters and the new economics of competition’, Harvard Business Review, pp. 77–90. ——(1998b). ‘Location, clusters, and the ‘new’ microeconomics of competition’, Business Economics, pp. 7–13. ——(2000a). ‘Locations, clusters and company strategy’, in G. Clark, M. S. Gertler and M. P. Feldman (eds) The Oxford Handbook of Economic Geography. Oxford: Oxford University Press, pp. 253–74. ——(2000b). ‘Location, competition and economic development: local clusters in a global economy’, Economic Development Quarterly, pp. 15–34. Porter, M. and M. Sakakibara (2004). ‘Competition in Japan’, Journal of Economic Perspectives, pp. 27–50. Porter, M. and Ö. Sölvell (1998). ‘The role of geography in the process of innovation and the sustainable competitive advantage of firms’, in A. Chandler, P. Hagström and O. Sölvell (eds) The Dynamic Firm. Oxford: Oxford University Press, pp. 440–57. Posner, M. (1961). ‘International trade and technical change’, Oxford Economic Papers, pp. 323–41. Potts, J. (2001). ‘Knowledge and markets’, Journal of Evolutionary Economics, pp. 413–31. Prais, S. (1981). Productivity and Industrial Structure. Cambridge: Cambridge University Press. Pratten, C. (1971). Economies of Scale and Manufacturing Industry. Cambridge: Cambridge University Press. ——(1988). ‘A survey of the economies of scale’, in Studies on the Economics of Integration, Research on the Costs of Non-Europe, Vol. 2. Brussels: European Communities, pp. 11–165. President’s Commission on Industrial Competitiveness (1985). Global Competition: The New Reality (Volumes 1 and 2). Washington, DC: US Government Printing Office. Puga, D. (1999). ‘The rise and fall of regional inequalities’, European Economic Review, pp. 303–34. ——(2002). ‘European regional policies in light of recent location theories’, Journal of Economic Geography, pp. 373–406. Puga, D. and A. Venables (1997). ‘Preferential trading arrangements and industrial location’, Journal of International Economics, pp. 347–68. ——(1998). ‘Trading arrangements and industrial development’, World Bank Economic Review, pp. 221–49. ——(1999). ‘Agglomeration and economic development: import substitution vs trade liberalisation’, Economic Journal, pp. 292–311. Redding, S. and A. Venables (2004). ‘Economic geography and international inequality’, Journal of International Economics, pp. 53–82. Reuber, G. (1973). Private Foreign Investment in Development. Oxford: Clarendon Press. Robson, P. (1983). Integration, Development and Equity. London: George Allen and Unwin. ——(1987). The Economics of International Integration. London: George Allen and Unwin. Rodrik, D. (2004). ‘Industrial policy for the twenty-first century’, Faculty Research Working Paper, RWP 04–047, John F. Kennedy School of Government, Harvard University, Cambridge, MA.

Bibliography

404

Romer, P. (1986). ‘Increasing returns and long-run growth’, Journal of Political Economy, pp. 1002–37. ——(1994). ‘New goods, old theory and the welfare costs of trade restrictions’, Journal of Development Economics, pp. 5–38. Rosenthal, S. and W. Strange (2004). ‘Evidence on the nature and sources of agglomeration economies’ in V. Henderson and J. Thisse (eds) Handbook of Regional and Urban Economics. Amsterdam: Elsevier, pp. 2119–71. Rousslang, D. (1997). ‘International income shifting by US multinational corporations’, Applied Economics, pp. 925–34. Rubin, S. (1970). ‘The international firm and the national jurisdiction’, in C. Kindleberger (ed.) The International Corporation. Cambridge, MA: MIT Press, pp. 179–204. Rugman, A. (1985). ‘The behaviour of US subsidiaries in Canada: implications for trade and investment’, in D. Fretz, R. Stern and J. Whalley (eds) Canada/United States Trade and Investments Issues. Toronto: Ontario Economic Council, pp. 460–73. ——(2002). ‘Multinational enterprise and the end of global strategy’, in J. Dunning and J. Mucchielli (eds) Multinational Firms: The Global–Local Dilemma. London: Routledge, pp. 3–17. Rugman, A. and R. Hodgetts (2001). ‘The end of global strategy’, European Management Journal, pp. 333–43. Rugman, A. and A. Verbeke (2003). ‘Regional and global strategies of multinational enterprises’, Kelley School of Business, Indiana University, Bloomington, IL, mimeo. Sabrianova Peter, K., J. Svejnar and K. Terrell (2004). ‘Distance and the efficiency frontier and FDI spillovers’, IZA Discussion Paper No. 1332, Institute for the Study of Labour, Bonn. Sala-i-Martin, X. and J. Sachs (1992). ‘Fiscal federalism and optimum currency areas: evidence for Europe from the United States’, in M. Canzoneri, V. Grilli and P. Mason (eds) Establishing a Central Bank: Issues in Europe and Lessons from the US. Cambridge: Cambridge University Press, pp. 195–219. Sapir, A. (1992). ‘Regional integration in Europe’, Economic Journal, pp. 1491–506. Schiff, M. and A. Winters (2003). Regional Integration and Development. Washington: World Bank. Schmalensee, R. (1988). ‘Industrial economics: an overview’, Economic Journal, pp. 634–81. Schmitz, H. (1998). ‘Fostering collective efficiency’, Small Enterprise Development, pp. 4–11. ——(1999). ‘Collective efficiency and increasing returns’, Cambridge Journal of Economics, pp. 465–83. Schmitz, H. and K. Nadvi (1999). ‘Clustering and industrialization: introduction’, World Development, pp. 1503–14. Schumpeter, J. (1911). Theorie der wirtschaftlichen Entwicklung. Leipzig: Duncker & Humblot. (Translated from German as The Theory of Economic Development (1934). Cambridge: Harvard University Press.) ——(1939). Business Cycles. New York, McGraw-Hill. ——(1983). The Theory of Economic Development. New Brunswick: Transaction Books. ——(1987). Capitalism, Socialism and Democracy. London: Unwin. ——(2005). ‘Development’, Journal of Economic Literature, pp. 108–20. Shakespeare, W. (1998). Julius Caesar. Oxford: Oxford University Press. ——(2000). Hamlet. Harlow: Longman. Sharp, M. and K. Pavitt (1993). ‘Technology policy in the 1990s: old trends and new realities’, Journal of Common Market Studies, pp. 129–51. Smarzynska Javorcik, B. (2004). ‘Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages’, American Economic Review, pp. 605–27. Smarzynska Javorcik, B. and M. Spatareanu (2005). ‘Disentangling FDI spillover effects: what do firm perceptions tell us?’, in T. Moran, E. Graham and M. Blomström (eds) Does Foreign

Bibliography

405

Direct Investment Promote Development? Washington: Institute for International Economics, pp. 45–71. Smith, D. (1981). Industrial Location – An Economic Geographical Analysis. New York: John Wiley. Smith, D. and R. Florida (1994). ‘Agglomeration and industrial location: an econometric analysis of Japanese-affiliated manufacturing establishments in automotive-related industries’, Journal of Urban Economics, pp. 23–41. Stiglitz, J. (2002). Globalization and Its Discontents. London: Allen Lane. Sternberg, R. and C. Tamasy (1999). ‘Munich as Germany’s no. 1 high technology region: empirical evidence, theoretical explanations and the role of small firm/large firm relations’, Regional Studies, pp. 367–77. Storper, M. (1995). ‘The resurgence of regional economies, ten years later’, European Urban and Regional Studies, pp. 191–221. Storper, M., C. Yun-chung and F. de Paolis (2002). ‘Trade and location of industries in the OECD and European Union’, Journal of Economic Geography,pp.73–107. Sturgeon, T., and R. Florida (2004). ‘Globalization, deverticalization, and employment in the motor vehicle industry’, in M. Kenney and R. Florida (eds) Locating Global Advantage. Stanford: Stanford University Press, pp. 52–81. Tanzi, V. and H. Zee (2000). ‘Tax policy for emerging markets: developing countries’, National Tax Journal, pp. 299–322. Tappi, D. (2005). ‘Clusters, adaptation and extroversion’, European Urban and Regional Studies, pp. 289–307. Temple, J. (2002). ‘The assessment: the new economy’, Oxford Review of Economic Policy, pp. 241–64. Tharakan, M., D. Greenaway and J. Tharakan (1998). ‘Cumulation and injury determination of the European Community in antidumping cases’, Weltwirtschaftliches Archiv, pp. 320–39. Trimbath, S. (2002). Mergers and Acquisitions: Changes across Time. Boston, MA: Kluwer. Tyson, L. (1987). ‘Comments on Brander’s “Shaping comparative advantage”: creating advantage, an industrial policy perspective’, in R. Lipsey and W. Dobson (eds) Shaping Comparative Advantage. Toronto: C. D. Howe Institute, pp. 65–82. ——(1992). Who Is Bashing Whom? Trade Conflict in High-technology Industries. Washington, DC: Institute for International Economics. UNCTAD (1992). World Investment Report: Transnational Corporations as Engines of Growth. New York: United Nations. ——(1993). World Investment Report: Transnational Corporations and Integrated International Production. New York: United Nations. ——(1995a). ‘Incentives and foreign direct investment’, TD/B/ITNC/Misc. 1, 6 April. ——(1995b). World Investment Report: Transnational Corporations and Competitiveness. New York: United Nations. ——(1997). World Investment Report: Transnational Corporations, Market Structure and Competition Policy. United Nations: New York. ——(1998). World Investment Report: Trends and Determinants. United Nations: New York. ——(1999a). World Investment Report: Foreign Direct Investment and the Challenge of Development. United Nations: New York. ——(1999b). Investment-Related Trade Measures. New York: United Nations. ——(2000a). World Investment Report: Cross-border Mergers and Acquisitions and Development. United Nations: New York. ——(2000b). ‘Tax incentives and foreign direct investment’, ASIT Advisory Studies No. 16, UNCTAD/ITE/IPC/Misc.3. ——(2001). World Investment Report: Promoting Linkages. New York: United Nations. ——(2002). World Investment Report: Transnational Corporations and Export Competitiveness. United Nations: New York.

Bibliography

406

——(2003). World Investment Report: FDI Policies for Development – National and International Perspectives. United Nations: New York. ——(2005). World Investment Report: Transnational Corporations and the Internationalisation of R&D. United Nations: New York. ——(2006). World Investment Report: Translational Corporations, Extractive Industries and Development. United Nations: New York. ——(2007). World Investment Report: FDI from Developing and Transition Economies. United Nations: New York. UNCTC (1991a). The Impact of Trade-related Investment Measures on Trade and Development. New York: United Nations. ——(1991b). World Investment Report: The Triad in Foreign Direct Investment. New York: United Nations. ——(1991c). Government Policies and Foreign Direct Investment. New York: United Nations. Ulltveit-Moe, K. (2007). ‘Regional policy design: an analysis of relocation, efficiency and equity’, European Economic Review, pp. 1443–67. Vanhove, N. and L. Klaasen (1987). Regional Policy: A European Approach. Aldershot: Avebury. Veblen, T. (1898). ‘Why is economics not an evolutionary science?’, Quarterly Journal of Economics, pp. 373–97. ——(2003). Imperial Germany and the Industrial Revolution. Kitchener: Batoche Books. Venables, A. (1985). ‘Trade and trade policy with imperfect competition: the case of identical products and free entry’, Journal of International Economics, pp. 1–19. ——(1994). ‘Economic integration and industrial agglomeration’, Economic and Social Review, pp. 1–17. ——(1995). ‘Economic integration and the location of firms’, American Economic Review (Papers and Proceedings), pp. 296–300. ——(1996a). ‘Localization of industry and trade performance’, Oxford Review of Economic Policy, pp. 52–60. ——(1996b). ‘Equilibrium locations of vertically linked industries’, International Economic Review, pp. 341–59. ——(1998). ‘The assessment: trade and location’, Oxford Review of Economic Policy, pp. 1–6. ——(1999). ‘Regional integration agreements’, World Bank Policy Research Working Paper No. 2260, World Bank, Washington, DC. ——(2001). ‘Geography and international inequalities: the impact of new technologies’, Centre for Economic Performance, London School of Economics, London. ——(2003). ‘Winners and losers from regional integration agreements’, Economic Journal, pp. 747–61. ——(2006). ‘Shifts in economic geography and their causes’, London School of Economics, Jackson Hole Symposium, mimeo. Venables, A. and N. Limão (2002). ‘Geographical disadvantage: a Heckscher–Ohlin– von Thünen model of international specialisation’, Journal of International Economics, pp. 239–63. Vermeij, G. (2004). Nature: An Economic History. Princeton: Princeton University Press. Vermulst, E. (2005). ‘The 10 major problems with the anti-dumping instrument in the European Community’, Journal of World Trade, pp. 105–13. Vernon, R. (1966). ‘International investment and international trade in the product cycle’, Quarterly Journal of Economics, pp. 190–207. Vertova, G. (ed.) (2006). The Changing Economic Geography of Globalization. London: Routledge. von Thünen, J. (1826). Der Isolierte Staat in Beziehung auf Landtschaft und Nationalökonomie. Hamburg: Perthes. (Translated from German by C. Wartenberg, von Thünen’s Isolated State (1966), Oxford: Pergamon Press.) Warsh, D. (2006). Knowledge and the Wealth of Nations. New York: Norton.

Bibliography

407

Wasylenko, M. (1991). ‘Empirical evidence on interregional business location decisions and the role of fiscal incentives in economic development’, in H. Herzog and A. Schlottmann (eds) Industry Location and Public Policy. Knoxville, TN: University of Tennessee Press, pp. 13–30. Weber, A. (1909). Über den Standort der Industrien. Tübingen: J. C. B. Mohr. (Translated from German by C. Friedrich, Theory of the Location of Industries (1962), Chicago: University of Chicago Press.) Weder, R. (1995). ‘The Swiss dyestuff industry’, in M. Enright and R. Weder (eds) Studies in Swiss Competitive Advantage. Bern: Peter Lang, pp. 24–60. Weisbrot, M., D. Rosnick and D. Baker (2004). ‘NAFTA at ten: the recount’, CEPR Briefing Paper, Centre for Economic Policy Research, London. Wheeler, D. and A. Mody (1992). ‘International investment location decisions: the case of US firms’, Journal of International Economics, pp. 57–76. Witt, U. (2003). ‘Evolutionary economics and the extension of evolution to the economy’, in U. Witt (ed.) The Evolving Economy. Cheltenham: Edward Elgar, pp. 3–34. Wolf, M. (2004). ‘Globalization and global economic governance’, Oxford Review of Economic Policy, pp. 72–84. ——(2005). ‘Will globalization survive?’, Third Whitman Lecture, Washington: Institute for International Economics. World Bank (1993). The East Asian Miracle. Oxford: Oxford University Press. World Bank Commission on Growth and Development (2008). The Growth Report. Washington: The World Bank Wren, C. (1990). ‘Regional policy in the 1980s’, National Westminster Bank Quarterly Review, pp. 52–64. WTO (2004). Trade Policy Review: European Communities. Geneva: WTO; WT/ TPR/S/136. ——(2007). Trade Policy Review: European Communities. Geneva: WTO; WT/TPR/ S/177/Rev.1. Yamawaki, H. (1993). ‘Location decisions of Japanese multinational firms in European manufacturing industries’, in K. Hughes (ed.) European Competitiveness. Cambridge: Cambridge University Press, pp. 11–28. Yuill, D., K. Allen, J. Bachtler, K. Clement and F. Wishdale (1994). European Regional Incentives, 1994–95. London: Bowker. Zysman, J. and A. Schwartz (1998). ‘Reunifying Europe in an emerging world economy: economic heterogeneity, new industrial options, and political choices’, Journal of Common Market Studies, pp. 405–29.

Index adjustment costs 46, 177; central and eastern Europe 187; European Free Trade Association (EFTA) 187; evolutionary economic geography 187; General Agreement on Tariffs and Trade (GATT) 187, 249; intervention 188; neoclassical equilibrium model 186–87; non-interference policy 188; reallocation of resources and adjustment costs 186–88; trade liberalisation 187–88; World Trade Organisation (WTO) 187; see also international economic integration and location; regional policy aerospace industry 32, 217, 225, 302, 303; Airbus 230, 246–48, 302, 303, 348, 350, 367; Boeing 146–47, 230, 246, 303, 367, 378; Britain 230, 246; Italy 147; mergers 257; military aircraft 298; politicised industry 247–48; United States 146–47, 246, 302, 303 agglomerations 22, 45, 47, 76, 129, 180, 192; centripetal forces 81; cities 110; economic integration 82, 194; evolutionary economic geography model 174; factors that influence agglomeration tendencies 165; France 76; good/bad agglomerations 81; Industrial Revolution 80; labour mobility 174; multiple equilibria 22, 174; principle 85; public institutions 85–86; reasons for agglomeration 111; and spread forces 192–93; see also cities; clusters; regions agreements, conventions, rounds: Agreement on TRIMs 296, 377;

Index

409

Canada-US Free Trade Agreement 249, 277, 278; Cotonou Agreement 213; Kyoto Customs Convention (1973) 277, 281; Multilateral Agreement on Investment (MAI) 331, 341, 387–88, 396; national and international regulation 385–88; Treaty of Rome 187, 282, 292, 294, 305; Uruguay Round 20, 29, 259, 277–78, 285, 296; WTO 1995 TRIMs Agreement 377, 395, 404; see also European Free Trade Area (EFTA); General Agreement on Tariffs and Trade (GATT); North American Free Trade Agreement (NAFTA) agriculture: agricultural population and cities 114; agriculture-related machinery 77, 425; gardening 24, 156; share of labour related to capital 109; von Thünen’s model 23–24; see also land Africa: crude oil 348–49, 403; Sub-Saharan Africa 397; transnational corporations (TNCs) 397 Albania 26 Amazon.com 139, 258 Angola 397 antidumping 283–88; antidumping duties 285; barriers to trade 286; countervailing duties 286; cumulation rules 286–87; dumping 286; European Union 285; injury 286; negative aspects of 287–88; predatory pricing 271; United States 285; ‘zeroing’ 304; see also tariffs Apple 62, 71, 221, 246, 260, 297 Arab world 239; brain drain 239; innovation 300; Saudi bribes scandal 399; see also Islamic world Argentina 129, 204, 354; International Monetary Fund (IMF) 331, 406 Aristotle 54; De Caelo 54 art 166, 427, 428; auction houses 427; Florence 166, 427, 428;

Index

410

location of valuable classical painting 166; Mantegna, Andrea: Opportunity and Penitence 78, 79 Asia: Asian crisis 314–15, 354, 387, 408; Association of Southeast Asian Nations (ASEAN) 287; competition model 217; confiscation 242; convergence 213; cotton textile industry 162; ‘East Asian miracle’ 373; Euro-Asia 238–39; globalisation 348; high-technology production 40; International Monetary Fund (IMF) 331; Nike 375; reallocation of manufacturing 19, 20; Thai baht devaluation 71 assets 112, 153, 270, 311, 312, 315–16, 353, 361; asset bubbles 308; complementary assets 153, 234; domestic assets 366, 392; intangible 95, 96, 112, 144, 232, 311–12, 388–89; physical 112; region 121, 128 Association of Southeast Asian Nations (ASEAN): Free Trade Area 287 AT&T 62, 218 Austria 192, 273; café bars 149; political reforms 301; tax system 323, 324; transport costs 152; Turkish occupation (1529–1683) 149; Vienna 149 Australia 155; antidumping duties 285; convergence 198; patents 231; peripheral location 26, 175; property rights 155; research and development (R&D) 299–300; university 300; wine industry 153, 169 automotive industry 170, 190, 249, 255, 264–65, 426; BMW 257; Britain 148–49; Chrysler 249, 257, 351; Delphi 370; Europe 190; Ferrari 148–49, 158; FIAT 95;

Index Ford 235, 300; France 304; Germany 125–26; GM 95, 235, 265; Honda 77, 140, 169, 414; Hyundai (KIA) 157, 398; Italy 294; Japan 277; Japanese cars in Europe 293–94; Lancia 148; lock-in effect 66; mergers 257; Mitsubishi 404; motor car racing 148–49; Nissan 304; Peugeot-Citroën 157, 398; rules of origin 280, 304; Slovakia 157, 398; Toyota 66, 95, 169, 252, 265, 381–82, 395, 397; tyre-making 162; Volkswagen 95, 157, 257, 265, 398; Volvo 265 Bangladesh: textile industry 359–60 banking: changes 138–39; clusters 102; England 102; evolutionary geographical chain 102; Holland 102; Italy 102; security 169; Switzerland 141, 423; United States 102 barriers: entry barriers 167, 269–70, 390; exit barriers 167, 269–70; see also antidumping; free trade; market structure and location of production; non-tariff barriers (NTBs); tariffs; trade barriers Beinhocker, E. 57–59; adaptive walk 57, 58, 59; random jump 57–58, 59 Belgium 22, 192, 273; Antwerp 102, 119–20, 128, 166; Belgium Walloon region 66; Brussels 111, 121, 213;

411

Index

412

liquid fuel 321; research and development (R&D) 299–300; tax system 324 biology and economics 40–41, 42, 43, 137, 169, 233, 412; ‘biological Internet’ 41; biological evolution 56; changes in climate 12, 39, 41, 57, 118, 131, 168; economic biology 12; and evolution 12–16; evolutionary biology 43, 131, 133, 168; ‘first-mover advantages’ 133; Journal of Bioeconomics 13; Journal of Evolutionary Economic Geography 13; punctuated equilibrium 131, 168; similarities and differences between biology and economics 13, 412; transnational corporations (TNCs) 317; see also evolution; evolutionary economic geography book industry 139; Amazon 139; Encarta 139; Encyclopaedia Britannica 139; Wikipedia 139 brands 54, 97, 153, 170, 258, 291, 400 Brazil 72, 204; Brasilia 112–13; global economy 159; International Monetary Fund (IMF) 331, 406; World Social Forum 332 Britain 91–92, 162, 354; aircraft industry 230, 246; banking 138–39, 364; beef and BSE 295, 305; British ASDA Stores Ltd 290; British East India Company 139; British Imperial preference scheme 279; City of London 92, 100; clusters 164; communication industry 218; corporate income taxes 320–21; corruption 399; cotton textile industry 162; cumulation of knowledge 245; economic growth 240–41; educational policy 129; Encyclopaedia Britannica 139; European Union (EU) 201; face-to-face contact 91–92; financial industry 165; First Industrial Revolution 164, 229, 242; free trade 33;

Index

413

Imperial Chemical Industries (ICI) 65; important cities 115–16; industrialisation 302; lock-in effect 65, 161, 162; long-distance trade 409; Magna Carta (1215) 242; mergers 257; motor car racing 148–49; Nissan 304; Northern Rock financial fiasco 165; parallels between United States and United Kingdom 401; patents 231; path dependency 162; rail wagons 161; relocation of business 204; research and development (R&D) 299–300; shipbuilding industry 212; tax system 347, 364; technological progress 302; Thatcher, Margaret 320–21; traditional regional policy 202; transnational corporations (TNCs) 317; university 300; see also England; Scotland Buridan, Jean 54; Buridan’s ass 54 Cambodia 422; Angkor 108 Canada: antidumping duties 285; Canada-US Free Trade Agreement 249, 277, 278; competition model 217; convergence 198; Foreign Investment Review Agency (FIRA) 385–86; international economic integration and location 203–5; Massey Ferguson 77; national economic core 29; national health insurance system 382; North American Free Trade Agreement (NAFTA) 186, 203, 249; patents 231; resources-rich countries 385, 405; university 300; US-Canada Auto Pact 249 capital 354–56, 360, 392; accumulation 129; capital-labour ratio 47, 159; capital market 259, 329, 337, 360, 361, 388; capital mobility 307–8, 360, 391, 396; capital stock 128, 129; cities 109–10,111;

Index

414

clusters 85; direct investment 307; international capital flow 308; liberal capital mobility 320; low/high technology industries in countries at different stages of development 354–55; national endowment of factors 17; plain capital 308; portfolio investment 307–8; see also foreign direct investment (FDI); International Monetary Fund (IMF); World Bank capitalism 332; ‘alliance capitalism’ 83; clusters 83; globalisation 332, 337, 397; see also economic liberalism; globalisation changes in the economic system 3, 13, 41, 48, 116, 130–31, 151, 168; banking 138–39; business 132–33; cities 131; continuous changes 65; differential change 199–200; entrepreneurs 130; Middle Ages 131, 132, 168; neoclassical equilibrium economics 73, 130–31; outcomes 14; policy-induced changes 20; social progress 132; technological changes 13, 18, 48, 92, 94, 123, 131, 135, 138, 139, 177, 187, 224, 228, 241, 253, 262, 265, 343, 413; see also biology and economics; education, evolution; innovation chemical industry: AKZO 271; Germany 146, 245; Holland 271; Imperial Chemical Industries (ICI) 65; Switzerland 147, 424 China 19, 98, 208, 239–40, 242–43, 299, 301, 354–55; antidumping duties 286, 287, 304; Christmas trees 286; clusters 82, 106; comparative advantage 252–53; concentration of innovation 236; convergence 198; crude oil 348–49, 403; culture 373; domestic market 23, 315; domestic regional imbalance 315; economic development 178, 239–40;

Index

415

economic growth 239–40, 301, 315, 348, 414–15; education 212, 299; export-led growth 208; financial services 103; foreign direct investment (FDI) 347; free economic zones 371; global economy 159; globalisation 348; government 415; Great Wall 146; growth model for developing world 414; herd behaviour 314–15; hight-echnology production 40; Hong Kong 55, 103, 116, 325, 327; institutions of learning 242–43; inter-regional labour mobility 174; knowledge and innovations 242–43; lack of primary resources 408; low labour costs 315; manufacturing 23, 150, 299, 315, 348, 396, 408; patents 231; Pearl River delta 315, 398; policies 414–15; property rights 402; reallocation of manufacturing 19, 20, 21, 150, 170, 348, 408; religion 238; research and development (R&D) 230, 299, 365–66, 367; routes to 120; silk production 239; society of obedience 236, 243; society of will 243–44; Sun Tzu 235, 300; textile industry 19, 20, 21, 55, 293; transnational corporations (TNCs) 314–15, 348–49, 403; World Trade Organisation (WTO) 20 Chile 204, 205; food industry 278 choice of outcomes 72–73, 162 Church 168, 223–24, 236–37; 299, 400; Christianisation 337, 338; Galileo Galilei 237; Inquisition 119, 237, 422; Pope John Paul II 237; religion/science debate 237–38 cities 80, 107–21, 412; agglomeration gains 110; Angkor 108; Antwerp 119–20; Berlin 166; Brasilia 112–13; Brussels 111;

Index

416

business climate/people climate 117; capital 109–10, 111; changes 131; City of London 115; commuting 116; competition 115; Corby 116–17; creation 112–13, 114, 133–34, 165; Crewe 117; definition 108; diversity 115, 166; dying cities 166; economies of scale 110, 165; equilibrium spatial location of activities 109; Eridu 108; external shocks 120–21; factors that influence agglomeration tendencies 165; foundation mechanisms 114–17; Germany 118; health 165; Hiroshima 120–21; incentives for urbanisation 111; industrial systems in 114–15; Internet 116; labour 109–10, 111, 112; lifestyle 112, 118; location of 109, 112, 113; Mexico City 117; Milton Keynes 116; Nagasaki 120–21; neoclassical equilibrium economics 107–8; New Orleans 118–19; Paris 166; Plato 109; proximity-productivity effect 110, 113; reasons for agglomeration 111; royal courts 133–34; Silicon Valley 114–15; Singapore 116; size 114 (optimum city size 117; and productivity 111; and sprawl 117–18); spatial economics 112, 113–14; specialisation 114–15, 166; specialisation/diversification 115–16, 166; Stevenage 116–17; Sumerian civilisation 108; Tikal 108; Tokyo 121; United States 112, 165; Ur 108;

Index

417

urban and rural population of the world 110; urban slums 110; urbanisation 109–10, 165; Uruk 108; Venice 120; Washington DC 111, 113; Zipf’s Law 117; see also agglomerations; clusters; regions clusters 22, 76, 80–107, 151, 164, 409, 412, 423–27; advantages and disadvantages of 103–4; agglomeration principle 85; ‘alliance capitalism’ 83; Bangalore 213; benefits 86, 87, 104; capital 85; ‘chaotic concept’ 105; China 82, 106; cluster policy 99–105; common features 106–7; competition 95, 96, 99; consequences of clustering 92; core region 80, 82; creation, origins 9, 100, 101, 104, 106; definition 80, 83, 105; Denmark 98; differences between evolutionary theory and neoclassical equilibrium economics 45, 47, 174; economic gains 83; economic integration 82, 194; education and training 106; efficiency considerations 85; Ellison-Glaeser concentration index 104, 165; England 84, 86, 91–92, 99, 106; evolution xvi, 101–2; face-to-face contact 90–92, 95, 164; Finland 98, 106; foreign direct investment (FDI) 106; France 28, 99, 157–58; general operations within a cluster 96–97; geographical boundary 83; Germany 84, 96, 106, 430; Gini concentration index 5, 104, 165; globalisation 82, 85, 97, 105; heterogeneous firms 88; identical firms 87–88; industrial boundary 83; Industrial Revolution 80; innovation and institutions 92–94; intra-firm trade 310; Ireland 97, 106; Italy 86, 94, 99, 106, 136–37, 425, 429–30; knowledge externalities 95;

Index

418

liberal trade 80, 106; lobbyist clusters 111; location 112, 136; lock-in effect 92, 93; managing clusters 100; Marshall, A. 105; Massachusetts General Hospital (Boston) 85; a mature cluster 102; meaning and motives 83 –88; Mexico 86; multiple equilibria 85; natural evolution 84–85; negative externalities 190; organisations and externalities 94–98; Pakistan 84; path dependency 92; periphery region 80–81, 82; policy 99–105, 174; policy implications 106–7; power concentration 86, 163; public institutions 85–86; reasons for choosing a cluster as location 83–84; recap 105–6; Ruhr region 66, 67, 68, 106, 126; Silicon Valley (California) 84, 98, 99, 100, 102, 104, 107; spatial barriers 85; specialisation 81, 92, 107; sports goods 84; spread of information and knowledge 86, 88–92, 95, 163, 168–69; success, factors of 84, 95; success in clusters of related firms 94–95, 96, 97–98; surgical instruments 84; Switzerland 107, 423–25; talented people 97; taxes 323–24; textile manufacturing 84; training 86–87; underlying forces (centripetal/centrifugal) 80–82, 85; university 98–99, 100; US clusters 89; US film industry 88; see also agglomerations; cluster policy; firms; knowledge and experience cluster policy 99–105, 106; creation of clusters 100, 101, 104, 106; goal 100; policy implications 106–7; see also clusters communication industry 225, 228; Britain 218;

Index

419

France 217; globalisation 338; India 157; long-distance telephone call market 225; mobile phone 138, 157, 218, 228, 299, 340, 417; monopoly 217–19; Motorola 157, 157, 315, 367, 383; Nokia 157; United States 217, 218; Switzerland 218–19; see also electronics industry; Internet; software communication technology see communication industry; high-technology industry; Internet Communist Manifesto 337, 401 comparative advantage 129, 172, 179, 269, 273, 379–80, 415, 418; comparative advantage model 8, 17, 24, 55, 151, 193; cumulation 244; engineered comparative advantages 379; greatest comparative advantage 252; international economic integration 188; neoclassical equilibrium economics 267; Ricardo, David 8, 55, 151 competition 115, 153, 215, 261–62, 409, 416; anti-competitive business practices 260; appraisal 296–97; benefits of competition 226–27; cities 115; clusters 95, 96, 99, 172; competition among expectations 138–45, 167, 169; competition and geographical changes of production 141–43; competition models 216, 217, 297; competition policy 215–16, 218–19, 234, 297; competitive advantages 60, 66, 103, 105, 107, 383, 409; competitive international industry model (exchange of threats) 315–16; competitive markets 298; ‘creative destruction’ 222; definition 215; economies of scale 267; education 140–41; effects on innovation 226; EU rules on competition 286; ‘first-mover advantages’ 133; flat world hypothesis 140; free competition 225, 247, 265; functions 215; geography of production 177; global competitiveness 151, 170, 262; goal 216; imperfect competition 7, 8, 53, 82, 125, 151, 194, 206, 247, 412;

Index

420

innovation 234; internal competition 265; Japan 170; local 165, 170; location and global competition 329; perfect competition 30, 111, 163, 216–17, 222, 250, 254, 263, 267, 388; predatory pricing 270–72; predatory pricing/vigorous price competition 271; pricing policy 260–61; Red Queen race 139–40, 412, 419; regional competitiveness 67, 128, 162; tax competition 320–28, 347, 398; transnational corporations (TNCs) 357, 394; see also market structure and location of production complex systems 38, 39, 52–75, 412; Arthur, Brian 52–53, 159, 419; characteristics 52–75; complex system analysis 154; differences between complex and complicated system 15; effectiveness 15, 56–64; and evolutionary economic geography 52–75; Holland, John 38; indecision and ambiguity 54; inflexibility (lock-in effect) 64–71; International Institute for Applied Systems Analysis (IIASA) 52; multiple equilibria 53 –56; equilibria 53–56; non-linear behaviour 38; non-predictability 53–56; path dependence, strong history and non-ergodicity 71–73; potential inefficiency (absence of optimality) 56–64; properties 15–16; Santa Fe Institute 52; stable/unstable system 80; time delays 73–75; see also complexity theory; evolutionary economic geography complexity theory xiv, 9, 36, 154; complex system analysis 154; see also complex system complicated system 15; differences between complicated and complex system 15 convergence hypothesis 158 coordination failure 14, 17, 54, 160–61; India 160; multiple equilibria 54, 55 core competences 100, 153; irreproducible capabilities 397; transnational corporations (TNCs) 311 corruption 30, 204, 327, 341, 350, 374, 399; anti-corruption 384; bribery 399;

Index

421

definition 399; Global Compact principles 384–85; see also red tape costs 85, 160, 267; average costs 160; cost curves with constant, increasing and decreasing returns to scale 263–64; cost minimisation 10, 43, 316; distribution of manufacturing and transport costs/wages 172–74; fixed costs 53, 108, 160, 388, 405; marginal costs 160; opportunity costs 76, 160; reallocation of resources and adjustment costs 186–88, 213; social costs 180, 190; start-up costs 266, 304; transport costs 97, 152, 170, 192, 193, 335, 402; variable costs 160; see also adjustment costs; sunk costs; trade costs; transport costs Costa Rica 205; Intel 370–71 crisis 162, 182; 1930s 124; Asian crisis 314–15, 354, 387, 408; British banking 165; debt crisis 142; global 72; Latin America 405; Tequila crisis 204 culture 372–74; China 373; elements and norms that support economic development 374; globalisation 340–41; trans-border business activities 350, 372–74; wealth 374 cumulation: cumulation rules 286–87, 304; Pan-European Cumulation System (PANEURO) 287, 304 cumulative causation 14, 44, 110, 125, 127, 129, 167, 423 customs unions 185, 204, 250, 275–76, 280; Community Customs Code (CCC) 281; economies of scale 266, 268, 269; fiscal frontiers 318; free trade 275, 280; Kyoto Customs Convention (1973) 277, 281; rules of origin 275–82; welfare in a trade-diverting customs union 185; see also tariffs Czech Republic 397–98; Hyundai 381; spillovers 403

Index

422

Darwin, Charles: On the Origin of Species 12 Dell Computer Corporation 61, 77–78, 235, 315, 414 Delos 320 demand 11, 17, 22, 74, 106, 252, 270, 299, 388; demand-driven approach 84, 148, 348; demand structure 255; domestic demand 22, 423; innovation 223; local demand 22, 136, 317, 368, 395; national demand 23; supply and demand 63; time delay 74–75 Denmark 98; research and development (R&D) 299–300; traditional regional policy 202; wooden furniture 98 deregulation 218, 248, 399, 405 deterministic process, way 14, 150, 155; non-deterministic system 15 developed/developing countries 141–42, 152, 193, 272, 354–55, 399; competition 262; debts 141–42; Global Compact principles 385; globalisation 335, 342, 344; government intervention 182; growth 124; low/high technology industries in countries at different stages of development 354–55; migration 214; Multilateral Agreement on Investment (MAI) 387–88; neoclassical equilibrium economics 355–56; preferential trade 279; trade-related investment measures (TRIMs) 296; transfer pricing 365; transnational corporations (TNCs) 309, 313, 327, 369, 370; see also globalisation; transnational corporations (TNCs) development 30, 31, 105, 178; assets 112; chaos/order 127; education 162; elements and norms that support economic development 374; economic development 374; endogenous/exogenous 105; evolutionary development 127; ‘high road’/’low road’ 105; history-dependent development 130; human capital 141; of manufacturing industry 165; national levels 31;

Index

423

see also developed/developing countries development path 72, 163; education 162; Finland 162; Sweden 162; see also lock-in effect; path-dependency disequilibrium 4, 38–39, 158, 410; the norm in evolutionary economic geography 42; regional disequilibria 122; Schumpeter, Joseph 41, 42; see also equilibrium Dubai Port World 156 DVDs 65; Sony’s Blu-ray format 61; Toshiba’s HD-DVD 61; VCRs 65; see also electronics industry economic geography 3, 7, 9; core-periphery model 315; differences between evolutionary economic geography and 409; history 16; trade costs 156, 408; trade within industries 253, 256; see also evolutionary economic geography; geography and economics; geography of production; spatial economics economic growth 128–29, 140, 167, 169, 198–99, 240, 244, 253, 302, 414; China 239–40, 315; elements and norms that support economic development 374; European Union (EU) 199; export-led growth 208; first ‘modern’ corporation-like business in Europe 168; GPT 143, 169, 198; India 240; Industrial Revolution 223, 298; innovation 224, 228; Latin America 204–5; Middle Ages 131, 133–34, 168; non-linear effects 174; production function 224; productivity growth 144, 169; sources 131, 167; TNCs and the growth process 356–57; United States 144–45, 205; Wal-Mart effect 144, 169; see also changes in the economic system economic integration 97, 98, 172, 208–12, 226;

Index

424

agglomerations 82; continental integration 29; economic adjustment 177; effects on regional development 179–80; foreign direct investment (FDI) 347, 402; globalisation 338, 402; innovation 225; integrated areas 82, 98, 193, 196, 226, 247, 307, 377; non-linear effects 174; reaction to 152; regionalisation/integration 347; see also European Union (EU); globalisation; international economic integration and location; Single European Market; standards, standardisation economic liberalism 152, 159; economies of scale 329; liberalisation 329, 346; liberalism/controlled interventionism 339; neoclassical equilibrium economics 46; neo-liberalism 405; United States 216–17, 297; Washington Consensus 389, 405–6, 415; see also capitalism; competition; globalisation; liberal trade economics 36; biology, economics and evolution 12–16; interdisciplinarity 12, 13; mathematics 12; neuroeconomics 13, 47; physics 12, 34–41; psychology 13, 47; thermodynamics, first law 37; thermodynamics, second law 37 economies of scale 2, 9, 11, 18, 22–23, 151, 165, 292, 304, 309, 378, 405, 412, 416; approach to 264; characteristics 266; cities 110; clustering 22; concentration business 216; cumulative causation 127, 167; customs union 266, 268, 269; economic growth 129; economies of scope 268; Europe 196; European Union (EU) 263–64; government policy 269; impact on location 153; impact on production 18;

Index

425

internal/external to individual firms 266–67; investment 266, 304; liberalisation 329; market imperfections 267, 269; modelling of 268; neoclassical equilibrium economics 268; perfect/imperfect competition 267; plant efficiency 265–66; product differentiation 255; regional development 179–80; sources 18; specialisation 266, 267, 269, 349; technology 266; see also returns to scale; specialisation and returns to scale; standards, standardisation economy: closed/open system 38–39; new economy 143–44; principal moving forces 41 education 31, 48, 416; Bavaria 126; China 212; clusters 85, 106; development 162; educational policy 129, 140–41, 194, 418; government 418; importance of 85, 123, 140–41, 152; innovation 234; social potentials 125; transnational corporations (TNCs) 392; United States 212; see also knowledge and experience; learning, training; research and development (R&D); technology Egypt: society of obedience 236 electricity 228; Enron 103, 248, 258, 394, 399; from water power to electricity 68, 70–71; general-purpose technology 228, 230; Westinghouse Electric 139 electronics industry 66, 100–101, 169, 235, 246, 273; Betamax 61, 161; changes in 138, 139; colour-TV broadcasting 78; Czech Republic 398; Dell Computer Corporation 61, 77–78, 235, 315, 414; Digital 97; DVDs 61, 293; Flextronics 315;

Index

426

Italy 78, 421; Japan 225, 234, 246; lock-in effect 66; Mexico 361; microchips 283, 285; Polaroid 246; radios 100–101; Seagate 314; Silicon Valley (California) 84, 98, 99, 100, 102, 104, 114–15; Sony 61–62; Taiwan 361; Toshiba 61–62; VCR 245; VHS 61, 161; see also communication industry; DVDs; IBM; Internet; Macintosh; Microsoft; new economy; software; Sony Ellison-Glaeser concentration index 104, 165 employment 81, 144, 164, 179, 181, 198, 213, 375–76; employability 140; European employment and spatial planning laws 169; rate 191; see also labour force; unemployment; wages endogenous growth theory xiv, 154, 156; and neoclassical equilibrium economics 31; production function 224 endowment: factor endowment 250, 251, 266; factor endowment model (Heckscher–Ohlin model) 17–18, 19, 126, 250, 252, 253, 256; national endowment 250; national endowment of primary factors of 16–18 energy 65, 170; coal 65, 67, 123, 138, 146, 170; fuel, oil 20, 142, 170; steam 40, 65, 68, 131, 164, 229, 240, 268, 299, 342, 343; United States 144–45; see also electricity England 22, 103, 156, 302, 341; auction houses 427; banking 102, 103; Birmingham 115–16; Cambridge 99; Cambridge University 99;

Index

427

capital cities 166; Cavendish Laboratory 141; centralised state 86; Chadwick, James 40; City of London 100, 103, 106, 115, 165; clusters 84, 99, 427; cutlery 133; Corby 116–17; Cotton Exchange 84; Crewe 117; disincentives for regional expansion 183; financial services 103; Leicester Square 427; Lloyd, Edward 427; London 115–16, 131, 427; Manchester 115–16; metal production 133; Milton Keynes 116; Navigation Act (1660) 165; pan-European regulation of finance 103; Peace of Westphalia 149–50; Shakespeare, William 79; Soho 427; St. Martin’s Court 427; Stevenage 116–17; textile manufacturing 84, 234; universities 99; see also Britain Enron 103, 248, 258, 394, 399 entrepreneurs 41, 76–77, 85; Bavaria 126; changes in the economic system 130; clusters 90, 95; competitive entrepreneurship 60; entrepreneurship 60, 126, 213, 222, 421; lock-in 60; ‘tariff factories’ 307; uncertainty forces 38 entropy 37, 38–39, 159 environmental issues 156, 283, 284, 313; China 240; environment-related industries 67–68; environmental standards 351, 397, 403; environmentally friendly technology 78–79, 377; Germany 78–79; Global Compact principles 384–85; globalisation 333; harmonised standards 292; India 240; innovation 223; pollution in clusters 92, 240; precautionary principle 295;

Index

428

protectionist policies 17, 156; transnational corporations (TNCs) 313, 351, 356, 397, 403 equations: equation for the right location 11; homogeneous production functions 263; intra-industry trade index 251–52; production function 224; thermodynamics, first law 37; thermodynamics, second law 37; Weber’s model equation 24 equilibrium 12, 36, 38, 42, 159; evolutionary theory 55–56; general equilibrium point 34; neoclassical equilibrium 12, 30, 32, 35, 52, 55–56; optimal equilibrium point 158; punctuated equilibrium 130, 131; quasi-equilibrium 130; static analysis 167; stationary equilibrium 30, 35, 39, 52; see also disequilibria; multiple equilibria; neoclassical equilibrium economics ergodicity 162; ergodic system 162; non-ergodicity 44, 71, 83, 125, 319; path dependence, strong history and non-ergodicity 71–73 Europe 79, 105, 144; adjustment costs 187; Airbus 230, 246–48, 302, 303, 348, 350, 367; clusters 189–90; comparison between the US and European mergers and acquisitions 303; comparison between the US and European performance of firms 253–54, 303; comparison between the US and European spatial structure of industry 190–91; concentration of innovation 236; core location 175; culture 373–74; Eastern Europe 186; economic growth 189, 243, 301; Euro-Asia 238–39; European export of capital to the US 355–56; First Industrial Revolution 239; geographical advantage 238–39; imports 156; independent institutions 243; Japanese cars in Europe 293–94, 305; local, regional development 105; manufacturing 253, 254, 266; medieval universities 242; Middle Ages 131, 132–34, 168, 236; openness to novelties 240–41; pan-European production strategy 273; pan-European regulation of finance 103;

Index

429

patents, patent families 230–31; policy changes 174; political process 240–41; production geography 189–90; R&D 243; reallocation of resources and adjustment costs 186–88, 213; rules on competition 286; Second Industrial Revolution 125, 298; society of will 236; standards 273–74; trade barriers 189; see also European Union (EU); international economic integration and location; war and the location of firms European Commission 196, 201, 202, 274; antidumping duties 285; ‘Made in the EU’ label 291; the Microsoft case 220–21, 298; mutual recognition principle 292; precautionary and prevention principles 295, 305; public procurement 288; rules of origin 283, 285; tax competition 324; see also European Union (EU) European Court of Justice 221, 281, 289, 290, 305 European Free Trade Association (EFTA) 187, 275, 277, 278, 279, 304 European Union (EU) 185–86, 213, 247, 273; adjustment costs 187; antidumping duties 285; automotive industry 157, 293–94; brain drain 231; Brussels 213; capital mobility 396; Cohesion Policy 175, 197–98, 201, 206, 207, 213; Committee of the Regions 213; competition, competition models 217, 249, 297; contemporary regional policy 202–3; convergence 198–99; Cotonou Agreement 213; differential change 199–200; economic geography 82; ‘economic patriotism’ 274, 386, 389; Economic and Monetary Union (EMU) 197; economies of scale 263–64; European Constitution 346, 402; European Council (2000) 231; European Food Safety Authority 296; European Free Trade Association (EFTA) 187, 275, 277, 278, 279, 304; European Monetary System and the euro 132, 346; European Regional Development Fund (ERDF) 201, 213; eurozone 194, 211, 346, 396; foreign direct investment (FDI) 347, 396;

Index

430

free trade 20; geographical concentration 195–96; Japanese cars in Europe 293–94, 305; knowledge-based economy 231; labour mobility 212; high-technology industry 28; industrial concentration 195, 199, 200; industrial structure 199, 200; internal EU competition 262; international economic integration and location 196–203; inter-regional labour mobility 174; intervention 177; Ireland 194; lead-time advantages over competitors 232, 300; liberal trade policy 274; Lisbon declaration 231, 266, 402; lobbies 368, 404; ‘Made in the EU’ label 291; mergers 256, 259, 303; military industry 305; monetary union 346; national and international regulation 385–88; non-preferential rules of origin 283, 285; non-tariff barriers (NTBs) 263, 274, 282; Official Journal of the European Community 288, 289; Pan-European Cumulation System (PANEURO) 287, 304; pan-European production strategy 273; pan-European regulation of finance 103; periphery location 26, 175, 199, 211, 213; policies 261; political EU goal 207; Portugal 194; precautionary and prevention principles 294–96; product differentiation/product specialisation 249; public procurement 288–89, 304–5; Reform Treaty 402; regional policy 176–77, 201–3, 213; regional policy assessment 208–12; regional problems 183, 201, 203; research and development (R&D) 230, 231, 299–300, 368; rules of origin 281; rural/urban areas 199, 201; Spain 194; specialisation 199, 200; standardisation 273–75, 291–92, 304, 305; strategic technology alliances 368; system of mutual recognition 274–75, 292, 304; success of regional policy 203; tax evasion 324–25; tax system 319, 324, 398–99; textile industry 20, 28;

Index

431

trade liberalisation 250; traditional regional policy 202; transnational corporations (TNCs) 395–96; Treaty of Rome 187; Turkey 204; Union of European Beverages Association (UNESDA) 342; value added tax (VAT) 319; see also economic integration; Europe; European Commission; international economic integration and location; regional policy; Single European Market events 29, 64, 71, 100, 153, 158, 418, 420–22, 423–27; accordions 421; banking 102–3; Cannes 420–21; chocolate 422; creation of cities 114; cumulative causation 127, 167; First World War 72; Long Term Capital Management 159–60; marriages 422; mathematical model 162; the Portuguese in Latin America 420; QWERTY keyboard 60, 63–64; Second World War 72; serendipity 71, 73, 227–28, 299, 410; see also war and the location of firms evolution xv, xviii, 4, 12–13, 41, 56–58, 130, 153, 411; Beinhocker, E. 57–59; biological evolution 56, 130, 131–32, 168, 410, 412; clusters 84–85; Darwin, Charles 12, 41, 42 (On the Origin of Species 12); see also biology and economics evolutionary economic geography xiv, 3, 36, 41–52, 130, 137, 139, 152, 153, 378–79, 400, 407–18; address model 50; adjustment costs 46; agglomerations 174; analytical tools 412; an art 418; assumptions 44, 55; Beinhocker, E. 57–59; biggest problems of 72, 410; capital-labour ratio 47; challenges 414; competition among locations to attract firms and industries 45; competition on the market for goods and services 45; complex systems and evolutionary economic geography 52–75; coordination failure 47; creation of clusters 101;

Index

432

criticism of Neoclassical equilibrium economics xvi, 43; decision-making 48; disequilibrium 4, 38–39, 42, 58; differences between economic geography and xv, 409; differences between neoclassical equilibrium economics and 2, 4–5, 31, 32, 43–48, 49–50, 151, 174, 410, 412; divergence school 125–26; economic policy 46, 49, 50; economic rents 32, 158; effectiveness 46; endogenous changes 48; European Monetary System and the euro 132; evolutionary biology 43, 130, 412; evolutionary concepts 13–14, 43, 411; evolutionary process 13, 14, 58; An Evolutionary Theory of Economic Change 42; an experiment 48; facing, predicting, outsmarting, shaping the future 59–60; geography 75; globalisation 409; goal 1–2; history 75; history-dependent development 130; history matters 49, 127, 128, 133, 137–38; impersonal forces shaping 416, 419; incentive 32; innovation 46, 48, 49; interdisciplinarity 2, 3, 411; issues of study 3, 4, 412; Johnson-Krauss’ law 47; location 126–27, 411; long-term changes of the structure 48; market imperfections 32; Marshall, A.12, 42–43; micro/macroeconomic planes of the economy 45; Middle Ages 131, 132, 168; monopoly 61, 416; new goods and services 388; new knowledge 67–68; non-linear effects 174; open disequilibrium system 38–39, 47; openness to ideas outside economics and mathematics 47; patents 233; physics and thermodynamics 38–39, 47; players 48, 49, 159; pluralism 413–14; policy advices 414–16; policy making xvi, 31, 411–12, 414–16; political/professional influences 56; Red Queen race 139–40, 412, 419; rivalry 49; Schumpeter, Joseph 41, 42, 159;

Index

433

a science 418; size of firms 46, 159; spatial economics xvii, 6; spatial location xix; specialisation 45; structuralist-evolutionary theory 10; tax principles 319; technological changes, phases of 224; topics for further research 412, 418–19; trade barriers 45; trade pattern 46; and traditional model 4; treatment of mind in choice-making 47; uncertainty 43, 48; unpredictability 14, 155; Veblen, Thorsten 41; welfare effect of trade 46; workhorse 109; see also biology and economics; complex systems; economic geography; history and competition among expectations; multiple equilibria; neoclassical equilibrium economics; non-linear dynamics (‘chaos theory’); physics and economics; spatial economics exchange rate 47, 250, 271, 311, 316, 319, 320, 360, 405; devaluation 72, 208, 316; discriminatory 167; fixed 154, 155, 312, 360; flexible 154; Thai baht 72 external economies 111, 113, 117, 133, 187; see also Marshall, A. externalities 192; definition 357; evolutionary economic geography 125, 153; externalities of innovation 133; gains 127; linkages, externalities and spillovers 354–63; negatives 117, 122, 190; positives 181; regional development 179–80; regional disparities 122; rents 158 factors: effects of innovation on the use of factors 225–26; employment of factors 179; factors that influence agglomeration tendencies 165;

Index

434

mobility of factors 179; OLI factors 312–13; personal factors for location 426; policy-related factors regarding location 11, 134–36, 426; success, factors of 84, 95; supply of factors 222; see also endowment; mobility of factors feedback: negative 161; positive 161 film industry 61; Cannes 421; clusters 88; Hollywood 228; US film industry 88, 245; Warner Brothers 62 financial services 82, 116, 189, 208, 375, 392, 395, 404; international 103; pan-European regulation of finance 103; see also banking; services Finland 68, 192, 211; change in industrial structure 200; clusters 98, 106; knowledge-based economy 162; Linux 244; peripheral location 26, 175, 211; pulp and paper industry 98; resource-based economy 162; shipbuilding industry 212 firms 260; antidumping duties 285, 286; characteristics 76; clusters 98, 136; clustering of identical/heterogeneous firms 87–88; competition 234–35, 260, 303; differences between national firms and TNCs 308–9; dominant market position 260–61; economies of scale 266–67; geographical spread of 82; innovation 260, 303; international firm: theory 308–17; intra-firm trade 309–10; licences 261; local/foreign 353–54, 403; location 113, 136; major problem for 137; mergers 256; mergers and acquisitions 219, 256–59;

Index

435

monopoly 260; monopolistic/oligopolistic firms 222–23; ownership of a firm 397; performance, reasons for higher/inferior 253; price competition 235; pricing policy 260–61; providers/recipients of information 90; reallocation of 76; risk-loving firm 64, 260, 270, 304; success in clusters of related firms 94–95, 96, 97–98; trade unions 136; see also international firms; transnational corporations (TNCs) fixed exchange rates 154 –55; see also flexible exchange rates Flanders: Antwerp 102, 119–20 flexible exchange rates 154–55; see also fixed exchange rates flexibility 15, 28, 93, 125, 209, 215, 310, 413, 421; cluster 105; inflexibility 52, 64, 256, 310; policy 153; in prices 4, 151; technology 179; see also look-in effect food industry 400; Chile 278; Coca-cola 400; counterfeit food 291; European Food Safety Authority 296; GM food 295; Mexico 278; Nestlé 141, 364, 384; precautionary and prevention principles 294–96, 305; Switzerland 259 Ford 235, 249, 253, 300, 395; Ford’s model T 53 spillover effect 356 foreign direct investment (FDI) 7, 25, 105, 307–28, 353, 406; capital mobility 308; China 347; clusters 106; control 308, 397; culture matters 350, 372–74; definition 308; determinants for FDI location 349–54; developing countries 313–14; economic integration 347, 402; European Union 347, 396; FDI flows 311, 313, 316, 342, 347, 349, 382, 388, 391, 397; general considerations 307–8; globalisation 329, 347, 402;

Index goal 308; international firm: theory 308–17; ‘investment creation effect’ 307; Japan 277, 347; location of firms and taxes 322, 326–28, 346–47; Mexico 204; multilateral agreement 396–97; national and international regulation 385–88; ownership of a firm 308, 397; property rights 402; resource-oriented FDI 346; risks 397; Sub-Saharan Africa 397; tax competition 320–28, 347, 398; taxes and fiscal incentives 317–20; theories of FDI 307, 309–17, 397–98; trade liberalisation 280; and transnational corporations 307–28; Turkey 204; United States 347; see also globalisation; International Monetary Fund (IMF); international trade; trans-border business activities; transnational corporations (TNCs); World Bank; World Trade Organisation (WTO) France 19, 22, 157, 170, 314; agglomeration 76; Airbus 230, 246–48, 302, 303, 348, 350, 367; automotive industry 304; Cannes 420–21; capital cities 166; Cassis de Dijon 289, 290; clusters 28, 99; competition model 217, 297; core location 175; disincentives for regional expansion 183; EU regional policy 209; Franks 132; free trade 33; growth poles 28, 157–58; labour reform 300–301; mirror/glass-making craft 233–34; national and international regulation 386, 405; Paris 166; patents 23, 233–34; Peace of Westphalia 149–50, 170; Peugeot-Citroën 157; Poitiers case 293; political process 301;

436

Index

437

political reforms 300–301; prevention principle 295, 305; regional ‘imports’ 177–78; Sophia-Antipolis 99; Thimonnier, Barthelemy 241; traditional regional policy 202; transnational corporations (TNCs) 317; University of Nice 99; Wanadoo 271, 304; wine industry 169 free trade 20, 33, 124, 184, 192, 307; customs union 275, 280; free/preferential trade areas 275–76; free trade agreement 275, 276; geographical trade deflection 275–76, 280; north-south trade relations 279–80; rules of origin 275–82; see also European Free Trade Area (EFTA); non-tariff barriers (NTBs); North American Free Trade Agreement (NAFTA); tariffs Galileo Galilei 150, 237 Gates, Bill 62–63, 140, 426; IBM 71–72; strategies 62–63; see also Microsoft General Agreement on Tariffs and Trade (GATT) 25, 187, 282, 296, 305; Agreement on Trade in Civil Aircraft 302; non-tariff barriers (NTBs) 282; predatory pricing 271; see also non-tariff barriers (NTBs) General Electric 68, 364, 365, 367 geographical remoteness 25–26, 151, 156; core region 156; globalisation 33; peripheral region 156; see also transport costs geography and economics 7–10, 145; criticism between 7–8; competition and geographical changes of production 141 –43; production 141–43; ‘cultural turn’ 7, 154; ‘geographical turn’ 8; history of (economic) geography 9; Journal of Economic Geography xv, 8; new research tools 8–9; publications regarding 8; see also economic geography; geography of production; spatial economics

Index

438

geography of production xviii, 10, 16, 17, 145, 150, 151, 253, 256, 306, 407; change of 11–12; competition and geographical changes of production 141–43, 169, 177; core-periphery model 315; European Union (EU) 199; geographical concentration 195–96, 201, 234; geography of innovation/geography of production 245–46; international economic integration 188; Linder’s model 256; production geography: reduction in trade barriers 193–96; production geography: with trade barriers 192–93; spatial concentration 171, 407; United States 188–89; see also economic geography; geography and economics; location; spatial economics; theories of location Germany 19, 22, 23, 157; Aldi 341; automotive industry 125–26; Bavaria 125–26; Berlin 166; BMW 257; Boss, Hugo 426; Cassis de Dijon 289, 290; chemical industry 146, 245; city sizes 118; clusters 84, 96, 106, 430; comparative advantage 252; competition model 217, 297; core location 175; cumulation of knowledge 245; DaimerChrysler 136; educational policy 129; Eighty Years War (1568–1648) 149; electronics 234, 300; environment-related industries 67–68; environmentally friendly technology 78–79; Falkenhagen, Hans 40; First World War 146; Franks 132; free trade 33; invalidation of German patents by the Allies 147, 424; Japan 136; Lidl 341; Mercedes Benz 148, 149, 381, 404; motor car racing 148–49; patents, patent families 230–31; Peace of Westphalia 149–50;

Index

439

political reforms 301; Porsche 148; precautionary principle 295; public procurement 289; rail wagons 161; Reinheitsgebot (1516) 290; religion 373–74; retirement patterns 155; Reutlingen 84; Ruhr region 66, 67, 68, 106, 126; Second World War 118, 121, 166; Siemens 96, 136, 404; SMEs 96; surgical instruments 84, 148; tax system 323, 324, 398; technical standards 289; textile industry 252, 426; Thirty Years War (1618–48) 149; traditional regional policy 202; transnational corporations (TNCs) 317; von Thünen’s model 24; Wittenberge 166; Zollverein 125–26 Gini concentration index 5, 104, 165 globalisation 150, 328–48, 390, 408, 409, 418; advertising 341, 402; anti-globalisation movement 332–33, 385; Asia 348; asymmetries, economic and social 345; background 328–33; clusters 82, 85, 106; Communist Manifesto 337, 401; criticism 328–29, 332, 336–39, 345, 402; culture values 340–41; definition 333–36, 401; economic integration 338, 402; environment 333; eras 32–33; evolutionary economic geography 409; financial services 103; first and second globalisation waves 342–44, 345–46; for and against 333, 336–39, 345; foreign direct investment (FDI) 329, 338, 402; global economy 159; ‘global left’ 332; global market 21; global newspapers 88; global standards 340–42; global textile market 55; global TV stations 88; government 329, 333; imperialism 335;

Index

440

information, telecommunication technologies 338; instruments of global control 330–31; International Monetary Fund (IMF) 330–31, 332, 360, 400; Internet 333, 339–40, 402; jobs 338; liberalism/controlled interventionism 339; location and global competition 329–30; monetary union 346; Multilateral Agreement on Investment 331, 341, 387, 396; negative outcomes 334, 336–39; neoclassical equilibrium economics 329; neo-communist risk 335–36; neo-liberalism 405; new technology 329, 333, 342; patriotic, protectionist wave 386–87; process of 329; product of technology and TNCs 329; profits 390; rebottled old stuff 242–46; regional integration 338; regionalisation 330, 340, 347; regionalisation/integration 347; regionalism/multilateralism 338–39; risks and disruptions 335, 345; rules of origin 282; ‘spatial/deepening’ connotations 333; strands 346–48; taxes 364; trade 338, 402; transfer pricing 364; transnational corporations (TNCs) 329, 331, 332, 336–37, 397; transport costs 97, 335, 402; United States 335; urbanisation 345–46; Washington Consensus 389, 405–6, 415; weaknesses 332; World Bank 330–31, 332, 400; World Social Forum 332; World Trade Organisation (WTO) 330–31, 332, 400; see also economic integration; foreign direct investment (FDI); trans-border business activities; transnational corporations (TNCs) goods: address models of goods differentiation 254–55; counterfeit goods 290–91; economies of scale 255; fixed 388; ‘hybrid goods’ 280, 304; new goods 388, 405; product differentiation 254–55; product differentiation/product specialisation 249, 250;

Index

441

see also intra-industry trade Google 133, 139, 169, 221, 236, 414 government 400, 414–16; activities 31; adjustment costs 188; antidumping 283–88; cluster policy 99–105, 164; control over cyberspace 340; economies of scale 269; education policy 418; efficient government 347; First Industrial Revolution 164; globalisation 329, 338, 390; grants 380–81; intervention 174, 351, 354, 377–82, 388, 415–16; involvement in international trade 284; Japan 182; Korea 182; liberalisation 329; licences 261; lobbying 371, 418; local content 377; location of firms 352; military-related government expenditure 229–30; national and international regulation 385–88; national regional policy 176; non-tariff barriers (NTBs) 282; over/under-investment 380; policy 32, 49, 128, 159, 182, 209, 347, 418; political reforms 301; politics and vested interests 239–41, 300–301; predatory pricing 270; public procurement 288–89; reasons for government intervention 175; regional integration 338–39; research and development (R&D) 299; subsidies 11, 55, 76, 100, 166, 175, 176, 180, 302, 303, 379, 380–81; tournaments 369–70; transfer of profits 403; transnational corporations (TNCs) 312, 332, 351, 352, 370–71, 377–82, 390, 397; see also regional policy; policy Greece 428; Cohesion Fund 213; labour force 353; peripheral location 175; public procurement 289; research and development (R&D) 299–300; traditional regional policy 202 Greenspan, Alan 143, 144; ‘irrational exuberance’ 143 Guatemala:

Index

442

Tikal 108 health 31, 190; Coca-cola 342, 400, 401; harmonised standards 292; health care 41, 155, 184, 217, 245, 345, 405, 424; Pepsi 342, 401, 402; political/professional pressure on medical technologies 56, 413; precautionary and prevention principles 294–96; standards 148, 284, 289, 290, 291; Switzerland 424; urbanisation 165; World Health Organisation 332 herd behaviour 131, 314, 350; core-periphery model 315 high-technology industry 28, 40, 55, 60, 194, 196, 200, 245, 265; Cambridge 99; China 40; Germany 96; Ireland 194; low/high technology industries in countries at different stages of development 354–55; Siemens 96; Silicon Valley (California) 84, 98, 99, 100, 102, 104, 114–15; SMEs 96; United States 245; see also electronics history and competition among expectations 126–45; background 126–28; background ideology in US policy-making 134–36; competition among expectations 138–45, 167; competition and geographical changes of production 141–42; education 140–41; evolutionary development 127; expectations 169; external economies (backward-looking dynamics) 111, 113, 117, 133, 187; ‘first-mover advantages’ 133; flat world hypothesis 140; historical lock-in effect 127, 128–38; history-dependent development 130; history matters 49, 127, 128, 133, 137–38, 199; innovation 141; neoclassical equilibrium economics 167; non-linear probability theory 128; Red Queen race 139–40, 412, 419; regional competition 128; see also changes in the economic system; location; lock-in effect Honda 77, 140, 169, 414 human agency 14, 100, 153, 155;

Index

443

innovation 299; location of firms 126 human capital 89, 97, 141, 164, 179, 245, 253, 392, 426; accumulation 90; China 415; cities 112; education and training 176, 179; embodied/disembodied knowledge 164; health of 416; most precious economic factor 141, 176, 245; social capital 91; Switzerland 424; United States 102 human rights 17, 283, 330, 344, 384; Global Compact principles 384–85; globalisation 337–38; women and children 337 Hungary: research and development (R&D) 299–300 Hyundai 157, 381, 398, 405 IBM 77–78, 132, 221, 235, 246, 272, 303, 366, 386; ‘complexity catastrophe’ 78; Gates, Bill 71–72; look-in 66; MS-DOS 62; PC 66, 71; see also software Iceland: research and development (R&D) 299–300 import 17, 391; import quotas 20–21; see also international trade; foreign direct investment (FDI); transnational corporations (TNCs) income 17; China 259; disparities 111, 172, 207; high-income countries 203; income distribution 11, 155, 188; income-generating 300, 312, 353, 354, 403; income taxes 319, 320, 326, 371, 398; increase in 17, 106, 161, 223, 229, 266, 273, 336; level of 16, 24, 122, 123, 183, 199, 206; location 23; low-income countries 203; per capita income 152, 192, 197, 203, 206, 252 indecision and ambiguity 54; Aristotle 54; Buridan, Jean 54; Morton, John 54

Index

444

India 20, 21, 98, 208; antidumping duties 287; automotive industry 66; Bangalore 208, 213; clusters 106, 213; Coca-Cola 401; coordination failure 160; domestic market 23; economic development 178; economic growth 240; global economy 159; globalisation 348; information technology 208; manufacturing 348; mobile phone industry 157; non-governmental organisations (NGOs) 401; Pepsi 401; reallocation of services 150; research and development (R&D) 367; society of obedience 236, 243; society of will 243–44 Indonesia 21; globalisation 337 industrial organisation 7, 153, 154, 222, 419 Industrial Revolution 18, 299, 302; clusters 80, 164; First Industrial Revolution 164, 223, 229, 239, 242, 298, 383; pre-Industrial Revolution 223; Second Industrial Revolution 125, 298 industry 195; barriers to enter and exit an industry 269–70; change in industrial structure 200; concave, convex and spherical distributions of potential industry location 75–76; concentration of industries 27–28, 104–5, 165, 195, 199, 200, 217–18; concentration of specialised industries 133; economies of scale 195; footloose industries 14, 53, 66, 110, 184, 211, 267, 326, 379; industrial organisation 7, 153, 154, 222, 419; industrial structure 200; industrial systems in cities 114–15; inter-industry trade 194; intra-industry trade (IIT) 18, 28, 152, 194; location 112, 133–34, 200; low/high technology industries in countries at different stages of development 354–55; multiple equilibria 160; see also international economic integration and location; intra-industry trade (IIT); regional policy inflexibility see lock-in effect information 88–89; globalisation 338;

Index

445

major strength of clusters 90; spillover effects 356; spread of information and knowledge 86, 88–92, 95, 97, 112, 114, 163, 166, 168–69, 356; see also education; knowledge and experience; learning, training information technology 208, 218, 228; see also communication industry; electronics industry; Internet innovation 14, 20, 141, 151, 222–59, 261, 302, 411; background 222–32; benefits of past/contemporary innovations 224; clusters 92–94; competition 260, 303; concentration of innovation 229, 236–48; economic growth 224; education and learning 234; effects 411; effects of competition on innovation 226; effects of innovation on the use of factors 225–26; evolutionary economic geography 46, 48; externalities of innovation 133; geography of innovation/geography of production 245–46; impact of innovation 228, 244; incentives for 222; individual creativity 302; innovation activities 229, 299; innovation process 227, 232; innovators’ rights 232–36; international economic integration 229; institutions 223–24, 241–44; intra-industry trade 249–56; levels of 228, 299; lifestyle 228; lock-in effect 60; mergers and acquisitions 256–59; the Microsoft case 219–22; monopoly 217–19, 260; non-profit-making innovators 232; past/contemporary innovations 224; patents, patent families 230–31; process of invention 224, 299; production function 224; ‘productivity paradox’ 228–29; ‘prospect theory’ 227; research and development (R&D) 164, 222, 225, 230–31, 243, 299–300, 302; returns from innovation 232; risk 227; serendipity 227–28; sources of novelty 223–24;

Index

446

success, principal tool for 236; technical innovations, legal and policy innovations 225; technological changes, phases of 224; technological revolution 228, 229, 299; uncertainty 222, 227; welfare 229; see also changes in the economic system; innovation, concentration of; research and development (R&D); technology innovation, concentration of 229, 236–48; commercialisation of innovations 245–48; cumulation 229, 244–45; institutions 241–44; politics and vested interests 239–41, 300–301; size of the local market 248; values and culture 236–39; see also innovation intellectual property rights 219, 297, 298, 329, 367; foreign ownership of domestic inventions 367 institutions 9, 154, 241–44; clusters 92–94; formal/informal 241; innovation 92–94, 241–44; labour market institutions 300–301; lock-in effect 94; non-profit institution 232; organised interests 9, 154, 241–44; path dependency 163; public institutions 85–86; types 93; uncertainty reduction 93–94 insurance market 158, 267, 382, 408 Intel 61, 91, 212, 221, 235, 302, 370–71; Costa Rica 370–71; human capital 176 international economic integration and location 172, 174, 175, 181, 184–208, 263; adjustment tools 197; agglomeration and spread forces 192–93; basic theory of 184–85; beneficial effects of 205; beneficiaries 201; clusters 188–91, 194; Cohesion Policy 175, 197–98, 206, 207, 213; comparative advantage 188; convergence 197, 198–99, 206, 213; core-periphery problems 179, 194–95, 212; divergence 198, 206; education 194; Europe 182, 189–91, 196–203; European Regional Development Fund (ERDF) 201; export-led growth 208;

Index

447

geographical concentration 195–96; geography of production 188, 195–96; globalisation 177, 208; industrial concentration 195; innovation 229; integration and location of production 195–96; Ireland 206; market-oriented approach 206; non-interference policy 188; non-linear effects 174; North America 203–5; North American Free Trade Agreement (NAFTA) 29, 186, 203, 204; potential gains 171, 185–86; production geography: reduction in trade barriers 193–96; production geography: with trade barriers 192–93; reallocation of resources and adjustment costs 186–88, 213; rules of origin 275, 280; Single European Market 29, 180, 190, 194, 195, 200, 202; Spain 194; spatial economics 195; economics 195; trade barriers 192; trade liberalisation 172, 192; transnational corporations (TNCs) 390; transport costs 192, 193; United States 188–89, 190–91, 193–94, 197; welfare 184–86; see also economic integration; Europe; European Union (EU); international firms; regional policy; trade liberalisation; United States international firms 306–406; benefits and concerns 388–97; foreign direct investment and transnational corporations 307–28; globalisation 328–48; international firm: theory 308–17; trans-border business activities 248–88; see also firms; international economic integration and location; international trade International Labour Organisation 331, 332 International Monetary Fund (IMF) 142, 331, 360; Argentina 331, 406; Brazil 331, 406; globalisation 332, 360, 400; instrument of global control 330–31;

Index

448

Latin America 331; policy interventions 31; Washington Consensus 389, 405–6, 415; World Economic Outlook 205; see also capital; World Bank; World Trade Organisation (WTO) international trade: essentials determining 17; flows 251; international trade and location of production 16–23; mobility of factors 179; neoclassical equilibrium economics 179; protectionist policies 17; Standard International Trade Classification (SITC) 251, 256; see also foreign direct investment (FDI); non-tariff barriers (NTBs); international firms; transnational corporations (TNCs) Internet 143, 145, 339–40; Amazon.com 139, 258; ‘borderless world’ 339; cities 116; control over cyberspace 340; dot-coms 159; globalisation 339–40; Google 133, 139; impact of 228; Microsoft 220; negative outcomes 339, 402; pre-Internet era 40; Wikipedia 139; YouTube 139; see also communication industry; electronics industry; software intra-industry trade (IIT) 18, 28, 152, 249–56, 379, 380; description 249, 303; developed countries 251; incentives to 252; inter-industry trade 250; intra-industry trade index 251–52; monopolistic/oligopolistic competition 254, 303; negative influences on 252; neoclassical equilibrium economics 250; perfect monopolistic competition 254; product differentiation 254; product differentiation/product specialisation 249, 250; Standard International Trade Classification (SITC) 251, 256; statistical aberration/real phenomenon 251 Iraq: Mesopotamia 108;

Index

449

Sumerian civilisation 108 Ireland 68, 194, 211; change in industrial structure 200; clusters 97, 106; Digital 97; European Union (EU) 194; Ferguson 77, 78; market-oriented approach 206; peripheral location 26, 175, 211, 213; Reform Treaty 402; spillover 358–59; tax system 321; transnational corporations (TNCs) 357–58, 382, 403 Islamic world 374; concentration of innovation 236, 300; confiscation 242; economies 243; globalisation 334, 337, 400; independent institutions 243; Islamisation 337, 338; knowledge and innovations 242, 243; religion 238, 300, 334; Turkey 334, 400–401; see also Arab world Italy 22, 106, 146, 190; accordions 421; aircraft industry 147; Alfa Romeo 148; Associazione Prosciutto di Parma 290, 305; banking 102; clusters 86, 94, 99, 100, 106, 113, 136–37, 421, 425, 429–30; competition 95; competition model 297; core location 175; cumulation of knowledge 245; electronic industry 78, 421; EU regional policy 209; family-run business 94; Ferrari 148–49, 158; FIAT 95; Fibonacci, Leonardo 241; Florence 102, 121, 166, 427, 428; footwear 113; Galileo Galilei 237; Genoa 102; Japanese cars in Europe 294, 305; Lancia 148; leather production 136–37; manufacturing 429–30; mirror/glass-making craft 233 –34; Modena 99, 425;

Index

450

motor car racing 148–49; Murano 233; patents 231, 233; Piaggio 147; Pirelli 404; plague 121; political reforms 301; Renaissance 427, 428; Rome 121, 166; Sassuolo 425; tax system 398; da Vinci, Leonardo 146, 241; Venice 102, 120, 166, 233; Vespa 147 Japan 21, 23, 145, 170, 208, 243, 273, 305, 354, 418; antidumping duties 285–86; automotive industry 293–94, 305, 313, 314, 398; Canon 304; comparative advantage 252; convergence 198; Czech Republic 397–98; electronics 145, 225, 234, 314, 398; export-led growth 208; foreign direct investment (FDI) 277, 347, 395; Germany 136; global competitiveness 170; government intervention 182, 354; hardware 145; herd behaviour 314, 397–98; Hiroshima 120–21; Honda 77, 140, 169, 414; intra-firm trade 310; Japanese cars in Europe 293–94, 305; labour content 352; microchips 283, 285; Mitsubishi 404; Nagasaki 120–21; Nissan 304; Organisation for Economic Cooperation and Development (OECD) 230; patents, patent families 230–31, 234; peripheral location 26, 175; Pioneer 61; Poitiers case 293; research and development (R&D) 230, 231, 299, 404; shipbuilding industry 212; Sony 61–62; Tokyo 120–21; Toshiba 61–62, 235, 366; Toyota 66, 95, 169, 252, 381–82, 395, 397; transnational corporations (TNCs) 170, 307, 317, 352, 395, 397–98, 404, 418

Index

451

Johnson-Krauss’ law 47 journals 36; Cambridge Journal of Economic Geography xv; Economic Journal 159, 419; Journal of Bioeconomics 13; Journal of Economic Geography xv, 8; Journal of Evolutionary Economic Geography 13 Kazakhstan 157 Kenya 156; transport costs 170 knowledge and experience 89, 108, 163; accumulated knowledge 151; ‘casual Fridays’ 91; cities 114; clusters 83, 88–92, 97, 114; creation of 90; cumulation of 244–45; developed personal networks 91, 164; educational policy 129; embodied/disembodied knowledge 89, 164; evolutionary economic geography 67–68; face-to-face contact 90–92, 95, 164; importance of 85, 89, 94, 95, 123, 140–41; informality 90, 91; information 88, 90; institutions 242; knowledge-based economy 231, 232; knowledge flows 163; knowledge externalities 95; knowledge spillovers 86, 89, 95, 114, 115, 164; linkages, externalities and spillovers 354–63; lock-in 67–68, 73; the most important resource 95; new knowledge 67–68, 73, 172, 219; non-market forms of transmission of 92; non rivalry of knowledge 96, 164, 244; personal contacts 90, 91–92, 164; ‘private good’ 91; production function 224; a returns asset 96; social potentials 125; spread of information and knowledge 86, 88–92, 95, 97, 112, 114, 163, 166, 168–69, 356; standardised knowledge 408; tacit knowledge 89, 90, 91; transnational corporations (TNCs) 392; see also education, learning, training; patents; research and development (R&D); technology Korea:

Index

452

government intervention 182; Hyundai 157, 381, 398, 405; North Korea 174; shipbuilding industry 212; South Korea 174, 208, 231, 337, 341, 415 Kuwait: crude oil 20 labour force 159, 413; capital-labour ratio 159; child labour 148, 337, 383; China 315, 353; cities 109–10, 111, 112; education 162; employment 375–76; Global Compact principles 384–85; Greece 353; illegal workers 291; Italy 353; Japan 352; low/high technology industries in countries at different stages of development 354–55; national endowment of factors 17; Portugal 353; rights 17; skills 40, 112; standards 283, 384; transnational corporations (TNCs) 352–54, 356, 375–76, 393; wages 352–53; see also employment; learning, training, wages labour mobility 172, 192, 213, 214, 262, 345, 413; brain drain 11, 143, 144, 231, 239, 300, 350; European Union 212; inter-regional labour mobility 174; regional policy 183–84; transnational corporations (TNCs) 358 land: land rents 23, 24, 81, 92; location 16, 17; national endowment of factors 16, 17; von Thünen’s model 23–24; see also agriculture Latin America 155, 239; Christianity 337; economic growth 204–5; International Monetary Fund (IMF) 331; North American Free Trade Agreement (NAFTA) 29, 186, 203, 204–5; policy interventions 31; the Portuguese in Latin America 155, 420; property rights 155; Washington Consensus 389, 405–6, 415

Index

453

learning, training 14, 48; Bavaria 126; cities 114; clusters 86–87, 94, 114, 123; cumulative causation 127, 167; domestic education 20; educational policy 129, 140–41; horizontal/oblique transmission 86, 163; importance of 85, 123, 140–41, 152; learning by doing 18; learning by watching 303; learning effects 18; the most important process 95; skills 40, 140–41; social potentials 125; transnational corporations (TNCs) 392; see also education, knowledge and experience; research and development (R&D); technology liberal trade 20, 80, 124, 182; clusters 106; reaction to 152; see also economic liberalism; trade liberalisation Lithuania: spillovers 403 lobbies, lobbying 207, 229, 368, 371, 392, 404; business lobbies 290; and corruption 30; European Union 368, 404; government 229, 418; industrial lobbies 276; lobbyist clusters 111; manufacturing lobbies 291, 294; political/professional pressure on medical/military technologies 56, 413; protectionists 265; transnational corporations (TNCs) 371 local market 138, 394; economic integration 22; globalisation 346, 402; research and development (R&D) 366, 367; size of the local market 248, 322, 346, 349; transnational corporations (TNCs) 328, 353, 357, 359; see also globalisation; transnational corporations (TNCs) location 14, 49, 150–53, 154, 167, 306, 408, 410; address model 50; basic concepts 10–16; biology, economics and evolution 12–16; business climate/people climate 117; cities 109, 112, 113, 165; comparative advantage and location of economic activities 354–55;

Index

454

competition 96, 99; concentration of specialised industries 133; determinants for FDI location 349–54; developed personal networks 91, 164; of economic activity 1, 125, 167; equation for the right location 11; equilibrium spatial location of activities 109; evolutionary dynamics 126–27, 411–12; of firms and industries xix, 2, 10, 16, 113, 136; first/second nature geography 112; Gini index 5; globalisation 33; government policy 134, 351, 352; history matters 49, 127, 133, 137–38, 199; home-biased procurement policies 28–29; location of cities 109; ‘location tournaments’ 111, 127; locational ‘catastrophe’ 54; lumpiness 49, 159; multiple equilibria 54, 127, 160; non-linear effects 174; patterns of 126; personal factors 426; physical conditions of the area 133–34; policy-related factors regarding 11, 134–36, 426; social factors regarding 11; spatial distribution of cities 26; taxes and location of firms 322; theories of location 16–52; unpredictability 54; of valuable classical painting 166; variables when considering the right location 11; see also clusters; education; history and competition among expectations; international economic integration and location; locational possibilities; market structure and location of production; spatial economics; theories of location; war and the location of firms locational possibilities 75–79; concave, convex and spherical distributions of potential industry location 75–76; see also location; special economics; theories of location Long Term Capital Management 159–60 look-in effect 15, 52, 64–71, 151, 155–56, 411; Belgium Walloon region 66; British rail wagons 161; British textile industry 162; causes 64;

Index

455

clusters 92; coexistence of various technologies (products) 68; development path 72; from water power to electricity 68, 70–71; General Electric 68; German Ruhr region 66, 67; historical lock-in effect 128–38; IBM 66; Imperial Chemical Industries (ICI) 65; India 66; innovation 66; institutions 94; lock-out 162; negative lock-in 162; new knowledge 67–68, 73; path dependence 65, 67, 92, 162; positive lock-in 67; positive/negative sides 65; preventing lock-in 60, 65; productivity/application curves of new technologies over time 68–70; QWERTY keyboard 60, 63–64; regional competitiveness 67; technological lock-in 65; Toyota 66; tyre-making 162; see also flexibility; history and competition among expectations; path dependency lumpiness 49, 53, 159 Luxembourg 192, 367; liquid fuel 321; tax system 324 Macintosh 61, 62, 63, 161, 272; see also software Malaysia 21 Malthus, Thomas: An Essay on the Principles of Population 12 Mantegna, Andrea: Opportunity and Penitence 78, 79 manufacturing 165, 204; allocation 172; ‘American System of Manufacturing’ 253–54, 303; China 23, 150, 299, 315, 348, 396, 408; distribution of manufacturing and transport costs/wages 172–74; Europe 253, 254, 266; home-biased procurement 28–29; India 348; Italy 429–30; lobbies 291, 294; reallocation of manufacturing 19–20, 21, 150, 170, 348, 408;

Index

456

United States 193–94 market: after-sales service 234; 300; characteristics in neoclassical equilibrium economics 32; competitive markets 298; distortions 17, 30; domestic markets 22–23; dominant market position 259–61; home market effect 22, 23, 156; imperfections 10, 30, 31, 32, 123, 181, 219, 225, 269, 319–20, 328, 354, 379, 413; market potential 21, 113, 114; market pre-emption 272, 351; self-reinforcing/snowball/herding effect 22; size of the local market 248; states 259–60; see also competition; market structure and location of production market structure and location of production 215–305; American competition model 216, 217, 297; antitrust policy 216, 221, 222, 261, 265, 269; appraisal 296–97; background 216–22; barriers to enter and exit an industry 269–70; benefits of competition 226–27; competition 215, 261–62; competition models 216–17; competition policy 215–16, 218–19, 234, 297; competitive markets 298; concentration of industries 217–18; concentration of innovation 236–48; ‘creative destruction’ 222; dominant market position 259–61; economic integration 226; effects of competition on innovation 226–27; firms 234–35, 260; France 217; free competition 225, 247, 265; free market 216; imperfect competition 7, 8, 53, 82, 125, 151, 194, 206, 247, 412; innovation 222–59, 60; innovators’ rights 232–36; intellectual property rights 219, 297, 298, 329; internal competition 265; intra-industry trade 249–56; licences 261; market pre-emption 272; mergers and acquisitions 219, 256–59; the Microsoft case 219–22, 298; monopoly 217–19, 234, 254, 260; neoclassical competition model 216–17; non-tariff barriers 282–96; perfect competition 30, 111, 163, 216–17, 222, 250, 254, 263, 267, 388;

Index

457

predatory pricing 270–72; price competition 235, 261; pricing policy 260–61; rules of origin 275–82; Sherman Act of 1890 218; specialisation and returns to scale 262–75; states 259–60; see also competition; innovation; innovation, concentration of; market; non-tariff barriers (NTBs); specialisation and returns to scale Marshall, A. 12, 42–43; clusters 105; concentration of specialised industries 133; external economies (backward-looking dynamics) 111, 113, 117, 133, 187; multiple equilibria 53; Principles of Economics 12; see also clusters Marx, Karl 38, 159; Communist Manifesto 337, 401; Das Capital 126 Massey Ferguson 77, 78 Mauritius 55; textile industry 55 maximisation 30, 44, 56, 256, 318; profit maximisation 10, 43, 349 mergers and acquisitions 219, 256–59, 303; benefits 256; comparisons between the US and European mergers and acquisitions 303; definition 256; disadvantages 256; dominant position in the market 260; European Union (EU) 256, 259, 303; failure 257–58; light/heavy diversification 258; takeover (merger and acquisition) 375–76; transnational corporations (TNCs) 375–76, 393; United States 256, 257, 258–59, 303; Uruguay Round (1994) 259; see also firms Mexico: Cemex 383; chocolate 422; Chrysler 351; clusters 86; Cortés, Hernán 64–65, 239, 422; electronics industry 361; food industry 278; globalisation 337;

Index

458

integration with United States 203–4; Mexico City 117; mineral resources 369; national economic core 29; North American Free Trade Agreement (NAFTA) 186, 203–5, 249, 278, 304; Tequila crisis 204; textile industry 304 Microsoft 61, 62, 107, 133, 143, 161, 235–36, 260, 414, 426; antitrust violations 219, 221; dominant firm 261; Encarta 139; European Commission 220–21, 298; Gates, Bill 62–63, 71–72; IBM 71–72; Media Player software 219, 220, 222, 298; the Microsoft case 219–22, 298; monopolistic behaviour 219, 220; MS-DOS 61, 71–72, 161; transfer pricing 364; Windows 161, 219, 220, 222, 235, 298; see also Gates, Bill; software Middle Ages 131, 133–34, 236, 241, 424; Barbarians 132, 168; Battle of Poitiers 132; Byzantine Empire 132; Charlemagne 132; Franks 132; medieval universities 242; Roman Europe 132; royal courts 133–34 military industry 305; European Union 305; political/professional pressure 56, 413; United States 149 Mittal Steel 169 MNEs (multinational enterprise) 91; face-to-face contact 90–91 mobility of factors 4, 19, 47, 109, 113, 174, and agglomeration tendencies 126, 165; immobile factors 192; international mobility of factors 44, 179, 222, 402 monopoly 217–19, 222–23, 234, 260, 270, 412; American local cable monopoly 217; American local phone monopoly 217; geographical monopoly 76; innovation 260; intra-industry trade 254, 255, 303; market characteristics in neoclassical equilibrium economics 32; monopolistic competition 22, 254, 303; ‘natural’ monopolies 261; oligopolies 32, 222–23, 234, 254, 303, 315;

Index

459

predatory pricing 270–72; returns to scale 268; Sherman Act of 1890 218; sources 18; transnational corporations (TNCs) 310, 310, 369, 372; see also market structure and location of production Moore, Gordon 246, 302 Morton, John 54; Morton’s fork 54 most favoured nation (MFN) 279, 385, 397 Müller, Anton 241, 296, 301 multiple equilibria 21, 29, 151, 411, 417; agglomeration economies 22; centripetal/centrifugal forces 80–82, 85, 151; clusters 85; complex systems 53–56; coordination failure 54, 55, 160–61; evolutionary economic geography 48, 55; expectations 54; and history 53; industrialisation 160; location 127, 137; location theory 160; Marshall, A. 53; regional disequilibria 123; time delays 73; see also equilibrium neighbouring effect 14, 155 neoclassical equilibrium economics xvii, 30–34, 52, 55, 178, 378–79, 409–10, 412, 413, 417; adjustment costs 46; assumptions 44, 55, 63; capital-labour ratio 47; cities 107–8; closed equilibrium system 38, 47; comparative advantage 267; competition among locations to attract firms and industries 45; competition model 216–17; competition on the market for goods and services 45; convergence school 124–25; coordination failure 47; creation of clusters 101; criticism xvi, 33, 39–40, 49, 152; differences between evolutionary economic geography and 2, 4–5, 31, 32, 43–48, 49–50, 151, 174, 410, 412; economic policy 46, 49, 50; economies of scale 268; effectiveness 46; endogenous growth theory 31; entropy 39; factor endowment (Heckscher–Ohlin model) 17–18;

Index

460

geography and economics 7, 8; globalisation 329; innovation 222; intra-industry trade 250; Johnson-Krauss’ law 47; market characteristics in 32; micro/macroeconomic planes of the economy 45; models xviii; necessary adjustments 40; neoclassical equilibrium 12, 32; non-linear dynamic system 53; ‘old analytical box’ xvii, xviii, 9; openness to ideas outside economics and mathematics 47; perfect competition 263; physical concepts 32; physics and thermodynamics 38, 47; policy 30–31, 33; political/professional pressure 56, 413; principles 30; profit maximisation 43; regions 124; returns to scale 123, 263; size of firms 46; spatial location 10; specialisation 45, 267; static analysis 167; tax principles 319; theoretical dogma 410; thermodynamics, first law 37; time delays 73; trade barriers 45; trade pattern 46; transnational corporations (TNCs) 351; treatment of mind in choice-making 47; uncertainty 43; Washington Consensus 389, 405–6, 415; welfare effect of trade 46; see also equilibrium; evolutionary economic geography; QWERTY keyboard Netherlands: AKZO 271; Amsterdam 102, 120; banking 102; intra-firm trade 310; patents, patent families 230–31; Phillips 61, 245–46; research and development (R&D) 299–300; Royal Dutch Shell 156; taxes 322; traditional regional policy 202; transnational corporations (TNCs) 317;

Index

461

VCR 245 networks 137, 156, 272, 282, 308, 335, 368; clusters 95; computer networks 340; cumulative causation 127, 167; marketing networks 272, 346, 350, 351, 376, 389; MNC networks 404; network effects 18; service networks 158; social networks 91; spatial evolution xvi; trade networks 308 neuroeconomics 13, 47 new economy 143–44; information economy 144; productivity growth 144, 169; see also high-technology industry; software new economic geography xiv, 8, 154, 195; differences between evolutionary economic geography and xv; principal workhorse 109 New Zealand 278; peripheral location 26 Nigeria 142, 397 Nobel Prize 154, 159; Bohr, Niels 407; Einstein, Albert 232; Gell-Mann, Murray 36; Prigogine, Ilya 36 non-governmental organisations (NGOs) 331–32, 400; Global Compact principles 385; India 401; lack of accountability 332, 400; transnational corporations (TNCs) 331–32 non-linear dynamics (‘chaos theory’) 15, 38, 47, 53, 137, 417; complex systems 36, 53; evolutionary concepts 14; Holland, John 38; non-linear probability theory 128 non-tariff barriers (NTBs) 29, 282–96, 307; antidumping 283–88; Associazione Prosciutto di Parma 290, 305; Cassis de Dijon 289, 290; classification 283, 284; criticism 283; definition 282; elimination 282; European Union (EU) 263, 274, 396; government 282; Japanese cars in Europe 293–94; Poitiers case 293; precautionary and prevention principles 294–96, 305;

Index

462

public procurement 288–89; regulatory red tape 293; tariffs 282; technical standards 289–92; trade liberalisation 282; trade-related investment measures (TRIM) 296; Treaty of Rome 282; vested interests 282; see also antidumping; free trade; market structure and location of productoin; Single European Market; tariffs North American Free Trade Agreement (NAFTA) 29, 249; Canada 186, 203; Latin America 204; Mexico 186, 203, 204, 304; rules of origin 277–78, 279–80, 304; strategic technology alliances 368; United States 186, 203, 204; US-Canada Auto Pact 249; Western Hemisphere Free Trade Agreement 204; see also economic integration; free trade oil 142, 315–16, 321, 369; Angora 397; China 348–49, 386, 403; Kuwait 20; liquid fuel 321; Nigeria 397; Shell 315; Texaco 315 ‘old analytical box’ xvii, xviii, 9 optimality 357; absence of 56–64; principle of optimality 413; uncertainty 56 Organisation for Economic Cooperation and Development (OECD) 23, 205; Declaration on International Investment and Multinational Enterprises 385; Guidelines for Multinational Enterprises 385; Multilateral Agreement on Investment (MAI) 331, 341, 387–88, 396; research and development (R&D) 230 organisations: Church 168, 223–24, 299; firsts ‘modern’ corporation-like organisation 168, 299; ‘soft organisation’ 227, 299 organised interests 9, 154, 241–44; see also institutions painting 166;

Index

463

location of valuable classical painting 166; Florence 166, 427, 428; Mantegna, Andrea: Opportunity and Penitence 78, 79; see also art Pakistan 21; clusters 84; Sialkot Dry Port 148; sports goods 84, 147–48; surgical instruments 84, 147–48 patents 163, 230, 284, 297; appropriations 232; definition 300; foreign ownership of domestic inventions 367; France 233–34; individual creativity 302; innovators’ rights 224, 232–36; intellectual property rights 219, 297, 298, 329, 367; invalidation of 147, 424; Italy 233–34; Japan 234; law of 1844 423; lead-time advantages over competitors 232, 300; non-profit-making innovators 232; patent families 230–31; R&D 205, 366; secrecy about innovation 233; see also innovation path-dependency 30, 162, 411; clusters 92; cumulative causation 127, 167; institutions 163; lock-in 65, 67, 92; path dependence, strong history and non-ergodicity 71–73; regional development 179–80; sources of 73, 163; unresolved issues associated with 73, 74; see also development path; lock-in effect Penn World Tables 205 Peru 204 pharmaceutical industry 255, 329, 367, 368; Novartis 141, 367; patents 231; Pfizer 227–28; research and development (R&D) 366, 367, 368; Sulzer 23, 107; Switzerland 23, 107, 424; Viagra 227–28 Philippines 20 physics and economics 12, 34–41; closed/open system 37–38, 47;

Index

464

entropy 37, 38–39; Holland, John 38; Newton, Isaac 32, 41, 42, 150, 238, 240, 241, 417; non-linear physics and positive feedback economics 161; Pareto, Wilfredo 158; thermodynamics laws 37, 39; Walras, Léon 34, 158; see also evolutionary economic geography Plato 109; The Laws 109 Poland 157 policy xvi, 39, 50–52, 153, 154, 212, 413, 414–16, 426; antitrust policy 216, 221, 222, 261; background ideology in US policy-making 134–36; challenges 357; China 414–15; cluster policy 99–105; competition policy 215–16; economies of scale 269; educational policy 129; European Union (EU) 261; evolutionary economic geography xvi, 31, 33, 46, 49, 50; industrial 261; intervention 174, 351, 354, 377–82, 388, 415–16; liberal policy 46; liberalisation 346; liberalism/controlled interventionism 339; local content 377; national and international regulation 385–88; national regional policy 76, 176; neoclassical equilibrium economics 30–31, 46, 49, 50; non-retaliation policy 212; policy imperialism 33; policy interventions 30, 31, 33, 46, 49, 50, 124, 152, 177; policy-related factors regarding location 11; politics and vested interests 239–41, 300–301; pricing policy 260–61; protectionist policies 17, 20, 21, 156; skilful behaviour 51; spillovers and externalities 357, 362–63; subsidies 11, 55, 76, 100, 166, 175, 176, 180, 302, 303, 379, 380–81; tax competition 320–28; transnational corporations (TNCs) 329, 331, 332, 336–37, 397; United States 134, 261; Washington Consensus 389, 405–6, 415; see also government; regional policy Portugal 213; Cohesion Fund 213; European Union (EU) 194; peripheral location 175;

Index

465

the Portuguese in Latin America 420 predictability 37; see also entropy; unpredictability prices 261; antidumping duties 271; internal prices 403; intra-company pricing 403; manipulation in 363–64; predatory pricing 270–72, 286; predatory pricing/vigorous price competition 271; price competition 235, 261; pricing policy 260–61; retaliation against predatory pricing 271–72; taxes 364; transfer pricing 363–65, 403; see also transnational corporations (TNCs) private market-based firms 154 production: clustering 22; cost 24; economies of scale 18; innovation 106, 107, 409; locations patterns 52; national endowment of primary factors of 16–18; production function 224; productivity 46, 92, 143, 169, 191, 224, 359; productivity/application curves of new technologies over time 68–70; structure of production 16, 17, 145, 152, 185; see also clusters; education; history and competition among expectations; international economic integration and location; locational possibilities; market structure and location of production; spatial economics; theories of location; war and the location of firms profits 32, 247, 364, 365, 378–79, 387, 413; after-sales service 234; 300; cost minimisation 10, 43; globalisation 337; growth rate 257; monopoly profits 380; policy making process 271; profit maximisation 10, 43, 53, 309–10, 349; profitability 87, 92, 129, 177, 256, 267, 291, 371, 426; spatial location 10; supernormal profits (economic rents) 32, 158, 229, 247, 260, 377, 380, 389; transnational corporations (TNCs) 337, 403 property rights 155; foreign investment 402

Index

466

psychology 13, 47; ‘prospect theory’ 227 Puerto Rico 205 pulp and paper industry: Finland 98; Sweden 168 ‘quantitative revolution’ 7 quotas 55, 307; elimination of 286; export of textiles 20; import quotas 20–21; ‘quota jumping’ 55; Treaty of Rome 282 QWERTY keyboard 60, 63–64, 73, 410; Dvorak keyboard 60; lock-in effect 60; see also lock-in effect; neoclassical equilibrium economics raw materials 11, 21, 24; distribution 27; and population 27; see also resources red tape 204, 293, 350; see also corruption regional competitiveness 67, 128; lock-in effect 162 regional policy 99, 123, 171–214, 212; aim 174–75, 209; allocation of resources 180; assessment 208–12; background 171–77; beneficial effects of 181; challenges 206–8; comparison between the US and European spatial structure of industry 190–91; compensation 179–80; competition 172; contemporary regional policy 202–3; convergence 123–26, 172, 197, 198–99, 205, 206, 213; core-periphery geography 174, 175, 179, 194–95, 212; disincentives for regional expansion 183; distribution of manufacturing and transport costs/wages 172–74; divergence 123–26, 172, 198, 206; economic development 178; economic integration 172, 174, 177, 208–12; employment of factors 179; European Union 176–77, 201–3, 209; export-led growth 208; globalisation 177, 208; goals 182;

Index

467

government intervention 182, 209; human capital 176; impact of international economic integration on location 172, 174, 175, 184–208; instruments 182–84; intervention 124, 177, 178, 182; Ireland 206; labour mobility 172, 174, 183–84; market-oriented approach 206; non-interference policy 188; objectives and justification 177–82; patterns of regional specialisation 126, 171; political grounds 181; public expenditure 181; reasons for government intervention 175; regional integration 338–39; regional problem 122–23, 175, 183, 201, 203; regionalisation 330, 340; regionalisation/integration 347; regionalism/multilateralism 338–39; stabilisation 180–81; subsidies 175, 176, 180, 183; success of regional policy 203; trade liberalisation 172, 174, 177; traditional regional policy 202; wages 172–73, 177–78; see also European Union; international economic integration and location; regions; trade liberalisation; United States regions 80, 121–26; Bavaria 125–26; convergence and divergence 123–26, 172, 197, 198–99, 213; core region 80, 82, 175; core-periphery relations 179, 194–95, 212, 323–24, 398; definition 121–22, 167; neoclassical equilibrium model 124; openness 121, 167; periphery region 80–81, 82, 124, 175, 213; regional competitiveness 67, 128, 162; regional disequilibria 122, 123; regional ‘imports’ 177–78; regional integration 338–39; regional policy 99, 123; regional problem 122–23, 175, 177, 183, 201, 203; regional specialisation 80; regionalisation 330, 340; regionalisation/integration 347; regionalism/multilateralism 338–39; theoretical types of 123; Zollverein 125–26;

Index

468

see also agglomerations; cities; clusters; regional policy Renaissance 427, 428 religion: China 238; impact on economic prosperity 373–74; Muslim world 238; Protestant countries 373–74; religion/science debate 237–38; Roman Catholic countries 373; see also Church; science rents 32, 174, 184, 190, 225, 272, 300; cities/countryside 108, 111, 117; land rents 23, 24, 81, 92; supernormal profits (economic rents) 32, 158, 229, 247, 260, 377, 380, 389; von Thünen’s model 24 research and development (R&D) 410; Cambridge 99; determination of origin of goods 285; European Union (EU)230, 231, 299–300; foreign funding 299–300; foreign ownership of domestic inventions 367; geographical spread of R&D activities carried out by TNCs 367; government involvement 299; headquarters effects 365–66, 404; incentives to 368; innovation 164, 222, 225, 243, 299; internationalisation 366–67; Japan 230, 231, 299, 404; local market 366; orientation of 299; patents 232; R&D behaviour 76; Silicon Valley (California) 84, 98, 99, 100, 102, 104, 114–15; strategic technology alliances 368; subsidies 368; Switzerland 424; technology transfers 368; transnational corporations (TNCs) 365–68, 392; trends in R&D intensity by area as a percentage of GDP 230–31; United States 230, 231, 299; Washington DC 102; see also education, knowledge and experience; learning, training; patents; technology resources 22, 24, 138, 142, 174, 225, 248, 290, 350, 409; allocation of resources 180, 209, 215, 234, 323, 370, 389; immobile resources 177;

Index

469

mineral resources 369; mobile resources 177; natural resources 26, 27, 111, 141, 142, 165, 253, 321, 348, 359, 389, 405; non-renewable 225; primary resources 408; raw materials 11, 21, 24, 27; reallocation of resources 81, 177, 179, 188, 213, 267, 271; reallocation of resources and adjustment costs 186–88; resources-rich countries 385, 405; sources of supply 21, 27, 166, 188; tangible resources 48, 388 retirement patterns 11, 155 returns to scale xix, 8, 17, 18, 23, 53, 100, 171, 304; agglomeration 111, 192; constant, increasing and decreasing returns to scale 263–64; cost curves with constant, increasing and decreasing returns to scale 263–64; homogeneous production functions 263; i-activities 56; increasing returns 18, 22, 23, 52, 53, 76, 108, 263–64, 268; knowledge: a returns asset 96; meaning 262; monopolies 268; neoclassical equilibrium economics 123, 263; non-ergodic system 71; non-linear probability theory 128; plant efficiency 265–66; positive feedback economics 161; production function 224; returns from innovation 232–33; services 267; standards, standardisation 273; studies on 267–68; technologies with constant, increasing and decreasing returns to scale 263– 64; see also economies of scale; specialisation and returns to scale Rhodes 320 Ricardo, David: comparative advantage model 8, 17, 24, 55, 151, 193 risk 36, 37, 145, 158, 159, 416–17; innovation 227; insurable risk 30, 44 Romania 113, 181 rubber: rubber-based products 21 rules of origin 216, 275–82, 284, 287, 304; British Imperial preference scheme 279; controversies 282; criticism 278, 280, 282; determination of origin of goods 285; EU non-preferential rules of origin 283, 285; EU rules of origin 281;

Index

470

full cumulation system 287; globalisation 282; liberal/restrictive 277, 280; methods for the determination of origin of goods 276–77; Pan-European Cumulation System (PANEURO) 287, 304; protectionist intentions 276, 280; purposes 275–76; trade-diverting effect 277; trade in services 282; value-added rule 278, 281, 304; see also non-tariff barriers (NTBs); tariffs Russia 72, 98–99, 155, 169–70, 355; global economy 159; globalisation 337, 403; October Revolution 132; property rights 402; research and development (R&D) 367; Soviet Academy of Science 98; spillovers 403; universities 98; largest exporter of cut logs 372 Scandinavian countries: competition model 217, 297 science 242; Bohr, Niels 407; Einstein, Albert 232; Fibonacci, Leonardo 241; Gell-Mann, Murray 36; humanistic critique of 238; Newton, Isaac 32, 41, 42, 150, 238, 240, 241, 242, 417; Prigogine, Ilya 36; religion/science debate 237–38; da Vinci, Leonardo 146, 241; see also biology and economics; Church; evolution; physics and economics; scientific law scientific law 159; entropy 37, 159; thermodynamics, first law 37; thermodynamics, second law 37 Scotland: change in industrial structure 200 Second World War 77, 147; effect on cities 118, 120–21, 166; instruments of global control 330–31; see also war and the location of firms self-organisation 14, 155, 168;

Index

471

biological self-organisation 155; economic self-organisation 155 services 31, 44, 45, 83, 87, 148, 155, 208, 412, 419; after-sales services 45, 234, 250, 260, 300, 354; financial services 82, 103, 116, 189, 208, 375, 392, 395, 404; public services 111, 122, 175, 352, 403; services jobs 348; services sector 196, 229, 245, 267, 345; returns to scale 267; trade in services 403; transnational corporations (TNCs) 351, 403 Shakespeare, William 79 Siemens 96, 404; trade unions 136 Singapore 116, 314; Economic Development Board of Singapore 363; financial services 103; Flextronics 315; Seagate 314; transport costs 170 Single European Market 29, 180, 261, 274, 296, 307, 396; creation 247, 262, 396; economic integration 266; effects 201, 202, 248; elimination of NTB 194, 396; eurozone 194, 211, 346, 396; goal 247; home-biased procurement policies 29; impact on FDI 396; merges and acquisitions 259; mutual recognition and standardisation 290, 291; non-tariff barriers (NTBs) 274; outcomes 201, 202; regional disequilibria 202; Single Market Programme (1985–92) 180, 190, 195, 261, 263, 266, 282; see also economic integration; Europe; European Union; non-tariff barriers (NTBs) skills 129, 141; educational policy 129, 140–41; skilful behaviour 51; see also education; learning, training; knowledge and experience Slovakia 157, 398; Hyundai (KIA) 157, 398; Peugeot-Citroën 157, 398; Volkswagen 157, 398 SMEs (small and medium-size enterprise) 11, 96, 106, 210, 288, 315, 316, 398; aid for 180; see also clusters

Index

472

social issues 283; corporate social responsibility 383–85; economic growth and society 223; non-tariff barriers (NTBs) 283; regarding location 11; social capital 91; social costs 180, 190; social institutions 236; social needs as incentive for innovation 223; 298; social networks 91; social policy 211; social potentials 125; social progress 132; social security system 428; social technologies 16, 156, 216; societal values 82; societies of obedience/of will 236 software 65, 71–72, 107, 145, 169, 235–36, 261, 313, 354, 368; Bangalore 208, 213; Dell Computer Corporation 61, 77–78, 235, 315, 414; Digital 97; India 208; Linux 244; Silicon Valley 107; see also communication industry; electronics industry; IBM; Internet; Macintosh; Microsoft; new economy Somalia 26 Sony 61–62, 272, 315; Betamax 61, 161, 245–47; Blu-ray (DVD) 61; Nintendo Game Boy 247, 260; PlayStations 61, 260; Walkman 247 South Korea 174, 208, 231, 337, 341, 415; Hyundai 157, 381, 398, 405 Spain 22, 157, 175, 213; Cohesion Fund 213; education and regional policies 194; European Union (EU) 194; Madrid 166; periphery region 213; retirement patterns 155; shipbuilding industry 212; spillover 359; transnational corporations (TNCs) 317 spatial economics 16, 17, 19, 125, 150, 151, 152; academic field 6–7;

Index

473

cities 112, 113–14; divergence school 125–26; growth poles 28; industries 195; location of cities, clusters and industries 112, 195 (first/second nature geography 112); new research tools 8–9; spatial distribution of cities 26; see also economic geography; economics and geography; geography of production; location spatial units 79–126; cities 107–21; clusters 80–107; regions 121–26; see also agglomerations; cities; clusters; regions specialisation 4, 18, 21; bottom-up specialisation 152, 415; cities 114–15, 166; clusters 81, 92, 101, 107; economies of scale 266, 267, 269; European Union (EU) 199, 200; evolutionary economic geography 45; location 126, 133; lock-in effect 66; Marshall, A. 133; neoclassical equilibrium economics 45, 267; patterns of regional specialisation 126, 171; polarisation 81; regional development 179–80; and returns to scale 262–75; rigid specialisation 411; specialisation/diversification 115–16, 166; Switzerland 23, 147; United States 191, 194 specialisation and returns to scale 262–75, 304; barriers to enter and exit an industry 269–70; economies of scale 266, 267; meaning 262–69; neoclassical equilibrium economics 267; plant efficiency 265–66; predatory pricing 270–72; specialised/flexible plants 166; standards 272–75; see also economies of scale; market structure and location of production; returns to scale spillovers 356, 354–63, 393, 419;

Index effects 361–62, 403; industry-specific channels 357–58; Jacobs-type (across industries) 115, 357; knowledge spillovers 86, 89, 95, 114, 115, 164; linkages, externalities and spillovers 354–63; Marshal-Arrow-Romer-type 115, 357; policy regarding 362–63 sponsorship 56, 383 standards, standardisation 272–75, 284, 290; beginnings 272; definition 272; European Union 291–92, 305; global standards 340–42; globalisation 333; hidden barriers to trade 289; increasing returns to scale 273; mutual recognition and standardisation 290, 291; national standards 274, 284, 289, 290, 292; tariffs 289; technical standards 289–92; westerns standards 340–41; see also economic integration; economies of scale; Single European Market Sun Tzu 235, 300 sunk costs 2, 11, 18, 53, 158; competition 235; cumulative causation 127, 167; regional policy 179; see also costs Sweden 129, 253; farming-based economy 129; intra-firm trade 310; knowledge-based economy 162; patents, patent families 230–31; periphery region 213; pulp and paper industry 168; tax system 346–47 Switzerland 22, 140–41, 273, 423–24; communication industry 218–19; banking 141, 398–99, 423; Basle 423–24; chemical industry 147, 424; clusters 423–25; competition model 217; food industry 259; Geneva 424–25; human capital 424; Jura 425; liquid fuel 321; Nestlé 141, 364, 384; Novartis 141, 367;

474

Index

475

Paracelsus 424; patents, patent families 230–31; peripheral location 26; pharmaceutical industry 424; production 23; research and development (R&D) 23, 424; specialisation 23; Sulzer 23, 107; tax competition 231; tax system 324–25, 398–99; textile industry 423–24; transnational corporations (TNCs) 321; transport costs 152; university 300, 423 systems stable 80; unstable 80; see also complex systems Taiwan: electronics industry 361; export-led growth 208 Tanzania: transport costs 170 tariff: antidumping 283–88; difference between technical standards and tariffs 289; elimination of 286; on final goods 280; rules of origin 275–82; tariff discrimination 276; ‘tariff factories’ 280, 307; see also antidumping; customs unions; non-tariff barriers (NTBs) taxes 150, 242; adjustment tools 197; Cohesion Policy 207; concessions 183; controversies 325–28; core/periphery areas 323–24, 398; corporate income taxes 320–21, 325, 328; corruption 327, 399; Delos 320; European Union 319; evolutionary economic geography 319–20; fiscal frontiers 318; flat tax 326; general equilibrium model 319; globalisation 364; income taxes 319, 320, 326, 371, 398;

Index

476

liabilities 183; location of firms 322, 326–28, 346–47; Morton’s fork 54; principle of destination 317, 318, 319, 320; principle of origin 317–18, 319, 320; Rhodes 320; Switzerland 321; tax avoidance 364, 394, 403; tax collection 319; tax competition 320–28, 347, 398; tax evasion 318–19, 387, 398, 403; tax harmonisation 323; tax havens 320, 325, 398; tax inspection 364–65; tax revenue 319, 394; tax treaties 387, 405; taxes and fiscal incentives 314, 317–20, 326–28, 352; taxes levied 318, 319; on trade 156; transfer pricing 363–65, 387, 394, 403; transnational corporations (TNCs) 364–65, 394, 403; United States 197, 319; value added tax (VAT) 304, 318, 319, 364; see also policy; regional policy technology 161, 167, 272; coexistence of various technologies (products) 68; economies of scale 266; environmentally friendly technology 78–79, 377; general-purpose technology (GPT) 143, 169, 228, 230; globalisation 329, 333, 342; ‘hard’ physical technologies 16; hierarchy 68; human capital 245; levels of innovation 228, 299; mature technology 272; new technologies 143, 179, 228, 229, 333, 411; physical technologies 16; productivity/application curves of new technologies over time 68–70; protectionist policies 156; sacrifices 265; social technologies 16; ‘soft’ technologies 16; linkages, externalities and spillovers 354–63; spread of information 166; strategic technology alliances 368; technological changes 13, 18, 48, 92, 94, 123, 131, 135, 138, 177, 187, 228, 241, 253, 262, 265, 299, 302, 343, 413; technological changes, phases of 224; technological progress 244, 299, 302; technological revolution 228, 229, 299; technology transfers 368, 369, 415;

Index

477

transnational corporations (TNCs) 310, 311, 316–17, 351, 356, 392; see also education; innovation; knowledge and experience; learning, training; patents; research and development (R&D) textile industry 28; Bangladesh 359–60; bikinis 170; Bra-making 116; China 19, 20, 21, 55, 293; coordination failure 55; domestic jobs 20; England 84; global textile market 55; Mauritius 55; Multifibre Agreements 21; multiple equilibria 55; New York 116; non-tariff barriers 293; reallocation of manufacturing 19, 20; rules of origin 277–78; Second World War 170; Switzerland 423–24; Toyota 169; United States 116, 170 Thailand 21; globalisation 337; Thai baht devaluation 71 theories of location 16–52; domestic firms 23–30 (Christaller’s model or central place theory 26–27, 156, 157; growth poles 28; home-biased procurement 28–29; Moses-Weber’s model 25; von Thünen’s model 23–24; Weber’s model 24–25); evolutionary strand in economics 41–52; international trade and location of production 16–23 (economies of scale 22–23; factor endowment model or Heckscher–Ohlin model 17–18, 19; home market model 22–23, 156); neoclassical equilibrium economics 30–34; physics and economics 34–41; see also evolutionary economic geography; neoclassical equilibrium economics Thünen, Johann Heinrich von 9, 23; location of agricultural production 23–24; von Thünen’s model 24

Index

478

time delays 53, 73–75; multiple equilibria 73; neoclassical equilibrium economics 73 tipping point 29, 158; see also events Toshiba 61–62, 235, 366; HD-DVD 61 Toyota 66, 95, 169, 235; Canada 381–82; suppliers 95; taxes 364; training 235 trade: barriers to trade 282, 286, 289, 348; data 200–201; globalisation 402; international trade and location of production 16–23; intra-industry trade (IIT) 18, 28, 152, 249–56; liberal trade 20, 80; location 16; Ministry of Trade and Industry (MITI) 293–94; non-tariff barriers (NTBs) 282; patterns 18, 46; regulations 165; trade diversion 185; trade theory 21, 153; transnational corporations (TNCs) 356; welfare effect of trade 46; welfare in a trade-diverting customs union 184–85; see also free trade; international trade; intra-industry trade (IIT); liberal trade; non-tariff barriers (NTBs); trade barriers trade costs 4, 10, 150, 154, 195; trade costs matter 408–9; see also costs trade barriers 10, 192, 249, 282, 286, 289, 348; entry barriers 167, 269–70, 390; Europe 189; evolutionary economic geography 45; exit barriers 167, 269–70; neoclassical equilibrium economics 45; production geography: reduction in trade barriers 193–96; production geography: with trade barriers 192–93; spatial barriers 85; see also barriers; General Agreement on Tariffs and Trade (GATT); non-tariff barriers (NTBs)

Index

479

trade liberalisation 172, 192; economic adjustment 177, 187–88; economies of scale 329; effects on regional development 179–80; European Union 250; Foreign Direct Investment (FDI) 280; Mexico 204; non-linear effects 174; non-tariff barriers (NTBs) 282–83; see also economic liberalism; international economic integration and location; liberal trade trade-related investment measures (TRIMs) 296, 377, 395, 396; Agreement on TRIMs 296, 377; WTO 1995 TRIMs Agreement 377, 404; see also foreign direct investment (FDI); transnational corporations (TNCs) trade unions 134, 162, 223–24, 267; DaimerChrysler 136; political reforms 301; Siemens 136 trans-border business activities 248–88; corporate social responsibility 383–85; culture matters 350, 372–74; determinants 349–54; host countries and transnational corporations 369–72, 389–90, 391; intervention 377–82; linkages, externalities and spillovers 354–63, 393; local content 377; market entry 375–76; national and international regulation 385–88; organisation 375; research and development 365–68; theoretical aspects 348–49; transfer pricing 363–65; wages 376–77; see also foreign direct investment (FDI); globalisation; transnational corporations (TNCs) transnational corporations (TNCs) 16, 100, 127, 156, 170, 400; advertising 341, 402; availability of technology 316–17; benefits and concerns 388–97; capital mobility 308; clusters 310; companies 349; competition 357, 394; cost minimisation 316; definition 308; determinants for FDI location 349–54; developing countries 313–14, 369–72, 382, 389, 406;

Index

480

developed/developing countries 309, 313, 327, 369, 370, 385; differences between national firms and TNCs 308–9; environmental standards 351, 397, 403; export-oriented strategy 311; financial service 375, 404; free economic zones 371; geographical origin 352; globalisation 329, 331, 332, 336–37, 346, 400; government 312, 329, 332, 351, 352, 397; the greatest power of 369, 404; greenfield investment 375–76, 392, 393; and the growth process 356–57; herd behaviour 314–15, 397–98; host countries and transnational corporations 369–72, 389–90, 391; import substitution strategy 311; incentives for 390–91; instrument of global control 330; intangible assets 311–12; internal trade within 171–72, 394; international investment treaties 387, 405; intra-firm trade 309–11, 394; Ireland 194; labour force 352–54, 356, 375–76, 393; linkages, externalities and spillovers 354–63; local market 328, 353, 357, 359; local/foreign firms 353–54, 403; location of firms and taxes 322, 326–28, 346–47; monopoly 310, 369, 372; national and international regulation 385–88; neoclassical equilibrium economics 351; non-governmental organisations (NGOs) 331; OLI factors 312–13; ‘organisational capital’ 309; product-cycle model 313–14; risks 397; Royal Dutch Shell 156; Sub-Saharan Africa 397; Switzerland 321; takeover (merger and acquisition) 375–76, 386; taxes 364–65, 394, 403; tax competition 320–28, 347, 398; tax controversies 325–28; tax inspection 364–65; taxes and fiscal incentives 317–20, 326–28, 394; technology 310, 311, 351; technology transfers 368, 369, 404; TNCs theories 309–17, 328, 397–98 (competitive international industry model/exchange of threats 315–16; diversification of portfolios model 316; eclectic model 312–13; firm-specific assets and exchange rates 316;

Index

481

follow-your-leader model 314–15; general equilibrium trade model 311–12; internalisation model 309–11; market-power model 309; strong currency model 316); transfer pricing 363–65, 394, 403; transfer of profits 403; wages 352–53; Washington Consensus 389, 405–6, 415; see also foreign direct investment (FDI); globalisation; international trade; trans-border business activities transport costs 25, 97, 124, 152, 163, 164, 167, 170, 192, 193, 335, 402, 409; central-place model 27; core region 156; distribution of manufacturing 172–74; equilibria 173; factor endowment model 18; geographical remoteness 25–26, 156; globalisation 97, 335, 402; home market model 23; just-in-time delivery 95–96; landlocked/coastal countries 152, 170; peripheral region 156; von Thünen’s model 24; Weber’s model 24–25; see also costs Treaty of Rome 187, 282, 292, 294, 305 Turkey 334, 400–401; Adrianople 146; integration with the European Union 204; Turkish occupation (1529–1683) 149 uncertainty 32, 38, 56, 145, 151, 158, 159, 416–17; evolutionary economic geography theory 43, 46, 48, 56; innovation 222, 227; institutions and uncertainty reduction 93–94; neoclassical equilibrium economics 43, 56; serendipity 71, 73, 227–28, 299, 410; see also risks unemployment 180, 183, 198; education and training 184; high levels of 81, 99, 123; rates 184; regional policy 180, 182, 183; see also employment United Nations (UN) 331, 332; Global Compact 384–85 United States 19, 23, 68, 97, 145, 156, 273; 3M 227;

Index

482

adjustment tools 197; aircraft 146–47, 230, 246, 302, 303, 367, 378; ‘American System of Manufacturing’ 253–54, 303; antidumping duties 285, 288; antitrust policy 265; AT&T 62, 218; automotive industry 190, 249; Auto-Ped 170; banking 102; Boeing 146–47, 230, 246, 303, 367, 378; Boston 131; bra-making 116; Brady, Nicholas 142; Byrd Amendment 288; Canada-US Free Trade Area 249; Caterpillar 147; China 19; Chrysler 32, 249, 257, 351; city sizes 117–18; clusters 84, 98, 106, 188–89; Coca-Cola 147, 258, 337, 342, 383, 400, 401, 402; Committee on Foreign Investment in the United States (CFIUS) 386; communication industry 217, 218; comparative advantage 252; comparison between the US and European mergers and acquisitions 303; comparison between the US and European performance of firms 253, 303; comparison between the US and European spatial structure of industry 190–91; competition model 216, 217, 265, 297; concentration of industries 218; convergence 198; cotton textile industry 162; cumulation of knowledge 245; DaimlerChrysler 136; Department of Commerce 285; Department of Justice 220, 222, 298; Detroit 189, 190; economic geography 82, 188–89, 190; economic growth 144–45, 189; economic integration 194; ‘economic patriotism’ 386–87, 389; education 212; electronics industry 66, 100–101, 314; energy 144–45; Enron 103, 248, 258, 394, 399; European export of capital 355–56; Federal Trade Commission 222; film industry 88; financial services 103; First World War 146; Ford 235, 249, 253, 300, 356, 395; foreign direct investment (FDI) 347; foreign investors 128;

Index

483

free market 216–17; free trade 20, 33; from water power to electricity 68, 70–71; GM 95, 235, 249; globalisation 335; government policy 134; Greenspan, Alan 143, 144; herd behaviour 314; human capital 102; ideological distribution of the parties in the House 135; imperialism 335; India 20; industrial concentration 200; industrialisation 302; innovation activities 229, 299; intellectual property rights 219, 298; international economic integration and location 203–5; inter-regional labour mobility 174; internal competition 265; intra-firm trade 310; knowledge spillovers 164; litigious climate 103; local cable monopoly 217; local phone monopoly 217, 218; location of cities 113, 118, 165; Macintosh 61, 62, 63, 161, 272; manufacturing industry 193–94; Mars Bars 147; Massachusetts General Hospital (Boston) 85; McDonald’s 258, 303, 337, 341, 383; mergers 256, 257, 258–59, 303; Mexico 203; Microsoft 61, 62, 107, 133, 143, 161, 219–22, 298; military industry 149; Mississippi 119; Missouri 119; monopoly 218, 219; Motorola 157, 315, 367, 383; national and international regulation 385–88; Navigation Act (1660) 165; New Orleans 118–19; New York 97, 102, 103, 116, 131, 170, 427; Nike 375, 377, 383; Organisation for Economic Cooperation and Development (OECD) 230; parallels between United States and United Kingdom 401; patents, patent families 230–31; Pepsi 342, 401, 402; Philippines 20; policies 134, 261; policy-making 134–36; population 112; Procter & Gamble 272, 321, 364–65;

Index

484

production geography 188–89; reallocation of manufacturing 19–21; research and development (R&D) 230, 231, 299; rights 163; Route 128 (Boston) 84, 98, 102, 106; rubber-based products 21; Sarbanes-Oxley Act 103, 248; Seagate 314; Sherman Act of 1890 218; Silicon Valley (California) 84, 98, 99, 100, 102, 104, 114–15; Singapore 314; Smith & Wesson 258; software 145; specialisation 191, 194; tax system 197, 319; Texas Instruments 270–71; textile industry 19, 20, 21, 156, 170, 304; trade unions 134, 162; transnational corporations (TNCs) 314, 395; transport system 119; tyre-making 162; universities 98, 102, 300; US-Canada Auto Pact 249; VCR 245; Wal-Mart 144, 169, 341, 377, 426; Washington Consensus 389, 405–6, 415; Washington DC 102, 111, 113, 135–36; wine industry 153; Zipf’s Law 117; see also North American Free Trade Agreement (NAFTA) university 98–99, 100, 300; Czech Republic 398; medieval universities 242; patents 232; science park 98; Switzerland 300, 423 unpredictability 30; evolutionary economic geography 14, 155; multiple equilibrium 54; see also predictability Uruguay: Uruguay Round 20, 29, 259, 277–78, 285, 296 utilities 31, 118, 261, 325 Venezuela 204 Volkswagen 95, 157, 257, 265, 398 wages 376–77; automotive industry 157; China 315, 353;

Index

485

distribution of manufacturing and transport costs/wages 172–73; Greece 353; herd behaviour 314–15; Italy 353; Japan 352; Portugal 353; transnational corporations (TNCs) 352–53, 376–77; see also employment; labour force Walras, Léon 34, 158; Elements of Pure Economics 34 war and the location of firms 145–50; Adrianople 146; the ‘cockpit of Europe’ 145; Cold War 149, 235; Eighty Years War (1568–1648) 149; First World War 72, 146, 298; Germany 146; Great Wall 146; Iraq War 149; Italy 146, 147, 148; military aircraft 298; Pakistan 147–48; Peace of Westphalia 149–50; the ‘Quadrilateral’ 145; Second World War 72, 77, 118, 120–21, 166, 147, 148, 170; Switzerland 147; Thirty Years War (1618–48) 149; topography 146, 169–70; Turkish occupation (1529–1683) 149; United States 146–47, 149; see also location wealth 77, 140, 169, 174, 347, 355; distribution of 181; growth 338; impact on economic prosperity 373–74; knowledge 89; material wealth 77, 174, 223, 373, 374; wealth creation 111–12, 165 welfare: competition 216, 219; international economic integration 185–86; welfare in a trade-diverting customs union 184–85 Westinghouse Electric 139 wine industry 153, 169 World Bank 142, 143, 331; globalisation 332, 400; Growth Report 4114; instrument of global control 330; North American Free Trade Agreement (NAFTA) 204–5; Washington Consensus 389, 405–6, 415; World Bank Commission on Growth and Development 414;

Index see also capital; International Monetary Fund (IMF) World Economic Forum 297; Scandinavian competition model 297 World Social Forum 332 World Trade Organisation (WTO) 20, 25, 29, 331; adjustment costs 187; aircraft industry 303; antidumping duties 304; Byrd Amendment 288; Doha Round 265, 330, 331; globalisation 332, 400; intellectual property rights 329; instrument of global control 330; Millennium Round 330; national and international regulation 387, 388; non-tariff barriers (NTBs) 282, 283; predatory pricing 271; rules of origin 279, 280, 281; WTO 1995 TRIMs Agreement 377, 395, 404; see also foreign direct investment (FDI); transnational corporations (TNCs)

486