Endogenous Regional Development

Endogenous Regional Development
STIMSON PAGINATION (M2469).indd i 20/12/2010 15:12

NEW HORIZONS IN REGIONAL SCIENCE
Series Editor: Philip McCann, Professor of Economic Geography, University of Groningen, the
Netherlands and Professor of Economics, University of Waikato, New Zealand
Regional science analyses important issues surrounding the growth and development of
urban and regional systems and is emerging as a major social science discipline. This series
provides an invaluable forum for the publication of high-quality scholarly work on urban and
regional studies, industrial location economics, transport systems, economic geography and
networks.
New Horizons in Regional Science aims to publish the best work by economists, geogra-
phers, urban and regional planners and other researchers from throughout the world. It is
intended to serve a wide readership including academics, students and policymakers.
Titles in the series include:
Entrepreneurship, Industrial Location and Economic Growth

Edited by Josep Maria Arauzo-Carod and Miguel Carlos Manjn-Antoln
Creative Cities, Cultural Clusters and Local Economic Development
Edited by Philip Cooke and Luciana Lazzeretti
The Economics of Regional Clusters
Networks, Technology and Policy
Edited by Uwe Blien and Gunther Maier
Firm Mobility and Organizational Networks
Innovation, Embeddedness and Economic Geography
Joris Knoben
Innovation, Agglomeration and Regional Competition
Edited by Charlie Karlsson, Brje Johansson and Roger R. Stough
Technological Change and Mature Industrial Regions
Firms, Knowledge and Policy
Edited by Mahtab A. Farshchi, Odile E.M. Janne and Philip McCann
Migration and Human Capital
Edited by Jacques Poot, Brigitte Waldorf and Leo van Wissen
Universities, Knowledge Transfer and Regional Development
Geography, Entrepreneurship and Policy
Edited by Attila Varga
International Knowledge and Innovation Networks
Knowledge Creation and Innovation in Medium Technology Clusters
Riccardo Cappellin and Rdiger Wink
Leadership and Institutions in Regional Endogenous Development
Robert Stimson and Roger R. Stough with Maria Salazar
Entrepreneurship and Regional Development
Local Processes and Global Patterns
Edited by Charlie Karlsson, Brje Johansson and Roger R. Stough
Endogenous Regional Development
Perspectives, Measurement and Empirical Investigation
Edited by Robert Stimson, Roger R. Stough and Peter Nijkamp
STIMSON PAGINATION (M2469).indd ii 20/12/2010 15:12

Endogenous Regional
Development
Perspectives, Measurement and Empirical
Investigation
Edited by
Robert Stimson
Professor of Geographical Sciences and Planning, University
of Queensland, Australia
Roger R. Stough
Vice President for Research and Economic Development,
NOVA Endowed Chair and Professor of Public Policy, George
Mason University, USA
Peter Nijkamp
Professor of Regional, Urban and Environmental Economics,
VU University Amsterdam, the Netherlands
NEW HORIZONS IN REGIONAL SCIENCE
Edward Elgar
Cheltenham, UK Northampton, MA, USA
STIMSON PAGINATION (M2469).indd iii 20/12/2010 15:12

Robert Stimson, Roger R. Stough and Peter Nijkamp 2011
All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, electronic,
mechanical or photocopying, recording, or otherwise without the prior
permission of the publisher.
Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK
Edward Elgar Publishing, Inc.

William Pratt House
9 Dewey Court
Northampton
Massachusetts 01060
USA
A catalogue record for this book

is available from the British Library
Library of Congress Control Number: 2010927669
ISBN 978 1 84980 456 1
Typeset by Servis Filmsetting Ltd, Stockport, Cheshire

Printed and bound by MPG Books Group, UK
03
STIMSON PAGINATION (M2469).indd iv 20/12/2010 15:12

Contents
List of contributors vii
Preface xi
Robert Stimson, Roger Stough and Peter Nijkamp
1 Endogenous regional development 1

Robert Stimson, Roger Stough and Peter Nijkamp
2 The economists perspective on regional endogenous
development 20
Kenneth Button
3 Endogenous regional theory: a geographers perspective
and interpretation 39
Michael Taylor and Paul Plummer
4 Endogenous rural development from a sociological
perspective 59
Frank Vanclay
5 Rural, urban or regional endogenous development
as the core concept in the planning profession 73
Edward Blakely
6 Diversity and endogeny in regional development:
applying appreciative intelligence 83
Tojo Thatchenkery and Jessica Heineman-Pieper
7 An exploratory approach to model determinants
of endogenous regional growth performance 111
Robert Stimson and Roger Stough
8 A theory of entrepreneurial rents in endogenous
growth: implications for regional innovation policies 142
Zoltan Acs and Mark Sanders
9 Foreign direct investment, knowledge assets and the
economic geography of growth in the Asian BRIICS
countries 160
Tomokazu Arita, Chie Iguchi and Philip McCann
10 Implications of European Union structural assistance
to new member states on regional disparities: the
question of absorption capacity 182
Daniela Constantin, Zizi Goschin and Gabriela Dragan
STIMSON PAGINATION (M2469).indd v 20/12/2010 15:12

vi Endogenous regional development
11 Macroeconomic and territorial policies for regional

competitiveness: theory and empirical evidence from the EU 204
Roberto Camagni and Roberta Capello
12 Endogenous employment growth and decline in Australian
capital city statistical divisions 237
Alistair Robson
13 A case study approach to investigating local development
initiatives in rural small towns in Victoria 268
John Martin
14 Economic development incentives and the measurement
of local endogenous growth: is there a need for modeling
adjustment? 286
Terry Clower
15 Regional growth and development theories revisited 301
Roberta Capello and Peter Nijkamp
Index 325
STIMSON PAGINATION (M2469).indd vi 20/12/2010 15:12

Contributors
Zoltan J. Acs is University Professor at the School of Public Policy and
Director of the Center for Entrepreneurship and Public Policy at George
Mason University, USA. He is also a Research Scholar at the Max Planck
Institute for Economics in Jena, Germany, and Scholar-in-Residence at
the Kauffman Foundation. He is co-editor and founder of Small Business
Economics, the leading entrepreneurship and small business publication in
the world. Acs is a leading advocate of the importance of entrepreneurship
for economic development.
Tomokazu Arita is Associate Professor in the Department of Social
Systems and Management, Graduate School of Systems and Information
Engineering, at the University of Tsukuba, Japan. After earning a PhD at
the Regional Science Department of the University of Pennsylvania, he
worked for twelve years at the Japanese Ministry of Land, Infrastructure,
Transport, and Tourism (the former Ministry of Construction) before
being appointed at Tsukuba.
Edward J. Blakely is Urban Policy Professor at the United States Studies
Centre at the University of Sydney, Australia, and formerly the Executive
Director for Recovery in New Orleans. He has over 30 years as an expert
and academic in the fields of regional and local economic development.
He is a Fellow of the American Academy of Public Administration and a
former Guggenheim Fellow.
Kenneth Button is University Professor and Director of both the Center
for Transportation Policy, Operations and Logistics and the Center for
Aerospace Policy Research in the School of Public Policy, George Mason
University, USA. He is also Professor in Civil Engineering at the University
of Porto, Portugal and Visiting Professor in Applied Economics at the
University of Bergamo, Italy. Button is a Fellow of the Chartered Institute
of Logistics and Transport, a Fellow of the Institution of Highways and
Transportation and an Academician of the Academy of Social Sciences.
He is past President of the Transportation Research Forum.
Roberto Camagni is a Professor of Urban Economics at the Politecnico
di Milano, Italy. He is a past President of the European Regional Science
Association.
vii
STIMSON PAGINATION (M2469).indd vii 20/12/2010 15:12

viii Endogenous regional development
Roberta Capello is Professor of Regional Economics at Politecnico

di Milan, Italy. She is the current President of the Regional Science
Association International, is Editor-in-Chief of the Italian Journal of
Regional Science, and co-editor of Letters in Spatial and Resource Sciences.
Terry Clower is Director of the Center for Economic Development and

Research Associate Professor of Applied Economics at the University of
North Texas, USA. He is Editor (Americas) of Regional Science Policy
and Practice and a board member of the Council for Community and
Economic Research.
Daniela Constantin is a Professor of Regional Economics at the Academy

of Economic Studies of Bucharest, Romania, where she is also Director
of the Research Centre of Macroeconomic and Regional Forecasting.
She is the President of the Romanian Regional Science Association and a
member of the European Regional Science Association Council. She is the
Director of the Romanian Journal of Regional Science.
Gabriela Dragan is a Professor of European Integration at the Bucharest

Economic University, Faculty of International Business and Economics
and the General Director of the European Institute of Romania. She
is a member of the Romanian Regional Science Association, the Team
Europe, the European Commissions panel of independent conference
speakers, and Director of the Romanian Journal of European Affairs.
Zizi Goschin is a Professor of Statistics, Econometrics and Regional

Statistics at the Academy of Economic Studies of Bucharest, Romania,
and a Senior Scientific Researcher at the Institute of National Economy
of the Romanian Academy. She is a member of the Romanian Regional
Science Association Council and the Editor-in-Chief of the Romanian
Journal of Regional Science.
Jessica Heineman-Pieper is an Assistant Professor in the Master of Science

program in Organization Development and Knowledge Management in
the School of Public Policy, George Mason University, USA. She has con-
sulted to the National Institute of Mental Health (USA) on science policy,
has worked as a management consultant with Booz Allen Hamilton, and is
on the Executive Committee of the Critical Management Studies Division
of the Academy of Management.
Chie Iguchi is Assistant Professor of Global Strategic Management at

Rikkyo University College of Business in Tokyo. She gained her PhD
at the University of Reading Business School. Before joining Rikkyo
University she previously held a full faculty appointment at Ritsumeikan
STIMSON PAGINATION (M2469).indd viii 20/12/2010 15:12

Contributors ix
University, Kyoto, and was also a Visiting Research Fellow at Rutgers

University.
John F. Martin is Professor and Director of the Centre for Sustainable
Regional Communities at LaTrobe University in Bendigo, Australia. His
research interests include the role of local institutions in the sustainability
of rural and regional communities across Australia.
Philip McCann is Special Adviser to Johannes Hahn, the European
Commissioner for Regional Policy and currently holds the University of
Groningen Endowed Chair of Economic Geography in the Faculty of
Spatial Sciences at the University of Groningen in the Netherlands. He
is also a part-time Professor of Economics at the University of Waikato,
New Zealand. He has won academic awards in five countries, is a keynote
speaker in conferences around the world, and advises major international
organizations and governments.
Peter Nijkamp is at present Professor of Regional, Urban and Environmental
Economics at the VU University in Amsterdam, the Netherlands. He has
published extensively in the area of regional development and innovation.
For many years, until 2009, he has served as President of the Netherlands
Organization for Scientific Research (NOW). He is also a past President of
the Regional Science Association International.
Paul Plummer is a Professor at the University of Calgary, Canada. He
has published widely on the foundations of analytical political economy,
competing theories of spatial competition and the dynamics of regional
growth and change.
Alistair Robson has worked as an economist in both academia and gov-
ernment in Australia. Currently he is an economist in New South Wales
Treasury. Prior to that he was a Research Fellow at the Institute for Social
Science Research at the University of Queensland, where he also gained
his doctorate.
Mark Sanders is Assistant Professor in Economics at Utrecht University,
the Netherlands, and research fellow of the Max Planck Institute of
Economics in Jena, Germany. He conducts research in innovation eco-
nomics, entrepreneurship and economic development, with a special focus
on the role of entrepreneurial activity in the introduction of renewable
energy technologies.
Robert J. Stimson is Professor of Geographical Sciences and Planning
at the University of Queensland, Australia, where he was also Director
of the Urban and Regional Analysis Research Program in the Institute
STIMSON PAGINATION (M2469).indd ix 20/12/2010 15:12

x Endogenous regional development
for Social Science Research, and was the Convenor of the Australian
Research Council Research Network in Spatially Integrated Social
Science (ARCRNSISS) until 2009. He is a Fellow of the Academy
of the Social Sciences in Australia, a past President of the Regional
Science Association International, and the current Chair of the Applied
Geography Commission in the International Geographical Union.
Roger R. Stough is the Vice President for Research and Economic
Development at George Mason University, USA, where he also holds
the NOVA Endowed Chair and is an Eminent Scholar and Professor of
Public Policy. He is a past President of the Regional Science Association
International and holds an honorary doctorate from Jonkoping University
in Sweden.
Michael Taylor is Professor of Human Geography at the University of
Birmingham, UK. He has held academic positions in New Zealand and
Australia, and also worked for the Australian Federal Administration in
Canberra. His research is focused on enterprise and regional economic
development and particularly on the theory of the firm. He is Chair of
the International Geographical Union Commission on the Dynamics of
Economic Spaces.
Tojo Thatchenkery is Professor and Director of the Masters in Science
program in Organization Development and Knowledge Management
at the School of Public Policy, George Mason University, USA. He is a
member of the NTL (National Training Laboratory) Institute for Applied
Behavioral Science, the Taos Institute, the Editorial Board of the Journal
of Applied Behavioral Sciences, the Journal of Organizational Change
Management (JOCM), and the book review editor of JOCM.
Frank Vanclay is Professor of Cultural Geography at the University of
Groningen, the Netherlands. He is a past President of the International
Rural Sociology Association (IRSA).
STIMSON PAGINATION (M2469).indd x 20/12/2010 15:12

Preface
The role of endogenous processes in regional economic development and
as potential explanatory factors in accounting for spatial variations in
the pattern of regional growth and decline has been attracting increasing
interest from regional scientists and from regional development policy-
makers and practitioners.
Among researchers there are a range of perspectives on endogeneity
and endogenous regional growth processes. There are challenges in devel-
oping suitable measures of endogenous regional growth and in building
and implementing operational models to analyse endogenous regional
development and growth across spatial systems. It is also important for
research to generate in-depth analysis of endogenous processes at the scale
of the local region to enhance our understanding of the different contex-
tual bases in which endogenous processes occur in local regional economic
development.
The chapters in this book address these important issues. They are
based on papers presented at an international workshop on Regional
Endogenous Development: Measurement, Models and Empirical
Investigation held at The University of Queenslands Customs House facil-
ity in downtown Brisbane, Australia, in February 2008. It was hosted by
the Australian Research Council Research Network in Spatially Integrated
Social Science, with financial assistance from the Australian governments
Bureau of Infrastructure, Transport and Regional Economics. The work-
shop brought together researchers in regional science from around the
world to address these issues.
The workshop was one of many held over the last decade in which a
small international group of about 20 to 25 established and emerging
researchers come together to discuss topics relating to regional eco-
nomic development, with an emphasis on innovation and entrepreneur-
ship. The workshops have been held under the banner of the Tinbergen
Institute (named after the Nobel Laureate in Economics) and are normally
held in Amsterdam. The workshops are sponsored by the Tinbergen
Institute; the Department of Spatial Economics at The Free University,
Amsterdam; the School of Public Policy at George Mason University,
Fairfax, Virginia, USA; and the Australian Research Council Research
Network in Spatially Integrated Social Science.
xi
STIMSON PAGINATION (M2469).indd xi 20/12/2010 15:12

xii Endogenous regional development
This book is, we believe, a unique contribution to the growing literature

that addresses endogenous issues in regional economic development in
that it deliberately presents a series of different disciplinary perspectives
on endogeneity and the nature and role of endogenous factors in regional
development and growth. Those contributions come from the perspective
of an economist, a geographer, a sociologist, a planner and a researcher
in organizational management/social psychology. In addition, the book
includes chapters in which researchers explore new frameworks to model
endogenous regional development, focusing on a range of factors includ-
ing regional resource endowments, entrepreneurship and institutional
factors. Contributions to the book also include empirical investigations
conducted at a variety of levels of spatial scale, from cross-national mega-
regional level of scale, to the national level of scale, to the city level of scale
and the local small town level of scale.
We thank the participants in the Brisbane workshop for their partici-
pation in the processes leading to the publication of this book, which we
hope will attract the interest of researchers, students and practitioners in
regional science and regional development. The financial and other assist-
ance of the sponsors of the Brisbane workshop is gratefully acknowledged.
And we are indebted to the Edward Elgar team for their encouragement to
produce this book and for that publishers ongoing commitment to pub-
lishing research in regional science.
Robert Stimson, Roger R. Stough, Peter Nijkamp

February, 2010
STIMSON PAGINATION (M2469).indd xii 20/12/2010 15:12

1. Endogenous regional development
Robert Stimson, Roger Stough and
Peter Nijkamp
INTRODUCTION
Over the past few decades the emphasis in regional development theory
has shifted from a focus primarily on exogenous factors to an increasing
focus on endogenous factors. Traditional regional economic development
approaches were erected on neoclassical economic growth theory, based
largely on the Solow (1956, 2000) growth model. The emerging endogen-
ous approaches while recognizing that development is framed by
exogenous factors attribute a much more significant role for endogenous
forces. In that context, a suite of models and arguments that broadly
convey the new growth theory have been directed towards endogenous
factors and processes (see, for example, Johansson et al., 2001).
Those developments are of great interest to regional development ana-
lysts and also to practitioners for several reasons, including the recogni-
tion of the importance of cities and regions in the development process,
and also because they introduce an explicit spatial variable into economic
development and growth theory, which was a mostly ignored element
in neoclassical thinking. This evolutionary development is particularly
significant as the importance of regions in national economies and in
particular the role of many of the worlds mega-city regions has changed
considerably since the 1970s as a result of globalization, deregulation and
structural change and adjustment. Understanding these recently recog-
nized processes of change are crucial for analysing and understanding dif-
ferent patterns of regional economic performance and in formulating and
implementing regional economic development planning strategy.
This book focuses explicitly on endogenous regional development,
attempting to provide a range of disciplinary perspectives on the nature
of endogeneity and what endogenous regional development is about, how
it might be conceptualized, measured and modelled, and how empirical
analysis taking an endogenous perspective may be conducted at different
levels of spatial scale.
STIMSON PAGINATION (M2469).indd 1 20/12/2010 15:12

2 Endogenous regional development
THE NATURE OF REGIONAL DEVELOPMENT

A Product and a Process
Stimson et al. (2006, p. 4) observe that it is often difficult in regional eco-

nomic development planning strategy formulation and implementation
to match desired outcomes of regional economic development with the
processes that create them. That gap in understanding the relationship
between the apparent causes and effects of development poses a dilemma
for those responsible for managing regional economic development in the
making of policies and strategies, and the implementation of plans. The
dilemma they face is how to achieve some form of congruence between
desired outcomes and appropriate and acceptable economic development
tools and processes. The dilemma is further compounded by the frequently
unstable and changing nature of economic environments, where exter-
nalities or exogenous factors (such as exchange rates, new technologies,
foreign competition) increasingly impact the decision-making processes
that influence economic policy and strategy in cities and regions.
Blakely (1994) emphasizes how regional economic development needs
to be viewed as both a product and a process but often not by the same
groups or actors in the development milieu. For example, economic agents
that live, work and invest in regions are those most concerned with eco-
nomic development outputs or products such as job and wealth creation,
investment, quality of life or standards of living, and conditions of the
work environment. Contrary to this view is the more process orienta-
tion of regional scientists, development planners and practitioners where
concern focuses on the creation of infrastructure, labour force prepara-
tion, human capital and market development. So it is important when
considering regional economic development to maintain an awareness of
its product and process aspects.
Regional economic development is also known in terms of quantitative
and qualitative attributes. In that context, and with respect to the benefits
it creates, our concern has typically been with the quantitative measure-
ment of such factors as increasing/decreasing wealth and income levels,
job creation or employment levels, the availability of goods and services
and improving financial security. At the same time and especially in
recent times our concern has also been with such qualitative considera-
tions as generating creative capital, creating greater social and financial
equity, achieving sustainable development, creating a spread in the range
of employment and gaining improvements in the quality of life. Thus, the
regional economic development process needs to be informed by both
quantitative and qualitative information.

Endogenous regional development 3
This multidimensional aspect of economic development led Stimson

et al. (2006) to propose the following definition of regional economic
development:
Regional economic development is the application of economic processes and

resources available to a region that result in the sustainable development of,
and desired economic outcomes for a region and that meet the values and
expectations of business, of residents and of visitors. (Ibid., p. 6)
Changing Paradigms
Policy for economic development and regional planning strategy has

undergone a series of evolutionary changes since World War II, driven
by different paradigms of economic thought as discussed by Stimson et
al. (ibid., pp. 1117). In the post-World War II era there was emphasis
on Keynesian approaches to economic policy with an associated focus
in regional development planning on master planning and then goals
and objectives planning approaches. The era was very much embedded
in the notion that differential regional development largely reflected dif-
ferences in the comparative advantage/disadvantage of regions. Then,
from the 1970s and into the 1980s there was a shift in economic policy
to embrace monetarist thought and then rationalist economic thought in
the 1980s and into the 1990s. That era was also associated with a refocus-
ing in regional development planning on structure and strategic planning
approaches. That reflected the notion that the enhancing of regional com-
petitive advantage as a driver of regional development was an important
objective. But over the last two decades there has been a further shift to
a focus on the principles of sustainable development in regional develop-
ment and planning, with planning strategies being based on an integrated
approach whereby a strategy would seek to advance the capacity and
capability of a region to better utilize its resource endowments through
a focus on endogenous processes encouraging collaborative advantage
across the private, public and community sectors.
Those changing paradigms have shaped the way regional and local com-
munities and people think and plan for the future. But much thinking on
regional economic development still remains embedded in the paradigms
of the 1970s, because of an inherent reluctance of many regions and local
communities to proactively embrace change. Subsequently, as suggested
by Stimson et al. (ibid., p. 11):
many regions are not re-equipping themselves fast enough to compete effectively
in the global age of business and technology of the post-industrial economy.
To compete successfully in the global economy, regional organizations and

businesses need to understand the implications of the paradigm shifts occur-

ring in economic policy and strategy, and to build the flexible strategic
infrastructure to do so.
We now turn to a more detailed consideration of the emergence of the

so-called new growth theory in regional development, with its increasing
emphasis on endogenous factors and processes.
THE EVOLUTION OF THE NEW GROWTH THEORY

APPROACH
During the 1980s by which time the focus in economic policy paradigms
had shifted from a focus on Keynesian thought and the focus in regional
development thinking on the Solow (1956) model emphasizing exogenous
forces, to a new paradigm represented by monetarism and economic
rationalist thought there had been a shift from concerns about develop-
ing a regional comparative advantage to developing a regional competitive
advantage. There had also been a shift in regional development planning
strategy from master planning and structural planning to strategic plan-
ning paradigms. Thus a new way of conceptualizing regional economic
growth and development had begun to emerge, which today is known as
the new growth theory.
As early as the late 1970s, Rees (1979) had proposed that technology
was a prime driver in regional economic development, and since then
over the ensuing two to three decades the regional science literature
has shown how technology is directly related to traditional concepts of
agglomeration economies in regional economic development. Economists
such as Romer (1986, 1990), Lucas (1988), Barro (1990), Rebelo (1991),
Grossman and Helpman (1991) and Arthur (1994) sought to explain tech-
nical progress in its role as a generator of economic development as an
endogenous effect rather than accepting the neoclassical view of long-term
growth being due only to exogenous factors. In macroeconomic models
of endogenous growth, technological progress was mainly seen as an
endogenous process in an economic system, where knowledge is generally
embedded in human capital that is enhanced through education, training,
creativity and R&D. Thomas (1975) and later Erickson (1994), among
others, showed how technological change was related to the competitive-
ness of regions. Norton and Rees (1979), Erickson and Leinbach (1979)
and Markusen (1985) showed how the product cycle, when incorporated
into a spatial setting, demonstrated that some regions could be seen as
the innovators, while others become the branch plants or recipients of the

innovation, and those might even then become innovators via endogenous
growth. The concept of innovative milieu was formulated to explain the
how, when and why of new technology generation. That notion linked
back to the importance of agglomeration economies and localization
economies that had been viewed as leading to the development of new
industrial spaces (Scott, 1988; Porter, 1990). In particular, research by
Krugman (1991, 1995, 1996) and Krugman and Venables (1996) led to a
greater emphasis on knowledge as a tacit and primarily local good and the
recognition of it as a driving endogenous self-reinforcing mechanism for
regional development.
There has also been an emphasis on the importance of investment in
human capital and its role in regional development, as demonstrated in
work by the OECD (2000, 2001). In addition, writers such as Fukuyama
(1995) suggest that it is not just economic but also value and cultural
factors including social capital and trust that are important in the rise
of technology agglomerations as seen in the Silicon Valley phenomenon,
where collaboration among small and medium-size enterprises through
networks and alliances and links with universities forge a powerful R&D
and entrepreneurial business climate. How to enhance regional eco-
nomic development through engendering regional collaborative advan-
tage (Huxam, 1996) began to be seen as a desirable component of regional
development planning strategy.
There has emerged as well a considerable emphasis on the role of leader-
ship and institutions as factors that can enhance or even act as a catalytic
effect in endogenous regional development, as demonstrated by Stimson
and Stough (2009). As Rees (2001) has pointed out, technology-based
theories of regional economic development need to incorporate the role of
entrepreneurship and leadership, particularly as factors in the endogenous
growth of regions, and it is the link between the role of technology change
and leadership that can lead to the growth of new industrial regions and to
the regeneration of older ones (p. 107).
Especially in the period since the mid-1970s, the processes of globaliza-
tion have resulted in the emergence of an increasingly borderless economic
world with increasingly unrestricted mobility of capital and labour and
increasing freedom of trade in merchandise and services. Seemingly, the
influence of the nation state was reduced in a world where cities and par-
ticularly mega-city regions assumed increasing importance as strategic
hubs and as the drivers of creativity, innovation and entrepreneurial activ-
ity and as they increasingly became the dominant engines of economic
growth (Knight and Gappert, 1989; Ohmae, 1995; Prudhomme, 1995;
Florida, 2002). That created new stresses for both nations and for regions
and their governments in developing strategies to find a competitive edge

in a globalized economy and a highly competitive and rapidly changing

world. There was a considerable shift in regional development plan-
ning strategies towards the notion of enhancing regional self-help. More
recently the emergence of concern for achieving sustainable development
has diversified the goals for regional development and intensified competi-
tive pressures. And it is presenting new challenges for institutional reform,
leadership and governance.
Thus, the new growth theory models have allowed for and indeed have
implied the importance of both agglomeration effects (economies of scale
and externalities) and market imperfections, with the price mechanism
not necessarily generating an optimal outcome through efficient alloca-
tion of resources. And there has been a considerable emphasis on factors
such as leadership, institutions, creativity, innovation and entrepreneur-
ship, which might be regarded as the endogenous intangibles that may
enhance the performance of cities and regions.
The processes of capital accumulation and free trade have not necessarily
led to convergence of wage and price levels between regions, with positive
agglomeration effects tending to often concentrate activity in one or a few
regions in many nations through the self-enforcing effects that attract new
investment. That process may be mediated positively by the endogenous
intangibles we have referred to. The new growth theory has actually
allowed for both concentration and divergence in regional development.
Most importantly, as the spatial distribution of knowledge and its
spillovers are now considered to be important success factors in regional
development in framing and implementing regional development strate-
gies, it is now seen as being crucial for a city or region to fully understand
the nature of the geographical patterns of knowledge diffusion and the
barriers to access to knowledge as they relate to creativity, innovation and
entrepreneurship as being catalysts for employment and wealth generation
(see, for example, Keeble and Wilkinson, 1999; Acs et al., 2002; Dring
and Schnellenback, 2006).
THE CONTEMPORARY CHALLENGES
We certainly live in a rapidly changing and increasingly competitive world

in which uncertainty and risk are considerable. As discussed by Stimson et
al. (2006), in many countries the challenge facing economic development
planners in contemporary times has been how to formulate economic
policy that will respond to both global dynamics and sometimes (or maybe
even more often than not) the existence of a national vacuum in adopting
a regionally oriented macro-policy.

At one time regions were protected from outside competition, and

to some extent their economies could be manipulated by national gov-
ernments. But that ability has been overwhelmingly compromised as
the economic rationalism pursued by many national governments left
many cities and regions to fend for themselves. Many cities and regions
have continued to look to higher levels of government for support and
resources to provide economic direction and investment to stimulate
economic development. Unfortunately, many cities and regions have
failed to understand that globalization has left those higher levels of gov-
ernments relatively weak when it comes to using their inherent power to
apply economic and policy mechanisms to enhance the competitiveness
of regional economies.
In the literature a number of key themes have emerged regarding what
constitutes regional growth and development and what drives regional
competitiveness. Not surprisingly there have been differences of views
among regional development scholars, and some of those differences relate
to the relative focus given to the roles of exogenous forces on the one hand
and the roles of endogenous processes and factors on the other. But there
does now seem to be an almost universal realization of what Garlick et al.
(2007) have referred to as the institutional embeddedness of endogenous
processes and factors in regional development.
Of course exogenous factors are likely to remain important to a regions
economic performance and how it develops over time. But increasing
importance is being placed on endogenous forces as determinants of
a regions competitiveness. However, regional economic development
policy initiatives now tend to be more oriented towards measures that
enhance local capacity and capability for a city or region to develop and
cope with rapid change in an increasingly competitive global environment.
While endogenous growth theory makes mention of leadership, entrepre-
neurship and institutional factors, little systematic analysis has occurred
to thoroughly conceptualize or, even more, measure their roles as endog-
enous factors in the development process.
But as discussed by Stimson et al. (2006), in the contemporary policy
era of the last decade or two, it would seem that it has been more and
more up to regions to develop and use their own devices to compete inter-
nationally in order to survive. Thus, it had become increasingly common
in regional development planning strategy for there to be a reliance on
endogenous processes, and typically that has been espoused in regional
economic development policy. To do that a region would need first to
have understood what the factors were that set the dynamics of the new
economic age that had emerged in the late twentieth century. In the wake
of the current global financial crisis and recession conditions, it will be

interesting to see whether these much-changed macro-circumstances will

set the conditions for a rethinking of that regional self-reliance philosophy
and usher in a new era of innovation in institutional arrangements that
might incorporate more interventionalist policies in regional development
strategy planning.
In the regional growth literature there is no doubt that the strategic
importance of knowledge for innovation and entrepreneurship has been
increasingly recognized. That has built on the notion of the learning
region as proposed by Simmie (1997). As discussed by Capello and
Nijkamp (2009), in a neoclassical framework of analysis, long-range
factors such as education, R&D and technology have played a criti-
cal structural role in the context of the spatial mobility of production
factors, which could remove disparities (for example, in terms of per
capita income) in the long run and, as a result, may equalize factor
productivity across a nations regions. And in the endogenous growth
literature we have seen how knowledge spillovers and institutional
arrangements in local regions are widely acknowledged as factors explain-
ing how knowledge spillovers are spread (as growth spillovers), with those
knowledge spillovers representing pure externalities that produce non-
compensating advantages for the receivers (Nijkamp and van Hemert,
forthcoming). But Capello (2009) has pointed to a discrepancy between
the private and social optimum that creates the emergence for ad hoc
policy interventions.
In the current economic climate of the global financial crisis and reces-
sion, Nijkamp and van Hemert (forthcoming) have suggested that in
trying to capture the catalytic effect of creativity, innovation and R&D
in generating knowledge growth spillovers: more than ever there is a
role for government in focusing strong and directed efforts to boost the
translation of scientific ideas into useful technologies, and to reinforce the
base of science skills that drives this innovation (p. 1). They go on to say:
Currently, there are different forces at play in the science domain that
need attention and support from governments. Besides tensions between
local and regional demands, the current crisis has highlighted the growing
frictions between the individual and societal needs (ibid.).
The challenges today include the need to revolutionize transport
technologies, meet climate-change targets and secure diversity of energy
supply. On a national level, that will require more directed research,
education and training innovation to develop the required skills to enact
the new technologies, and the active participation of industry in govern-
mentscience relations to help encourage innovation. This changing
socio-political environment, Hertz (2009) suggests, will require different
research disciplines to work together more than ever.

In the context of regional development, Taylor (2009) has referred to

the ability to capture ideas and discoveries that flow from research as the
main test of whether the UK can recover growth and prosperity. He says
that at present the UK does not have the workforce needed to enact new
technologies to address the challenges just mentioned, and that is also the
case across many if not all countries. While it is a major policy challenge
it does, nonetheless, represent an opportunity for local initiatives to be
taken to boost investment in education and R&D, particularly in science
and technology.
The notion is that through what has been termed the triple helix sce-
nario (Etzkowitz and Leydesdorff, 1996, 1997, 2000), whereby investment
in innovation and R&D inputs will lead to greater innovation outputs
when they originate from local sources, cities and regions might be able
to catalyse future economic growth. That notion affirms the existence
of a spiral pattern of relations and links between the three major institu-
tional actors in a local environment industry, university and research
institutes, and government in which the education and research sector
tends to have a critical part to play in the context of economic growth and
regional development in the contemporary knowledge-based economy
and in helping societies to address the technological and policy challenges
they face with respect to issues such as climate change and achieving more
sustainable development. Thus, as Nijkamp and van Hemert (forthcom-
ing) say: Concentrations of outstanding scientific facilities and activities
are very important to create challenging and attractive working conditions
and opportunities for talented people (p. 6). That reinforces what Florida
(2002) has suggested in his work on the creative class and the emergence
of some cities as centres of creativity. Understanding the institutional
barriers that mitigate against achieving this creativity and the associ-
ated economic dynamism of a city or region and how to unlock those
barriers for the emergence of a learning region is an obvious priority in
regional development strategy planning if the pentagon model proposed
by Nijkamp (forthcoming) is to be pursued.
A big issue will be the degree to which regional development and growth
will converge or diverge as a result of the institutional embeddedness of
endogenous processes to which Garlick et al. (2007) refer (and which we
have discussed in this chapter) and what will be the nature of the jumps
and anomalies in urban and regional systems that Nijkamp (forthcoming,
p. 6) refers to. Endogenous growth theory can help us to understand the
complexities of a dynamic space-economy (including shock and bifurca-
tions), but contextual drivers and government policies offer another cause
of unexpected dynamics. Consequently, the evolution of complex spatial
systems will partly remain an unresolved mystery.

TOWARDS A SUSTAINABLE DEVELOPMENT

MODEL
In the contemporary context of a focus on sustainable development in
regional economic development strategy it becomes even more necessary
to place concerted attention on harnessing endogenous factors in pursuing
regional growth and development. Towards that end, from the writings
of Nijkamp et al. (1994) and Capello et al. (1999) it is possible to propose
a production function for sustainable innovative development based on
a pentagon model of five critical success factors (CSFs) as illustrated in
Figure 1.1.
The five elements of that model that need to be mobilized for the
regional development process are:
1. The availability of productive capital (PC): this corresponds to neo-

classical production theory where output is determined by the tradi-
tional production factors labour and capital.
2. The presence of human capital (HC): this refers to the quality of
labour input obtained by means of education, training or new skills
(for example, in ICTs) and may be seen as a productivity-enhancing
PC
EC HC
SID
CC SC
Source: Compiled by author Peter Nijkamp.
Figure 1.1 A pentagon model of creative forces for sustainable regional

development

factor. Clearly a balanced distribution of human capital over people

is of great importance.
3. The access to social capital (SC): this condition comprises interaction
and communication between people, socioeconomic bonds, social
support systems, business networks (formal and informal), relations
based on trust, and so on.
4. The usage of creative capital (CC): this may be seen as a great ability
to cope with challenges and new opportunities, and is reflected in
entrepreneurial spirit, new ways of thinking and acting, trend-setting
artistic expressions, innovative foresights, and so forth. Such a factor
is often found in a multicultural urban melting pot.
5. The existence of ecological capital (EC): this condition takes for
granted that a favourable quality of life, an ecologically benign con-
dition in a city, presence of green space and water, or an attractive
living climate (for example, recreation and entertainment possibilities)
contribute significantly to the innovative and sustainable potential of
a region.
The pentagon factors can in principle be measured and quantified,

and then be put in an explanatory econometric model (for an empirical
estimation see Capello et al., 1999).
THE ORGANIZATION OF THIS BOOK
The remainder of this book is arranged around a number of themes. These

are outlined below.
Different Disciplinary Takes on Endogenous Regional Development
First there are five chapters in which scholars from different disciplinary
backgrounds provide their take on endogenous regional development:
from an economist, a geographer, a sociologist, a planner and a behav-
ioural/management scientist.
Kenneth Button provides an economists viewpoint (Chapter 2) in
which he refers to the powerful tool-kit of powerful techniques economists
have for analysing and understanding why some countries and regions
grow faster than others. He refers to a world without endogenous growth
and the well-known exogenous growth models in neoclassical economics
as illustrated by the Solow model before turning to discuss the emer-
gence of the new or endogenous growth theory from the 1980s, while
also reminding us of important precursors to the new growth theories

such as Schumpeters discussion of entrepreneurs in the 1930s. The Romer

model in particular is discussed, along with the contributions of Arrow
and Lucas, with technology, R&D and knowledge being considerations
of key importance in the development process. And there is a discussion
of the important issue of the empirical testing for regional convergence
and diversity in regional development. Button raises the question that if
indeed there are endogenous growth effects and if there is endogeneity in
the development process, then the opportunity for wider policy options
certainly opens up.
The geographers perspective is provided by Michael Taylor and Paul
Plummer (Chapter 3) in which the discussion of regional development is
placed firmly within the contemporary process of globalization and the
transition from the Fordist to a post-Fordist production regime. They
refer to the new economic geography and the new regionalism dis-
courses, focusing the discussion mainly on the latter and plugging force-
fully for the adoption of both quantitative and qualitative approaches to
analysis in bringing these discussions to bear in the empirical investigation
of regional development. There is discussion of the abstract nature of
theoretical models of regional growth and criticism of what are referred to
as stylized facts about local economic development and a dependence on
proxy variables in regional econometrics. At the same time the authors are
critical of the simplistic and under-theorized approach often taken in the
geographic literature. There is discussion of the notion of embeddedness
to explain the dynamics of local growth, especially as it related to learning,
innovative milieu and regional innovation systems in the new regionalism
debate, with the added emphasis on the advantages of geographical prox-
imity. Importantly there is a detailed discussion of how mixed-method,
but theoretically informed empiricism, research might be of use in framing
policy advice for regional development, with the authors providing a case
study of a project in which they were involved in Australia with a focus on
institutional capability in human capital development to enhance regional
development.
A sociologists perspective is provided by Frank Vanclay (Chapter 4)
in which he sets the discussion specifically in the context of rural devel-
opment in a local situation where the focus is explicitly on place-based
dimensions and initiatives. In keeping with much of the concern of the
sociologist, the emphasis is on understanding the natural, human and cul-
tural attributes of place. Vanclay refers to the notion of development being
about how a place might realize its potentialities to create sustainable
activity. Much consideration is placed on values and the achievement
of multifunctionality and a concern about the subsidiarity principle to
renew and revitalize the countryside and its community. In many ways the

approach outlined by Vanclay is very micro in scale and takes a cultural

economy approach to local development, with detailed reference being
made to the operation of the EUs LEADER (Liaison Entre Actions de
Dveloppement de lEconomie Rurale) programme.
Edward Blakely (Chapter 5) provides a planners perspective that
focuses on how to derive development outcomes from local resources, the
harnessing of which is necessary to produce results that are tangible. He
discusses how planning relies on what he calls indigenous activities and/
or endogenous resources to shape outcomes in space and place. To assist
this it is necessary to formulate a Plan, as plans (about which there is a
lengthy discussion) are descriptive instruments but they need implemen-
tation. Thus, there is a considerable emphasis on what might be called
institutional factors. Not surprisingly for a planner, Blakely places con-
siderable emphasis on land as a resource which needs to be capitalized
on and its use as an endogenous base in local or regional development.
There is also discussion of the importance of planning strategy that seeks
equitable and sustainable wealth outcomes. Institutional capability and
capacity building is crucial for the process to succeed. Blakely provides a
discussion of how for many places the challenge is how to overcome the
product cycle trap through harnessing their endogenous resources, and
of how social capital can play an important role.
A behavioural/management science perspective is provided by Tojo
Thatchenkery and Jessica Heineman-Pieper (Chapter 6) in which the
emphasis is on human capital, cultural and organizational dimensions.
They show how an endogenous ecology coupled with technology inno-
vation may spontaneously create cascading entrepreneurial activity, as
highlighted in the case of Silicon Valley and the social and cultural
embeddedness of its actors. The authors present a framework they call
appreciative intelligence and discuss how this concept may be transfer-
able as an endogenous process in a wider context of regional development.
The discussion of this perspective also incorporates the broader regional
science literature on knowledge spillovers and industrial districts and the
long-established notion of agglomeration.
Measuring and Modelling Endogenous Regional Growth and Development
The second theme that is addressed in this book relates to the measure-
ment and modelling of endogenous regional growth and development.
Somewhat surprisingly there are few examples of models that have been
explicitly developed and empirically tested to actually measure endogen-
ous regional growth or to investigate the determinants of spatial variations
in regional performance on an endogenous growth measure.

In Chapter 7 Robert Stimson and Roger Stough demonstrate how they

have used the regional (differential) shift component derived from a shift-
share analysis of regional employment change, standardized by the size
of the regional labour force at the beginning of the period, as a surrogate
measure of endogenous regional growth (or decline). The authors present
a new model framework for endogenous regional growth. They then
provide summary results from two exploratory attempts to model regional
endogenous employment change using spatial econometrics regression
modelling, first in a study of non-metropolitan government local areas
across Australia, and second in a study of metropolitan regions in the US.
In the second application of the model framework the authors are able to
incorporate proxy variables for institutional and leadership factors and
measures of entrepreneurial activity, in addition to a large set of resource
endowment measures, as explanatory variables in their model.
A specific attention is paid to entrepreneurship in regional development
in the chapter by Zoltan Acs and Mark Sanders (Chapter 8). The focus is
on analysing rents and endogenous entry in regional growth. The authors
present a model in which incumbent firms finance R&D for a clear profit
motive but entrepreneurs capture the rents of commercializing the oppor-
tunities that this R&D generates through regional intra-temporal knowl-
edge spillovers. The model formalizes the knowledge spillover theory of
entrepreneurship, predicting that in a decentralized equilibrium R&D will
be under-funded and that there is room for welfare gains through policy.
But a delicate balance now needs to be found when R&D and entrepre-
neurship compete over the same resources, and the social optimum is to
stimulate both R&D and entrepreneurship and invest resources in facili-
tating the knowledge spillover to entrepreneurs. This modelling approach
is seen to turn the Romer argument on incentive structure on its head.
Acs and Sanders show that their results can carry over to the regional
level when considering the impact of limited geographical labour mobil-
ity, transport costs and communication costs. That has important policy
implications at the aggregate and regional levels, but the challenge is to
ascertain the exact channels through which such spillovers arise.
Addressing Endogenous Factors at Different Levels of Scale
There follows a series of chapters in which endogenous factors in regional

development are considered at a number of different levels of scale.
Taking a national economy perspective, in Chapter 9 Tomokazu Arita,
Chie Iguchi and Philip McCann address the role that foreign direct invest-
ment plays in three of the worlds largest newly industrializing countries:
China, India and Indonesia. The focus of their analysis is on an evaluation

of institutional factors that are operating in the context of globalization

and rapidly improving information technologies. The combination of
these features produces complex interrelationships between economic
geography, economies of scale and global trade and investment, and the
authors review those interrelationships in the three rapidly developing
case study countries. The authors highlight how the economic growth
of these three countries is dominated by specific large urban regions
where agglomeration economies appear to be critical, likely exacerbating
regional disparities. But the authors identify differences in roles played by
localized technology spillovers.
In a study of disparities in inter-regional performance across the EU,
Daniela Constantin, Zizi Goschin and Gabriela Dragan (Chapter 10)
focus on the performance of ten new member states and evaluate the
institutional capabilities affecting their absorption capacity. The discus-
sion focuses on the implications of the EUs structural assistance pro-
grammes (a part of the social cohesion policy) on regional disparities for
the new member states and the regions within them. The authors use Gini
coefficients and the covariation, weighted by population, to investigate
the degree to which convergence is occurring, showing how divergence
has actually increased. Institutional inefficiencies and ineffectiveness are
highlighted as potential endogenous barriers or impediments.
In Chapter 11, Roberto Camagni and Roberta Capello also discuss the
issue of differentials in the performance of regions across the EU nations.
Their concern is with investigating regional competitiveness through the
concept of territorial capital. They provide a set of theoretical perspec-
tives on the regional convergence/divergence debate, and discuss the rela-
tionships between regional performance and exogenous and endogenous
factors. Using the regional economic growth simulation model known as
MASST, the authors investigate the impact of macroeconomic policy at
the regional level across EU regions, the impact of EU regional policies
such as the structural funds and the role of specific elements of territorial
capital on regional growth patterns.
Moving to the level of scale of the city, in Chapter 12 Alistair Robson
provides a comparative analysis of endogenous regional employment
growth performance for each of Australias five largest cities Sydney,
Melbourne, Brisbane, Perth and Adelaide over the decade 1996 to 2006.
He conducts a shift-share analysis of employment change and focuses in
particular on the regional shift component as an indicator of endogenous
performance. The analysis includes segmenting the analysis by industry
sectors. Marked differences are apparent in the performance of the cities.
At a much more local level of scale, in Chapter 13 John Martin looks
at the performance of four small towns in two rural shires in the state of

Victoria in Australia in the wider context of population and economic

decline that is apparent widely across small towns. Martin focuses on the
role of local leadership in particular the role of local government and
social capital in those four small communities through a sociometry
framework. The author demonstrates how leadership and institutional
factors, including strategies enhancing social capital, have played a signifi-
cant role in the local planning process to try to enhance the survival and
functioning of some of those small towns in a rural region.
In Chapter 14 Terry Clower investigates the influence of economic
development incentives through government industry assistance pro-
grammes as an institutional factor in the states of Alabama and Texas
in the US. These are common initiatives in response to plant closures at
the local level. The author provides a detailed case study of a food manu-
facturing firm relocation to Grayson County in the Sherman-Denison
Metropolitan Area of Texas under the Texas Enterprise Fund. The dis-
torting effect of such a one-off exogenous incentive on the results of a
regional employment shift-share analysis is clearly demonstrated. There is
also a discussion of what might be better endogenous incentives to grow
and attract industry in local development policy.
A Way Ahead?
In the final chapter (Chapter 15) Peter Nijkamp and Roberta Capello
point out that regional growth theory covers a little more than half a
century and during that time we have deepened our knowledge and
understanding of the complex backgrounds and impediments to balanced
regional development. In that context endogenous considerations are of
great importance.
Nijkamp and Capello provide a provocative review of approaches to
investigate regional development, which begins with a discussion of the
vexed question of convergence hypothesis in the debate on regional dis-
parities. In particular they draw attention to the policy issues involved,
referring to the need for regions to achieve a role in the international
division of labour and maintain it over time, and pointing to the implica-
tions of regional growth trajectories in coping with sustainability issues
and scarcity of environmental resources, such as environmental quality,
natural resource security, urban quality of life, or spatial biodiversity. The
authors debate the need for theories of regional growth and development
to be more realistic and to better incorporate the complex and interac-
tive behaviour and processes that take place in space and to understand
regional competitiveness in terms of endogenous factors. In particular
they refer to the importance of the spatial distribution of knowledge and

its spillovers as important success factors for regional development in an

open competitive economic system and how they are critical for regional
welfare creation.
Nijkamp and Capello argue strongly for a more dynamic approach to
theories and models while recognizing the difficulties of the complexity
of time with spatial analysis through non-linear approaches. Finally, the
authors write of the institutional challenges for regions pursuing a sustain-
able development path and the crucial role of innovation processes.
The chapter concludes with the prophetic statement that regional
growth is thus a race without a finish; and that also applies to regional
growth theory. Hopefully, the chapters in this book contribute to that
race.
REFERENCES
Acs, Z.J., Anselin, L. and Varga, A. (2002): Patents and Innovation Counts as
Measures of Regional Production of New Knowledge, Research Policy, 31(7),
106985.
Arthur, W.B. (1994): Increasing Returns and Path Dependency in the Economy,
University of Michigan Press, Ann Arbor.
Barro, R.J. (1990): Government Spending in a Simple Model of Endogenous
Growth, Journal of Political Economy, 98(5), S103S125.
Blakely, E.J. (1994): Planning Local Economic Development: Theory and Practice,
2nd edition, Sage Publications, Thousand Oaks, CA.
Capello, R. (2009): Spatial Spillovers and Regional Growth: A Cognitive
Approach, European Planning Studies, 17(5), 64360.
Capello, R. and Nijkamp, P. (eds) (2009): Handbook of Regional Growth and
Development Theories, Edward Elgar, Cheltenham, UK and Northampton,
MA, USA.
Capello, R., Nijkamp, P. and Pepping, G. (1999): Sustainable Cities and Energy
Policies, Springer-Verlag, Berlin.
Dring, Th. and Schnellenback, J. (2006): What Do We Know About Geographical
Knowledge Spillover and Economic Growth?, Regional Studies, 40(3), 37595.
Erickson, R.A. (1994): Technology, Industrial Restructuring and Regional
Development, Growth and Change, 25(3), 35379.
Erickson, R.A. and Leinbach, T. (1979): Characteristics of Branch Plants
Attracted to Non Metropolitan Areas, in: Lonsdale, R. and Seyter, H.L. (eds):
Non Metropolitan Industrialization, Winston, Washington, DC.
Etzkowitz, H. and Leydesdorff, L. (1996): The Future Location of Research:
A Triple Helix of UniversityIndustryGovernment Relations, II, EAAST
Review, 15(4), 2025.
Etzkowitz, H. and Leydesdorff, L. (1997): Universities and the Global Knowledge
Economy, a Triple Helix of UniversityIndustryGovernment, Pinter, London.
Etzkowitz, H. and Leydesdorff, L. (2000): The Dynamics of Innovation: From
National Systems and Mode 2 to a Triple Helix of UniversityIndustry
Government Relations, Research Policy, 29(2), 10929.

Florida, R. (2002): The Rise of the Creative Class: And How Its Transforming
Work, Leisure, Community and Everyday Life, Basic Books, New York.
Fukuyama, F. (1995), Trust: The Social Virtues and Creation of Prosperity, Free
Press, New York.
Garlick, S., Taylor, M. and Plummer, P. (2006): An Enterprising Approach to
Regional Growth: The Role of VET in Regional Development, NCVER (National
Centre for Vocational Education Research), Australian National Training
Authority, Adelaide.
Grossman, G.M. and Helpman, E. (1991): Innovation and Growth in the Global
Economy, MIT Press, Cambridge, MA.
Hertz, N. (2009): Cooperation Must Rule, Nature, 457(7232), 9623.
Huxham, C. (1996): Collaborative Advantage, Sage, Thousand Oaks, CA.
Johansson, B., Karlsson, R. and Stough, R.R. (eds) (2001): Theories of Endogenous
Regional Growth, Springer, Berlin.
Keeble, D. and Wilkinson, F. (1999): Collective Learning and Knowledge
Development in the Evolution of Regional Clusters of High-technology SMEs
in Europe, Regional Studies, 33(4), 295303.
Knight, R.V. and Gappert, G. (eds) (1989): Cities in a Global Society, Sage
Publications, Newbury Park.
Krugman, P. (1991): Geography and Trade, MIT Press, Cambridge, MA.
Krugman, P. (1995): Development, Geography and Economic Theory, MIT Press,
Cambridge, MA.
Krugman, P. (1996): The Self-Organising Economy, Blackwell, Oxford.
Krugman, P. and Venables, A. (1996): Integration, Specialization and
Adjustment, European Economic Review, 40(35), 95967.
Lucas, R.E. (1988), On the Mechanics of Economic Development, Journal of
Monetary Economics, 22(1), 342.
Markusen, A. (1985): Profit Cycles, Oligopoly and Regional Development, MIT
Press, Cambridge, MA.
Nijkamp, P. (forthcoming): E Pluribus Unum, Region Direct.
Nijkamp, P. and van Hemert, P. (forthcoming): Knowledge Infrastructure and
Regional Growth, Regional Symbiosis.
Nijkamp, P., Vleugel, J., Maggi, R. and Masser, I. (1994): Missing Transport
Networks in Europe, Ashgate, Aldershot.
Norton, R.D. and Rees, J. (1979): The Product Cycle and the Decentralization of
North American Manufacturing, Regional Studies, 13(2), 14151.
OECD (2000): Links between Policy and Growth: Cross-country Evidence, draft
paper for Working Party 1, Economics Department, OECD.
OECD (2001): The Well-Being of Nations: The Role of Human and Social Capital,
Centre for Educational Research and Innovation (CERI), OECD Publishing,
Paris.
Ohmae, K. (1995): The End of the Nation State: The Rise of Regional Economics,
Free Press, New York.
Porter, M.E. (1990): The Competitive Advantage of Nations, MacMillan, New York.
Prudhomme, R. (1995): On the Economic Role of Cities, in Cities and the New
Global Economy, papers from an international conference presented by the
OECD and the Australian Government, Australian Government Publishing
Service, Canberra, 3, 72842.
Rebelo, S. (1991): Long Run Policy Analysis and Long Run Growth, Journal of
Political Economy, 98(5), S71S102.

Rees, J. (1979): State Technology Programs and Industry Experience in the USA,
Review of Urban and Regional Development Studies, 3(1), 3959.
Rees, J. (2001): Technology and Regional Development: Theory Revisited, in:
Johansson, B., Karlsson, Ch., Stough, R.R. (eds): Theories of Endogenous
Regional Growth, Springer-Verlag, Heidelberg, pp. 94110.
Romer, P.M. (1986): Increasing Returns and Long-run Growth, Journal of
Political Economy, 94(5), 100237.
Romer, P.M. (1990): Endogenous Technological Change, Journal of Political
Economy, 98(5), S71S102.
Scott, A.J. (1988): New Industrial Spaces: Flexible Production Organization and
Regional Development in North America and Western Europe, Pion, London.
Simmie, J. (ed.) (1997): Innovation, Networks and Learning Regions?, Jessica
Kingsley, London.
Solow, R.M. (1956): A Contribution to the Theory of Economic Growth,
Quarterly Journal of Economics, 70(1), 6594.
Solow, R.M. (2000): Growth Theory: An Exposition, Oxford University Press, New
York.
Stimson, R.J. and Stough, R.R. (with Salazar, M.) (2009): Leadership and
Institutions in Regional Endogenous Development, Edward Elgar, Cheltenham,
UK and Northampton, MA, USA.
Stimson, R.J., Stough, R.R. and Roberts, B.H. (2006): Regional Economic
Development: Analysis and Planning Strategy, revised edition, Springer, Berlin.
Taylor, I. (2009): Learn to Convince Politicians, Nature, 457(7232), 9589.
Thomas, M.D. (1975): Growth Pole Theory, Technological Change and Regional
Economic Growth, Papers of the Regional Science Association, 34(1), 325.

2. The economists perspective on
regional endogenous development
Kenneth Button
There must be a purely economic theory of economic change which does not
merely rely on external factors propelling the economic system from one equi-
librium to another. (J. Schumpeter, Preface to the Japanese edition of Theorie
der Wirtschaftlichen Entwicklung, 1911, transl. 1934)
Economic progress, in capitalist society, means turmoil. And, ... in this turmoil
competition works in a manner completely different from the way it would
work in a stationary process, however perfectly competitive. Possibilities for
gains to be reaped by producing old things more cheaply are constantly materi-
alizing and calling for new investments. These new products and new methods
compete with the old methods not on equal terms but at a decisive advantage
that may mean death to the latter. (J. Schumpeter, Capitalism, Socialism, and
Democracy, 1942)
INTRODUCTION
Economists have developed a tool-kit of very powerful techniques for ana-

lyzing a wide range of social phenomena but they have, beyond gaining a
very general understanding, been singularly unsuccessful in unanimously
agreeing on an explanation of why some regions or countries economies
grow faster than others.1 This is an important intrinsic question, but it also
raises issues about the roles of institutions that have been established to
narrow the gap between the rich and the poor.
Those involved in trying to explain economic growth have included the
luminaries of their generations, and their political leanings range from the
capitalist market orientation of Joseph Schumpeter to the centralism of
Karl Marx. Indeed, if one accepts Milton Friedmans (1953) criteria for a
good model, namely its ability to produce good forecasts, then economic
growth models have been sadly inefficient, although, one might add, that
by the standards of other social science disciplines, the notions of precision
used by economists are particularly demanding.
The expansion of material development and enhanced human welfare
20

The economists perspective 21
has always been at the core of modern economics. Adam Smith, David
Ricardo, Malthus and other early classical economists offered a variety
of explanations for improving economic efficiency, focusing on the ben-
efits of the division of labor in one way or another, but they were largely
focused on squeezing more production from a given resource base than
economic growth per se.2 Their concern was essentially with what we now
consider comparative statics rather than dynamics.
More recently the importance of technological progress in economic
growth has attracted attention. Initially, and following the pioneering
work in the 1920s of Frank Ramsey (1928), strenuous intellectual efforts
have been made to incorporate into economic models the changes in tech-
nology that have manifestly affected economic growth since the Industrial
Revolution. The basic point being that capital and labor augmenting
technology continually increases returns to factors of production, leading
to additional investment, allowing per capita output to increase seemingly
indefinitely. The challenge has been to understand where this technical
progress comes from, how it becomes embedded in production processes
and how it will, in quantitative terms, influence the economic growth rates
of different regions. While this has largely been treated as a question of
macroeconomic interest, there are important spatial implications espe-
cially because technological progress is not ubiquitous or cannot always
be easily and quickly transferred between regions.
BACKGROUND: A WORLD WITHOUT

ENDOGENOUS GROWTH
In the 1960s the UKs Royal Economic Society and the American Economics
Association commissioned a series of survey papers by leading figures in their
fields. Two of these, that by John Meyer (1963) on Regional Economics,
and by Frank Hahn and Robin Matthews (1964) on The Theory of
Economic Growth provide a useful basis for considering how and why
endogenous growth theory has developed over the past 45 years or so.
While there was concern at the time amongst regional economists over
spatially differential growth patterns, the emphasis of academic work, and
possibly reflecting the dominance of Keynesian economics at the time,
was more on short-term considerations. In particular there was interest
in the tracing through via multiplier and input-output analysis the local
implications of expanding a regions export base.
The neoclassical world of macroeconomic growth theory before the
1980s thought that growth was the result of forces that impinged from
the outside rather than being a product of the economic system itself.

Per capita income growth was seen as dependent on saving rates and on
exogenous shocks to the steady-state growth path that resulted.
The empirical work of the time focused on extracting the inputs of the
factors of production de facto, labor inputs from the growth process
and then treated differences in national or regional growth as an account-
ing residual. To be fair, however, in part this was as much due to data
weaknesses as to intellectual understanding.3 As we see below, endogenous
growth theory goes beyond this and seeks to look at the private and public
sector choices that cause these residuals to vary.
Comparative Statics
Traditional macroeconomic, long-run growth theory at the time these

papers were written was largely divided between theories that assumed
economic growth could occur without technical progress and those that
focused, albeit in a particular way, on technology impacts. The former
were very much in the classical tradition of Smith with a focus on growth
in the labor force. In equilibrium, the growth of output is limited by the
growth of the labor force and that implies a constant per capita income.
Where regional, as opposed to macro, growth (at any level of aggrega-
tion) models differed from this to some extent was to treat labor supply as
both dependent on the internal demographics of a region and upon migra-
tion between regions. Basically, it allows for the spatial mobility of factors
of production. Hence, a specific region may grow when there is capital
abundance by importing labor from other regions. This, combined with
the natural growth in labor supply within the region, can lead to growth
through more efficient use of the labor stock although in the long run there
will be convergence in per capita income.
Figure 2.1, developed from Harts (1975a, 1975b) synthesizing work,
offers a simple illustration of the point. We assume there to be two regions,
A and B. The former enjoys high-income levels and low unemployment
whilst B is the mirror image of this.4 There are decreasing returns to factors.
The classical economic model assumes that with zero costs of migration
and a homogeneous labor force, labor will move from B to A seeking
work and higher pay whereas, on the assumption of uniform commercial
risk across regions, capital will move from region A to B where it can be
combined with abundant, cheap labor to maximize returns. Wages will fall
in A, unemployment will increase, and the return on capital will rise as the
labor supply increases and capital becomes scarcer. The additional amount
of capital and the decreasing size of the labor force in region B will push
down the marginal return on investment and concurrently push up wages.
The process continues until labor costs and unemployment levels are

Region A Region A
Capital Labor Capital Skilled

Labor
Region B Region B
Classical Model New GrowthTheory Model
Note: I = income; U = unemployment.
Source: Developed form Hart (1975a and 1975b).
Figure 2.1 Simplified theories of migration
equalized.5 This equalization is achieved in a world of zero transpor-

tation costs and full information about employment and investment
opportunities. There are no legal impediments to factor mobility, and
racism is absent. Essentially the model is devoid of any real geographical
consideration.
Exogenous Growth Modes
The neoclassical theory can be treated as an extension to the well-known

Harrod-Domar model6 of economic growth developed in the 1940s that
includes an additional term reflecting productivity growth. The broad
theory is typified by the work of Robert Solow (1956) and the Australian
Trevor Swan (1956) on changes in the factor endowment of a country.
In their neoclassical models, the long-run rate of growth is exogenously
determined in other words, it is determined outside of the model. A
common prediction of these models is that an economy will always con-
verge towards a steady-state rate of growth that depends on the rate of
technological progress and the rate of factor accumulation. A country
with a higher saving rate, for example, will experience faster growth.
In the Harrod-Domar framework, steady-state growth is unstable; it
is on a knife-edge in the sense that any deviation from that path would
result in a further, and subsequently cumulative, move away from that

path, either up or down. However, Solow, Swan and others argued that
the capitaloutput ratio of the Harrod-Domar model should not be
regarded as exogenous. They proposed a growth model where the capital
output ratio was precisely the adjusting variable that would lead a system
back to its steady-state growth path.
The key assumption of the neoclassical economic growth model is that
capital is subject to diminishing returns. Given a fixed stock of labor, the
impact on economic output of the last unit of capital accumulated will
always be less than the one before. Assuming for simplicity no techno-
logical progress or labor force growth, diminishing returns implies that
at some point the amount of new capital produced is only just enough to
make up for the amount of existing capital lost due to depreciation. At this
point, because of the assumptions of no technological progress or labor
force growth, the economy ceases to grow.
Assuming non-zero rates of labor growth complicates matters at the
regional level, where migration may be large, somewhat, but the basic
logic still applies in the short run the rate of growth slows as diminishing
returns take effect and the economy converges to a constant steady-state
rate of growth (that is, no economic growth per capita).7
Including non-zero technological progress is very similar to the assump-
tion of non-zero workforce growth: a new steady state is reached with con-
stant output per worker-hour required for a unit of output. However, in
this case, per capita output is growing at the rate of technological progress
in the steady state (that is, the rate of productivity growth).
More formally, an exogenous growth model can be formulated by
taking a simple Cobb-Douglas production function of the form, Y = Y =
A(t)K1b Lb (where Y is net national product, K is the capital stock, L is the
labor stock and A is the level of technology). The fact that A is a function
of time (t) indicates the standard neoclassical assumption that technology
only improves over time for reasons external to the model.
ENDOGENOUS GROWTH MODELS
Adherents to endogenous economic growth theory cite three main limita-

tions of the neoclassical model:
1. It relies heavily upon technological change to supply growth in per

capita income but has no mechanism for explaining the sources for
such change.
2. It offers only a very rudimentary framework for assessing the effects
of government policy, and while government actions may not be

able to raise long-run growth rates, government interventions do

affect behavior and this, in aggregate, affects the growth path (be it
positively or negatively).
3. The model has limited capabilities for analyzing trade between regions
or countries and the links between such trade and economic growth.
According to the new or endogenous growth theory that began to

emerge in its current form in the 1980s, economic growth can be under-
stood as a process of learning-by-doing, within a firm, within an industry
and within a given spatial jurisdiction such as a region or metropolis.
While there were earlier attempts to indigenize technical progress much
of the credit for the modern formulation of endogenous growth theory is
attributed to Paul Romer (1986, 1990a) and Robert Lucas (1988, 1993).
By way of a counterpoint, while exogenous growth theory sees rising
output per capita as resulting from externally given increases in the quanti-
ties of labor and capital with no internal technological or organizational
change, and no economies of scale, but only constant returns in an envi-
ronment altered solely by investment levels and labor force growth, endog-
enous growth models see economic growth over time entailing increasing
returns to scale for a metropolis or a national economy. A proportion-
ate increase in labor and capital gives rise to more than proportionate
gains in output. The explanation lies in better recipes, as Romer terms
innovations, and in spillovers that operate over time, enhancing skill and
productivity levels throughout the economy. This approach to growth has
the advantage of supplementing the neoclassical model, without destroy-
ing it, by refining the ways in which technology change is stimulated and
absorbed in economies.
The need for a more complete theory of technical change was there in
the 1980s, as pointed out by Romer (1994), but earlier efforts aimed at
bringing technology in from the cold had not been analytically developed
or fully embraced in the theorizing.
Precursors to Modern Endogenous Growth Theory
In his Capitalism, Socialism, and Democracy, Schumpeter (1942), albeit

not in a mathematically rigorous fashion, laid the foundations for much of
the subsequent work on endogenous growth theory. The basis of his anal-
ysis, however, had more to do with the nature of general political systems,
and their implications for national welfare than with per capita income per
se, and it was not developed at the regional level.8 Nevertheless, the idea
was that within a capitalist system the continual pressure on profit margins
of incumbent businesses generated by the pressures of entrepreneurs and

imitators led to an inevitable push within business to be at the cutting

edge of technology. Put another way, in Schumpeters interpretation of
capitalism, innovative entry by entrepreneurs was the force that sustained
long-term economic growth, even as it destroyed the value of established
companies that enjoyed some degree of monopoly power.
Part of the subsequent stimuli to develop a more technically rigorous
endogenous framework in the 1950s and 1960s also came from reassess-
ments of earlier empirical findings.9 According to calculations by Solow
(1957), for example, 87.5 percent of growth in output in the US between
1909 and 1949 could be ascribed to technological improvements alone,
suggesting that the Solow residual was large. One of the first reactions
of the neoclassical economists was to argue that by reducing much of
that influence to pure capital improvements in the calculations, capital-
intensity seemed to play a larger role than suggested by the arithmetic,
although, Solow (1960) did later argue that increased capital-intensive
investment embodies new machinery and new ideas as well as increased
learning for even further economic progress.10 Nevertheless, there were
still significant issues to be addressed.
Nicholas Kaldor was really the first major post-war economic theo-
rist to consider endogenous technical change as a potentially important
element in explaining differences in regional and national growth rates. In
a series of papers, including the oft-cited contribution with James Mirrlees
(Kaldor and Mirrlees, 1962), Kaldor posited the existence of a technical
progress function and that per capita income was indeed an increasing
function of per capita investment. Thus, learning was regarded as a
function of the rate of increase in investment. However, Kaldor held that
productivity increases had a concave nature (i.e., increases in labor pro-
ductivity diminish as the rate of investment increases). This proposition
falls short of Solows insistence on constant returns.
Kenneth Arrow (1962) took the view that the level of the learning
coefficient is a function of cumulative investment (i.e., past gross invest-
ment). Unlike Kaldor, Arrow sought to associate the learning function
not with the rate of growth in investment but rather with the absolute level
of knowledge already accumulated, a stock rather than a flow concept.
Because Arrow claimed that new machines are improved and/or are more
productive versions of those in existence, investment does not only induce
productivity growth of labor on existing capital (as Kaldor would have it),
but it would also improve the productivity of labor upon all subsequent
machines made in the economy.
The trick is to utilize the concept that while firms face constant returns,
the industry or economy as a whole takes increasing returns into account.
This can be formalized using the Cobb-Douglas production function:

Y 5 AKaL1 2a
where there are constant returns to scale for all inputs together (i.e., a 1
(1 2 a) 5 1). Therefore, it seems as if output per capita, and with it con-
sumption per capita, does not grow unless the exogenous factor, A, grows
too.
To indigenize A, the Cobb-Douglas production function for each
individual firm is defined as:
Yi 5 AiKai L1i 2a
where the output of an individual firm is related to capital and labor as

well as the augmentation of labor by Ai. Arrow assumed that Ai, while it
looks specific to the firm, is in fact related to knowledge in the economy.
This knowledge and experience is common to all firms.
Arrow argued that knowledge accumulation arises from the past cumu-
lative investment (G) of all firms, that is, the technical augmentation factor
is related to economy-wide aggregate capital in a process of learning-by-
doing. In other words, the experience of the particular firm is related to
the stock of capital in the economy, G, by Ai 5 Gz. Hence, as the physical
capital stock G accumulates, knowledge used by a particular firm also
accumulates by a proportion z such that 0 , z , 1. Transferring to the
production function for an individual firm yields:
Yi 5 G 2zK ai L1i 2a
where only G does not have a subscript i, it is a productive force external

to the firms and assumed a quasi-public good. Thus, constant returns to
scale are maintained at the firm level, but in the aggregate G 5 K because
it is only the accumulated stock of capital for the economy. The economy-
wide aggregate production function is therefore:
Y 5 K a 1 zL1 2a
Arrow assumed that a 1 z , 1; hence increasing only capital does not

lead to increasing returns. Increasing returns to scale can be obtained by
a 1 z 1 (1 2 a) 5 z . 0, but capital and labor must both expand.
However, by adding this restriction, Arrows original model meets Solows
condition by exhibiting non-increasing returns to scale in aggregate if the
rate of growth in an economy is steady.

The More Recent Models
The 1980s saw an upsurge of interest in economic growth theory partly

because major structural shifts were occurring as the service sector, and
particularly information-based industries, began to grow in importance
and as entities such as the EU began to fulfill some of their potential.
Essentially, changes were taking place at both the technical and institu-
tional levels that were seen as potentially affecting economic growth.
Paul Romer (1986), for example, went to great lengths to disqualify
the restriction retained by Arrow decreasing returns to capital. His
main initial thesis, however, was that long-term economic growth stems
from the accumulation of knowledge; knowledge being seen as a factor of
production embodied in capital and labor.11
Taking Kenneth Arrows idea of disembodied knowledge, Romer con-
cludes that there indeed could be constant returns, but argues that the rate
of growth of K alone may yield increasing returns; de facto he assumed
(a 1 z) . 1 was possible. With this externality, the growth rate of capital
gK equals the growth rate of per capita consumption gc 5 gC 2 gL such
that gK 5 gc . 0; that is, there is constant positive growth in per capita
consumption and capital. It is the externality, from learning-by-doing,
that gives a positive growth rate for consumption and output. In addition,
the higher the level of disembodied knowledge, the more soil exists upon
which innovation (i.e., increases in productivity) can work and the higher
the rate of technical progress.12
There is a problem with Romers model. The size of the labor force
enters as an argument. Thus, if L increases, then the growth rate of con-
sumption increases the Malthusian effect. This is an unattractive feature
that could be solved if we had originally divided the accumulated knowl-
edge equation by L. However, the notation in the model gets unnecessarily
complicated after this and is not included here.13
Finally, the social planners solution to the overall Arrow-Romer model
would inevitably be different from the competitive solution presented.
Arrow assumed that any firm is so small relative to the overall economy
that its investment decisions cannot have any important effect on the
aggregate stock of capital or knowledge. In sum, the firm takes no account
of the second-round effects of its investments on technological progress,
but rather maximizes profits through its choice of labor and capital taking
the stock of knowledge as given. A social planners idea of the marginal
product of capital would be different, as she would take the externality
into account and that, in turn and if it could be done effectively, would
yield a higher steady-state growth rate. Thus, the equilibrium growth rate
under a competitive system is smaller than the optimal growth rate under

a social planner because firms are not, individually, taking the externality
into account. In other words, with full information and an effective tool-
kit there is a role for public policy.
Learning-by-doing within a firm means current unit costs are a function
of experience (as measured by the firms cumulative past output). Given
the learning curve for a single firm, then imitation of successful firms
on the part of other firms in the industry spreads the learning around,
such that the industry as a whole can benefit from falling-forward supply
curves. The process links unit costs to cumulative industrial output within
a country or region. The ease of imitation and learning then increases
within spatial agglomerations, which in turn can be understood as what
are often called little nations benefiting from increasing returns.
Romer refines his work by looking at why private firms directly invest in
technical knowledge by undertaking R&D when it has the characteristics
of a quasi-public good; knowledge only being partially excludable means
that the firm will not reap the full benefits. Patents, trademarks and copy-
rights are assumed by Romer, as they had been in previous work by Joseph
Schumpeter (1942),14 to protect, at least for a time, part of the knowledge
acquired, but even without them a firm generally enjoys a first mover
advantage and can costlessly reproduce for sale or internal use a new
design. It can sell these new products for a time at prices above marginal
cost and earn super-normal profits, but even as the short-term monopoly
position is eroded there is the incentive to further innovate because of the
lowered R&D costs that stem from the previous experience.15 In effect,
R&D activities become a barrier to entry of competition into the market.
Robert Lucas (1988, 1993) takes a somewhat different approach to the
spillovers from investment, focusing on education and on-the-job training.
Essentially, the interest is more on human capital rather than in physical
capital and R&D as in Romers framework.
One approach followed by Lucas involves assuming that personal
human capital is built up by workers through their postponing current
consumption in the hope that education will allow for greater consump-
tion in the future. To develop an endogenous growth effect from this, he
assumes that there is an externality from this activity because the actions
of these individuals will enhance the overall productivity of the labor
force, including those that do not defer consumption. Lucas found empiri-
cal support for this in the spatial agglomeration that can be observed,
especially in cities.
A second approach is to assume that knowledge is obtained on the job
rather than through explicit formal educational programs. Hence, labor
does not leave the workforce to acquire and develop knowledge, but gains
it from having to produce an ever-increasing range of more sophisticated

products.16 Lucas supports this line of argument by looking at the experi-

ences of some of the newly industrializing countries of Eastern Asia that
have experienced rapid growth in per capita income through selling to
diverse and continually changing export markets despite a much slower
expansion in the formal education system than his earlier work suggested
would be needed.
In sum, interpreted variously in terms of Arrows learning curves,
Romers recipe for growth, or Lucass vector-autoregressive (VAR)
time-series specifications, the key ideas underlying endogenous economic
growth are learning-by-doing and cross-fertilization over time. There are
effective feedback-loops in the economic system.
IS THERE A SPECIFIC REGIONAL CONTEXT?
As we have seen, one implication of Solows neoclassical growth model

is that each country or region should converge on to its own steady-state
growth path at a predictable rate. If one is willing to assume that technol-
ogy is identical across countries, then one should expect any two countries
with identical saving rates and population growth rates to converge to the
same level of income. Similarly, all countries should converge to a level of
income that can be predicted by its saving rates and population growth
rate. Regions, in theory, simply mirror this process, although the greater
flexibility in capital and labor markets within a country would suggest
that intra-national convergence would be more rapid than international
convergence. Empirical evidence for convergence has repeatedly been
interpreted, at either level of aggregation, as providing support in favor of
the neoclassical model and against endogenous growth models, which we
move onto, in both the national and regional context.17
The new growth theory, in contrast to the neoclassical framework, is
consistent with divergence in economic growth along the traditional lines
of Myrdals (1957) ideas of circular and cumulative causation, albeit for
somewhat different reasons. The situation can be illustrated by returning
to Figure 2.1.18 Taking the initial starting positions for two regions, this
approach argues that not only will equalization of real wages and employ-
ment levels not be attained but that there may be cases where they diverge
further. Labor mobility between regions A and B may be impeded by the
various costs of migration embracing social and search costs as well as
simple financial costs and heterogeneity in the labor market the jobs
available in A not being compatible with the skills and knowledge of labor
in region B. This is a potentially serious constraint in knowledge industries
in either of the Lucas cases because of the costs of acquiring education in

his initial model, and the lack of appropriate on-the-job training in the
second.19
Equally, capital does not move from region A to B because of the higher
returns and less uncertainty that are to be found in regions that already
have a high level of prosperity and a pool of complementary labor. While
the earlier regional models of Kaldor, in particular, focused largely on
divergent growth rates in the context of traditional-style manufacturing
industries that still dominated Western economies, the more contem-
porary form of the theory pays particular attention to the endogenous
growth that occurs in regions that have an established high-skilled work-
force, and the ability to further develop their knowledge-based industries.
Again, there is no substantive difference in the modeling whether it is con-
ducted at the national level or for smaller regions, although institutional
and cultural factors may influence the speed and detail of the growth paths
of either.
EMPIRICAL TESTING FOR ECONOMIC

CONVERGENCE
Testing the validity of the alternative theories, in the absence of easily

quantifiable counterfactuals, has frequently involved looking at secondary
evidence, and in particular at situations that shed light on whether there
is convergence in the economic growth paths of regions or, at the macro-
level, nations.
The empirical question that is explored at the specific regional level of
aggregation becomes one of whether there is convergence in economic
growth rates in, generally, per capita income as is a natural outcome
of the neoclassical model, but only possible with endogenous growth
under rather particular circumstances.20 The aim here is not to even try
to provide a comprehensive critique of the numerous studies of growth
paths for the variety of economic variables (e.g., income, employment
and prices) that have been conducted across a range of different levels of
spatial aggregation, locations and time periods, but rather to offer a flavor
of what has emerged at the regional level, from some of the more cited,
and thus influential, pieces of analysis.
The extensive empirical analysis that has emerged has been assisted
by improved data sets made available in recent years, as well as new
modeling frameworks, enhanced econometric techniques and better
understanding of how to measure convergence. In particular, there
has been the development of the concept of b-convergence measures
(Barro and Sala-i-Martin, 1992; Sala-i-Martin, 1994) that has allowed

a more rigorous analysis of economic convergence than the more tradi-

tional sconvergencemeasure.21The estimation of possible bconvergence
involves a mean-reversion calculation and it occurs if there is a negative
relationship between the growth rate of income per capita and the level
of initial income.22 Much of this type of analysis has been done at the
national level, but it, and the more limited body of analysis that has been
conducted at the regional level, does offer some insights into the validity of
the idea of endogenous economic growth.
Much of the early work on spatial economic convergence relied upon
aggregate, national data sources and focused on sconvergence meas-
urements (e.g., Abramovitz, 1986; Baumol, 1986; Maddison, 1987). The
findings indicated that labor productivity, and with it per capita income
in the world, was converging in the long run and thus rendered support to
the neoclassical growth theory. The difficulty with this work was that the
data sets used only contained countries that had already industrialized,
and even for those there were periods of divergence (De Long, 1988).23
Improved data, most notably the Heston-Summers panel data set that
embraces a large group of countries, subsequently indicated a lack of any
general economic convergence (Romer, 1994).
The more recent work that has made use of bconvergence measures,
and embraces a number of subnational studies, tends to find little support
for overall convergence. Barro and Sala-i-Martin (1991) in a number of
studies that, for example, have examined the economies of US states and
European Union regions find that while there is evidence of convergence
in per capita income, it is slow about 2 percent per annum and well below
the 12 percent or so that neoclassical theory would suggest. These are also
conditional convergence measures that allow for homogeneity between,
for example, the regional economies within an EU economy but diversity
between countries that would suggest potential differences in steady-state
growth rates for the regions within them.24 Hence, overall there is little
support in this body of work for the exogenous growth idea.
POLICY IMPLICATIONS
From a policy perspective, if there were indeed endogenous growth effects

this would seem to provide decision-makers with some opportunity to
intervene to stimulate growth and to combat spatially divergent growth
paths, in effect overcoming the frustration with neoclassical theory
expressed by Schumpeter at the beginning of the chapter. This contrasts
to the Solow model where only a change in the savings rate could gener-
ate long-run growth in per capita income. When in disequilibrium, the

neoclassical model does allow for fairly limited public policy interventions
that would de facto lubricate the system and facilitate a more rapid move
to a steady-state growth path. It would not, however, produce movement
along it or shift it. In the context of migration, for example, this may
involve improved information and enhanced transportation services to
allow existing resources to migrate and be used more effectively along
Adam Smiths lines of argument of greater divisions of labor.
If there is endogeneity in the growth process then the policy options are
somewhat wider. Since knowledge is important, then diffusion of ideas
and broader national policies for R&D can be deployed to bring lagging
regions up to the production frontier enjoyed by the leading regions. To
stimulate a nations growth, Romer, for example, argues for a reduction in
the USs federal deficit to reduce interest rates that would in turn increase
the amount of human capital devoted to R&D by raising the discounted
value of any given stream of future revenues associated with a new design.
The Romer framework would also suggest subsidies for R&D because of
the currently uncompensated external benefits that it generates; in contrast
the Lucas models suggest that these subsidies should largely go to the
education and training of workers.25
There is also a case for freer trade in that it allows for the more rapid
diffusion of knowledge and thus breaks down the monopoly of those
regions and countries that currently enjoy its ownership. More generally,
it releases knowledge workers to invent new designs rather than for those
in the lagging regions having to expend energies on catching up and effec-
tively continually having to reinvent the wheel. If one pursues the Richard
Florida (2005) line of argument that the creative classes are attracted
and retained by the larger environment in which they live and work, then
investment in various forms of social and economic infrastructure become
important.26
CONCLUSIONS
In a way there is no lack of economic theories seeking to explain why

some regions or countries grow faster than others. Indeed there may be
too many theories. What the recent developments in endogenous growth
theory have done is to introduce the less tangible elements of labor and
capital (knowledge in short) much more systematically into the analysis
of national and regional economic growth modeling. The accompany-
ing body of econometric and statistical work has helped fine-tune these
models and provide numerical methods for policy development and
assessment. But the understanding of variations in economic growth paths

is still far from complete. In part this is because the economic modeling
remains controversial and this in turn can partly be explained by gaps in
the work that has been completed by complementary disciplines.
While there are now relatively powerful economic and programming
techniques available to examine macro- and meso-data our understanding
of some of the fundamental micro-forces that lead to economic growth is
still relatively poor.27 To more completely understand the motivations for
companies to undertake R&D, and for individuals to invest differentially
in enhancing their knowledge bases, requires more systematic analysis by,
in the first case, management scientists, and in the latter by sociologists
and social psychologists. Even at the more macro level there appear to be
fundamental cultural differences that affect the speed and nature of uptake
of new technologies, and especially those that emerge in other regions that
are not normally seen as within the scope of economics. Similarly, there
appear to be variations in the flexibility to which nations or regions can
deal with shocks to their economic systems even within what superficially
appear to be relatively similarly educated populations (Pack, 1994). As the
old saying goes, The devil is in the detail, and useful analyses of many
of these details require a broader set of tools than those conventionally
deployed by traditional economists.
NOTES
1. We generally discuss development here as an economic phenomenon, and thus in terms

of regional incomes. Of course, there are many other elements to development, includ-
ing such things as the distribution of this income, the condition of the regional environ-
ment and a plethora of cultural factors that must be embraced in any holistic approach.
The excuse for the narrow approach is, in part, because the author is an economist, but
also it allows a focus on the more tangible elements of development.
2. Adam Smith (1776) in explaining the Wealth of Nations posited a supply-side-driven
model of growth. Taking the simplest of production functions,
Y = g(L, K, T)
where, Y is output, L is labor, K is capital and T is land, so output is related to labor

and capital and land inputs, output growth (gY) is driven by population growth (gL),
investment (gK) and land growth (gT) and increases in overall productivity (g):
gY = f(gL, gK, gL, gT).
The growth in population is endogenous depending on the sustenance available to

accommodate an increasing workforce. Investment is also endogenous, being deter-
mined by the rate of savings; land growth was dependent on conquest of new lands or
technological improvements of fertility of old lands. Technological progress could also
increase growth. Smiths thesis that the division of labor improves growth was funda-
mental to his argument. The Smithian and Ricardian models implicitly had technical

change arising from profit-squeezes or, in the particular case of Smith, arising because
of previous technical conditions. This was understandable. At the time of his writing,
the UK economy was moving into an extended period when economic growth reached
about 1 percent per annum compared with the 0.5 percent per decade that had previ-
ously been the norm (i.e., effectively peoples living standards did not change over their
life-times) and Smith was, in a way, groping to understand this new world.
3. Harry Richardson (1978, p. 18) says of the empirical work on regional growth up to the
1980s that, The models developed so far have been handicapped by severe data limita-
tions, too few observations, the reliance on static models, their recursive structure, the
heavy dependence on national economic variables, and the neglect of space.
4. Because of its market-clearing assumptions the pure classical model would be driven
only be income differentials.
5. Strictly, with full, instantaneous market clearing there is no unemployment in this type
of model, labor movements being determined by real relative wages. The unemploy-
ment effect is added to indicate possible imperfections in the short-term labor markets
in the two regions.
6. This model synthesizes the work of Evsey Domar (1947) and Roy Harrod (1939).
7. The Malthusian model, because of its implied capital productivity function, may be
seen as an extreme case of this whereby society cannot rise above the subsistence level
of per capita income.
8. McCraw (2007), for example, argues that part of Schumpeters motivation for the style
of presentation in the book was due to a degree of chagrin felt over the popular success
of John Maynard Keynes General Theory that had overshadowed his own analysis of
trade cycles. Strict modeling along Schumpeterian lines did not emerge until the 1990s.
9. Another of the stimuli for the interest in developing a more complete understanding
of economic growth in the 1950s and 1960s was institutional and founded on concerns
about the economic performance of the economies of the newly independent, ex-
colonial nations and the role that institutions such as the World Bank should play in
assisting them. This was at a time when it became clear that developing countries were
not catching up with higher-income nations.
10. For Solows later thoughts on the roles between endogenous and exogenous growth
theory see Solow (1994).
11. Romers assumption that knowledge and physical capital are related stemmed from
empirical observations that there were greater correlations between investment and
output than the neoclassical model would predict.
12. Romer (1990b) provides a series of anecdotal examples to support this view.
13. How the stock of knowledge is to be measured is a moot point (Steedman, 2001). There
are various elements or components to knowledge that need weighting to determine
the stock, even if they can be defined and measured. There are also questions as to the
extent that knowledge is non-rival rather than being unique to individuals.
14. Joseph Schumpeter provided a general model of economic growth in which through
the forces of creative destruction in oligopolist markets, entrepreneurs were stimulated
to develop new technologies. His discussion of endogenous growth was couched very
much in terms of a critique of Karl Marx but now emerged into a rigorous model of the
type Romer and others later developed.
15. This raises issue of whether endogenous growth is, in effect, an assumption in the model
or a result of the model.
16. The model is not closed, however, because Lucas only assumes that new goods are
continually being produced without having an explicit model of why.
17. Studies of economic convergence provide a pragmatic way of indirectly testing endog-
enous growth theory without the need to specifically define and measure knowledge.
18. For a much more rigorous analysis of endogenous growth involving two regions or
countries see, Segerstrom et al. (1990).
19. Richard Florida (2005) has added to these ideas by arguing that creative classes having
preference for where they live, and that the attributes required to attract and retain

them are more likely found in regions with an established high-income and educated
labor force. In either the older or more recent formulations of this framework, the
implications are often circular-and-cumulative causation; essentially richer regions get
richer and poorer regions, poorer.
20. Theoretical divergence is not an automatic outcome of endogenous growth (Kelly,
1992). Robert Tamura (1991), for example, by making the simple assumption that an
idea is more difficult to discover than it is to learn, and following from this that there is
a higher payoff from education for people with a lower basic level of education, dem-
onstrates that convergence can occur. Put another way, if a person is at the technology
frontier, movement out requires new knowledge whereas for someone within the fron-
tier it is possible, and easier, to move out by acquiring existing knowledge.
21. This approach essentially looks for changes in standard statistical indicators of
dispersion normally the variance.
22. A standard estimating equation takes the form D(yit) = a + b(yitj) + DiG + e it, where
the yit is that being tested for convergence, Di is a (n k) matrix of k variables capturing
spatial economic heterogeneity, G is a (k 1) vector of parameters, a and b are constants
and eit is a white noise error term. The condition b > 0 is a necessary but not sufficient
condition for b-convergence (Bernard and Durlauf, 1996). A critique of the empirical
work using this and similar tests for convergence is provided by Howard Pack (1994).
23. Bradford De Long and Lawrence Summers (1991) make the general point in their
specific analysis of investment in equipment, to which they credit much of the positive
spillover effects on economic growth, that markets are not perfect and that bad policy
decisions (government failures) inevitably occur and will influence the nature of the
growth path for any economy. Some studies seek to embrace such effects in the analysis,
for example, Barro (1991) includes political stability, potential rent seeking, and other
political variables in his calculations while Button and Pentecost (1995) embrace mon-
etary policy within their work on the European Union.
24. Button and Pentecosts (1995, 1999) work on the regional economies of the EU find
that there is evidence of divergence when no allowance is made for possible national
steady-state growth path differences. The national boundaries used by Barro and Sala-
i-Martin would not seem to correspond to natural economic units.
25. There would seem to be another policy difference in terms of the types of industry that
central policy-makers may seek to encourage to locate in economically slow-growing
regions. The tradition of seeking any form of economic activity to stimulate Keynesian
multiplier effects is alien to both approaches. Romers framework suggests that there is
a need to locate more knowledge-based elements of firms and industry in such regions
to stop their cumulative decline. Lucas, in contrast, because of the perceived impor-
tance of learning-by-doing, may well argue that provided industry achieved this it
would set in motion an effect to stimulate faster long-term growth.
26. The evidence that investment in more traditional infrastructure in slow-growing regions
can stem the tide of divergence is less certain (Button, 1998).
27. This is not to say that these techniques are always fully exploited. As Pack (1994)
argues, there are likely strong interactive effects between the factors that lead to eco-
nomic growth but there is a tendency for econometric analysis to miss these interactions
in the specifications used.
REFERENCES
Abramovitz, M. (1986), Catching up, forging ahead, and falling behind, Journal of
Economic History, 46(2), 385406.
Arrow, K.J. (1962), The economic implications of learning by doing, Review of
Economic Studies, 29(3), 15573.

Barro, R.J. (1991), Economic growth in a cross-section of countries, Quarterly

Journal of Economics, 106(2), 40743.
Barro, R.J. and Sala-i-Martin, X. (1991), Convergence across states and regions,
Brookings Papers on Economic Activity, 22(1), 10782.
Barro, R.J and Sala-i-Martin, X. (1992), Convergence, Journal of Political
Economy, 100(2), 22351.
Baumol, W. (1986), Productivity growth, convergence and welfare: what the long-
run data shows, American Economic Review, 76(5), 107285.
Bernard, A.B. and Durlauf, S.N. (1996), Interpreting tests of the convergence
hypothesis, Journal of Econometrics, 71(12), 16173.
Button, K.J. (1998), Infrastructure investment, endogenous growth and economic
convergence, Annals of Regional Science, 32(1), 14562.
Button, K.J. and Pentecost, E.J. (1995), Testing for convergence of the EU
regional economies, Economic Inquiry, 33(4), 66471.
Button, K.J. and Pentecost, E.J. (1999), Regional Economic Performance within
the European Union, Edward Elgar, Cheltenham, UK and Northampton, MA,
USA.
De Long, B. (1988), Productivity growth, convergence and welfare: comment,
American Economic Review, 78(5), 113854.
De Long, B. and Summers, L.H. (1991), Equipment investment and economic
growth, Quarterly Journal of Economics, 106(2), 445502.
Domar, E.D. (1947), Capital expansion, rate of growth and employment,
Econometrica, 14(2), 13147.
Florida, R. (2005), The Flight of the Creative Class: The New Global Competition
for Talent, Harper Collins, New York.
Friedman, M. (1953), The methodology of positive economics, in M. Friedman
(eds) Essays in Positive Economics, University of Chicago Press, Chicago.
Hahn, F.H. and Matthews, R.C.O. (1964), The theory of economic growth: a
survey, Economic Journal, 74(296), 779902.
Harrod, R.F. (1939), An essay in dynamic theory, Economic Journal, 49(193),
1433.
Hart, R.A. (1975a), Interregional economic migration: some theoretical considera-
tions (Part I), Journal of Regional Science, 15(2), 12738.
Hart, R.A. (1975b), Interregional economic migration: some theoretical consid-
erations (Part II), Journal of Regional Science, 15(2), 289305.
Kaldor, N. and Mirrlees, J.A. (1962), A new model of economic growth, Review of
Kelly, M. (1992), On endogenous growth with productivity shocks, Journal of
Lucas, R.E. (1988), On the mechanics of economic development, Journal of
Lucas, R.E. (1993), Making a miracle, Econometrica, 61(2), 25172.
Maddison, A. (1987), Growth and slowdown in advanced capitalist economies:
techniques of quantitative assessment, Journal of Economic Literature, 25(2),
64998.
McCraw, T.K. (2007), Prophet of Innovation: Joseph Schumpeter and
CreativeDestruction, Belknap Press of Harvard University Press, Cambridge,
MA.
Meyer, J.R. (1963), Regional economics: a survey, American Economic Review,
53(1), 1954.

Myrdal, G. (1957), Economic Theory and Underdeveloped Regions, Gerald

Duckworth, London.
Pack, H. (1994), Endogenous growth theory: intellectual appeal and empirical
shortcomings, Journal of Economic Perspectives, 8(1), 5572.
Ramsey, F.P. (1928), A mathematical theory of saving, Economic Journal, 38(152),
54359.
Richardson, H.W. (1978), The state of regional economics, International Regional
Science Review, 3(1), 148.
Romer, P.M. (1986), Increasing returns and long-run growth, Journal of Political
Economy, 94(5), 100237.
Romer, P.M. (1990a) Endogenous technological change, Journal of Political
Economy, 98(5), S71S102.
Romer, P.M. (1990b), Are nonconvexities important for understanding growth?,
Romer, P.M. (1994), The origins of endogenous growth, Journal of Economic
Perspectives, 8(1), 322.
Sala-i-Martin, X. (1994) Cross-sectional regressions and the empirics of economic
growth, European Economic Review, 38(34), 73947.
Schumpeter, J.A. (1942), Capitalism, Socialism, and Democracy, Harper and
Brothers, New York.
Segerstrom, P.S., Anant, T.C.A. and Dinopoulos, E. (1990), A Schumpeterian
model of the product life cycle, American Economic Review, 80(5), 107791.
Smith, A. (1776), An Inquiry into the Nature and Causes of the Wealth of Nations
(Oxford University Press edition, Oxford, 1976).
Solow, R.M. (1956), A contribution to the theory of economic growth, Quarterly
Solow, R.M. (1957), Technical change and the aggregate production function,
Review of Economics and Statistics, 39, 31220.
Solow, R.M. (1960), Investment and technical progress, in K.J. Arrow, S.
Karlin and P. Suppes (eds) Mathematical Models in Social Sciences, Stanford
University Press, Stanford.
Solow, R.M. (1994), Perspective on growth theory, Journal of Economic
Steedman, I. (2001), On measuring knowledge in new (endogenous) growth
theory, paper presented to the Growth Theory Conference, Pisa.
Swan, T.W. (1956), Economic growth and capital accumulation, Economic
Record, 32(63), 33461.
Tamura, R. (1991), Income convergence in an endogenous growth model, Journal
of Political Economy, 99(3), 52240.

3. Endogenous regional theory:
a geographers perspective and
interpretation
Michael Taylor and Paul Plummer
INTRODUCTION
The question posed in this chapter is: How do we understand and inter-
vene in processes of regional economic growth and change?
Globalization is eroding the old economic certainties, increasing com-
petition, reconfiguring the demand for skills, forcing the pace of tech-
nological and managerial change, transforming inter-firm relationships
and changing (through commodification) the role of finance. Some
would argue that there has been a sea change in the workings of national,
regional and local economies with the end of Fordism and the coming of
post-Fordist flexible specialization and mass customization. For policy-
makers, the pressing issue is how can appropriate and workable policies
and programmes be developed to cope with the internationalization of
production and consumption leading to rapid and radical change, foster
sustainable and robust local growth and enhance local capacities beyond
providing short-term palliatives for unemployment and its associated
social problems?
From an economic geography perspective, two bodies of discourse
currently dominate our understanding of the processes driving local
economic change: (1) the discourse of endogenous growth in the new
economic geography of the economists; (2) the new regionalism dis-
course along with its social constructionist and relational underpinnings in
economic geography proper and equivalent areas in other social sciences.
Based upon seemingly contradictory criteria of success, derived from
either the stylized facts of endogenous growth theory or the contingency
of new regionalism, these bodies of discourse have become reified into a
discourse of policy and practice based upon stereotypical understandings
of the mechanisms that are supposed to characterize the development of
local economies in a world of globalization. Here we argue that, despite
39

different conceptual foundations, and seemingly contradictory criteria of

empirical success, increasingly these two bodies of work are beginning to
overlap in terms of the substantive forces that each hypothesize to drive
local economic growth. The critical difference between those discourses is
placed into sharp relief when we are asked to interpret the causal efficacy
of their respective explanatory claims and translate those claims into
effective and implementable policies and practice.
On the one hand, drawing on the theoretical and methodological presup-
positions of neoclassical economics, there is regional endogenous theory.
According to its leading proponents this represents a new perspective on
understanding the process of regional growth and change that is capable
of providing clarity, rigour and relevance to a field of research that has
hitherto been characterized by anti-clarity, anti-rigour and anti-relevance
(Krugman, 1991; Overman, 2004). Typically, the new geographical eco-
nomics is dismissed by many economic geographers as being based on
under-socialized ill-conceived theory constructs and superficial empirical
regularities derived from extensive research designs that are considered
to be inappropriate means with which to identify the causal mechanisms
driving local economic growth. According to Plummer (2007), quantita-
tive methodologies are tolerated in economic geography only to the extent
that they are capable of providing background information as a prelude
to the real, and preferred, business of the intensive qualitative research.
Faced with the apparent inadequacies of superficial extensive meth-
odologies, many economic geographers have simply abandoned quan-
tification completely in favour of a suite of supposedly deeper causal
explanations provided by qualitative methodologies (Sayer, 1984; Massey
and Meegan, 1985; Clark, 1998). As an alternative discourse, economic
geography proper has constructed its own new regionalism that blends
ideas on institutions, individual agency and social regulation (IAR) in an
approach, drawing heavily on Polanyi (1957) and Granovetter (1985),
which argues that all economies are socially constructed. Eschewing
the constraints of theory construction through mathematical modelling
and quantitative reasoning based upon stylized facts, this new region-
alism attempts to establish the empirical adequacy of its explanatory
claims through qualitative description, close dialogue and fine-grained
case studies. In contrast with the economists emphasis on economic
mechanism and quantitative analysis, this more discursive methodol-
ogy celebrates the supposed inherently complex construction of societal
processes and the contingency of the geographical world. In practice, this
means exploring the workings of socially constructed regional economies
through the lived experience of individuals living and working within it.
The perspective that we adopt in this chapter is at odds with both the

A geographers perspective and interpretation 41
new geographical economics and economic geography proper. We are, in

effect, advocating a critically engaged pluralist approach to understand-
ing local economic growth and change (Plummer and Sheppard, 2007).
Unlike many geographers, we reject what we believe to be an unsustaina-
ble dualism between quantitative and qualitative economic geographies in
favour of a meaningful and grounded multi-method research design that
allows us to triangulate the insights of both economists and geographers
ways of knowing. Put differently, as economic geographers, we reject the
tout court dismissal of extensive research designs of the type employed
by endogenous regional theory. Instead, we argue that the way forward
is to blend the economism of endogenous theory, and its extensive
research strategy, with the social constructionism of new regionalism
and its intensive research strategy. We want to suggest that multi-method
research strategies can provide us with critical understanding and explana-
tion of social reality, which can both empirically evaluate the explanatory
claims and inform policy debate (King et al., 1994).
CONTRASTING RESEARCH STRATEGIES:

CARICATURES AND STYLIZATIONS
It can be argued that both endogenous regional theory and geographys

new regionalism are caricatures of functioning regional economies, both
of which posit processes that are necessary to developing an understand-
ing of economic change at the regional scale, but neither of which offers
sufficient explanation on its own. Put differently, all conceptualizations
are partial representations of complex real world processes; where they
differ is over which aspects of geographical reality they choose to prioritize
and how those characteristics that are identified are thought to be related
to one another. It is in this sense that all theories or models are stylized
descriptions of imitations or local economies, and hence caricatures.
The trick, however, is to avoid the exaggerating, comic and/or grotesque
characteristics that are typically associated with caricatures!
Endogenous Regional Theory
With its origins in contemporary economics, endogenous regional theory

is driven by a methodological commitment to accounting for a set of
established stylized facts using abstract mathematical reasoning. To
constitute an adequate explanation of these stylized facts, economists
models of local economic growth need to be formulated in a particular
style of mathematical reasoning in which representative agents maximize

their inter-temporal objective function, subject to resource constraints.

Ideally, the actions of these representative agents are supposed to ensure
an optimal and stable equilibrium configuration whose properties act as
a guide to empirical work (see Sheppard, 2000; Plummer and Sheppard,
2006). Notwithstanding the existence of path dependence and multiple
equilibria that are characteristic of models derived from the core model
of geographical economics (Brakman et al., 2001), there remains an onto-
logical commitment to the proposition that geographical reality can be
understood using equilibrium-based models.
Within the constraints of optimum-and-constraints theorizing,
endogenous regional theory is grounded in the logic of aggregate produc-
tion functions relating per capita output (income) to the degree of capital
intensity and technological change. In its classical Solow-Swan formula-
tion, long-run growth of per capita output (income) is sustained through
a constant and smooth, but exogenous, process of technological change
(Fingleton, 2003). As a corollary, technological change represents the
untheorized determinant of long-run local economic growth and change.
In contrast, the new regional growth theory attempts to endogenize
technological change by explicitly modelling processes of learning-by-
doing, knowledge spillovers and/or Schumpeterian creative destruction
(Martin and Sunley, 1998). Typically, endogeneous technological change
is introduced either through the ad hoc augmentation of a conventional
production function with factors such as social and/or human capital
or in the form of a knowledge production function, which is derived from
the decision of a rational profit-seeking capitalist to invest in knowledge
and innovation (Jones, 1998). These theoretical specifications form part
of a wider class of models, which form the core of geographical economics,
and have replaced the conventional assumption that competitive markets
exhibit constant returns to scale with the assumptions specifying increas-
ing returns and imperfect competition based upon product differentiation
(Fujita et al., 1999; Brakman and Heijdra, 2004).
As has been pointed out by many observers, the abstract nature of the
theoretical models proposed by endogenous regional theory, coupled with
the unrealism of the assumptions that are necessary to generate closed-
form solutions to these abstract models, means that empirical testing has
lagged behind theoretical developments (Neary, 2001; Fingleton, 2007).
Nonetheless, recently we have witnessed an explosion in empirical models
attempting to evaluate the explanatory claims of endogenous regional
theory using third-way growth econometrics (Fingleton, 2001). The new
growth econometrics represents a particular research strategy that has
become the work-horse of economists attempts to account for both the
convergence that is hypothesized to operate between regions, resulting from

the inter-regional mobility of factors of production and commodities, and

the set of structural conditioning factors that are hypothesized to produce
persistent, and perhaps divergent, regional growth differentials. Much
of the literature emphasizes five stylized facts determining, to varying
degrees, aspects of change at the regional level, including (Glaeser, 2000):
technological change and innovation;

human capital, embracing research and education;
agglomeration and externalities;
knowledge spillovers, including entrepreneurship and new firm
formation;
sectoral specialization/diversification.
Fundamentally, we should remain sceptical about the existence of the

economists stylized facts about local economic growth. Indeed, it may
well be the case that diminishing returns have set in regarding the empiri-
cal efficacy and policy relevance of growth econometrics. In studies that
have adopted the growth regression approach, over 50 different proxy
variables have been identified as potential conditioning variables, ranging
across societal preferences, techniques of production, rates of population
growth and government policy regimes (Durlauf et al., 2004). Whilst ad
hoc proxy variables may fit satisfactorily, the significance of individual
variables is not robust, and tends to be sensitive to the set of other condi-
tioning variables that are included in the model (Durlauf and Quah, 1999).
Furthermore, it may be difficult to give these proxy variables a mean-
ingful interpretation in terms of the economic reasoning underlying the
economists growth theory, especially when mechanistic variable selection
searches are driven by data availability rather than theoretical relevance
or plausibility. It is, perhaps, ironic that economists insist on employ-
ing sophisticated mathematics to ensure that the parameters related to
a local economys production function are derived from economic rea-
soning based upon inter-generational rational choice models, yet they
seem willing to rely on casual empiricism when selecting the appropriate
(assumed exogenous) set of conditioning variables.
At the same time, from the perspective of both theoretical and empirical
research in geography, many of the relationships posited in the endogen-
ous regional approach appear simplistic and under-theorized, merely a set
of control variables that allows us to focus on the convergence amongst
and between local economies. For example, uncritically, agglomeration is
assumed to be a source of external economies, reducing transaction costs
when there is empirical evidence that agglomeration offers not cheaper
production, even in transaction cost terms, but simply easier production

in purely behavioural terms (Taylor, 1975). Similarly, the presence of

knowledge in a place is assumed to lead to spillover effects without any
real conceptualization of the transmission mechanism through which
knowledge is supposed to diffuse between individuals and/or regional
economies. Even casual empiricism would suggest that the facts that:
contract law is one of the fastest-growing components of the legal profes-
sion; the protection of IPR is a growing corporate concern (reflected in the
inimitability version of the competencies theory of the firm); and that the
spat between Ferrari and McLaren teams in Formula 1 racing in 2007 was
so acrimonious when an employee moved between the two organizations
and took technical data with him that it brought a multi-million dollar fine
all point to knowledge spillover effects being a far more complex process
than endogenous regional theory would imply.
The point we want to make here has been made before (Clark, 1998);
that endogenous regional theory is based on stylized facts that need seri-
ously to be unpacked. Even within the terms of reference of endogenous
regional theory, in order to understand both the existence and persistence
of local growth differentials, we need to understand the nature of the
mechanisms that are hypothesized to diffuse technological change through
a regional economic system. Certainly, this form of analysis offers a way
to prioritize the theorized drivers of regional economic change that new
regionalism lacks. However, the whole comes together not as a consist-
ently articulated theory of local economic growth but as a series of con-
ceptual and/or empirical explorations of the components of what might
be developed into something fuller. In practice, attempts to bridge the gap
between abstract, but untestable, models and the messy contingency of
the social world using growth econometrics has produced some insights,
but can be more broadly caricatured by the use of sophisticated techniques
and poor data to make inferences and gain insights into the processes
driving local economic growth (Plummer, 2001).
New Regionalism and Embeddedness
In the past decade, an increasing body of research has built upon

Granovetters (1985) concept of embeddedness to explain the dynamics
of local economies. This perspective emphasizes the role of social rela-
tions in economic transactions, holding that anonymous markets are
non-existent and that economic life and transactions are built on social
connections. The concept of embeddedness has given rise to a powerful
model of local economic growth that draws on a range of complemen-
tary literatures on new industrial spaces, learning regions, innovative
milieus and regional innovation systems (e.g., Storper, 1997; Braczyk

et al., 1998; Porter, 1998; MacKinnon et al., 2002) now more generally
captured under the term new regionalism (Rainnie and Grobbelaar,
2004). Building directly on the ideas of Granovetter, the new regionalist
writing suggests that economic growth is linked not only to market condi-
tions, but also to repeated inter-firm interaction and knowledge exchange,
collaborative long-term buyersupplier relationships, the creation of
social capital (including trust, reciprocity and loyalty) and a supportive
tissue of local institutional thickness (see Putnam, 1993; Cumbers et al.,
2003; Malmberg and Maskell, 2006).
The new regionalism and embeddedness ideas have prompted renewed
interest in the advantages of geographical proximity between firms in
related industries (Keeble and Nachum, 2002) and it is a significant shift
away from traditional understandings of agglomeration based on input
output relations and transaction costs, focusing instead on the social and
institutional drivers of growth. Whilst the approach has theorized new
drivers and has added richness to the local growth debate, it has been less
rigorous in how we might measure those processes or detect their impact
on long-run economic growth. It is also a model of economic growth with
great policy appeal, but the application of this new orthodoxy has often
been implemented uncritically and in the absence of detailed empirical
evaluation (i.e., evidence-based policy).
The new regionalism approach may be popular but it also has sig-
nificant and prominent limitations (see also Taylor, 2005). First and
most importantly, the mechanisms of local economic growth transmis-
sion are not clearly articulated in the model, and this limitation has four
interconnected components:
1. The model lacks a meaningful explanandum. It does not measure

growth or change, just success in that the qualitative data on which
the model is based are drawn from apparently successful regions and
places, with the processes identified within them being taken to account
for that success. As a consequence, sample selection bias is combined
with a truncated and censored distribution to increase the threats
to validity when attempting to draw inferences about the processes
driving local economic growth from self-elected case studies.
2. Time, too, is poorly incorporated into the model, and local firm net-
works are seen to translate unproblematically and instantly into local
structures of knowledge mobilization and exchange, learning, innova-
tion and social capital that create growth. This is a caricature of trans-
mission mechanisms, leaving us without any knowledge of how and
when economic potentialities are actualized within a local economy.
All that the model alludes to is a mass (or morass) of contingency.

3. New knowledge is assumed to translate unproblematically into new

business ventures, neglecting processes of coalition formation, third
party referral, the commercial translation of ideas and knowledge, the
securing of finance, for example. Transmission mechanisms are again
ignored.
4. The model fetishises proximity (Oinas, 1999). As a consequence, it
underplays geographically more extensive transmission mechanisms
and networks that might be as, if not more, important than those of a
particular locality (see Glckler, 2007).
Second, the new regionalist model fails to recognize the importance of

unequal power relations between firms; inequalities that privilege some
and peripheralize others within both local and geographically more exten-
sive networks (Taylor, 2000).
Third, the model also fails to deal adequately with the capitalist impera-
tives of profit generation, the price mechanism and wage discipline as
mechanisms of personal wealth creation (Hudson, 1999).
Finally, much of the work on new regionalism is based on theoretical
speculation, rather than empirical research, thus providing only a limited
understanding of processes at work (Maskell and Malmberg, 1999).
Empirical research has focused on high-technology clusters and dynamic
growth regions, neglecting those localities, places and regions experienc-
ing decline, industrial restructuring, or socioeconomic marginalization.
Taken together, these limitations raise the spectre that without clearly
articulated theories specifying the manner in which posited mechanisms
operate both within and between local economies, the new regionalism
is in danger of degenerating into a morass of contingency, the qualita-
tive analogue of over-fitting econometric models. It is not clear whether
this observed contingency is the result of local capacities and competitive
mechanisms working themselves out in different contexts or whether there
are simply different drivers of local economic growth that operate idi-
osyncratically in each local economy at any time. As a corollary, the new
regionalism fails to address the methodological questions of how, and in
what ways, local economic capacities might be actualized in a particular
empirical context.
New regionalism, therefore, offers a differently limited caricature of
regional economic processes. It is a caricature built on over-theorization
in which layers of contingency have been laid one over another, obscur-
ing as much as clarifying processes of growth and change. In part, this
accounts for its popularity, especially in policy circles, because it can be
all things to all people. Again, it is built on stylized facts, in this case relat-
ing to trust, networks, institutional support and knowledge transfer, for

example. However, as more theorizing has been added to the model, it has
become evident that there is no mechanism to say what the importance of
the separate, postulated processes is in promoting or retarding regional
economic growth. It is here that the modelling strategy of endogenous
regional theory has much to offer.
ANALYTICAL CAPABILITIES AND POLICY ADVICE

There is, perhaps, a more targeted and appropriate way to develop a richer
understanding of growth and change at the level of the local economy that
goes beyond the dualism of stylized facts and contingency, and yet is
capable of providing policy advice that can be used to formulate effective
strategies that are capable of being implemented in practice. We advocate
a mixed-method research design that is grounded in critically engaged
pluralism (Taylor and Plummer, 2003). This necessitates moving beyond
the types of econometric modelling strategies employed by endogenous
regional theory and the intensive case studies of qualitative geographers
to triangulate our knowledge claims about local economies. In practice,
this entails a research design that can accommodate and use a blend of
qualitative and quantitative information to articulate local capacities and
transmission mechanisms and evaluate empirically the likelihood that
these capacities are actualized in any particular context (Plummer and
Taylor, 2001a, 2001b). In short, we reject neither approach to regional
analysis but suggest that their combination offers important synergies
for both analysis and policy formulation. It is the combination of these
approaches that we illustrate in the analyses of the remaining sections of
this chapter.
A recently completed study relating to policy advice on vocational edu-
cation and training undertaken for the National Centre for Vocational
Education Research (NCVER) in Australia (Garlick et al., 2007) explored
the potential and limitations of such a mixed-method research design.
This study was motivated by the need to develop evidence-based policy
outcomes in the context of vocational education and training (VET) in
Australia. We can use our experience of trying to implement this research
project to illustrate the virtues of a pluralist methodology when develop-
ing analysis to inform policy advice. Based upon these methodological
norms, this study produced controversial and provocative conclusions
that, we would contend, are essential if new policy approaches are to be
developed that go beyond repeating the tired and increasingly hackneyed
initiatives suggested in the one-size-fits-all guru theorizing on clusters
and creatives.

Theoretically Informed Empiricism
The Garlick et al. (2007) study of vocational education and training (VET)
provision in Australia focused on providing advice on policy alternatives
that might better target VET to foster regional economic growth that was
consistent with the long-run processes shaping those regional economies.
The study asked two key questions: (1) What processes are driving eco-
nomic change at the regional scale in Australia? (2) How can VET work
sympathetically and in a targeted manner to modify those processes, create
growth and, therefore, more fully realize national economic potentials?
From an institutionalist perspective, successful local economies are said
to depend on complex processes of integration and embedding built on
trust, reciprocity and loyalty that create social capital. They involve
the exchange of information, ideas and innovation through mechanisms
of quasi-integration that creates new knowledge through processes of
situated learning. In addition, the support of local institutions (labelled as
institutional thickness; Amin and Thrift, 1994) is seen as bolstering these
local economic processes. Enmeshed in webs of global coordination and
value transfer, these local economies are seen as nodes of untraded inter-
dependencies (Storper, 1997) that are tapped into by large corporations
and TNCs, which, for their part, act as global information arbitrageurs
(Economist, 2003).
To reflect institutionalist thinking on the propulsive elements of local
economic growth, eight drivers were distilled from the broad range of
theories, each of which prioritized different sub-sets of drivers as operating
in different ways, depending on local context. These were:
technological leadership at the enterprise level;

knowledge creation and access to information;
the local integration of small firms their propensity to deal one
with another within a locality;
infrastructure support and institutional thickness;
human resources the skills and capacities of a local workforce;
the power of large corporations the assets and authority they are
able to wield;
local demand and inter-regional trade;
local sectoral specialization the existing specialist skills and knowl-
edge of a locality or region (see Plummer and Taylor, 2001a).
To perform the quantitative element of the multi-method project,

these drivers were incorporated into a growth econometrics model con-
sistent with existing models derived from endogenous regional theory.

The eight drivers distilled from theory were measured empirically using
a unique Australian data set generated across Australia between 1984
and 2002, with the country divided into 94 functional regions. The nature
and origins of the data set are outlined in Appendix 3A.1 at the end of
this chapter. These empirical measures were incorporated into a conven-
tional gap convergence (mean reversion) econometric model (Plummer
and Taylor, 2001b; Taylor and Plummer, 2003) in which regional growth
(measured in terms of unemployment relativities) is broken down into
three components:
1. transition dynamics: the speed at which a regions growth rate returns

to a long-run equilibrium after a shock; that is, mean reversion;
2. structural characteristics: the extent to which there are growth
differences between neighbouring regions;
3. random shocks: the unanticipated and unpredicted factors that can
impact on the regional growth rate (Martin and Sunley, 1998).
The modelling exercise is fully explained in Garlick et al. (2007). Here,

however, it is the final model specification that is important. The findings
are quite stark. Regional growth in Australia in the last two decades has
been driven by four mechanisms:
1. human capital: fundamentally the local stock of skilled and educated

people;
2. access to high technology and technological leadership at the
enterprise level;
3. industry specialization: reflecting local stocks of specialist knowledge;
while, interestingly,
4. government intervention: through direct institutional support held
back regional growth.
Very clearly, human capital and enterprise are the life blood of regional
economic growth, especially when there is appropriate government
intervention.
Qualitative Information and Facilitated Workshops
However, the study recognized that not only must positive drivers of
growth be found at the regional scale as the quantitative modelling shows,
but, to achieve success, it is also important that growth transmission pro-
cesses at that scale are faced by as few impediments as possible. To address
this aspect of the local growth process, qualitative information was

Table 3.1 Australian case study regions and their characteristics
Region (State) Growth Characteristics

Orange (NSW) A rural city with high growth
Western Sydney/ A peri-urban metropolitan area with high-technology
Penrith (NSW) enterprise
Wollongong (NSW) A provincial city with static growth
Mount Isa (Qld) A remote mining centre with low growth
Wide Bay Burnett A region with rural cities, and low and declining growth
(Qld)
Shepparton (Vic.) Centred on a rural city and shows virtually no growth
Horsham (Vic.) Centred on a rural city, and with relatively high growth
Burnie (Tas.) A rural city with declining growth
North-East Adelaide A peri-urban metropolitan area with low growth
(SA)
Pilbara/Port Hedland Remote mining communities with rapidly declining
(WA) growth
Alice Springs (NT) A remote city with recent relatively high growth
Source: Authors compilation.
collected through facilitated workshops in 11 case study regions (Garlick

et al., 2007), and this information adds significantly to the interpretation
of growth dynamics across Australias regions.
Workshops were held in regions across the spectrum of growth rates
in Australia, spread across states and territories and embracing a mix of
urban and rural regions. These regions and their growth characteristics are
listed in Table 3.1. Between 12 and 30 people attended each workshop, with
participants from VET and technical and further education (TAFE), local
government, state government, business, schools, universities, regional
development organizations and social development bodies. There were
managers from small and large business, institutional managers, students,
teachers, social entrepreneurs and others.
A number of salient findings from these workshops are disturbing.
In particular, there was a general inability to fully appreciate the global
dynamic influencing business development. Also, there was an apparent
naivety about the way companies use their capital and labour and the way
networks are used for corporate rather than community gain. There was
a penchant for off-the-shelf, quick-fix solutions brought in from outside
the region, and a parallel lack of local strategic thinking. In some regions,
with economic problems, there was a disturbing tendency towards self-
delusion. Though local economic problems were recognized there was a

tendency to argue that, of course, things had changed and that the future
promised growth. This could only be interpreted as the optimism of the
immediate. It highlights, however, the potential limitations of qualitative
analyses and, possibly, the unreal expectation that consultants can create.
Not all regions were the same, and in some outstanding initiatives had
been taken. However, where plans were developed, there was a tendency
for people to be transfixed by the plan and not to move to implementa-
tion. It was concluded that 20 years of bottom-up regional development
policy and practice in Australia had produced disturbingly low levels of
local action and that government-mandated regional leadership might
not be the most effective way to encourage local growth that depends on
free-flowing ideas and enthusiasm across the whole of a region (Garlick et
al., 2007).
On a more positive note, the workshops did recognize a range of
impediments to growth transmission. In particular, they recognized a
general complacency and lack of dynamism in some regions. They pointed
to human capital being exported to major metropolitan regions across the
country, with this exodus being seen as normal. Key industries were not
seen as producing business spin-outs at the regional level, and that facili-
tative regional planning to realize opportunities appeared to be disorgan-
ized. At the same time, there was a recognition that the education system
was unable to inspire a regions human capital to be enterprising. What is
more, the workshops saw an overemphasis on specific sectoral winners,
pushed by institutions and consultants from outside the region, as the
silver bullets to achieve local economic growth rather than building on
unique local capacities, attributes and abilities (ibid.).
Enterprising Human Capital and Regional Engagement
When the extensive and intensive analyses of regional economic growth

in Australia are combined, a clear picture of its dynamics and transmis-
sion emerges that highlights the roles of enterprising human capital and
regional engagement (ibid.). It is a picture that neither research strategy
could generate independently of the other.
Enterprising human capital, we would contend, is created by enter-
prising people who take ideas and turn them into outcomes using the
attributes, competencies and capabilities at their disposal. This is very dif-
ferent from the creative human capital that Richard Florida (2002) has
emphasized that sees people with skills and knowledge generating new
ideas. The translation of new ideas into socially and commercially viable
enterprises is a non-trivial process. Rather, it is a process at the very heart
of the regional growth process, which needs to create wealth, income

and jobs at least to maintain and possibly to enhance the well-being and
standard of living of entire communities that are currently facing global
competitive pressures. It involves having a strong grasp of, amongst
many other things, markets and market potentials, sourcing, financing,
business planning, accountancy, cost control, timing, coordination and
negotiation. It involves outcome-orientated people who, at their most
successful, are able to work in teams that blend complementary skills.
The enterprising outcomes of these people are not necessarily just eco-
nomic in the sense of creating new businesses. They may also emerge as
new social, cultural and environmental outcomes that reshape the very
nature of the communities within which they operate and may, in turn,
create a climate for further enterprising in that place. All the case study
regions pointed to the lack of a local enterprising culture within them
and none had mechanisms in place to facilitate the creation of such an
enterprising capacity.
Essential to the creation of enterprising human capital are mutual and
creative connections among a regions human capital what more broadly
can be called regional engagement. It is, at its most fundamental, a frame-
work of mutual cooperation and dialogue that complements and extends
the processes of enterprising human capital. However, it is more than a
process. Regional engagement is also about achieving results and generat-
ing outcomes. It is the community equivalent of the enterprising human
capital of individuals.
Towards a Policy
The question posed by these empirical findings on regional growth dynam-

ics in Australia is: how do you translate these ideas into a policy that can
be used to guide the development of vocational education and training
(VET) in Australias regions? The report we prepared (Garlick et al., 2007)
suggested four new roles for VET, building on local knowledge and skills,
consistent with the processes shaping the countrys regional economies.
Those roles are described below:
1. VET should aim to develop competencies within communities by build-

ing enterprising skills. Consistent with the theoretically informed
analysis, this role might be achieved in two ways: by developing
locally the skills to enable enterprising opportunities of all types to
be turned into outcomes; and creating greater knowledge and aware-
ness of how the global economy works. The first of these approaches
to competency building is essentially entrepreneurship education.
It is hard, if not impossible, to teach people to be entrepreneurial.

However, it is possible to equip those who are keen to try to take risks
with the skills that can help them succeed, including:
issues involved in identifying commercial and other enterpris-

ing opportunities;
business planning;
marketing contracts their drawing up and enforcement;
accessing venture capital and development finance;
entrepreneurialism;
networking;
risk management;
communication.
The second approach to competency building is a broader educational

provision that discusses the workings of the global economy, the
competition it brings and the opportunities and limitations it offers
together with the mechanisms to facilitate change. Here it is important
to understand the inter-company relationships in supply chains, and
the unequal power relationships that operate within them.
2. VET can assist local capacity building by taking a lead in developing
coalitions of people with enterprising ideas in competitive areas, and
equipping them with connections and synergies with other businesses
and institutions. Firms are essentially temporary coalitions of people
who deploy assets in an effort to achieve anticipated commercial out-
comes. As yet, our understanding of the dynamics of those coalitions
is relatively underdeveloped (Taylor, 2006), but it is an area where
intensive case study evidence can be used to great effect.
3. VET needs to target 40+ year-olds to provide them with the skills to
be enterprising and to build businesses, and realize their capacities
and capabilities. In this context, vocational education and training has
the opportunity to realize the otherwise lost potential of this section of
society as was pointed out in the workshop element of the Australian
study.
4. VET has an important role in bringing together information of a
regions knowledge resources. This is a significant additional facilita-
tive role that can speed up processes of enterprising.
The reaction to these policy proposals for VET has proved very conten-
tious. It proposes a very different educational agenda from the one that has
traditionally been followed. It makes educators uncomfortable because it
disturbs peoples comfort zones and involves doing something new!

CONCLUSION
In developing a geographic perspective on endogenous processes operat-
ing in regional and local economies, this chapter has sought to highlight
the differences in methodology in this discipline compared with econom-
ics: between the intensive qualitative approach in geographys new
regionalism and the extensive modelling strategy of economics endog-
enous growth theory. Both are distinctive caricatures of local economic
processes built on sets of differently stylized facts.
It has been suggested that the two approaches are complementary and,
through integration, have the potential to bring a fuller and more nuanced
perspective on local economic growth processes and issues that are
capable of reflecting the unique characteristics and capabilities of places
rather than generating a one-size-fits-all policy prescription. As such it is
an approach that has the potential to suggest new policy initiatives.
A multi-method analysis of Australian regional change using theoreti-
cally informed empirical modelling coupled with qualitative information
collected through facilitated workshops in case study regions, threw new
light on regional growth processes in this economy. Focused on a review
of VET in Australia, this approach was shown to have the capacity to gen-
erate new policy perspectives that fall outside the usual comfort zone of
bureaucratic policy-makers. From this it might be suggested that a multi-
method approach to the analysis of endogenous processes in regional
economies might be a useful way of developing new regional policies and
programmes.
REFERENCES
Amin, A. and Thrift, N.J. (1994), Living in the Global, in A. Amin and N.J.
Thrift (eds), Globalization, Institutions and Regional Development in Europe,
Oxford University Press, Oxford, pp. 122.
Braczyk, H.J., Cooke, P. and Heidenreich, M. (eds) (1998), Regional Innovation
Systems: The Role of Governances in a Globalized World, UCL Press, London.
Brakman, S. and Heijdra, B. (2004), The Monopolisitic Competition Revolution in
Retrospect, Cambridge University Press, Cambridge, UK.
Brakman, S., Garretson, H. and Von Marrewijk, C. (2001), An Introduction to
Geographical Economics, Cambridge University Press, Cambridge, UK.
Clark, G. (1998), Stylized Facts and Close Dialogue: Methodology in Economic
Geography, Annals of the Association of American Geographers, 88(1), 7387.
Cumbers, A., MacKinnon, D. and Chapman, K. (2003), Innovation, Collaboration
and Learning in Clusters: A Study of SMEs in the Aberdeen Oil Complex,
Environment and Planning A, 35(9), 1689706.
Durlauf, S.N. and Quah. D. (1999), The New Empirics of Economic Growth,

in J.P. Taylor and M. Woodford (eds), Handbook of Macroeconomics, North-

Holland, Amsterdam, pp. 235308.
Durlauf, S., Johnson, P. and Temple, J. (2004), Growth Econometrics, in P.
Aghion and S.N. Durlauf (eds), Handbook of Economic Growth, North-Holland,
Amsterdam.
Economist, The (2003), The New Geography of the IT Industry, The Economist
Newspaper, 368(8333), 535.
Fingleton, B. (2001), Equilibirum and Economic Growth: Spatial Econometric
Models and Simulations, Journal of Regional Science, 41(1), 11748.
Fingleton, B. (ed.) (2003), European Regional Growth, Springer-Verlag, New York.
Fingleton, B. (ed.) (2007), New Directions in Economic Geography, Edward Elgar,
Cheltenham, UK and Northampton, MA, USA.
Florida, R. (2002), The Rise of the Creative Class: and How Its Transforming
Work, Leisure, Community and Everyday Life, Basic Books, New York.
Fujita, M., Krugman, P. and Venables, A. (1999), The Spatial Economy: Cities,
Regions, and International Trade, MIT Press, Cambridge, MA.
Garlick, S., Taylor, M. and Plummer, P. (2007), An Enterprising Approach to
Regional Growth: The Role of VET in Regional Development, report prepared
for the National Centre for Vocational Education Research (NCVER), funding
provided by the Australian National Training Authority (ANTA).
Glaeser, E. (2000), The New Urban and Regional Growth, in G. Clark, M.
Gertler and M. Feldman (eds), Handbook of Economic Geography, Oxford
University Press, Oxford, pp. 8399.
Glckler, J. (2007), Geography of Reputation: The City as the Locus of Business
Opportunity, Regional Studies, 41(7), 94961.
Granovetter, M. (1985), Economic Action and Social Structure: The Problem of
Embeddedness, American Journal of Sociology, 91(3), 481510.
Hudson, R. (1999), The Learning Economy, the Learning Firm and the Learning
Region: A Sympathetic Critique of the Limits of Learning, European Urban and
Regional Studies, 6(1), 5972.
Jones, C. (1998), Introduction to Economic Growth, Norton, London.
Keeble, D.E. and Nachum, L. (2002), Why do Business Service Firms Cluster?,
Transactions of the Institute of British Geographers, 27(1), 6790.
King, G., Keohane, R. and Verba, S. (1994), Designing Social Inquiry: Scientific
Inference and Qualitative Research, Princeton University Press, Princeton, NJ.
Krugman, P. (1991), Geography and Trade, MIT Press, Cambridge, MA.
MacKinnon, D., Cumbers, A. and Chapman, K. (2002), Learning, Innovation
and Regional Development: A Critical Appraisal of Recent Debates, Progress
in Human Geography, 26(3), 293311.
Malmberg, A. and Maskell, P. (2006), Localized Learning Revisited, Growth and
Change, 37(1), 118.
Martin, R. and Sunley, P. (1998), Slow Convergence?: The New Endogenous
Growth Theory and Regional Development, Economic Geography 74(3),
20127.
Maskell, P. and Malmberg, A. (1999), Localised Learning and Industrial
Competitiveness, Cambridge Journal of Economics, 23(2), 16790.
Massey, D. and Meegan, R. (1985), Politics and Method: Contrasting Studies in
Industrial Geography, Routledge, London.
Neary, P. (2001). Of Hype and Hyperbolas: Introducing the New Economic
Geography, Journal of Economic Literature, 39(2), 53661.

Oinas, P. (1999), Activity-Specificity in Organizational Learning: Implications for

Analysing the Role of Proximity, GeoJournal, 49(4), 36372.
Overman, H. (2004), Can We Learn Anything From Economic Geography
Proper?, Journal of Economic Geography, 4(5), 50116.
Plummer, P. (2001), Vague Theories, Sophisticated Techniques, and Poor Data,
Environment and Planning A, 33(5), 7614.
Plummer, P. (2007), Economic Geography, by the Numbers, in A. Tickell,
E. Sheppard, J. Peck and T. Barnes (eds), Politics and Practice in Economic
Geography, Sage, London.
Plummer, P. and Sheppard, E. (2006), Geography Matters: Agency, Structures
and Dynamics at the Intersection of Economics and Geography, Journal of
Economic Geography, 6(5), 61937.
Plummer, P. and Sheppard, E. (2007), Toward Critically Engaged Pluralism in
Geographical Debate, Environment and Planning A, 6(5), 127581.
Plummer, P. and Taylor, M. (2001a), Theories of Local Economic Growth:
Concepts, Models and Measurement, Environment and Planning A, 33(2), 21936.
Plummer, P. and Taylor, M. (2001b), Theories of Local Economic Growth:
Model Specification and Empirical Validation, Environment and Planning A,
33(3), 38598.
Plummer, P. and Taylor, M. (2003), Theory and Praxis in Economic Geography:
Enterprising and Local Growth in a Global Economy, Environment and
Planning C, 21(5), 63349.
Polanyi, K. (1957), The Great Transformation, Beacon Press, Beacon Hill, MA.
Porter, M. (1998), On Competition, Harvard Business School Press, Boston, MA.
Putnam, R. (1993), Making Democracy Work: Civic Traditions in Modern Italy,
Princeton University Press, Princeton, NJ.
Rainnie, A. and Grobbelaar, M. (eds) (2004), New Regionalism in Australia,
Ashgate, Aldershot.
Sayer, A. (1984), Method in Social Science: A Realist Approach, Hutchinson, London.
Sheppard, E. (2000), Geography or Economics?; Conceptions of Space, Time,
Interdependence, and Agency, in G. Clark, M. Gertler and M. Feldman (eds),
Handbook of Economic Geography, Oxford University Press, Oxford, pp. 99125.
Storper, M. (1997), The Regional World: Territorial Development in a Global
Economy, Guilford Press, New York.
Taylor, M. (1975), Organizational Growth, Spatial Interaction and Location
Decision-Making, Regional Studies, 9(4), 31323.
Taylor, M. (2000), Enterprise, Power and Embeddedness: An Empirical
Exploration, in E. Vatne and M. Taylor (eds), The Networked Firm in a Global
World, Ashgate, Aldershot, pp. 199234.
Taylor, M. (2005), Embedded Local Growth: A Theory Taken Too Far?, in R.
Boschma and R. Kloosterman (eds), Learning from Clusters: A Critical Assessment
from an Economic-Geographical Perspective, Springer, Dordrecht, pp. 6988.
Taylor, M. (2006), The Firm: Coalitions, Communities and Collective Agency, in
M. Taylor and P. Oinas (eds), Understanding the Firm: Spatial and Organizational
Dimensions, Oxford University Press, Oxford, pp. 87116.
Taylor, M. and Plummer, P. (2003), Promoting Local Economic Growth: The
Role of Entrepreneurship and Human Capital, Education and Training, 45(8/9),
55864.

APPENDIX 3A.1 THE NATURE AND ORIGINS OF

THE DATA SET
The variables used in this analysis were generated initially in the
1980s and early 1990s for a series of studies published through the
Australian Government Publishing Service (AGPS) in the Office of Local
Governments (OLG) Australian Regional Developments series. The pub-
lications were part of a large database describing the economic, social
and accessibility vulnerability and resilience of 94 Australian regions
(see Australian Regional Developments monograph A Regionalisation of
Australia for Comparative Economic Analysis, 1988). The data are unique,
developed from initially unpublished data sources and major empirical
analyses. Where possible, they continue to be updated.
Technological leadership at the enterprise level (HITECH)

The data surrogate for this measure is the regional significance of
R&D-intensive industries. The Department of Industry, Technology and
Commerce (DITAC) identified the following ASIC industry groups and
classes as Measures of Science and Innovation (1987, pp. 3789) containing
high-technology components: (1) Pharmaceutical and Veterinary produce;
(2) Aircraft Manufacturing; (3) Photographic, Professional and Scientific
Equipment: (4) Data Processing Services; (5) Research and Scientific
Institutions. The HITECH measure is calculated as the proportion of
employment in each region in these four-digit ASIC industries using
the unpublished data from the Integrated Regional Information System
(IRIS) database.
Knowledge creation and access to information (INFOACC)

This measure is based on a simple interaction model in which the size
of information activity at a place is measured as employment in profes-
sional and managerial jobs in each region, and the distance between
pairs of regions measured as time-distance by the quickest means avail-
able. The measure is more fully described in the OLG Working Paper,
Accessibility and Remoteness (1992, pp. 214), which describes six
elements of regional accessibility (access to goods and services, access
to intermediate goods markets, the cost of access to intermediate goods
markets, access to export ports, access to information and the cost of
access to information).
The local integration of small firms (MLOCN)

This measure is the inverse of the proportion of establishments in a region
that belong to multi-location enterprises, taken from the Australian

Bureau of Statistics IRIS database. This enables the removal from the
region of establishments with the weakest local affiliations.
Infrastructure support and institutional thickness (PROT)

This measure is the regional effective rate of industry protection drawn
from the OLG Working Paper (1992) The Regional Impact of Changing
Levels of Protection in Australian Industries. A weighted averaging
method has been used to allocate industry-effective rates of protection to
regions using sectoral employment levels using the IRIS database.
Local human resource base (DEGREE)

The measure used here is the proportion of the population in each region
without university degrees. This is a surrogate not only for local skill levels
but also for issues of income and, indirectly, for the local availability of
capital.
The power of large corporations (TOTPOP)

An index of corporate control in the region was constructed using the head
office regional address for companies listed in the top 1000 in the Business
Review Weekly magazine. Employment for each business was expressed as
a quotient of the total employment of the particular region in which it is
located using IRIS data.
Inter-regional trade (MKTACC)

The measure for this driver for regional growth is taken as a regions acces-
sibility to intermediate goods markets within Australia and has been taken
from the OLG Working Paper Accessibility and Remoteness (1992,
p. 16). It is expressed as a function of the employment level in each pair of
regions in intermediate goods industries such as manufacturing and con-
struction, and the direct road distance between the pairs of regions.
Local sectoral specialization (SPEC)

This is a simple measure of the numbers of business establishments taken
from the IRIS database for each industry category in each region.

4. Endogenous rural development
from a sociological perspective
Frank Vanclay
DEFINING ERD SOCIOLOGICALLY
For many sociologists and some social geographers1 endogenous regional

development is synonymously endogenous rural development (both
abbreviated ERD). It is a multidimensional concept, a multi-level,
multi-actor and multi-faceted process (Van der Ploeg et al., 2000, p. 391).
Much more than a statement about the origin of development initiatives,
in the sociological perspective ERD has layers of meaning. Above all,
ERD is a new paradigm, worldview and philosophy about appropri-
ate development, rural renewal and a multifunctional, post-productivist
countryside (Slee, 1994; Van Broekhuizen et al., 1997; Ray, 1997; Van der
Ploeg et al., 2000; Marsden, 2003; OECD, 2006).
At a very basic level, if exogenous development is development that is
initiated outside a local region (i.e., externally) then endogenous develop-
ment is the opposite. In other words, ERD is development that is initiated
and controlled by the local community. However, from a sociological
perspective, ERD implies much more than this. Whereas exogenous
development is seen as modernist, Fordist and top-down, ERD is seen
as bottom-up and a reaction to modernization, or as resistance to it
(Bassand et al., 1986). To some extent, post-modernism and post-Fordism
are characteristics of ERD in that it accommodates niche-marketing of
value-added product and flexible specialization, while at the same time
valorizing local culture, tradition, artisanal production and regional
typical food.
In general, ERD refers to the utilization and celebration of local and
place-based dimensions of a region as the basis of its economic activity
and livelihood. The emphasis in most sociological approaches to ERD
is in understanding the characteristics (natural, human and cultural) of a
place that makes it special and/or distinctive (different from other regions),
and how these may become the focus of sustainable economic activity.
Some sociologists critique the concept and especially how it has been
59

appropriated by the state to devolve not power and control but responsi-
bility and costs (see, for example, Herbert-Cheshire, 2000).
ERD has had much policy significance in the European Union (EU)
because of the realization that conventional subsidies to agriculture were
not effective in promoting growth or development in rural areas and that a
new approach focusing on local potentialities was needed (OECD, 2006).
Considerable interest in ERD occurred because of the EU LEADER
programme (discussed later in this chapter), which was conceived as an
integrated and endogenous approach to rural development (ibid., p. 90).
There is not a singular coherent position to ERD in the sociology dis-
cipline, but a range of approaches. Furthermore, there is no agreed-upon
definition (Ray, 1999; Van der Ploeg et al., 2000). Nevertheless, perhaps
Long and Van der Ploeg (1994) best sum up the sociological perspective.
For them ERD is founded mainly, though not exclusively, on locally avail-
able resources, such as the potentialities of the local ecology, labour force,
knowledge, and local patterns for linking production to consumption, etc
(pp. 12). ERD and exogenous development should be seen as a dualism of
ideal types that blend in the development strategies of regions rather than
as a mutually exclusive dichotomy (Cristovao et al., 1994; Kneafsey et al.,
2001). They are, in essence, a heuristic device (Van der Ploeg et al., 2000).
CHARACTERISTICS AND VALUES OF ERD
Despite the lack of a common definition, there is general agreement around

the characteristics of ERD in the sociological perspective. Expanding on
Slee (1994) and Bowler (1999), the key elements are:
a goal to create diversified, resilient and sustainable local economies;

local determination of development options;
local control over the development process;
retention of benefits locally;
utilization of locally available resources (natural, human and
cultural);
valorization of the local and place, especially what is locally
unique or special, and respect for local values;
awareness of the rural as being post-productivist that is, being a
site of consumption as well as a site of production;
appreciation of multifunctionality.
The multifunctionality element recognizes that rural areas (and agricul-

ture) provide a range of non-market (as well as market) goods and services

A sociological perspective 61
such as environmental protection, landscape management, preservation

of biodiversity and habitat protection, ecosystem services, carbon sinks,
maintenance of cultural heritage, employment and livelihoods for rural
people and food security.
The social values implied by ERD in a sociological approach prima-
rily relate to development as a social concept rather than to economic
growth (Brugger, 1986). Perhaps in contrast to some other chapters in
this book, ERD should not be taken to infer regional economic growth!
Sustainability, in its broadest possible understanding, is a key value
(Murdoch et al., 1994). Local control and self-determination are also
fundamental ERD highly values the rights of people to have a say in
the things that affect their lives. The principle of subsidiarity that deci-
sions should be taken as closely as possible to the citizen and that larger
groupings should encourage the autonomy of smaller groupings (Carozza,
2003) is firmly embodied in the sociological and policy approach to
ERD. Notions of equity, capacity building, community development and
building resilient communities that have vitality, viability and health, are
all part of the concept.
CONTRASTING ENDOGENOUS AND EXOGENOUS
ERD arose because of the failing of exogenous development to deliver

benefits to rural areas. Exogenous development, with its focus on econo-
mies of scale and concentration of activity, has led to the industrialization
of urban growth poles, despite regions attempting to attract multinational
corporations by providing incentives, lowering taxes and sometimes
reducing other regulations in a race to the bottom. In that model, the role
of rural areas was to provide cheap food, and as a result they were increas-
ingly marginalized (Lawrence, 1986; Vanclay and Lawrence, 1993, 1995;
Vanclay, 2003). In the words of Ward et al. (2005, p. 4; italics original):
[Exogenous development] was criticised as dependent development, reliant on

continued subsidies and the policy decisions of distant agencies and board-
rooms. It was seen as distorted development, which boosted single sectors,
selected settlements and certain types of business (e.g. progressive farmers) but
left others behind and neglected the non-economic aspects of rural life. It was
cast as destructive development which erased the cultural and environmental dif-
ferences and considered to be dictated development devised by external experts
and planners from outside local rural areas.
There has been some criticism of ERD along the lines that it is a nice
ideal but not practical, that local areas will never be free of external

influences especially in a globalizing world, that rural will always exist in

juxtaposition to urban, that local will always interact with extra-local, and
that exogenous and endogenous processes need to be balanced (Lowe et al.,
1995; Ray, 2001a; Ward et al., 2005). As a result, Ray (2001a) coined the
term, neo-endogenous (rural) development to mean endogenous-based
development in which extra-local factors are recognised and regarded
as essential but which retains a belief in the potential of local areas to
shape their future (p. 4). Rays term neo-endogenous development, has
generated much discussion in the academic literature and at conferences.
Personally I am not fully convinced of its merits as indicated earlier,
the concepts were never intended to be a mutually exclusive dichotomy.
However, perhaps the popularity of the new term can be attributed both
to the need for increased emphasis that an integrated approach is neces-
sary, and the erroneous assumption that ERD was necessarily in opposi-
tion to extra-local factors. ERD is meant to stimulate local business to
supply local markets, and also to create external interest in local places
and stimulate place-based consumption, for example through ecotourism,
cultural tourism, farm-stays, rural B&Bs and the purchase of place-based
products for consumption in situ or at home. External interest is thus vital
to an ERD approach.
THE CULTURE ECONOMY APPROACH
In ERD the focus of attention of development initiatives is local resources,

not as traditional inputs for large extra-region or multinational firms, but
as the basis of the economic activity of local small to medium firms. The
concept of what a resource or asset can be thus expands from natural
and human capital to cultural capital. Ray (1998) calls this the culture
economy approach to rural development. With the valorization of local
culture, many things potentially become of interest to people with spare
time and/or disposable income in an information-rich, post-Fordist, post-
modern society.
Expanding considerably from Ray (1998, 2001a), the potential local cul-
tural assets and how they could be utilized by regions to promote tourism
and/or add value through place-based marketing are listed in Appendix
4A.1 at the end of the chapter. Regions that have highly significant fea-
tures, for example that might qualify for UNESCO World Heritage status,
have a clear advantage. The intention with the culture economy approach
is for all regions to identify the particular strengths they each have that
could be capitalized on, even though they may not be globally signifi-
cant. The list in Appendix 4A.1, long as it is, nonetheless remains a small

selection of the many possible cultural assets regions can identify to make
their region special. Many regions may also have natural resource assets
that would also attract interest. Perhaps it should be noted that there has
been some critique of this approach. Graham Day (1998), for example,
argues that it shows a misunderstanding of the concept of culture in that
it focuses on the material artefact rather than on social relationships and
shared understandings, meanings and values.
THE SIGNIFICANCE OF THE EU LEADER

PROGRAMME
LEADER is an acronym deriving from the French expression Liaison

Entre Actions de Dveloppement de lEconomie Rurale,2 which means
links between actions for the development of the rural economy
(European Commission, 2007). The LEADER programme was conceived
as an integrated and endogenous approach to rural development within
the European Union (OECD, 2006, p. 17). Ray (2001b, p. 280) suggests
that:
the official rationale of the LEADER intervention portrays it primarily as a

tool with which to redirect the trajectory of the rural economy so as to meet
concerns about the failure of regional economic convergence, the fragility
of primary sectors of the economy (especially agriculture) and the financial
burden of the [Common Agricultural Policy].
It was targeted at territories of less than 100 000 inhabitants.
Phases
There were three phases to the LEADER programme:
LEADER1 from 1991 to 1994;

LEADER2 from 1994 to 1999;
LEADER+ from 2000 to 2006.
The LEADER programme is now regarded as having reached a level of

maturity enabling rural areas to implement the LEADER approach more
widely in mainstream rural development programming (European Union,
2005, Clause 50). From 2007 on, the LEADER approach will continue
to persist, but has been integrated into mainstream rural development
programmes (European Commission, 2006). LEADER 1 with a budget
of EUR417 million was implemented in 217 territories. LEADER2 was

applied in approximately 1000 territories with a budget of EUR1755

million (LEADER European Observatory, 2001, p. I.14). There were
893 LEADER + local action groups in 2004 prior to enlargement (i.e.,
admitting the new member nations) with a budget of EUR2143 million
(European Commission, 2007, pp. 2, back cover).
Components
There are seven components to the LEADER approach (condensed from

European Commission, 2007, pp. 23):
1. Area-based approach: this entails defining a local development policy

on the basis of an areas particular situation, strengths and weak-
nesses. Areas should be fairly homogeneous and characterized by
internal social cohesion, shared history and traditions, a sense of
common identity, and so on.
2. Bottom-up approach: this aims to encourage participatory decision-
making at local level for all development policy aspects. Its objective
is the involvement of local players, including the community as a
whole, economic and social interest groups and representatives of
public and private institutions. The bottom-up approach relies on two
major activities, animation (facilitation of activities) and training
of local communities. It is important that the project is initiated by
local actors and that the public concerned with the action has been
consulted.
3. Partnership approach and the local action group: the LAG is a body
of public and private actors, united in a partnership that identifies a
joint strategy and a local action plan for developing an area. The LAG
represents a model of organization that can influence the institutional
and political balance of the area concerned. The LAGs provide appro-
priate mechanisms for participation, awareness raising and organiza-
tion of local actors in favour of rural development.
4. Innovation: LEADER demands that proposed actions are innova-
tive. They may be: actions to promote local resources in new ways;
actions that are of interest to local development but not covered by
other development policies; actions providing new answers to the
weaknesses and problems of rural areas; or actions that create a new
product, new process, new forms of organization, or a new market.
Innovation is also embodied in the programmes pedagogical and
networking components: disseminating information to other groups
of players wishing to gain inspiration from achievements elsewhere,
or to carry out joint projects.

5. Integrated approach: the actions and projects in the local action plan
should be linked and coordinated as a coherent whole. Integration may
concern actions within a single sector, all programme actions or specific
groups of actions, or, most importantly, links between the different
economic, social, cultural, environmental actors involved in the area.
6. Networking and cooperation between areas: by facilitating the
exchange and circulation of information on rural development poli-
cies and the dissemination and transfer of good practice and innova-
tive strategies and actions, the LEADER network aims to limit the
isolation of LAGs and to create a source of information and analysis
of the actions. Some LAGs have spontaneously organized themselves
into informal networks. Another core part of LEADER is the coop-
eration between rural areas. Cooperation between areas can be trans-
national but may equally take place between areas within the same
member state.
7. Local financing and management: delegating a large proportion of
the decision-making responsibilities for funding and management to
the LAG is a key element of the LEADER approach. However, the
degree of autonomy varies considerably depending on the member
states specific mode of organization and institutional context.
Significance
LEADER was significant in many ways. The foundations of LEADER

directly addressed the sociological critique of earlier models of develop-
ment and incorporated the sociological advocacy for a more participatory,
cultural economy approach. Sociologists were directly involved in devis-
ing LEADER as an EU-level framework and often at national and local
levels. They were involved in evaluating individual LEADER projects
and the LEADER programme as a whole. The various evaluations of
LEADER at project, member-nation and programme-wide (EU) level,
although identifying a number of specific shortcomings (see, for example,
Barke and Newton, 1997), overwhelmingly praised the approach.
While Ray (2000) somewhat cynically argued that the funding levels
were a very small proportion of the EU funding allocated for rural
development and that this reflected a lack of commitment, he also admit-
ted that LEADER was of considerable sociological interest. From an
Australian perspective, a country that eschews rural subsidies, the fact
that the approach potentially can be applied with a low level of govern-
ment funding is important. It was feasible in Europe to divert finances
from a very expensive system of payments to farmers under the Common
Agriculture Policy to fund LEADER, especially since this was expected to

have a greater return on investment (benefit to rural areas) and because

this partly addressed World Trade Organization concerns about protec-
tionism and trade distortion. Whether such an approach would be feasi-
ble, politically and otherwise, in countries without Europes resources and
past experience of government investment in rural areas is another ques-
tion. Therefore, the fact that this programme can be carried out virtually
without money (ibid., p. 165) is a virtue.
CONCLUSION
Endogenous rural development (ERD) is a powerful way of renewing

and revitalizing the countryside. Its benefits have been documented in the
OECD (2006) report, The New Rural Paradigm, and in much literature
emanating from the LEADER programme (e.g., European Commission,
2007). Because of a focus on local resources and on mobilizing (animat-
ing in LEADER language) local community activity, ERD has far greater
potential to be sustainable than, typically, do exogenous forms of develop-
ment (Gralton and Vanclay, 2006, 2009). The benefits of ERD extend far
beyond what is normally considered by traditional economic indicators or
measured as growth. They include, for example:
increased pride in where people live;

increased sense of being part of a community;
increased interest in participating in community activities;
increased social networks (social capital);
an increased sense of place (Vanclay, 2008).
All of these can be summarized as increased social well-being and quality

of life. Most likely this translates into improved health as people lead more
active, engaged lives.
Despite the practical experience provided by LEADER, ERD is prima-
rily a philosophy about appropriate development at the local level. ERD
argues for a shift in thinking about development initiatives and what local
resources might entail, and how endogenous and exogenous development
can work together for sustainable regional development.
NOTES
1. As I have been requested to take a sociological perspective, I will use the words sociol-
ogy, sociologist and sociological. However, I point out that a sociological perspective

is not necessarily distinguishable from a more generic social science position and cer-
tainly blends with and draws heavily on the social and human geography literature.
Further, although I have been a member of the Australian Sociological Association
for over 25 years and am a former President of the International Rural Sociology
Association, I also identify as a human geographer and am a member of the Institute of
Australian Geographers and a fellow of the Royal Geographical Society in the UK. In
general, I take an eclectic approach and ascribe to transdisciplinarity.
2. It is curious that the OECD report The New Rural Paradigm (OECD, 2006, p. 101) has
a peculiar wording of Liaison Entre Activits du Dveloppement de lEconomie Rurale
rather than what is most likely correct, Liaison Entre Actions de Dveloppement de
lEconomie Rurale, at least according to the majority of sources consulted including the
FAQ website for LEADER+ http://ec.europa.eu/agriculture/rur/leaderplus/faq_en.htm
and the French language version of the EU portal: www.welcomeurope.com (both
accessed 19 May 2010.).
REFERENCES
Barke,M. and Newton, M. (1997), The EU LEADER Initiative and Endogenous

Rural Development: The Application of the Programme in Two Rural Areas of
Andalusia, Southern Spain, Journal of Rural Studies, 13(3), 31941.
Bassand, M., Brugger, E.A., Bryden, J.M., Friedman, J. and Stuckey, B. (eds)
(1986), Self-reliant Development in Europe: Theory, Problems, Actions, Gower,
Aldershot.
Bowler, I. (1999), Endogenous Agricultural Development in Western Europe,
Tijdschrift voor Economische en Sociale Geografie, 90(3), 26071.
Brugger,E.A. (1986), Endogenous Development: A Concept between Utopia and
Reality, in M.Bassand et al. (eds), Self-reliant Development in Europe, Gower,
Aldershot, pp. 3858.
Carozza, P. (2003), Subsidiarity as a Structural Principle of International Human
Rights Law, The American Journal of International Law, 97(1), 3879.
Cristovao, A., Oostindie, H. and Pereira, F. (1994), Practices of Endogenous
Development in Barroso, Northern Portugal, in J.D. Van der Ploeg and
A. Long (eds), Born from Within: Practices and Perspectives of Endogenous
Rural Development, Van Gorcum, Assen, pp. 3858.
Day, G. (1998), Working with the Grain?: Towards Sustainable Rural and
Community Development, Journal of Rural Studies, 14(1), 89105.
European Commission (2006), The EU Rural Development Policy 20072013
(Fact Sheet), Office for Official Publications of the European Communities,
Luxembourg.
European Commission (2007), A Selection of Leader + Best Practices 2007/1,
European Observatory of Rural Areas, European Commission, Brussels.
European Union (2005), Council Regulation (EC) No. 1698/2005 of 20 September
2005 on support for rural development by the European Agricultural Fund for
Rural Development (EAFRD), Official Journal of the European Union L 277
(21 October 2005).
Gibson,C. and Connell, J. (2005), Music and Tourism: On the Road Again, Chanel
View Publications, Clevedon.
Gralton, A. and Vanclay, F. (2006), Quality Food Production, Diversity and
Sustainability: Opportunities for Small Towns, in M. Rogers and D. Jones

(eds), The Changing Nature of Australias Country Towns, VURRN Press

(Victorian Universities Regional Research Network Press), Ballarat, pp. 12638.
Gralton,A. and Vanclay, F. (2009), Artisanality and Culture in Innovative Regional
Agrifood Development: Lessons from the Tasmanian Artisanal Food Industry,
International Journal of Foresight and Innovation Policy, 5(1/2/3), 193204.
Herbert-Cheshire, L. (2000), Contemporary Strategies for Rural Community
Development in Australia: A Governmentality Perspective, Journal of Rural
Studies, 16(2), 20315.
Kneafsey, M., Ilbery, B. and Jenkins, T. (2001), Exploring the Dimensions of
Culture Economies in Rural West Wales, Sociologia Ruralis, 41(3), 296310.
Lawrence, G. (1986), Capitalism and the Countryside, Federation Press, Sydney.
LEADER European Observatory (2001), LEADER, From Initiative to Method: A
Guide to Teaching the LEADER Approach, available online at: http://www.fao.
org/sard/static/leader/en/index.htm; accessed 20 May 2010.
Long, A. and Van der Ploeg, J.D. (1994), Endogenous Development: Practices
and Perspectives, in J.D.Van der Ploeg and A.Long (eds), Born From Within:
Practice and Perspectives of Endogenous Rural Development, Van Gorcum,
Assen, pp. 16.
Lowe,P., Murdoch, J. and Ward, N. (1995), Networks in Rural Development:
Beyond Exogenous and Endogenous Models, in J.D. Van der Ploeg and
G. Van Dijk (eds), Beyond Modernization: The Impact of Endogenous Rural
Development, Van Gorcum, Assen, pp. 87105.
Marsden, T. (2003), The Condition of Rural Sustainability, Van Gorcum, Assen.
Murdoch, J., Ward, N. and Lowe, P. (1994), Sustainable Agriculture and
Endogenous Development: A Socio-Political Perspective, in J.D.Van der Ploeg
and A.Long (eds), Born From Within: Practice and Perspectives of Endogenous
Rural Development, Van Gorcum, Assen, pp. 26279.
OECD (2006), The New Rural Paradigm: Policies and Governance, Organisation
for Economic Co-operation and Development, Paris.
Ray, C. (1997), Towards a Theory of the Dialectic of Rural Development,
Sociologia Ruralis, 37(3), 34562.
Ray, C. (1998), Culture, Intellectual Property and Territorial Rural Development,
Ray, C. (1999), Towards a Meta-framework of Endogenous Development:
Repertoires, Paths, Democracy and Rights, Sociologia Ruralis, 39(4), 52137.
Ray, C. (2000), The EU LEADER Programme: Rural Development Laboratory,
Ray, C. (2001a), Culture Economies: A Perspective on Local Rural Development
in Europe, University of Newcastle, Centre for Rural Economy, monograph
available online at: http://www.ncl.ac.uk/cre/publish/books.htm; accessed 19
May 2010.
Ray, C. (2001b), Transnational Co-operation Between Rural Areas: Elements
of a Political Economy of EU Rural Development, Sociologia Ruralis, 41(3),
27995.
Slee, B. (1994), Theoretical Aspects of the Study of Endogenous Development,
in J.D. Van der Ploeg and A. Long (eds), Born From Within: Practice and
Perspectives of Endogenous Rural Development, Van Gorcum, Assen, pp. 18494.
Van Broekhuizen, R., Klep, L., Oostindie, H. and Van der Ploeg, J.D. (1997),
Renewing the Countryside: An Atlas with Two Hundred Examples from Dutch
Rural Society, Misset, Doetinchem.

Vanclay, F. (2003), The Impacts of Deregulation and Agricultural Restructuring

for Rural Australia, Australian Journal of Social Issues, 38(1), 8194.
Vanclay, F. (2008), Place Matters, in F.Vanclay, M.Higgins and A.Blackshaw
(eds), Making Sense of Place, National Museum of Australia Press, Canberra,
pp. 211.
Vanclay, F. and Lawrence, G. (1993), Social and Environmental Impacts of
Economic Restructuring in Australian Agriculture, International Journal of
Sociology of Agriculture and Food, 3, 97118.
Vanclay, F. and Lawrence, G. (1995), The Environmental Imperative: Ecosocial
Concerns for Australian Agriculture, Central Queensland University Press,
Rockhampton.
Van der Ploeg,J.D., Renting, H., Brunori, G., Knickel, K., Mannion, J., Marsden,
T., Roest, K., Sevilla-Guzman, E. and Ventura, F. (2000), Rural Development:
From Practices and Policies Towards Theory, Sociologia Ruralis, 40(4),
391408.
Ward, N. et al. (2005), Universities, the Knowledge Economy and Neo-
endogenous Rural Development, Centre for Rural Economy Discussion Paper
Series, 1, University of Newcastle upon Tyne.

APPENDIX 4A.1 POTENTIAL LOCAL CULTURAL

ASSETS
Regional typical food, traditional foods, bushfood (native or wild
food), place-based foods and new foods (see Gralton and Vanclay,
2006, 2009).
Winery trails, real-ale trails (and micro-breweries) and, for example
(in Scotland especially) promoting visits to malt whisky distilleries
and/or the sampling of malts at pubs near to where they are distilled.
Regional languages and dialects and their associated cultures.
Local folklore and, for example, developing and promoting commu-
nity museums (ecomuseums), festivals and fairs that celebrate local
culture and folklore.
Local arts and crafts, especially those of high quality and that are
inspired by and/or depict local settings and/or utilize local materials,
for example, the extensive craftwork using Tasmanian native timbers
(e.g., Huon Pine, myrtle, sassafras) that is on sale at the weekly
Salamanca Craft Market and in Tasmanian galleries and boutiques.
Local festivals, especially where they celebrate sense of place such as
the Mountain Festival in Hobart, Tasmania; or some other activity
with local relevance such as the Wooden Boat Festival in Tasmania,
which partly celebrates Huon pine as a timber that was once used to
make boats.
Sites that have architectural significance, either because of a distinc-
tive architectural style (e.g., Art Deco), or as examples of the works
of specific architects such as, for example, colonial Tasmanian
Georgian architect, John Lee Archer, or Gothic revival architect
and designer, Augustus Pugin, who designed the interior of the
English Houses of Parliament as well as numerous churches and
religious paraphernalia in Australia and especially in Tasmania.
Sites that have musical significance, such as the buildings where now
renowned composers or performers once played, the houses they
lived in, or their birthplaces, deathbeds and final resting places; or
significant concert halls, recording studios, opera houses, organs or
carillons around the world (see further discussion in Gibson and
Connell, 2005).
Sites that are depicted in the movies or on television, for example
New Zealand has promoted the settings used in the Lord of the Rings
films, Barwon Heads in coastal Victoria promotes itself as the loca-
tion of a popular Australian TV series, SeaChange.
Sites that have literary significance, for example locations men-
tioned by Wordsworth, Thomas Hardys Wessex, or Lorna Doone

in Exmoor, or Gundagai in NSW where the dog sits on the tucker-

box (a statue nearby).
Sites associated with particular artists, for example Constable
country on the border of Essex and Suffolk in the UK, Albert
Namatjira in Central Australia, or Hans Heyson in South Australia.
Sites that have religious significance, such as places where there may
have been apparitions, places along the pilgrim route(s) to Santiago
de Compostela in Spain (El Camino), the journeys of Saint Paul in
modern-day Turkey, or places visited by significant religious figures,
for example, Sister Mary MacKillop, the only Australian to have
been beatified. Another example is Walla Walla, a small town of
around 600 people in south-central NSW, which boasts not only
the largest Lutheran Church building in Australia, but a story of an
1890s trek of German Lutherans cross-country from the Barossa
Valley in South Australia.
Sites that have sporting significance, such as the locations of sport
museums, famous playing fields, locations where some champion
was born, raised, played as a youngster, had success (or failure),
lived or died. For example in 2008, the centenary anniversary of the
birth of cricketer Don Bradman, many regions in Australia adver-
tised their connection to him.
Sites that have political or ideological significance, or where there
were political struggles or massacres (e.g., Robben Island in South
Africa where Nelson Mandela was incarcerated); places where
political leaders were born or grew up, or where significant political
events took place.
Sites with military significance, such as Gallipoli, the battlefields of
Flanders, the Kokoda Trail in Papua New Guinea, or the sites of
various incidents during the Vietnam war, or for example, the loca-
tion of the memorial to the loss of the HMAS Sydney in Geraldton
Western Australia, or the many battlements installed around the
coast of Australia.
Sites with industrial significance, such as the locations of mines or
factories, especially ones that have become culturally iconic, such
as the Cadbury chocolate factory in Hobart, or the Line of Lode in
Broken Hill (NSW).
Sites with historical significance, for example Burra in South
Australia, which had the largest mine in Australia in the middle of
the nineteenth century, and even though now a small town of only
around 2000 people has become famous in the form of the Burra
Charter (The Australia ICOMOS Charter for Places of Cultural
Significance).

Sites with criminal or penal significance, for example buildings that

were built by convicts, that housed convicts, or places where bush-
rangers hid (e.g., Ben Halls cave in Grenfell NSW) or conducted
their hold-ups (e.g., Glenrowan Victoria where the Ned Kelly Gang
had their last stand).
Sites of current indigenous or archaeological/anthropological signif-
icance, for example any place where there are Aboriginal rock paint-
ings or engravings, sites where significant artefacts or middens have
been found, sites where indigenous people may have lived, burial
sites, massacre sites, spiritual or sacred areas. [Note that there will
need to be negotiations with traditional owners before any attempt
is made to promote these areas.]
Even locations that were notorious can make a feature of their
infamy.

5. Rural, urban or regional
endogenous development as the core
concept in the planning profession
Edward Blakely
INTRODUCTION
The basic concept of endogenous development or deriving outcomes

from local resources is based on the notion that the local resource is
primary or essential to organizing or producing any results or tangible
products such as a building location or an intangible, such as a public
policy. Planning in all forms relies on elements of indigenous activities
and/or endogenous resources because planning as a policy science is based
on the use of rules to shape outcomes in space and place. All communi-
ties in the world no matter how large or small declare themselves to
be unique in some dimensions. In some respects this is correct because no
bounded area, such as a metropolitan area, is identical with any other. On
the other hand, human habitats in systems do not vary markedly around
the world from the most primitive to the most advanced. There are similar
features in all landscape and social organizations. Nonetheless, the par-
ticular mix of these resources is endogenous that is, it is local-specific.
No matter what the prime resource is from a coal mine to a park or
housing development some rules have to be devised to determine how
the natural resource is extracted or modified or manipulated, requiring
decisions on how the resource is to be exploited to enhance benefit for the
local community.
Thus, planning is the means to provide order to protect the environ-
ment, public safety and movement of goods. The Plan is the base of the
resource exploitation. Plans are descriptive instruments. That is, a plan
in the sense of rural, urban or regional planning is contained in a specific
locality and meant to engage the resource of that locality for advance-
ment, enhancement or preservation. In a much larger sense, planning and
plans are ways to articulate the endogenous character, assembly and use of
human and physical resources. The one that they use frequently becomes
73

the regulatory and monitoring instrument that exemplifies the resources

depictions. Even building height limitations act to provide ways for the
visual uniqueness of a place to be preserved. For example, the sight-lines
to the US national capitol, Washington DC, are preserved so that it is the
central feature of and the manifestation of the US nation state.
The evolution of the planning discipline comes from these former
notions of The Plan as the defining instrument for the preservation and
articulation of the relationship between humankind and space. The notion
of plan has refined the ways in which a plan becomes a code or legisla-
tive instrument for local endogenous resource control. In many respects,
planning as a science rests on the concept of how well local spatial and
social/human resources are mobilized and utilized for endogenous devel-
opment. The plan is thus the mechanism that matches human and physical
resources for developing the better use of local resources in the develop-
ment of a high quality of life in a specific place. Planning styles arise from
various ways of understanding how local human and spatial endogenous
resources can be combined within a city or other jurisdictional form.
Planning as described here is as proclaimed by leading US planning theo-
rists Charles Hoch, Linda Dalton and Frank So (Hoch et al. 2000): A
communitys physical aspects its size, climate, location and geography
shape many of its planning issues. Other less immediately visible feature
factors such as a communitys history and social and demographic make
up also contribute to its uniqueness (p. 33). However, some scholars
consider planning as being a-spatial; that is, planning is organizing any
activity from running companies to organizing social organization. In this
chapter we focus on the professional practice of planning that creates the
link between organizing and controlling an indigenous land resource base
or creating ways to develop, organize and use resources within a land-
defined space that is in a rural, urban or metropolitan setting.
THE EVOLUTION OF PLANNING AS INDIGENOUS/

ENDOGENOUS ASSESSMENT
The earliest formal literature of settlement describes how the local earth
endowed and the resources generated a rationale for placing certain infra-
structure, from roads and bridges to aqueducts, in an orderly fashion
designed to maximize resource exploitation and protect or advance some
form of settlement scheme. This set of activities evolved into what we
now term professional planning. The earliest records indicate that the
Romans, the Egyptians, the Greeks and other early civilizations had a very
high regard for laying out settlements to advance specific purposes. The

The core concept in the planning profession 75
Pharaohs of early Egypt were particularly adept at settlement design and

planning for the use of natural local resources. The Bible extols the virtues
of layout and design of places like Babylon. In this respect the notion of
indigenous planning was central to the development of the mighty Inca
and Aztec empires as well as the guiding principles of Chinas great cities.
So, the idea of an indigenous base for plans is as old and as deep as the
notions of professional planning. Indigenous asset development in this
concept is the bedrock of the modern planning profession and manifests
itself in several contemporary ways that are a continuum of the fields
ancestral roots. Here we will only describe the current form of local and
parochial resources as they provide the template for the profession of
planning.
Planning needs to be regarded as being both a process and a form. As
a process, planners in the last century have placed as a central tenet the
notion of citizen engagement and assessment of local needs as the critical
bedrock of The Plan.
In essence, a plan can only hold legitimacy if it grows out of a collec-
tive engagement with locality (defined as the geographic area the plan
embraces) and deals with a collective understanding of the resources that
the plan purports to advance. As a result, the planning process considers
all local indigenous resources human, physical, architectural, historical
and cultural as necessary elements to form the base for the planning
process. This means the reification of local assets is part of the indigenous
resource base. Collective description of resources is as important as the
actual scientific investigation and revelation of any local indigenous asset.
That is, if the community does not see a local stream or creek as a value,
then the planners science is challenged by community ascription versus
real evidence.
One of the classic examples is the often cited instance where a naive
planner attempts to convince villagers in the Amazon to build a vegetable
garden to improve local nutrition. The villagers object because they will
not build any new projects until a monument is built honouring a local
earth god. In essence, the planner has to consider belief as central to the
plan as material goods. The villagers view plans as incidental to larger
cosmic forces over which humankind has little control. So, planning as a
practice as we know it recognizes the local endowments and is rooted in
the notion of human control over natural environments. This has led to
the sad consequences of climate change through too much planned carbon
use with automobiles and factories.
Clearly, communities can view cultural and related practices as indig-
enous assets even if they are contentious, such as religious historical roots
or community practices that may be at variance with larger mores or laws.

Several obvious examples of this are well documented by Tomlinson et al.

(2003) and others as they describe the use of planning processes to justify
racial segregation as valued as indigenous resources in South Africa and
Northern Ireland, Israel and Palestine. For many years, so-called Jim
Crow Laws were codified in planning codes in the US to spatially separate
blacks and whites. In essence, planning is a policy tool that incorporates
and institutionalizes human values spatially. Clearly, there is more illustra-
tion of the reification of local indigenous/endogenous cultural resources.
New Orleans, post-Katrina in 2005, is using its past indigenous cultural
base as the primary object in re-planning the city after the hurricane. In
the case of New Orleans, the old indigenous spatial separations were chal-
lenged in a post-storm setting when people could articulate another set
of values that were not based on race and class as the way space can and
should be organized.
LAND AND ITS USES AS ENDOGENOUS PLANNING

BASE
Thus, the larger and more profound aspect of indigenous resources for
planning is how land is used and controlled, ranging from parcel sizes and
heights to access and transportation. In this case the use of land for various
purposes is based on an assessment of how the in-place resources might be
used for various purposes and linked together to form new opportunities.
Planning a railroad to carry natural mineral resources is a good example.
In this case, the natural resource that is a collective resource is exploited as
a reasonable plan for current and future use. But as the exploitation takes
place other natural resources may well be jeopardized. So, the fact that
the plan is to use something indigenous is not in itself an attribute so the
planning process of collective assessment has to include all of the options,
opportunities, barriers and threats this resource requires. This is how
planning is based on the endogenous aspects of the setting.
As we enter a new millennium the local asset is no longer what the place
is but who is in the place and how they understand the use of place assets
in presenting the human knowledge base to the local region and the world.
Computer technology transforms the notion of endogenous advantage
since global interchange can potentially now take place from any loca-
tion thus making many activities completely footloose in that regard.
Until the advent of computer technology, places were valued by their
connectivity via water, rail, road or air. Now, the place as natural setting
or collection of human skills can be an endogenous value. For example,
call centres handling airline reservations around the world can be located

in a picturesque rural area that is attractive for lifestyle reasons for the
call centre workforce or because the area is less expensive but the quality
is very high. Many small rural communities in the US are now economi-
cally robust because they are attracting a cross-section of people who can
work from these locations locally and globally. In Australia, the so-called
sea-changers who move to coastal locations are founding new business in
small coastal villages and linking to the world (Gurran et al., 2007). So the
endogenous resource to be planned in this case is how to provide locally
access to the ITC infrastructure that is necessary for small- and medium-
sized businesses (including one person operations) to have access to the
web, so that a local endogenous advantage in this regard is equivalent to
that elsewhere where such infrastructure access is available.
ENDOGENOUS ECONOMIC DEVELOPMENT

PLANNING
Within planning the activity of local economic development planning is

based on endogenous processes. The basic theory associated with local
economic development planning is that capitalizing on local indigenous
resources is the springboard for creating a sustainable economy. Blakely
and Bradshaw (2002) have laid a strong foundation for the notion that
jobs and wealth are not produced by attraction or altering incentives for
firms but must be induced by the quality of the location and the use of
local resources to develop equitable and sustainable wealth. Blakely and
Bradshaw reject the old theories of development that based the process
of progress on the attraction and retention of firms to what they term the
Third Wave Theory:
1. It assumes that better and not more populations is the key to eco-
nomic improvement.
2. It bases economic growth not on attracting firms but attracting and
retaining high-quality human resources with the institutions, learning,
research and innovation to retain them.
3. It places all communities in the global economy and not competing
with one another within their regions for the same market share of a
diminishing resource base. It proposes multi-economic bases versus
single industrial or sector bases for all communities, both urban and
rural.
4. It views social equity as critical to community well-being and stability.
5. It places technological and social innovation at the forefront of
economic expansion. (Ibid. pp. 478)

Sustainable wealth cannot be imported as Blakely and Bradshaw

illustrate above but must rest on local endogenous resources that are not
transferable. Even if the form of the indigenous resource is altered that
is, the factory is not competitive the people and place can alter their
competitive strategy by reconfiguring their resources to compete and not
attempt to copy some other place or relocate a firm from another loca-
tion. As the base for manufacturing changes from machine and natural
resource, a community must maintain its leadership by increasing the
skills of its population or adding other value to the goods or services. In
this way the real assets, the people and place, are recreated by planning.
In this model Blakely and Bradshaw (ibid.) assert that local economic
development planning is not a one-time but a continuous process of
adjusting resources internally and developing new means to compete with
those resources in different ways. To illustrate this point, communities
that redefine their current position can use a combination of place features
and institutional resources. A community that was a fishing village and
has a picturesque setting can leverage the setting after the fish are gone by
repositioning itself with its local college as a new music and arts venue that
attracts artists and musicians all year round using the old fish factories
as studios for music and art as well as the teaching staff at the college as
the artists in residence for the programme. This type of transformation
might attract new year-round residents who retire to the place for continu-
ing artistic education for themselves or to be part-time teachers and thus
improve their own lives in the process of developing a new economic niche.
The Product Cycle Trap
Places that import economic activity become victims of the import rather
than the resource to build continuing wealth. Thus, it is no accident that
poverty and natural resource extraction are so closely linked because the
extraction of the natural resources usually degrades the human resources
as well. Of course, when the natural resource is gone, the human resources
are discarded. Factory towns are the prime example. In the early years, the
town has a competitive advantage with energy and or labour cost being
low. The wage, transport or resource arbitrage as it is called the differ-
ence between the costs and profit margin is reduced over time as other
places closer to the markets or with lower labour or energy cost compete
for the same goods production. As a result, the factory town dies and its
workers are stranded. The only way for places to remain competitive in
a global economy is to build on local resources and retain the wealth or
added-value components as close to the source of exploitation as possi-
ble. Similarly, manufacturing and related mass production activities as

standardized as Vernon (1966) in his classic work has shown will move
to the lowest-cost areas, leaving the last host denuded to the capacity to
adjust and use the same skilled workforce or other resources to maintain
economic perch.
The product cycle theory goes like this according to Vernon (ibid.,
p. 52):
Economic development is defined as the creation of new products and the
diffusion of standardized products. Development originates in the more devel-
oped region and is exported to the less developed region through trade and
investment. Establishing a new industry in the less developed region creates a
progressive force that can help eliminate the barriers to interregional equality.
Beyond the Product Cycle
Youngstown in Ohio is a classic example of the deindustrialization process.

Once a thriving rubber manufacturing centre that could boast that it pro-
duced the majority of rubber tyres in the world, it was to lose all its rubber
factories in the 1970s. Now Youngstown is a model of economic recovery
as it incubates with the local university to support new-technology firms in
rejuvenated old derelict downtown retail spaces. Similarly, Emeryville in
California, once a troubled industrial suburb of San Francisco, is home to
some of the worlds fastest-growing biotechnology firms because it created
good links with the nearby University of California at Berkeley, luring
firms to old factory spaces with local land use zoning for start-up firms in
biotechnology. So in these cases new resources were generated from old
endogenous resources that were thought to be of low or no value.
Planning endogenous resources is thus essential in local economic
development planning. Planning is the key in local economic develop-
ment. Plans must be based on the flexible and continuous improvement
of local resources so they meet the needs of changing times and markets.
Local economic development planning then is the epitome of endogenous
planning. Local resources put together in the correct manner can remain
competitive only through planning processes that recognize how these
resources can be positioned, organized and reorganized to meet altering
conditions regionally, locally and globally.
Moving to a new strategic approach does not neglect the older assess-
ments and engagement of community in determining local physical assets
but provides a new crucible in which these resources fit. This is a stage and
not a finish of indigenous asset incorporation into the planning process.
As human resources are increasingly important the social links among
people as discussed previously are being examined in new ways to help
understand how people can form the endogenous base for planning.

ENDOGENOUS PLAN AS SOCIAL SPACE

Social capital is a new vocabulary but an old idea. For many years,
the concepts of social capital have been understood. Cities and towns
have reputations with respect to how people feel included or excluded
and can play a beneficial or hostile part in the social dimensions of the
place. Physical endowments are sometimes described as drivers for social
interactions and the formation of social space. For example, people who
live in cold wet places are said to be less sociable than people who live in
warm sunny places. In this sense the climate is an endogenous resource
that planners acknowledge in the planning process that promotes or
impedes social capital formation. Furthermore, people who live in
some cities seem to be bound by traditions that impede new projects or
programmes.
As mentioned earlier, New Orleans has experienced many failures when
ideas that succeeded elsewhere were attempted there. The New Orleans
World Fair in 1984 is an illustration of this. The city newspaper and lead-
ership of the old guard did not like the idea so they conspired to let it fail
rather than embrace it. On the other hand, San Jose in California built a
downtown arena with no sports team and eventually got an ice hockey
team to play in the arena. The idea of a northern-temperature ice hockey
team surviving in sunny San Jose was derided nationally but the team and
arena led to a downtown renaissance because of the civic spirit that led
the effort (Henton et al., 2006). So, human civic spirit is an endogenous
resource that firms and global capital is beginning to recognize in the form
of places like Dubai that offer almost no other resource than a strong can
do attitude from its civic leaders.
As discussed earlier, natural assets are not as pivotal in planning.
Furthermore, social and cultural attributes are increasingly important as
human capital attractors. Social capital is an endogenous resource that
comes from social contacts that generate social contracts among a set of
people who usually bond by space. That is, bonds of social relationships
can be a resource as much as any hard fixed assets. Social capital is inte-
gral to planning since it is the interactions that create new ideas as well
as act as the glue for capital formation and wealth generation. Individual
reliability is enhanced with collective obligation. Thus, social capital in
planning means sharing opportunities and resources to maximize desired
outcomes. So, simple social capital might be thought of as in friends and
colleagues working to fix a group members home for habitation post a
disaster, and it can in fact extend across multiple geographic scales from
a local community to the nation to the world (Putnam et al., 2003). Some
communities treasure this form of community collective action. There are

several excellent examples of endogenous social capital as the base for

planning both economic and land use activities. The Basque Madrigon
region is a prime example of collective social capital infusing all aspects of
community life and planning all activities related to it. And a recent and
well-known illustration of planned social capital is the Grameen Bank pio-
neered in Bangladesh with many off-shoots around the world in various
formulations of micro-credit (Yunus, 2007).
These forms of social capital are not part of the lexicon of planning at
every level. The concept of restoring social capital as one of the essen-
tial pillars of planning is increasingly popular in the planning literature.
Communities that lack endogenous social capital cannot determine their
own destiny because they lack the social glue to articulate a direction
and the common sense of long-term vision and spirit to hold that vision
through time. No plan can survive if the society does not view the plan as
an articulation of common purpose and protect that purpose through an
array of social capital institutions.
SUMMARY AND TAKE-HOME POINTS
Regional is local in the sense of endogenous planning (Stimson et al.,

2002); that is, the base resources set in a locality, no matter how defined,
are essentially at the smallest organized units of economic and social
space. So a region is a collectivity of interlocking resources, human and
physical, that can articulate a common planned use of these resources. The
planning paradigm:
incorporates the notion of endogenous to assets or locates the

common regional resource base;
creates a vision for these resources that can be communicated
through an instrument called a plan;
generates the social cohesion/capital to forge a destiny based on the
local/regional resources;
creates an economic system from these resources that is in a continu-
ing state of planning so that the region/locality remains competi-
tive by inducing a consistent flow of ideas and resources to it and
through it.
In these ways endogenous planning is basically the bedrock of quality

social, physical and economic development. Without a plan to link and
forge common purpose new endogenous resource wont be sustainable
for long. Planners are not merely police officers of rules for the present

but crafters of the future by organizing and expressing how to use local/
regional resources optimally in ways that sustain any region or community.
REFERENCES
Blakely, E.J. and Bradshaw, T.K. (2002), Planning Local Economic Development,
Sage Publications, Thousand Oaks, California.
Gurran, N., Blakely, E.J. and Squires, C. (2007), Governance Responses to Rapid
Growth in Environmentally Sensitive Areas of Coastal Australia, Coastal
Management, 35(4), 44565.
Henton, D., Melville, J. and Walesh, K. (2006), Grassroots Leaders for a New
Economy: How Civic Entrepreneurs Are Building Prosperous Communities,
Jossey Bass Nonprofit and Public Management Series, San Francisco.
Hoch, C., Dalton, L. and So, F. (eds) (2000), The Practice of Local Government
Planning (3rd edn), Washington, DC: International City/County Management
Association, Municipal Management Series.
Putnam, R.D., Feldstein, L. and Cohen, D. (2003), Better Together: Restoring the
American Community, Simon & Schuster, New York.
Stimson, R.J., Stough, R.R. and Roberts, B.H. (2002), Regional Economic
Development: Analysis and Planning, Springer, New York.
Tomlinson, R., Beauregard, R.A., Bremner, L. and Mangcu, X. (2003), Emerging
Johannesburg, Routledge, London.
Vernon, R. (1966), The Myth of Urban Problems, Harvard University Press,
Boston.
Yunus, M. (2007), Creating a World Without Poverty, Public Affairs Books,
Washington DC.

6. Diversity and endogeny in regional
development: applying appreciative
intelligence
Tojo Thatchenkery and
Jessica Heineman-Pieper
INTRODUCTION
Endogenous processes in regional economic development have received

renewed attention by regional scientists and regional development policy-
makers, due in part to the increasingly apparent contradictions within
increasingly globalizing economic development. Globalization is a largely
homogenizing force, and so the diversity required for innovation must
come from residual uniqueness of local cultures and contexts (Sachs,
1992; Shiva, 2000; Thatchenkery, 2006). Endogenous vibrancy is both
threatened by and a requirement for the engines of globalization, which
are thus self-limiting (Sachs, 1992). Both as oases of possibility within a
globalizing world and as reservoirs of possibility for a post-globalizing
future, endogenous vibrancy is indispensable. Most importantly, it is a
fundamental value in its own right.
Focusing on human, cultural and organizational dimensions, this
chapter examines a case study in how an endogenous ecology spontane-
ously created cascading entrepreneurial activity (one of infinite possible
forms of vibrancy that could be studied, and a popular focus in economic
development studies). In particular, the chapter analyzes the endogenous
ecology of Silicon Valley in the US, and highlights some of the key cul-
tural, organizational and human factors contributing to the technological
innovation, creativity and entrepreneurial impact of the region. Many of
these factors can be situated within the framework of appreciative intel-
ligence to provide generative and transferable lessons. At the same time,
these lessons apply not at the level of content (replicating structures and
systems) but rather at the level of being or fundamental stance. At this
level, the message also challenges some deeply held theories and beliefs
about what leads to a thriving economy and a thriving society.
83

DIVERSITY AND ENDOGENY IN EXPLANATORY

PARADIGMS
Diversity and endogeny are important not just as the topical focus but also
at the level of underlying theoretical orientations and explanatory para-
digms. One of the recurring tensions in social science theory and research
comes from different views about the nature of social reality (Burrell and
Morgan, 1985; Gergen, 1994) between researchers in search of a unitary,
objective truth and those who treat social reality as constituted by mul-
tiple, often competing perspectives on reality, arising out of the subjective
interpretations and sense-making processes of a variety of actors and
coalitions. Some would even argue that the interpretation of reality that
usually seeks and gains dominance is the one actively promoted by those in
power and that the challenge of building an egalitarian social order would
be one of creating emancipatory spaces in which suppressed or muted
views are allowed room for expression. For example, Zinns (2003) classic,
A Peoples History of the United States, demonstrates the impossibility of
providing an objective account of US history and the importance instead
of listening to perspectives that are usually elided in dominant narratives.
Based on extensive archival data and interviews, Zinn argues for recogniz-
ing the salience of multiple histories, and he demonstrates empirically that
African American, Native American and European Caucasian settlers
perspectives on and experiences of US history differ dramatically from
each other in ways that cannot merely be assimilated one to the other. To
speak about US history as if it were a singular, objective representation of
facts would be simplistic, distorting and highly political.
These considerations are not limited to history but infuse the sciences as
well. In biology, a field that is analogous to economics in its combination
of historical and experimental methods with complex systems, Lewontin
(1994a, 1994b) has shown how the conceptual resources of the entire field
are shaped by organizing metaphors, which structure, frame and circum-
scribe the search for evidence while operating outside of and immune to
evidence. Lewontin demonstrates how these organizing metaphors are
also laden with values, ideology and politics.
At stake is not just representational correspondence but, much more
importantly, what kind of world we contribute to creating (Heineman-
Pieper, 2009). Ever since the publication of Kuhns (1962) The Structure
of Scientific Revolutions, a growing literature has demonstrated as unten-
able the earlier views of science as a neutral, value-free window onto the
world (Wimsatt, 1976; Galison, 1987; Gergen, 1994; Lewontin, 1994a,
1994b; Rosnow and Rosenthal, 1997). Biological, ecological, psychologi-
cal, social and cultural realities are open, dynamic and complex systems

Diversity and endogeny in regional development 85
(Wimsatt, 2007; Heineman-Pieper, 2009). Moreover, human beings

(including researchers) do not stand outside of these systems but rather
are participants in them. As a result, how we interact with these systems
makes a difference in what we come to know about them (Wimsatt,
1976; Rosnow and Rosenthal, 1997; Heineman-Pieper, 2009), especially
when our own human limitations and imperfections are further taken
into account (Wimsatt, 2007). In this context, it should not be surprising
that researchers and the results they produce are impacted by their inter-
ests, values, culture and disciplinary socialization, among other factors
(Haraway, 1990; Sachs, 1992; Rahnema and Bawtree, 1997; Gergen and
Thatchenkery, 2004). For example, Alvares (1992) has shown how at its
very core the concept of efficiency has been defined so as to focus on
factors that favor high-temperature industrial production over ambient-
temperature natural production while effacing considerations favoring
natural, ambient temperatures over industrial production, even though
from a long-term systems perspective this choice is leading to far greater
and potentially deadly inefficiencies (global warming, destruction of
natural ecosystems, dispossession of highly sustainable traditional people
and cultures by development projects, etc.; Rahnema and Bawtree, 1997;
Gowdy, 2000; Shiva, 2000; Farmer, 2004).
Much of the work in economics and regional science proceeds without
reflective attention to the underlying framework assumptions and values
that structure the research, while simultaneously reproducing and often
even enforcing adherence to the dominant frameworks. The dominant
framework assumptions and values that are usually reproduced occur both
at the level of the content and at the level of the rules of the game what
counts as doing good research in the field (Gergen and Thatchenkery,
2004; Wimsatt, 2007; Heineman-Pieper, 2009). For example, at the level
of the rules of the game, readers embedded in positivist methodol-
ogy have objected to our approach on the grounds that good scientific
research should either allow for verification (Hempel, 1965) or refutation/
falsification (Popper, 2002). These positivistic requirements on research
impose a particular set of values and assumptions that have cascading
effects (Heineman-Pieper, 2009) and ignore half a century of scholar-
ship across a variety of fields showing these strictures to be untenable,
misleading and distorting (Heineman, 1981; Gergen, 2009). For example,
Galison (1987) has shown how Hempelian verification and Popperian fal-
sification are unviable even in experimental particle physics (so how much
more in the social sciences!). Galison shows how, in the debate over the
existence of neutral current, the critical tests required by verificationist/
falsificationist models amounted to a subjective process of social con-
sensus building because the vital distinction between experimental fact

and experimental artifact can be decided neither in advance nor on the

basis of logic (ibid.). In biology, and on grounds equally applicable in
economics, Lewontin (1994a, 1994b) shows how, contrary to the claims
and requirements of verificationism and falsificationism, biologists are
never forced to abandon their theories but can always merely revise their
range of applicability and yet these same theories embed values of social
and political import. Kohn (1993) provides a parallel example of this in
psychology and economics regarding the inexhaustible faith in the value of
incentives, despite decades of evidence showing that incentives undermine
intrinsic motivation, reduce creativity, destroy relationships and have a
variety of other unintended consequences. Kohn shows how for 50 years
researchers continue to ignore or dismiss this and other evidence and,
determined not to admit to the detrimental effects of incentives per se,
continue to search for the right incentive schemes.
At the level of content, regional economics has many standard explana-
tory constructs, ranging from external economies of scale and agglomera-
tions (Marshall, 1890; Hoover, 1948, 1971; Blumenthal, 1955; Krugman,
1991) to industrial clusters (Porter, 1998), to regional competition (Isard,
1949; Begg, 1999; Camagni, 2002), to knowledge spillovers (Henderson,
2007), to high-technology networks (Audretsch and Stephan, 1996), to
market structure and firm size (Acs and Audretsch, 1987). This chapter
adds an important perspective to the literature by offering a novel way
of framing, understanding and supporting endogenous development
based on a principle that is incommensurable with a core framework
principle in economics and the existing literature on regional economic
development, namely: valuing people, communities and activities as
ends in themselves rather than as means to other ends (such as making
money). Rather than attempting to reconcile or homogenize these diver-
gent frameworks and explanatory strategies we will allow both to stand
on their own terms, like the multiple histories of the US (Zinn, 2003).
For the benefit of those who nonetheless insist on seeing their familiar
categories, we will briefly mention the following points. Alchians (1950)
account of adopted outcomes in regional economies relates only to suf-
ficiency and not optimality of outcomes and thus has no bearing on the
case studies that ensue. Other constructs operated in both Route 128 and
Silicon Valley (the comparison set for the case study) and thus are not suf-
ficient to explain the special vibrancy of Silicon Valley. Examples of these
constructs include: external economies of scale, knowledge spillovers and
academic/industry relationships. In the case of the more specific investiga-
tions of Acs and Audretsch (1987), the small-firm advantages they show
still do not explain why cooperative-culture firms outperformed standard
autocratic firms in the region (Baron and Hannan, 2002), and thus only

underline the contribution of the present chapter. The chapter does not
claim to provide a comprehensive account of all of the factors involved in
the vibrancy of Silicon Valley; however, it does claim to elucidate a very
important factor that is systematically overlooked by the intersecting
disciplines that constitute regional economics. The chapter further claims
that this factor has been overlooked for reasons related to systematic
biases in the values, framework assumptions and organizing metaphors
through which these disciplines structure, relate to and attempt to explain
reality.
CASE STUDY: ECOLOGY OF SILICON VALLEY

VERSUS ROUTE 128
High-tech regions, clusters and industrial districts have attracted serious

scholarly attention in the last two decades. Porters (1998) treatment of
industrial clusters highlights the importance of the local to the global:
enduring competitive advantages in a global economy lie increasingly in
local things knowledge, relationships, motivation that distant rivals
cannot match (p. 78). Even within high-technology clusters in the US,
different locales have had very different histories and trajectories. In this
vein, many scholars have contrasted Silicon Valley, California, and the
older technology corridor on the East Coast near MIT and Harvard in
Massachusetts, Route 128.
Numerous studies (Weiss and Delbucq, 1987; Saxenian, 1999, 2006;
Delbecq and Weiss, 2000; Florida, 2002, 2003; Lee et al., 2000; Shipley,
2006; Hulsink et al., 2007) have documented organizational, cultural
and regional differences between Silicon Valley and the long-standing
East Coast high-technology corridor, Route 128. Both of these regions,
Silicon Valley and Route 128, are characterized by a concentration of
high-technology firms and access to top tier universities (e.g., Stanford,
MIT and Harvard). However, Silicon Valley has bypassed the longer-
established Route 128 in number of viable firms created as well as in other
Western management measures of economic and innovative success. As
Saxenian (2006, pp. 278) writes:
The regions technology upstarts proved more adaptive than their older, verti-
cally integrated counterparts in an environment of fast-changing markets and
technological advances. In the 1960s and 1970s, Silicon Valleys semiconduc-
tor companies outperformed older East Coast corporations like RCA and
Sylvania, and in the 1980s the regions personal computer industry outper-
formed large, established companies including NCR, Honeywell, and Digital
Equipment Corporation.

Using data from interviews, questionnaires, lists of associations and

custom databases, researchers (Weiss and Delbecq, 1987; Saxenian, 1999,
2006; Delbecq and Weiss, 2000; Florida, 2002) have identified several
key factors as contributing to the high levels of IT entrepreneurialism
in Silicon Valley as compared with Route 128. Whereas Route 128 was
dominated by traditional, highly vertically integrated firms with hier-
archical management and control cultures seeking to hoard as much as
possible within the walls of their own firms (a zero sum game attitude),
in Silicon Valley there was a fluid flow of people and ideas among firms
and a relatively flat organization. Some of the early firms, such as Varian
Associates, were explicitly founded as an association of equals, and the
founders stated that We did not want to have the hierarchy of a company
owning facilities and employing employees ... [and ] ... We wanted to be
a cooperative. We wanted to create a cooperative organization (Lecuyer,
2007, p. 98). For example, employees switched firms regularly and without
penalty in Silicon Valley, whereas such movement was frowned on and
often contractually limited in Route 128 firms. Whereas Route 128 took
a judgmental attitude towards failures, in Silicon Valley, failures were
viewed as just part of a process of getting things right. Saxenian quotes
WebEx cofounder Min Zhu: The advantage of Silicon Valley is that you
can fail and learn and try again. Its the only place where you can screw up
once and try again (Saxenian, 2006, p. 31).
Researchers have also documented key differences in the cultures under-
lying Route 128 versus Silicon Valley. Whereas Route 128 is traditional,
stodgy, hierarchical and conformist, in Silicon Valley technological
fluency made relative equals of people of diverse ages, nationalities, back-
grounds, levels of social skill and various other ancillary characteristics
(Delbecq and Weiss, 2000). However, this egalitarianism is not absolute
but relative (to Route 128); limitations and exceptions existed within
Silicon Valley, such as the persistence of discriminatory limitations in
advancement for some groups (Saxenian, 2000).
In Silicon Valley, and especially in the early days, many people engaged
in innovation primarily as a form of self-expression and creativity, with
financial gain as merely a secondary benefit. This important and prevalent
quality can be seen at cultural, organizational and individual levels. For
example, at an organizational level, useful data can be found in Baron and
Hannans (2002) longitudinal study of nearly 200 Silicon Valley start-ups.
In addition to collecting a variety of other data, these authors conducted
interviews with founders and CEOs in order to excavate the assumptions
and mental models that guided their thinking about how to organize
employment relations and manage personnel (ibid., p. 10). The authors
found the notions about how work and employment should be organized

varied along three main dimensions ... each characterized by three or four
fairly distinct options or approaches (ibid.). The three main dimensions
are:
the basis of employee attachment to the company;

the criteria for coordinating and regulating employees;
criteria for selecting employees.
In the first of these, employee attachment, the authors noted three

bases envisioned by founders, which they called love, work and money
(ibid.; italics original). Work corresponds to founders recognition that
the primary motivator for their employees is the desire to work at the
technological frontier (ibid.). Love corresponds to creating a strong
family-like feeling and an intense emotional bond with the workforce
that would inspire superior effort and increase retention (ibid.). This is
a form of social motivation that could be seen as providing a supportive
relational context in which the employees could express themselves freely.
It is a culture that values people over money and creates an environment
of trust and broader human valuing that can encourage more courageous
self-expression. The final form of attachment is money an exchange
relationship of labor for money.
After coding the data into all possible combinations of subtypes across
the three main factors, the authors found that firms clustered into five
main types or organizational blueprints. Only one of these involved money
as a form of attachment, and this blueprint formed only 6 percent of the
distribution of organizational types (ibid., p. 13). Moreover, even when the
authors controlled for a variety of variables, the money-based blueprint
was the most likely to fail of all the blueprints, including a catch-all non-
type category blueprint. In addition, although the underlying statistics
are not sufficiently presented to clarify the precise meaning of this result,
the magnitude of the percentage difference in failure likelihood that the
authors report was greater between this worst-performing money-based
blueprint and the next worst-performing blueprint than existed between
the second-worst performing blueprint and the best-performing blueprint.
While it is true that the focus on money to attract and retain employees
was only one of three elements within this poorly performing blueprint,
the other two qualities in the other two dimensions selecting employees
based on specific skills and a hierarchical form of control are not inde-
pendent of the money-based mindset (as suggested by the fact that not all
of the possible cells represented by the intersection of the three dimensions
were occupied) and are part of the same world-view and gestalt.
The commitment to meaningful activity for its own sake, as a form of

self-expression, rather than to make money can also be seen at a cultural

level. Florida (2002) notes some of Silicon Valleys roots in 1960s counter-
culture. Unlike organized labor, which fought for powers and rights
within the framework of the existing economic system (p. 203; italics origi-
nal), the counter-culture in northern California included a wide spectrum
of views on work and economics, including operating outside of it:
Some in the hippie milieu preferred simply to ignore the world of work, perhaps
living by their wits or the generosity of friends or parents. Some sought to rob
the system, as described how-to-do-it style in Abbie Hoffmans Steal this
Book. For many the strategy was to grudgingly coexist with the system. Get a
job, even a haircut if you must; earn the money you need to do what you have
to do, but no more. (Ibid., p. 203)
Freed from the bonds of greedy materialism, the innovators of Silicon

Valley were free to focus on following what they most enjoyed, innovating
for the love of it, and making things work. Economic and capitalist ideol-
ogy pretends that people can be driven primarily by a profit motive and
then claims that this assertion represents either the base reality of human
nature, a useful lubricant for the market mechanism, or else a neutral
fact. The success of Silicon Valley offers a counterweight to that view and
shows that it matters why we do what we do and that theres no substitute
for doing things for their own sakes and treating people as ends in them-
selves rather than as means (e.g., rather than as in the labor for money
model in Baron and Hannan, 2002).
At an individual level, the importance and prevalence in Silicon Valley
of people motivated not by money but by love of what they do work as
play or hobby can be seen both in famous founding stories (e.g., Apple)
as well as in stories from the more everyday. For example, Google
Desktop came to be because an employee had decided to create that func-
tion for himself on his own computer in his spare time. Simultaneously,
at an organizational level, Google had structured the work week to give
employees a free day every week to pursue their own personal interests
and did not demand that anything productive or useful come of it in
recognition of the fact that employees were intrinsically motivated to
create and would do this best when they were given the freedom to follow
what they enjoyed, unconstrained by top-down agendas (Elgin, 2005).
Delbecq and Weiss (2000) write that people who would be considered
misfits in regions (such as Route 128) that are concerned with superficial
impressions such as those based on elegance of expression (versus un-
fluent English as a second language), elegance of social skills, or elegance
of physical appearance (note that many psychology studies have shown
the biasing effects in business and other environments of what mainstream

society identifies as canonical good looks) are given prime projects to

lead and prime resources to command in Silicon Valley. According to
Delbecq and Weiss (ibid.), Silicon Valley is much less concerned with
superficial impressions than are other parts of the US (such as Route 128):
In Silicon Valley, it is the quality of ideas, the willingness to creatively problem

solve, the ability to arrive at a breakthrough technical solution, which is valued.
I have heard senior managers listen with rapt attention in briefings to young
talent whose modality of presentation is rough. ... As one venture capitalist
expressed it, I dont care how the talent is wrapped (ibid., p. 40).
Also, these authors report that in Silicon Valley there is an ethos that gains
are shared not based on hierarchical position, as in Route 128 firms, but
instead based on contribution.
Fleming and Marx (2006) present the concept of small worlds, social
networks that cut across individual firm boundaries, in explaining how
the effective management of innovation in an environment of ongoing
knowledge exchanges was a critical factor underlying the success of Silicon
Valley. Small worlds, families of interconnected firms, are more likely to
be the source of innovation than individual firms (see also OBrien, 2007).
Saxenian (1999 and 2006) shows how this small worlds phenomenon
was also made especially effective by the many highly educated immi-
grants who came to Silicon Valley. Saxenian (1999, 2000, 2006) shows
how these families of interconnected firms went global as a result of the
social networks of immigrants/expatriates largely from Asian countries
(also reported in Adams, 2005). Saxenian documents the social networks
of Indian, Chinese, Taiwanese and other immigrant groups and how these
associations, interactions and knowledge-sharing not only substantially
advanced the development within Silicon Valley, but also created global
technological and business networks beyond what would otherwise have
been possible. In both of these contexts, domestic and international, the
immigrant communities significantly augmented the overall size of the
pie. While Route 128 firms were busy protecting their turf, Silicon Valley
firms and individuals, inspired by innovation itself and coming from a
mindset of abundance rather than scarcity, shared freely and thereby
created a better result at each level from the individual through to the
whole.
Across diverse metrics, Saxenian (2000) shows that immigrants, often
from Asian countries such as China and India, were a vital part of the
Silicon Valley technological and entrepreneurial success. For example,
approximately one-quarter of Silicon Valley technology firms founded
between 1980 and 1998 had Chinese or Indian executives (p. 252), and this
number is likely to be an underestimate because racism powered by and

instituted through restrictions on venture capital required many non-white

immigrant founders to choose white American CEOs (Saxenian, 2000).
The (likely underestimated) immigrant-founded firms created signifi-
cant wealth and employment in Silicon Valley, and in 1998 collectively
accounted for more than $168 billion in sales and 58 282 jobs ... [and the]
. . . rate of immigrant entrepreneurship in Silicon Valley has increased
significantly over time (ibid., p. 253).
The experience of immigrants in Silicon Valley points to a serious short-
coming of the culture and place. While it may have been more open than
Route 128, Silicon Valley was still a place of advancement limitations and
discrimination against the very immigrants who were in large part fueling
its success (ibid., p. 251). Saxenian cites a 1991 survey of Asian profes-
sionals in Silicon Valley that showed that two-thirds of those working in
the private sector believed that advancement to managerial positions was
limited by race, and this is further supported by the fact that Chinese and
Indian engineers remain concentrated in professional rather than mana-
gerial positions, despite superior educational attainment [to the white
Americans occupying equal or higher positions] (ibid.). Interestingly,
those surveyed attributed these limitations less to racial prejudice and
stereotypes than to the perception of an old boys network that excludes
Asians ... [and the] ... lack of role models (ibid., p. 25).
At the same time, overt discrimination also existed, as is demonstrated
by the requirement of many venture capitalists that non-white immigrant
founders choose a white American CEO instead of running the company
themselves. For example, Saxenian reports the story of David Lee, who
left Xerox in 1973 to start Qume after a less experienced outsider was
hired as his boss. Lee was able to raise start-up capital from the main-
stream venture capital community, but only on the condition that he hire
a non-Asian president for his company (ibid., p. 251).
Despite (or perhaps partly because of) these barriers, immigrants in
Silicon Valley created local ecologies within the larger Silicon Valley
ecology to support themselves and their compatriots. Immigrant groups
and subgroups founded associations that provided both social net-
works and professional contacts, as well as connecting people with role
models and sources of tacit knowledge and experience. For example,
David Lee, the founder of Qume described earlier, and a handful of others
like him, became community leaders and role models for subsequent
generations of Chinese entrepreneurs (ibid., p. 252). Putnam (2000) has
shown the importance of these forms of civic engagement in building
social capital and community (see also Beugelsdijk, 2003).
While the immigrant experience shows how the social networks were
racially divided, indicating that Silicon Valley was not quite as free

of racism as some accounts suggest, all of these accounts point to the

importance of social networks as a factor in Silicon Valley entrepreneur-
ship. Audia and Rider (2005) show how the garage legend creates a
false impression of entrepreneurs as individual founders, whereas these
authors document the importance of organizational ties even in the case
of canonical founder stories such as Apple and Hewlett-Packard. Ferrary
(2003) explicates one of the many modalities in which social networks are
important. He outlines three mechanisms of exchange that underlie the
circulation of economic goods:
the arms length transaction;

the power relationship;
the gift exchange.
Those correspond to the modalities of the market, the organization and

the network, respectively.
The social embeddedness of the actors in Silicon Valley (arising from
having worked in the same firm, graduating from the same university or
sharing a common ethnicity) resulted in a majority of the exchanges being
informal exchanges based on reciprocity and a strong sense of community
and modeled on the metaphor of the gift exchange. By this argument,
Ferrary (ibid.) would suggest that the social architecture that normally
accompanies urbanization forces the actors unwittingly toward more
legalistic and formal forms of exchange and may explain the difficulties in
mechanistically replicating Silicon Valleys success if the underlying culture
of the system is not oriented toward a sense of community (gemeinschaft-
fuhl), a feature that is much harder to copy through rational intention.
Another dimension of distinctiveness for the region is the way in which
the different strains of the underlying culture and counter-culture them-
selves in turn created a variety of organizational cultures. As we have
seen, Baron and Hannan (2002) contrasted five different blueprints of
human resource practices among Silicon Valley companies. These are:
the Star (marked by the attitude that We recruit only top talent, pay
them top wages, and give them the resources and autonomy they need
to do their job), the Commitment (represented by I wanted to build
the kind of company where people would only leave when they retire),
the Bureaucracy (exemplified by We make sure things are documented,
have job descriptions for people, project descriptions, and pretty rigor-
ous project management technique), the Engineering (captured by We
were very committed. It was a skunk-works mentality and the binding
energy was very high) and the Autocracy (implied by You work, you get
paid.). They found that firms whose founders practiced the Commitment

blueprint have performed better over the ensuing years compared with
ventures based on the other blueprints. The Commitment blueprint high-
lights a reliance on emotional connections and family-type relationships;
Baron and Hannan (ibid.) argue that perhaps given the rarity of such
signals in the corporate world, the few companies that send out these
signals end up enjoying a competitive advantage that is hard to imitate,
particularly because it runs counter to conventional corporate ideology
in ways that are humanistic and thus that people readily find valuable.
It should perhaps not be surprising that the most successful variant, the
Commitment blueprint, is one in which people are treated and valued as
people rather than instrumentally.
In short, researchers have identified a number of unique cultural fea-
tures of the Silicon Valley region that have contributed to its ability to
bypass the more traditional high-tech corridor, Route 128. These features
include but are not limited to:
vibrant social networks;

an underlying counter-culture;
intense diversity of national background;
relative (to Route 128 though still less than it might be) valuing of
persons as persons rather than as objects in the production process
and sources of income and other measurable results for distant
owners;
willingness to fail without passing judgment, to learn, and to try
again;
relatively flat organizational structures versus hierarchies.
These features will next be considered from the framework of appreciative

intelligence (Thatchenkery and Metzker, 2006).
APPRECIATIVE INTELLIGENCE
Howard Gardner (1993) famously demonstrated that, contrary to notions

generally believed at the time in the US, intelligence is not unitary. In the
US, there has been a long (and unfortunate) tradition of viewing intelli-
gence as springing from an essential, innate and fixed source in individu-
als, and measurable through standardized IQ tests. The resulting measure,
the G-factor, was presumed to reflect the quantity of intelligence in any
individual. The IQ assumptions and methodology have all been roundly
critiqued by a host of scholars (see, e.g., Block and Dworkin, 1976), who
have shown that:

the assumption of an innate fixed quantity of intelligence is entirely

unfounded;
intelligence is highly situated for all of us, such that each of us may
show high levels of a particular kind of intelligence in certain kinds
of situations or contexts and at certain times and not in others;
intelligence itself has myriad forms (as demonstrated by Gardner,
1993).
Thus, in this part of the chapter we consider a quality that has been
ignored in the literature, namely appreciative intelligence, or the ability
to focus not on problems and shortcomings but on the beauty and poten-
tial inherent in any moment, including moments of suffering (e.g., to see
the mighty oak in the acorn). This is the type of intelligence that our
best school teachers have shown when they saw and brought out more
in us than we may even have known at the time that we had. In contrast
to intelligence focused on problem-solving which inevitably thereby
everywhere sees problems and solutions appreciative intelligence is a
way of being in the world that always sees the constructive side of persons
and situations without being utopian or unrealistic. It is about seeing and
taking a stand for our own and each others best selves. When we take this
stand completely and holistically, rather than partially and with limits, it
gains enormously in power and authenticity. In other words, if we see an
opportunity to make money by exploiting others, the narrowness of our
appreciative focus, and especially the lack of appreciative intelligence
towards other persons, constitutes a blind spot that will in some way feed
back around to impact our own well-being.
One of the most powerful aspects of the appreciative intelligence shown
in Silicon Valley is the very fact that it was focused on creating possibilities
that were positive intrinsically and in multiple dimensions in how people
organized work together, in the authenticity of their intrinsic enjoyment in
creating helpful and useful technologies and in their love of what they did
rather than being focused on competing with others and winning. We have
already seen several examples of this, such as the case of Google seeking
ideas from anywhere in the organization (rather than just the top) and
providing its engineers with a day a week to work on whatever personal
pet projects they wish. This practice demonstrates both how the employ-
ees were motivated by the love of discovery and problem-solving, such
that the day a week to work on their pet projects was a perk of working
at Google, and the organizational wisdom of Google that people usually
do their best work when they are doing what they most enjoy. In the end,
the organization will be enhanced from allowing people to be maximally
self-expressive, whether or not that self-expression ever yields financial or

other measurable results. In allowing employees total freedom to choose

to engage in what they enjoy on those days and to pursue it for their
own personal interest and benefit, Google recognized and refused to be
caught in the irony that as soon as peoples work is held to the standard
of measurable results, the freedom to pursue and experiment with what
they truly and intrinsically enjoy has been destroyed along with the pro-
found breakthroughs (such as Google Desktop in the example previously
described) that it may or may not spawn.
Components and Qualities of Appreciative Intelligence
Appreciative intelligence has three components, and leads to four qualities

(Thatchenkery and Metzker, 2006). The three components or elements
are:
reframing;
appreciating the positive;
seeing how the future unfolds from the present.
The four qualities that result from appreciative intelligence are:
persistence;
conviction that ones actions matter;
tolerance for uncertainty;
irrepressible resilience.
The three components are discussed in turn.
Reframing
This is the psychological process whereby a person intentionally views or
puts into a certain perspective any object, person, context or scenario. One
of the most common examples of framing is that of calling a glass half
empty or half full (Thatchenkery and Metzker, 2006, p. 6).
We have tremendous choice in how we see situations, ourselves and
others: what we pay attention to; what we take to be important; whether
we see things as a liability or an opportunity for learning and growth, and
so on. Often life appears to us as a given reality, and older traditions
in Western social science reinforced that way of seeing (contemporary
social sciences and philosophy of science have seriously revised this view,
however). In fact, we can always choose how we make meanings out of
situations and how we respond, and people exercising their appreciative
intelligence do just that. Viktor Frankl (1963, p. 104), a concentration

camp survivor, famously wrote: everything can be taken from a man but
one thing: the last of the human freedoms to choose ones attitude in any
given set of circumstances, to choose ones own way.
Appreciating
This second element of appreciative intelligence builds on the first and is
the most constructive aspects of a situation in a realistic fashion. We can
see the power of this element most forcefully when we think about being
on the receiving end of it. If you go well out of your way to do something
nice for someone, and they focus only on flaws in the content of how you
selected to express this goal, you can see that they are missing the most
powerful and generative aspect of the situation your good intentions. If,
instead of focusing on the problems in the execution, they focused on the
beauty of your intentions and from there, in a space of alliance, requested
a different mode of expressing the intention, the experience and result of
the interaction, as well as of downstream results and interactions, would
be very different indeed.
In Silicon Valley, this positive valuing was largely present along some
dimensions and in some ways, and not in others. For example, relative to
Route 128, Silicon Valley often embraced people regardless of the superfi-
cial polish and sheen important in most traditional business environments
the pattern quoted earlier of not caring how the talent is wrapped. This
valuing sometimes also may have extended to seeking to value employ-
ees as persons rather than as means to ends, as is suggested somewhat
by the prevalent commitment blueprint of organizations described by
Baron and Hannan (2002). At the same time, this valuing was incomplete
in that immigrants faced discriminatory limitations in advancement in
many firms and from many venture capitalists. Moreover, the valuing was
somewhat limited in that, while technical insight may have functioned to
create greater equality across people who might have been discriminated
against based on other superficial dimensions, this valuing did not extend
universally to persons as persons such that people of all sorts and in all
professions (janitors, construction workers) were accorded the same levels
of social appreciation, support and opening of possibilities, along the lines
Paolo Freire sought to establish in his emancipatory educational institu-
tions (Freire, 1996). While much more common than in other parts of the
country, this valuing wasnt universal, as the example of PayPal shows,
where employment was based on being like the founders along many
superficial dimensions including aversion to sports (OBrien, 2007). At the
same time, these exceptions existed in a context in which organizations
themselves had diverse cultures and thus, across organizations, the result
would have been a valuing of a distribution of characteristics. While less

than holistic, this organization-level distribution still provides greater

opportunities for valuing more diverse people (or aspects of people) than
other parts of the country were doing at the time. In other words, while far
from perfect, Silicon Valley offered greater opportunity and valuing than
the other members of its reference class.
Seeing how the future unfolds from the present

This is the third element of appreciative intelligence. Since in fact, as the
second element has implied, useful, desirable or positive aspects already
exist in the current condition of people, situations or things (Thatchenkery
and Metzker, 2006, p. 33), the final element involves attunement to how to
nurture and amplify these constructive elements. In short, when people are
showing high levels of appreciative intelligence, they:
see the oak in the acorn. They also go beyond they plan[t] their acorns and
persevere to help them grow. While others may doubt the potential of the
acorns, these leaders believe in their own and others abilities to water and ferti-
lize the plants from sapling to tall oak. They deal with the risk and uncertainty
that comes with planting something new and hoping for growth. Finally, they
find a way for the oaks to survive and thrive despite unpredictable circum-
stances or a challenging environment. (Ibid., p. 33)
Relevance to Silicon Valley
How does all of this relate to the Silicon Valley experience? Examples
abound along many of the dimensions (reframing, appreciating the
positive, seeing the future unfold in the present) and qualities (resilience,
persistence, conviction that ones actions matter, and comfort with ambi-
guity) of appreciative intelligence. For example, we have just encountered
elements of a general Silicon Valley culture of allowing everyone to be
who they are, regardless of how the more traditional firms might view
them as misfits, socially inept, counter-establishment, or otherwise failing
to have the outer demeanor and sheen that traditional business culture
unfortunately trains people to value. In the words of the venture capitalist
quoted earlier from Delbecq and Weiss (2000), I dont care how the talent
is wrapped. There are, as we have seen, limitations and exceptions to this,
such as the fact that at the organizational level, PayPal was fairly intoler-
ant of various dimensions of difference internally, though those dimen-
sions were themselves offbeat as an early employee described it, Were
all a little weird (OBrien, 2007, p. 106).
Likewise, consider the application of appreciative intelligence to tech-
nological problems by the young techies. This was a culture of making
things work with whatever materials might be available in the garage

and through social networks (Saxenian, 2000; Audia and Rider, 2005).
It also shows a willingness to fail and to persevere in the face of failures.
YouTube and Yelp founders learned a valuable lesson from PayPal:
The first idea isnt always the best. Yelp was a convoluted e-mail referral
service before becoming a top review site. YouTube started as a video
dating play (OBrien, 2007, p. 106).
Googles Marissa Ann Mayer says that Googles innovation relies on
its fearlessness she launches products early and often without fear
of failure (Elgin, 2005). Referring to Apple Computer and Madonna,
Mayer says Nobody remembers the Sex Book or the Newton. Consumers
remember your average over time. That philosophy frees you from fear
(ibid., p. 90).
A founder of PayPal had so much confidence in his own actions that he
founded a financial transactions company even though when he started,
Peter Thiel didnt know what a chargeback was.... Thats one of the fun-
damental things of any credit card payment system. Chargebacks almost
killed the company (OBrien, 2007, p. 106). At PayPal, the executive team
made up for nonmastery of details with unwavering vision, which inspired
the troops (ibid., p. 106).
Moreover, a can-do approach of getting in and trying things and learn-
ing on the fly inspired other employees to found their own businesses. For
example, Chad Hurley, CEO of YouTube remembers his PayPal days as
an education in business. When he arrived in California with a degree
in art from Indiana University of Pennsylvania, building a successful
company seemed like something other people did. You never think it
could happen to you, Hurley says: But seeing Peter and Max and the
guys come up with ideas and seeing how to make things work gave me a
lot of insight. You may not have a business degree, but you see how to put
the process into effect (ibid., p.106).
We also should not underestimate the fact that appreciative intel-
ligence was frequently applied not individualistically to advance ones
own personal gain at the expense of others, but rather collectively,
holistically, and in a more communitarian way to appreciating, sup-
porting and valuing the success of others. Rather than a scarcity-based
zero-sum-game mindset, Silicon Valley was frequently characterized by a
mindset of sharing, abundance and mutual encouragement and support.
Founders of one firm would support former employees in founding their
own firms with capital, knowledge-sharing and other resources. This
happened within myriad social networks, from the immigrant asso-
ciations documented by Saxenian (2000), to the organizational support
in which employees would be supported by their current and former
employers to start companies (Audia and Rider, 2005; OBrien, 2007).

PayPal founders claim to have supported employees to such an extent

that the PayPal mafia now runs $30 billion in businesses (OBrien, 2007).
This even occurred sometimes in cases of separations under contentious
and acrimonious circumstances, as in the case of Musk at PayPal, who
continued to provide financial support to projects by people from PayPal
even after hed been removed by the board while on a flight to Australia
for his first vacation in years (ibid.).
TRANSFERABLE LESSONS
The history of management thought is full of examples in which a mete-

oric success is identified and everyone rushes to imitate often even as
the original inspiration for the success story is itself being indicted for
accounting fraud (e.g., Enron) or going extinct (e.g., Wang Labs). What
makes us think that Silicon Valley will not burn itself out and end up
worse off than if it had developed more quietly, moderately and sustaina-
bly? When we think about transferable lessons we must first make sure we
really want what we are seeking to transfer. A second important irony is
that Silicon Valley was not the result of deliberate planning. How then can
we expect to create by deliberate planning something that itself emerged
organically through a confluence of complex and historically contingent
factors? What, if any, lessons can be learned from Silicon Valley?
Metcalfe (1998, p. 123), 3Coms founder, pointed out that Silicon Valley
is the only place on Earth not trying to figure out how to become Silicon
Valley. There we argue that the most important, generative and trans-
ferable lessons from Silicon Valley are not at the level of replicating the
content or structure of so-called success factors. Instead, the most power-
ful lessons are to be found at the level of a fundamental stance. In par-
ticular, what worked in Silicon Valley was authentically and thoroughly
valuing, supporting and inhabiting its own positive uniqueness not trying
to be Route 128, not trying to be Detroit and not trying to connect with
its best qualities in order to make money. The example of Silicon Valley
forces us to recognize the deepest level of appreciative intelligence: seeing
and celebrating the most life-affirming aspects of each community, rather
than trying to import a set of best practices from one place to another.
When seen to operate at the level of a fundamental stance, the lessons
from Silicon Valley also challenge some deeply held theories and beliefs
about what leads to a thriving economy and a thriving society. As we
have seen, the culture in Silicon Valley combined intensive and vibrant
transnational diversity with a counter-culture that valued technological
work and innovation intrinsically rather than instrumentally in terms of

whether or how they would translate into economic gains (Saxenian, 1999,
2000, 2006; Florida, 2002). Of course, the opportunity to make money
was not squandered, and it may even have been primary for some firms,
but this was relatively rare in the organizational and underlying cultures,
and when it did occur in organizations they failed in disproportionately
high numbers (Baron and Hannan, 2002). In the core incubating culture
spawning the phenomenon of Silicon Valley, money was a secondary
rather than a primary driver for large segments of the population.
Baron and Hannan (2002) showed that the Commitment blueprint
treating employees as persons rather than instrumentally as means was
the most successful within Silicon Valley. This model had the least likeli-
hood of failure, while the autocratic model, with a central exchange rela-
tionship of working for money, had the greatest likelihood of failure. This
blueprint, along with some of the others focused on work for the joy of it,
can be seen as an extension to the organizational level of a counter-culture
that refuses to value money as the end and persons as the means, but
instead values relationships and authentic activities as valuable in their own
terms rather than instrumentally for their monetary potential. Again, this
highlights the fallacy of trying to replicate Silicon Valley in order to create
economic wealth: paradoxically, part of what made Silicon Valley eco-
nomically successful may have been the very fact that many of the original
innovators were not motivated because of money but instead as a result of
intrinsic enjoyment of the technology and the company created around it.
The Silicon Valley experience also challenges other framework assump-
tions of Western management. For example, traditional management
prizes asymptotic levels of efficiency, and yet one of the most important
sources of success in Silicon Valley was the willingness to fail and the
acceptance of failure as part of the process. From a traditional man-
agement perspective, failure is waste, and something that should pro-
gressively be eliminated. Not only has the cult of efficiency in Western
management been eroding personal lives and the ability of employees to
bring their whole selves to bear in the workplace, but also it is, like global
capitals destruction of the regional and local, self-defeating.
Traditional management theories also collude with the hierarchical
structure of salaries based on power and control within the company.
Management ideologies usually rationalize these modes of apportioning
wealth with patently false (and irrelevant: Kohn, 1993) claims that pay is
related to actual contribution to the organization. These claims have now
been dramatically exposed by the US financial meltdown, where CEOs
who ruined their companies continued to reap handsome bonus pay imme-
diately before, during and even after their firms received millions of dollars
in bailout funds from taxpayers or were bought as a distressed sale (as in

Bank of America buying Merrill Lynch as the latter verged on collapse).

Many of the early Silicon Valley firms at least partially and relatively
contravened these hierarchical and extorting structures and mindsets by
adopting a relatively flat organizational structure and apportionment
of gains. Again, however, this structure was rooted in the ecology of the
region. When other regions adopt a stance of appreciative intelligence
towards their own organizational structures, they will not merely import
either the hierarchical or the flat models, but will consider (authentically,
and not just based on the narrowly conceived self-interest of the powerful)
the kinds of organizational structure that can best support their highest
potential. At the same time, in most cases, when the interests of the power-
ful are not in the drivers seat, the most effective structure is indeed most
likely to be a relatively flat one in which everyone has the autonomy,
power and, thereby, ability to be their best and contribute fully to the
organization.
In identifying the transferable lessons from Silicon Valley, this chapter
raises a fundamental paradox: why are we seeking imitations in the first
place and what does that say about what we are choosing to value and
the relationship between what we are valuing and what those were imitat-
ing were valuing? Imitations of Silicon Valley abound because economic
growth has become the new global obsession, to the point that whole
countries are losing sight of invaluable treasures of social fabric, spiritual
depth and resilience, emotional intelligence in their own backyards. And
yet Silicon Valley achieved its economic growth in part precisely because
it was not focused on economic growth. It was focused on the joy of
innovating for the sake of innovating.
Thus, we have seen many examples of transferable lessons from Silicon
Valley, including the caution that the act of seeking transferable lessons
may already violate one of the key conditions that was important for
Silicon Valley to be Silicon Valley (e.g., that it wasnt trying to be Silicon
Valley). And yet, it is only human to find things elsewhere that we want
to incorporate into our lives and communities. Under these circumstances,
what are some things we should keep in mind?
When we analyze a phenomenon for transferable lessons, we create
a figure-ground distinction, to use the famous gestalt terminology (e.g.,
Rubin, 2000), in which we focus on the salient components of the model
that got our attention in the first place (the figure), and we tend to de-
emphasize or overlook the interdependencies of those aspects with the
other aspects of reality that have become part of the ground. Instead, we
need to pay attention to the deep ecology impacting the functioning of the
whole and parts, both in the original context as well as in the context into
which we seek to import. It is also important to consider other collateral,

unintended consequences an import may have on interrelated aspects of

the milieu into which we seek to place it. When we look outside and see
something we value, how or when might we or might we not want to try
to adopt it?
We will illustrate the question of transferrability in Silicon Valley with
an analogue that shows kinds of interdependencies and deep ecology that
should be considered. Twenty-five years ago Karl Weick (1983) intro-
duced the notion of reverse simulation into management literature. In
a typical management or organizational simulation the goal is to test the
relationships between variables in the controlled condition of a laboratory
or classroom, extract a few principles or theories and propose various
ways these lessons learned may be applied to real world organizations.
Following fields such as engineering, Weick suggested the opposite:
instead of trying to simulate reality, create new realities in the lab and
transfer them to the real world. What happens when we create in a labora-
tory (simulated environment) new forms of organizational processes and
structures that differ from what are available in our lives outside, and we
would like to incorporate some of those possibilities in our real lives? In
such situations, we often look too narrowly at the conditions that made
the possibility come alive in the particular context in which it occurred,
and we often similarly overlook the conditions in the new context that
might alter the meaning or valence of the import in the new environment
(Wimsatt, 2007). We should attend not just to the phenomenon and its
structure, but also to the nature of the broader relationships and being
of the totality of the context that allows a phenomenon to occur as it did
(Wimsatt, 1980, 1994, 2007; Thatchenkery et al., 1999). We should also
keep in mind a healthy humility about our ability to see and understand
these complexities or to anticipate or resolve unintended consequences.
Often our solution to our last problem ends up creating many more and
worse problems.
Consider a deliberately simplified example: after a few days of specially
structured group interactions, participants in a group dynamics labora-
tory learned several key principles of open communication, giving and
receiving feedback and influencing others, and so on. Participants learned
in the laboratory that when they spoke with authenticity and from their
hearts, others listened to them better than when they had related superfi-
cially. However, when they tried to apply the learning at work after a few
days it did not work, leading them to blame the too theoretical academic
knowledge as the root cause. Telling the boss that I feel intimidated
when you speak to me like that did not yield the type of open response
the participant had received in the group dynamics lab a few days earlier.
How can we apply the lessons learned about open communication to

the real world? If we are to apply the principle of reverse simulation with
an understanding of the biases of figure and ground phenomena, we will
focus more on the conditions that facilitated the emergence of the learning
than on the content of the learning as such (Thatchenkery et al., 1999).
For example, a certain level of trust had to develop in the small group lab
before a member was willing to give or receive feedback. Therefore, what
should really transfer to the outside world is not a technique of giving
feedback, but the conditions that allow trust to develop in a group. If suf-
ficient trust had been developed between the boss and the employee, the
feedback cited earlier might have been received more positively.
Behara et al. (2008) have applied Weicks notion of reverse simulation in
an organizational context. While designing and implementing two distinct
product delivery processes for a large financial services company in the
Organizational Learning Laboratory at George Mason University, which
one of the authors had founded, a new methodology called Empathic
Knowledge Management was developed. It used the appreciative inquiry
framework (Cooperrider and Srivastva, 1987) and utilized tacit knowledge
of participants for process redesign in a learning laboratory environment.
The focus in the lab was to create conditions that would allow staff from
dispersed units in the organization to come together and share knowledge
so that the process redesign time could be cut by half. Once a prototype
loan processing tool was designed in record time in the Organizational
Learning Lab, the focus was not on applying the new tool in the rest
of the organization, but on figuring out how to create and maintain in
their financial services organization the unique learning climate that had
flourished in the laboratory during their three weeks of stay. In any such
endeavor, it is important to keep in mind not just the similarities but also
the differences between the original and transferred context at every level
from content to process to the being of the facilitators, participants and
organizational actors and cultures. It is also important to remember that,
since human and social systems are far more interdependent and complex
than we will be able to theorize in our (necessarily) simplified views and
models of them, whatever we think we know is but a small and fallible
veneer on what we dont know.
When other regions see value in what Silicon Valley has done, they
need to look not just at the superficial level of success factors or salient
characteristics, but also at:
the deeper context of Silicon Valley that allowed those characteristics

to blossom;
the deeper contexts in their own region that are and are not resonant
with what they are valuing from Silicon Valley;

what additional collateral, cascading effects might result from

importing and valuing the Silicon Valley model within their own
community and how that fits or doesnt fit with the endogenous
values and vibrancy of their own community.
Perhaps a region that has not become highly commercialized can recog-
nize, celebrate and support the contexts and cultures in which people can
continue to express themselves and be motivated intrinsically as opposed
to extrinsically (e.g., in which people engage in activities fundamentally
as ends in themselves vs. instrumentally as means to other ends such as
making money; Kohn, 1993). Perhaps a citys residents can reframe and
examine their unique strengths and regional advantages to come up with
new services and innovation that will further support the community?
What does it look like to reframe the fear of failure into embracing failures
as learning opportunities?
Appreciative intelligence and the embeddedness of human social and
cultural realities suggest that the real message and transferable lessons
from Silicon Valley are the following:
1. Each locale should appreciate, value and embrace its own unique
character and beauty, and be fully, thoroughly and elegantly itself
rather than trying to imitate models plucked from a distant and
foreign ecology.
2. That when communities see aspects of life they appreciate in other
places, they should be circumspect and reflective in why and how
they may adopt these elements and what cascading unintended effects
adoption might have.
3. Value is not unitary but has many and diverse forms, and thus
economic riches are not necessarily even the ideal or goal for a
community to seek.
While the venture capitalist John Doerr may announce to the world
that Silicon Valley represents the largest legal creation of wealth in the
history of the planet, we must not forget that it is only one form of wealth
that he is noticing. If we extend our conception of wealth, making it more
diversified and pluralistic to include other forms of wealth such as spiritual
and relational wealth, there are a number of Silicon Valleys in the world
that we are not celebrating. It is only in our modern Western discourse
that the creation of material and technical wealth has become an obses-
sion held out for all humanity to replicate, and that the pursuit of its
accumulation and concomitant concentration in certain regions appears
to be a goal worth replicating (Sachs, 1992; Escobar, 1995). Moreover, a

focus on materialist wealth accumulation may in fact even be incompat-

ible with, and at the expense of, authentic relational and spiritual wealth,
since it usually involves treating ourselves and each other instrumentally
as means, rather than as ends.
Amid the scholarship extolling the success of Silicon Valley, a few dis-
senting voices also can be heard. Rogers and Larsen (1984) in their book
Silicon Valley Fever have raised fundamental questions about the social
and business conditions of that environment, lamenting the veneration of
the workaholic high-tech achiever lifestyle, the stinginess of the region in
supporting the arts, the dark side of life in the less affluent South County
and a variety of other downsides that are often overlooked or taken to be
unimportant when we extol a model for its successful outcomes. Florida
and Kenney (1990) also caution us against embracing Silicon Valley as
an unmixed blessing, reminding us that neither Silicon Valley nor Route
128 have been able to rectify the fissures that have become increasingly
apparent in the US economy and social structures, and that our infatu-
ation with Silicon Valley stems from the fact that their image of free-
wheeling, high-technology entrepreneurship and quick-shooting venture
capital fits in nicely with the free enterprise ideology. If there ever was
an Eden of Cooperation these authors suggest it has degenerated into a
fiery pit of destructive cut-throat competition when confronted with huge
potential sums unleashing equal levels of greed. When other regions look
to Silicon Valley for inspiration to promote economic and technological
breakthroughs, policy-makers must look not merely at the success but
also at the degree to which the spread of a highly mobile venture capital
mindset might erode age-old traditions and cultural norms that have
constituted the wealth of these societies. By teaching us to seek happiness
in an ever-receding horizon of material externals, rapid economic growth
may indeed cause more human unhappiness than the stable, sustainable
lifestyles usually disparaged as stagnant. In depicting the Silicon Valley
as an attractive model worth emulating, Gary Hamels description in his
award-winning article in the Harvard Business Review unintentionally cap-
tures this paradox simultaneously great and terrible, depending on ones
vantage point of Silicon Valley culture:
The Valley is the distilled essence of entrepreneurial energy. Its ethos is simple:
If its not new, its not cool; if its not cool, its not worth doing. If you dont
own shares, youre getting screwed. If youve been in the same job for more
than two years, your career is over. If you havent been through an IPO, youre
a virgin. This is where a $2 million house is a teardown. This is where a Porsche
is just one more compact car and sushis just another fast food. Never has so
much wealth been created in so little time by so few people. If the Valleys
residents pause to think about it for even a nanosecond, they know theyre as

blessed as those who lived in Italy during the Renaissance. Like the Florentines
and Venetians, theyre building a new age an age of virtual presence, of glo-
bally interconnected communities, of frictionless commerce, of instantly acces-
sible knowledge and stunningly seductive media. (Hamel, 1999)
In sum, through applying appreciative intelligence we can discern that

what made Silicon Valley the poster child for economic development best
practices manuals and the envy of the world was its own deep authenticity
and celebration of itself. The paradox of Silicon Valley is that attempts
by other regions to emulate it are therefore fraught with challenges and
lead to problematic choices. As suggested by, among other examples, the
reverse simulation illustration above, rather than trying to be another
Silicon Valley, regions should look deep within, find the many different
forms of value in which they are rich, and celebrate their own unique con-
stellation of strengths and possibilities for realizing a better world.
REFERENCES
Acs, Z.J. and Audretsch, D.B. (1987), Innovation, Market Structure, and Firm
Size, The Review of Economics and Statistics, 69(4), 56774.
Adams, S.B. (2005), Stanford and Silicon Valley: Lessons on Becoming a High-
tech Region, California Management Review, 48(1), 2951.
Alchian, A.A. (1950), Uncertainty, Evolution, and Economic Theory, The
Journal of Political Economy, 58(3), 21121.
Alvares, C. (1992), Science, in W. Sachs (ed.), The Development Dictionary, Zed
Books, London, pp. 21932.
Audia, P.G. and Rider, C.I. (2005), A Garage and an Idea: What More Does an
Entrepreneur Need?, California Management Review, 48(1), 628.
Audretsch, D.B. and Stephan, P.E. (1996), CompanyScientist Locational Links:
The Case of Biotechnology, The American Economic Review, 86(3), 64152.
Baron, J.N. and Hannan, M.T. (2002), Organizational Blueprints for Success
in High-tech Start-ups: Lessons from the Stanford Project on Emerging
Companies, California Management Review, 44(3), 836.
Begg, I. (1999), Cities and Competitiveness, Urban Studies, 36(56), 795809.
Behara, R., Thatchenkery, T. and Kenney, C. (2008), Empathic Knowledge
Management: Reverse Simulation Experiments in a Learning Laboratory,
International Journal of Information Technology and Management, 7(3), 283
314.
Beugelsdijk, S. (2003), Culture and Economic Development in Europe, Doctoral
Dissertation, Tilburg University.
Block, N.J. and Dworkin, G. (1976), The IQ Controversy: Critical Readings,
Pantheon Books, New York.
Blumenthal, H. (1955), The Economic Base of the Metropolis, Journal of the
American Institute of Planners, 21(4), 11432.
Burrell, G. and Morgan, G. (1985), Sociological Paradigms and Organizational
Analysis, Ashgate, Aldershot, UK.

Camagni, R. (2002), On the Concept of Territorial Competitiveness: Sound or

Misleading?, Urban Studies, 39(13), 2395411.
Cooperrider, D. and Srivastva, S. (1987), Appreciative Inquiry in Organizational
Life, Research in Organizational Change and Development, 1, 12969.
Delbecq, A.L. and Weiss, J. (2000), The Business Culture of Silicon Valley: A
Turn-of-the-Century Reflection, Journal of Management Inquiry, 9(1), 3744.
Elgin, B. (2005), Managing Googles Idea Factory, Business Week (3 Oct.), 8890.
Escobar, A. (1995), Encountering Development: The Making and Unmaking of the
Third World, Princeton University Press, Princeton, NJ.
Farmer, P. (2004), On Suffering and Structural Violence, in P. Farmer (ed.),
Pathologies of Power, University of California Press, Berkeley, CA.
Ferrary, M. (2003), The Gift Exchange in the Social Networks of Silicon Valley,
California Management Review, 45(4), 12038.
Fleming, L. and Marx, M. (2006), Managing Creativity in Small Worlds,
Florida, R. (2002), The Rise of the Creative Class, Basic Books, New York.
Florida, R. (2003), Entrepreneurship, Creativity, and Regional Economic
Growth, in D. Hart (ed.), The Emergence of Entrepreneurship Policy,
Governance, Start-ups, and Growth in the U.S. Knowledge Economy, Cambridge
University Press, New York, pp. 3953.
Florida, R. and Kenney, M. (1990), Silicon Valley and Route 128 Wont Save Us,
Frankl, V. (1963), Mans Search for Meaning, Washington Square Press, New
York.
Freire, P. (1996), Letters to Cristina, Routledge, New York.
Galison, P. (1987), How Experiments End, The University of Chicago Press,
Chicago.
Gardner, H. (1993), Multiple Intelligences: The Theory in Practice, Basic Books,
New York.
Gergen, K. (1994), Realities and Relationships, Harvard University Press,
Cambridge, MA.
Gergen, K. (2009, An Invitation to Social Construction, Sage, Thousand Oaks, CA.
Gergen, K. and Thatchenkery, T. (2004), Organization Science as Social
Construction: Postmodern Potentials, Journal of Applied Behavioral Science,
40(2), 22849.
Gowdy, J. (2000), Terms and Concepts in Ecological Economics, Wildlife Society
Bulletin, 28(1), 2633.
Hamel, G. (1999), Bringing Silicon Valley Inside, Harvard Business Review, 77(5),
7084.
Haraway, D. (1990), Primate Visions: Gender, Race and Nature in the World of
Modern Science, Routledge, New York.
Heineman, M. (1981), The Obsolete Scientific Imperative in Social Work
Research, Social Service Review, 55(3), 37192.
Heineman-Pieper, Jessica (2009), Reclaiming Responsibility: New Foundations for a
Science of and by Persons, Saarbrcken, VDM Verlag Dr. Muller.
Hempel, Carl G. (1965), Aspects of Scientific Explanation, Free Press, New York.
Henderson, J. Vernon (2007), Understanding Knowledge Spillovers, Regional
Science and Urban Economics, 37(4), 497508.
Hoover, E.M. (1948), The Location of Economic Activity, McGraw-Hill, New
York.

Hoover, E.M. (1971), Basic Approaches to the Study of Demographic Aspects of

Economic Development: Economic-Demographic Models, Population Index,
37(2), 6675.
Hulsink, W., Manuel, D. and Bouwman, H. (2007), Clustering in ICT: From Route
128 to Silicon Valley, from DEC to Google, from Hardware to Content (30 Oct.),
ERIM (Erasmus Research Institute of Management) Report Series, Reference
#ERS-2007-064-ORG.
Isard, W. (1949), The General Theory of Location and Space-Economy, The
Kohn, A. (1993), Punished by Rewards, Houghton Mifflin, Boston, MA.
Krugman, P. (1991), Increasing Returns and Economic Geography, Journal of
Kuhn, T. (1962), The Structure of Scientific Revolutions, The University of
Chicago Press, Chicago.
Lecuyer, C. (2007), Making Silicon Valley: Innovation and the Growth of High
Tech, 19301970, The MIT Press, Cambridge, MA.
Lee, C.-M., Miller, W.F, Hancock, M.G., and Rowen, H.S. (eds) (2000), The
Silicon Valley Edge: A Habitat for Innovation and Entrepreneurship, Stanford
University Press, Stanford, CA.
Lewontin, R.C. (1994a), Facts and the Factitious in the Natural Sciences, in
J. Chandler, A.I. Davidson and H. Harootunian (eds), Questions of Evidence:
Proof, Practice and Persuasion Across the Disciplines, University of Chicago
Press, Chicago, pp. 47891.
Lewontin, R.C. (1994b), A Rejoinder to William Wimsatt, in J. Chandler, A.I.
Davidson and H. Harootunian (eds), Questions of Evidence: Proof, Practice
and Persuasion Across the Disciplines, University of Chicago Press, Chicago,
pp. 5049.
Marshall, A. (1890), Principles of Economics, Macmillan, London.
Metcalfe, B. (1998), Asian Tour Provides Useful Insight on Silicon Valleys
Worldwide Internet Edge, Info World (2 March).
OBrien, J.M. (2007), The Pay-Pal Mafia, Fortune, 156(11), 96106.
Popper, Karl R. (2002), The Logic of Scientific Discovery, Routledge, New York.
Porter, Michael E. (1998), Clusters and the New Economics of Competition,
Harvard Business Review, 76(6), 7790.
Putnam, R.D. (2000), Bowling Alone: The Collapse and Revival of American
Community, Simon and Schuster, New York.
Rahnema, M. and Bawtree, V. (1997), The Post-development Reader, Zed Books,
London.
Rogers, E.M. and Larsen, J.K. (1984), Silicon Valley Fever, Basic Books, New
York.
Rosnow, R.L. and Rosenthal, R. (1997), People Studying People: Artifacts and
Ethics in Behavioral Research, Freeman, New York.
Rubin, E. (2000), Figure and Ground, in S. Yantis (ed.), Visual Perception,
Psychology Press, Philadelphia.
Sachs, W. (ed.) (1992), The Development Dictionary, Zed Books, New York.
Saxenian, A. (1996), Regional Advantage: Culture and Competition in Silicon
Valley and Route 12, Harvard University Press, Cambridge, MA.
Saxenian, A. (1999), Silicon Valleys New Immigrant Entrepreneurs, Public Policy
Institute of California, San Francisco.
Saxenian, A. (2000), Networks of Immigrant Entrepreneurs, in C.-M. Lee,

W.F. Miller, M.G. Hancock and H.S. Rowen (eds), The Silicon Valley Edge:
A Habitat for Innovation and Entrepreneurship, Stanford University Press,
Stanford, CA.
Saxenian, A. (2006), The New Argonauts: Regional Advantage in a Global Economy,
Harvard University Press, Cambridge, MA.
Shipley, C. (2006), What Makes Silicon Valley So Different? (19 Oct.), Wisconsin
Technology Network, available at: http://wistechnology.com/articles/3416,
accessed 23 May 2010.
Shiva, V. (2000), Stolen Harvest, India Research Press, New Delhi.
Thatchenkery, T. (2006), Organization Development in Asia: Globalization,
Homogenization, and the End of Culture-specific Practices, in B. Jones
and M. Brazzel (eds), The NTL Handbook of Organization Development and
Change: Principles, Practices, and Perspectives, Pfeiffer/Wiley, San Francisco,
pp. 387403.
Thatchenkery, T. and Metzker, C. (2006), Appreciative Intelligence: Seeing the
Mighty Oak in the Acorn, Berrett-Koehler, San Francisco.
Thatchenkery, T., Behara, R. and Kenney, C. (1999), Building Capabilities for
Change through Laboratory Simulations, Developments in Business Simulation
and Experiential Learning, 26, 1446.
Weick, K. (1983), Utilization as Reverse Simulation: Making the World More
Like the Laboratory, in R.H. Kilmann, K.W. Thomas, D.P. Slevin, R. Nath
and S.L. Jerrell (eds), Producing Useful Knowledge for Organizations, Praeger,
New York, pp. 494520.
Weiss, J. and Delbecq, A. (1987), High-technology Cultures and Management:
Silicon Valley and Route 128, Group and Organization Studies, 12(1), 3954.
Wimsatt, W. (1976), Reductionism, Levels Organization, and the Mind-Body
Problem, in G. Globus, G. Maxwell and I. Savodnik (eds), Consciousness and
the Brain: A Scientific and Philosophical Inquiry, Plenum, New York, pp. 20567.
Wimsatt, W. (1980), Reductionistic Research Strategies and Biases in the Units of
Selection Controversy, in T. Nickles (ed.), Scientific Discovery, Reidel, Boston,
pp. 21360.
Wimsatt, W. (1994), The Ontology of Complex Systems: Levels of Organization,
Perspectives, and Causal Thickets, Canadian Journal of Philosophy, 20(Suppl.),
20774.
Wimsatt, W. (2007), Philosophy for Limited Beings: Piecewise Approximations to
Reality, Harvard University Press, Cambridge, MA.
Zinn, H. (2003), A Peoples History of the United States, Harper-Collins, New
York.

7. An exploratory approach to model
determinants of endogenous
regional growth performance
Robert Stimson and Roger Stough*
INTRODUCTION
Regional economic development may be viewed both as a process and

a product (or outcome). It is a multi-dimensional phenomenon, involv-
ing many actors and being influenced by many factors. It seems to defy
precise definition, incorporating both quantitative and qualitative dimen-
sions. Further, the study of regional economic development incorporates a
concern not only with analysis and modelling, but also a concern for policy
and strategy that may facilitate the regional development process and
facilitate regional change. The process of regional economic development
is certainly dynamic.
Over time, various theoretical approaches to theorizing about and mod-
elling regional growth and development have evolved, and, during the last
couple of decades, there has been an increasing emphasis on the role of
endogenous factors.
Stimson et al. (2003) have proposed the notion of the virtuous circle
as a path to achieving sustainable regional economic development (see
Figure 7.1). They suggest that the virtuous circle may be maintained
through the mediating or intervening effects of factors such as effective
leadership as it might be used to change and adjust institutions in order
to adapt the structure, processes and infrastructure of a regional economy
that is appropriate and needed to meet and anticipate changing circum-
stances, and to facilitate the optimal use of the regions resource endow-
ments and to assist industries to tap their full market potential. Stimson et
al. (2003) put forward the proposition that:
strong leadership means a region will be proactive in initiat-

ing regional economic development strategy to monitor regional
performance;
111

Strong proactive
leadership
Good use and

tapping of
potentials Vision for future
Sustainable development
development
Resource Market
endowments conditions
Strategy, plans,
processes
Mechanisms for Facilitate institutional

using and tapping change to enhance regional
capacity and capability
Effective institutions and
regional infrastructure
Source: Stimson et al. (2003).
Figure 7.1 The virtuous circle for sustainable regional development
leadership will set a vision for its future development; and

the institutions in the region will implement processes and plans that
will facilitate institutional change.
That, in turn, might enhance the capacity and capability of the region to:
positively adjust to changing circumstances;

attain a good fit with market conditions; and
harness its resource endowments,
in order to maintain and improve its performance and to achieve sustain-

able development as a learning region and to be one that is competitive.
In this chapter we first provide a brief discussion of the evolution of
approaches to regional economic development, emphasizing endogen-
ous factors. We then outline an approach we have been developing
to formulate and operationalize a new model framework to measure
regional endogenous performance and model the determinants of spatial
variations in regional endogenous growth performance. Third, we briefly
summarize the results of those exploratory attempts to apply a model
framework.

An exploratory approach to model determinants 113
THE EVOLUTION OF REGIONAL ECONOMIC

DEVELOPMENT THEORY
The importance of endogenous factors in regional economic develop-
ment and growth (and decline) has long been recognized and is largely
the basis of the so-called new growth theory that has been popularized
over the last couple of decades. Theories of endogenous growth place
emphasis not only on regional resource endowments and human capital
which have always been viewed as important factors affecting economic
development but also on technology, entrepreneurship and institutional
factors, including the role of leadership. However, there is neither a stand-
ard definition of nor a method of measuring endogenous growth, nor the
consistent specification of the factors that determine variations in regional
endogenous growth and performance.
The evolution of theories on regional economic growth and development
has seen a shift in the emphasis placed on exogenous factors to endogenous
factors. Traditional regional economic development approaches were
embedded in neoclassical economic growth theory, based largely on the
Solow (1956, 2000) model. These have been replaced over the last couple
of decades by a suite of models and arguments that are commonly known
as the new growth theory where the focus is directed towards endogenous
factors and processes (see, for example, Johansson et al., 2001).
Some of the key contributions to this shift in emphasis towards endog-
enous processes as drivers in regional growth and development are dis-
cussed in a recent review by Stimson et al. (2009a). This includes, inter alia,
the following developments and contributions:
1. In the late 1970s, Norton and Rees (1979) and Rees (1979) pro-
posed that technology was a prime driver in regional economic
development.
2. From that time, theorists such as Romer (1986, 1990), Lucas (1985),
Barro (1990), Rebelo (1991), Grossman and Helpman (1991) and
Arthur (1994) sought to explain technical progress as it generates
economic development as an endogenous effect rather than accept-
ing the neoclassical view of long-term growth being due to exogenous
factors.
3. Thomas (1975) and later Erickson (1994), among others, showed
how technological change is related to the competitiveness of regions.
4. Norton and Rees (1979) and Erickson and Leinbach (1979) showed
how the product cycle, when incorporated into a spatial setting, may
impact differentially on regions through three stages, namely an
innovation stage, a growth stage and a standardization stage.

5. Markusen (1985) extended the product cycle theory of regional

development by articulating how profit cycles and oligopoly in
various types of industrial organization and corporate development
can magnify regional development differentials.
6. The concept of innovative milieu was formulated to explain the how,
when and why of new technology generation. That notion linked
back to the importance of agglomeration economies and localization
economies that may lead to the development of new industrial spaces
(Scott, 1988; Porter, 1990; Krugman, 1991).
7. Some theorists, such as Fukuyama (1995), have suggested that it
is not just economic but also value and cultural factors including
social capital and trust that are important in the rise of technology
agglomerations as seen in the Silicon Valley phenomenon.
8. The effect of human capital skills and income in explaining differen-
tials in levels of regional economic performance has received a lot of
attention (see, for example, Hanushek and Kimko, 2000; Goetz and
Rapasingla, 2001).
9. The effect of industrial structure on regional stability and growth
has been emphasized, with one argument being that industrial
diversity and a trend towards diversification of industry sector
employment enhances opportunities for growth and development
(see, for example, Henderson et al., 1995; Gordon and McCann,
2000), although some researchers claim that there is scant empiri-
cal evidence on that proposition (Kaufman, 1993; Lande, 1994;
Productivity Commission, 1998).
10. The power of urban scale and agglomeration on regional per-
formance is also emphasized (Taylor et al., 2002), with a study by
Duranton and Puga (2000) reviewing national urban systems sug-
gesting that larger cities are more diversified, that individual city-size
rankings and individual city specialization tend to be stable over
time, and that specialized and diversified cities co-exist across a
national urban system.
11. Rees (2001) points out that technology-based theories of regional
economic development need to incorporate the role of entrepreneur-
ship and leadership, particularly as factors in the endogenous growth
of regions market circumstances.
12. There now seems to be an almost universal realization of what
Garlick et al. (2006) refer to as the institutional embeddedness of
endogenous processes and factors in regional development.
The term new growth theory has been coined to describe the transi-
tions such as those referred to above in the evolution of thinking on

regional economic development over the last three to four decades, with
the increasing emphasis being placed on the role of and the importance
of, endogenous factors. Importantly, in the new growth theory models
that have emerged, allowance is made for both agglomeration effects
(economies of scale and externalities) and market imperfections, with the
price mechanism not necessarily generating an optimal outcome through
efficient allocation of resources. Also, the processes of capital accumula-
tion and free trade do not necessarily lead to convergence between regions,
with positive agglomeration effects concentrating activity in one or a few
regions through self-enforcing effects that attract new investment. Most
importantly, the new growth theory allows for both concentration and
divergence.
As discussed by Stimson, Stough and Roberts (2006), these evolution-
ary developments in regional economic growth and development theory
are welcome for regional economic development analysts and planners
because, among other things, they explicitly introduce a spatial dimen-
sion into economic growth theory, a dimension that was ignored in
neoclassical economic development theory. That evolution is particularly
important as the role of regions in national economies has changed signif-
icantly since the 1970s. This has been a result to some degree of the effects
of globalization and structural change and adjustment. Understanding
those processes of change is crucial for analysing and understanding
differentials in the patterns of regional economic performance and for
formulating and implementing regional economic development planning
strategy. But just as important is the need to carefully address endogen-
ous factors and processes to enhance the capacity and capability of a
region to proceed along a path of achieving sustainable development and
growth.
AN APPROACH TO MEASURING AND MODELLING

ENDOGENOUS REGIONAL GROWTH
A New Model Framework
The authors and their collaborators have been seeking to develop and
operationalize a new model framework (see Figure 7.2) to measure and
investigate the determinants of spatial variations in endogenous regional
growth performance. In particular, existing theories of regional economic
development tend to underplay the significance of leadership and other
institutional factors.
Stimson, Stough and Salazar (2003, 2005) have proposed a model

Quasi-independent Variables Intervening Variables Dependent Variable(s)
The dynamic interrelationships

that act to create the catalysis OUTCOME
for regional development A region that is:
Resource
(I) Institutions competitive
endowments and
market conditions (E) Entrepreneurship entrepreneurial
(RE, M) sustainable
(RED)
(L) Leadership
Direct effects
Measure and evaluation change
Indirect effects over time
Benchmark performance (e.g.,
regional shift component in
shift-share analysis)
Source: Stimson, Stough and Salazar (2003).
Figure 7.2 A new model framework for the regional economic

development process
framework conceptualizing regional economic development (RED) and to

explain spatial variations in regional endogenous growth, which specifies
three key intervening or mediating factors:1
leadership (L);
institutional factors (I);
entrepreneurship (E);
which are seen as creating both catalysts and vehicles for better utilization
of a regions resource endowments (RE) and its market fit (M). The model
may be specified as follows:
RED= f [(RE, M) mediated by (L, I, E)]
A Definition of Regional Endogenous Growth: A Proxy Measure for the

Dependent Variable
Despite the evolving focus on endogenous factors in the regional economic

development and growth literature, it is somewhat surprising that there

is no standard definition of endogenous regional economic growth (or

decline) in terms of the specification of an agreed variable that measures
it. Thus, a key issue is: what is an appropriate proxy measure of a regions
endogenous growth? Furthermore, it is also surprising that there is a
lack of operational models to measure the affect of factors such as those
discussed above on explaining spatial variations in regional endogenous
growth performance in a nation or state.
In general economic analysis, most studies derive and measure a vari-
able for endogenous growth by using Ordinary Least Squares,2 or more
recently, panel data analysis. However, the data required to do so is often
not readily available at the disaggregated regional level. As such, many
economic geographers and regional economists have turned to other
techniques.
Approaches used to measure regional endogenous economic growth/
development have been discussed by Stimson, Stough and Salazar (2003,
2005). It might take the form of a proxy or surrogate measure such as:
(1) the aggregated (across all industry sectors) regional differential shift
component value in a shift-share analysis, a common technique in ana-
lysing regional differential performance used by economic geographers
and regional economists; or (2) an employment scale weighted location
quotient change over time, standardized by the size of the regions labour
force. Such measures can then be used as the dependent variable in a
model of regional endogenous economic growth/development.
In the exploratory model applications discussed later in this chapter,
there are attempts to combine a range of endogenous factors to reflect
regional economic growth (which is measured by employment, given the
data constraints on measuring production, at the regional level) rather
than one measure, such as technology or level of savings. This is done by
using shift-share analysis as proposed (see above) by Stimson, Stough and
Salazar (2003b, 2005). The reason for adopting this approach is pragmatic
as secondary data tends to be readily available in most countries to perform
a shift-share analysis, and typically that may be achieved using census data
for industry employment in regions. The regional shift component is a rea-
sonable surrogate measure of the degree to which employment growth or
decline in a region is due to endogenous or within-region processes after
accounting for exogenous factors, such as national shift and the industry-
mix shift effects. Indeed, that is what the regional shift component is pur-
ported to measure. While the authors recognize that using the differential/
regional shift component derived from a shift-share analysis is not an ideal
method of measuring endogenous regional economic growth (or decline),
it is nonetheless considered to be the most optimal given the lack of data at
the regional level to operationalize other measurement approaches.

Potential Independent and Intervening Variables
The potential sets of variables proposed by Stimson, Stough and Salazar

(2003b, 2005) that might be appropriate as measures of the independent
and the mediating factors in the model3 are the following:
1. RE = resource endowments, which might be measured by a set of

variables such as:
area size of the region;

climate;
topography;
agglomeration of industry key sectors (measured by location
quotients for employment in industry sectors);
population size and rate of growth/decline;
education levels (a derived index of human capital) and literacy;
per capita income, income distribution and income distribution
change over time;
housing ownership;
investment in industrial and commercial construction, bench-
marked to the regions national share vis--vis its national share
of population;
infrastructure investment (per capita), such as on roads, schools,
hospitals, and so on;
industrial structure and change in industrial structure (meas-
ured by an industrial diversity index);
regional organizational slack.
2. M = market fit, which might be measured by a set of variables such as:
basic economic activity in major industry sectors (measured by

location quotients for employment in industry sectors);
airline connections with other regions/cities;
road freight in/out movements;
volume and value of exports in key products and services.
It would also be useful to use variables that measure the degree to

which the regions products fit with changing demand and related
markets, to ascertain the degree to which supply fits the local market,
and to evaluate the extent to which the local infrastructure provides
the necessary linkages to export markets.
3. L = leadership, which might be measured by a set of variables such as:

the degree of change/stability in local political leadership;

expert assessment of leadership quality;
corporate headquarters located in the region;
density of business and community organizations per 10 000
population;
size of regional organizations (public and non-profit) budgets/
employment.
4. I = institutions, which might be measured by a set of variables such as:
institutional thickness (corporate and community organizations

per 10 000 population);
layers of government/government fragmentation;
formal institutions of governance, measured by number of
public agencies per 10 000 population;
number of headquarters of major corporations (e.g., Fortune
1000 firms);
value foundation capitalization per 10 000 population;
government fragmentation;
level of regional organizations (number and budget level);
an index of social capital.
5. E = entrepreneurship, which might be measured by sets of variables

such as:
churn rate or business start-up rate;

venture capital activity;
corporate venturing activity;
patents issued per 10 000 workers;
location quotient of employment in symbolic analyst occupa-
tions.
Stimson, Stough and Salazar (2005) argue that RED is positively related
to RE, M, L, I and E, but that there are likely to be lead and lag effects
in the short to intermediate run, and perhaps cyclical effects in the longer
run. Thus:
REDt = REt1 + Mt1 + (It1 to It10/10) + Lt2 + Et2 + e
Exploratory Model Applications
So far there have been two exploratory attempts to apply this model
framework to analyse spatial differentials in regional endogenous growth

performance and identify factors that may explain those variable patterns
in regional endogenous growth performance: (1) In the first application,
Stimson, Robson and Shyy (2005, 2006, 2009b) have conducted an explor-
atory analysis using the OLS technique to estimate a model of regional
endogenous growth performance across non-metropolitan regions in each
of the five mainland states of Australia. (2) In the second application,
Stough et al. (2007) have conducted an exploratory study of variations
in endogenous regional growth across US Metropolitan Statistical Areas
(MSAs).
In both of these model applications, the dependent variable measure
of regional endogenous growth is the regional or differential shift com-
ponent across all industry sectors derived from a shift-share analysis of
employment change over a specified period of time, standardized by size
of the regional labour force. Both studies employ a standard OLS tech-
nique to estimate the model of regional endogenous growth. This depend-
ent variable is regressed on a set of endogenously developed regional
variables on demography and economy, and, where possible leadership,
institutions and entrepreneurship, that are selected as explanatory vari-
ables. Both static point-in-time values around the base year and dynamic
change-over-time values over the period of time being studied are used to
model these explanatory variables, which are assumed a priori to be the
possible determinants of regional endogenous growth.4 While stepwise
regression is the core of the analytical methodology, OLS is used to esti-
mate the fit at each iteration as variables are stepped out of the initial
or general starting model. In that modelling, consideration is given to
testing for spatial autocorrelation to address spatial proximity/spillover
effects.
In the sections that follow we provide a summary discussion of the
results of these two exploratory model applications.
THE AUSTRALIAN APPLICATION
The purpose of the Stimson, Robson and Shyy (2005, 2006, 2009b)
exploratory model application was to empirically examine and model pat-
terns of regional endogenous employment growth in Local Government
Areas (LGAs) across non-metropolitan areas of the five mainland states
of Australia and to identify which factors might explain the spatial vari-
ations in the level of endogenous growth performance over the decade
1991 to 2001. It was restricted to using secondary data on regional
demographic and economic characteristics derived from the census.

The five mainland states were modelled separately in order to generate

comparative data.5
Model Variables
The modelling used a set of 27 independent or explanatory variables (see

Table 7.1) both static and dynamic representing measures relating to
factors that are hypothesized to play a significant role in affecting regional
economic development and growth as suggested in the international and
local Australian literature appraised by the researcher. Those factors
relate to aspects of:
regional industrial structure and specialization/diversification;

population size and growth;
labour force participation and unemployment;
human capital and income; occupational structure;
the effects of proximity to the state capital city (that is the metropoli-
tan area and proximity to the coast.
Those 27 variables were included in an OLS general model, initially a

convention to estimate the model of regional endogenous growth across
regional LGAs of the five mainland states of Australia. Following that, a
stepwise approach was used to determine a specific model.6
Patterns of Endogenous Regional Growth Performance
Stimson, Robson and Shyy (2005) mapped and analysed the spatial
patterns of variation in the regional endogenous employment growth
(or decline) performance over the decade 1991 to 2001 of the non-
metropolitan regional LGAs in the five mainland states of Australia on
the REG_SHIFT dependent variable of endogenous growth. It was found
that:
in New South Wales, overall 45 of the regional LGAs experienced

positive endogenous growth, while 90 had experienced negative
growth;
in Victoria, only eight of the regional LGAs experienced positive
endogenous growth, while 31 LGAs displayed negative endogenous
growth performance;
in Queensland, 39 of regional LGAs had experienced endogenous
growth, while 79 LGAs displayed negative growth;

Table 7.1 Definition of variables used in the non-metropolitan regions

model for the five mainland states of Australia
Variables Definition of Variables

Dependent variable
REG_SHIFT Regional Shift (from 1991 to 2001)/Labour Force (1991)
(i.e., employed + unemployed)
Independent variables
SPEC_91 Specialization Index for 1991 across 17 industry sectors
SPEC_CH Change in Specialization Index from 1991 to 2001
SCI_91-01 Structural Change Index for 1991 to 2001
SCI_CH Change in the Structure Change Index from 19911996 to
19962001
L_INC_01 Log (Average Annual Income for 2001)
UNEMP_91 Unemployment Rate in 1991 (for all persons)
UNEMP_CH Unemployment Rate Change from 1991 to 2001 (for all
persons)
L_POP_91 Log (Population for all persons) in 1991
POP_CH Percentage Point Change in Population from 1991 to 2001
LQMAN_91 Location Quotient for Manufacturing Industry in 1991
LQMAN_CH Change in Location Quotient for Manufacturing Industry
from 1991 to 2001
LQPBS_91 Location Quotient for Property and Business Services
Industry in 1991
LQPBS_CH Change in Location Quotient for Property and Business
Services Industry from 1991 to 2001
LQPER_91 Location Quotient for Personal and Other Services
Industry in 1991
LQPER_CH Change in Location Quotient for Personal and Other
Services Industry from 1991 to 2001
UNIQUALS_91 Proportion of Population with a Bachelor Degree or
higher in 1991
UNIQUALS_CH Change in Proportion of Population with a Bachelor
Degree or higher from 1991 to 2001
TECHQUALS_91 Proportion of Population with a Technical Qualification
in 1991
TECHQUALS_CH Change in Proportion of Population with a Technical
Qualification from 1991 to 2001
ROUTW_91 Proportion of Total Occupations (all persons) as Routine
Production Workers for 1991
INPERS_91 Proportion of Total Occupations (all persons) as In-
person Service Workers for 1991
SYMBA_91 Proportion of Total Occupations (all persons) as
Symbolic Analysts for 1991

Table 7.1 (continued)
Variables Definition of Variables

ROUTW_CH Change in Proportion of Total Occupations (all persons)
as Routine Production Workers from 1991 to 2001
INPERS_CH Change in Proportion of Total Occupations (all persons)
as In-person Service Workers from 1991 to 2001
SYMBA_CH Change in Proportion of Total Occupations (all persons)
as Symbolic Analysts from 1991 to 2001
A_COAST Border is adjacent to coastline (dummy variable 0 = NO,
1 = YES)
P_METRO Is adjacent to the State Capital Statistical District (SD)
(dummy variable 0 = NO, 1 = YES)
Source: Stimson, Robson and Shyy (2006).
in South Australia, 26 of the regional LGAs experienced positive

endogenous growth performance, while it is negative for 23 LGAs;
in Western Australia, there was positive endogenous growth in 32
of the regional LGAs while 80 of them displayed negative growth.
The mapping approach developed by Stimson, Robson and Shyy 2006,

2009b) is novel, using GIS to visualize patterns of variations in regional
endogenous employment growth/decline performance according to the
population size category of the region. An example for the state of New
South Wales (the state with the highest population) is provided in Figure
7.3.
Model Results
The OLS general model run for all five of the mainland states pro-
duced high R-squared values ranging from top values of 0.98 for South
Australia, 0.96 for Victoria and 0.92 for Western Australia, to the slightly
lower values of 0.88 for Queensland and 0.85 for New South Wales. The
general model thus explains much of the variance in dependent variables
on regional endogenous growth measured by the regional shift in employ-
ment change over the decade 19912001 across the non-metropolitan
LGAs. The results from the general model indicated not only some com-
monality but also a degree of variability in magnitude and direction of the
effects of the independent variables in explaining the spatial variation in
levels of endogenous employment growth/decline across non-metropolitan
regional LGAs in the five mainland states analysed.

Figure 7.3 Patterns of endogenous growth performance: New South

Wales
For the application of the specific model, the adjusted R-squared values
derived are high, explaining as much as 98 per cent of the variance for
South Australia, 96 per cent for Victoria and 92 per cent for Western
Australia, 88 per cent of the variance for Queensland and 86 per cent for
New South Wales.
The test conducted to investigate the degree of spatial dependence
showed that only in South Australia is any significant degree of spatial
dependency apparent with the Morans I statistic being 0.154, which is
well below the maximum value of 1. For the other states the Morans I
statistic varies from 0.012 for Victoria to 0.051 for Western Australia.
Not surprisingly there are differences between the five mainland states
in the number of significant variables and the mix of independent variables

Table 7.2 Summary of specific model results: statistically significant independent variables explaining spatial
differences in regional endogenous employment performance (19912001) in the five mainland states,
Australia (direction of influence, positive/negative)
Independent New South Wales Victoria Queensland South Australia Western Australia
STIMSON PAGINATION (M2469).indd 125

variables
SPEC_91 SPEC_91 + SPEC_91 + SPEC_ 91 +
SPEC_CH SPEC_CH + SPEC_CH + SPEC_CH +
SCI_91-01 SCI_91-01 + SCI_91-01 +
SCI_CH SCI_91-01 +
L_INC_01 L_INC_01 +
UNEMP_91 UNEMP_91
UNEMP_CH UNEMP_CH UNEMP_CH UNEMP_CH
L_POP_91 L_POP_91 L_POP_91 + L_POP_91 + L_POP_91 +
125
POP_CH POP_CH + POP_CH + POP_ CH + POP_CH + POP_CH +
LQMAN_91
LQMAN_CH LQMAN_CH+ LQMAN_CH+
LQPBS_91 LQPBS_91 + LQPBS_91 + LQPBS_91 +
LQPBS_CH LQPBS_CH +
LQPER_91 LQPER_91 + LQPER_91 + LQPER_91 + LQPER_91 +
LQPER_CH LQPER_CH +
UNIQUALS_91 UNIQUALS_91 UNIQUALS_91 UNIQUALS_91
UNIQUALS_CH UNIQUALS_CH + UNIQUALQ_CH+ UNIQUALS_CH +
TECHQUALS_91 TECHQUALS_91 TECHQUALS-91 TECHQUALS-91 +
TECHQUALS_CH TECHQUALS_CH + TECHQUALS_CH +
ROUTW_91 ROUTW_91 ROUTW_91
ROUTW_CH ROURW_CH ROUTW_CH + ROUTW_CH +
INPERS_91 INPERS_91 INPERS_91 +
INPERES_CH INPERS_CH +
20/12/2010 15:12
Independent New South Wales Victoria Queensland South Australia Western Australia
variables
SYBBA_91 SYMBA_91 SYMBA_91 +
126
SYMBA_CH SYMBA_CH
D_COAST D_COAST
D_METRO D_METRO D_METRO
20/12/2010 15:12
0.04
Tweed
Residuals
0.03
Snowy River
Muswellbrook
0.02
Cobar Tamworth
Residuals
Broken Hill Bega Valley

0.01
Shoalhaven
Armidale Dumaresq
Queanbeyan
Dubbo Wagga Wagga Deniliquin
0 Albury Wollongong
0 20 40 Bathurst 60 80 100 Orange 120 140
Griffith Grafton
Glen Innes
0.01
Ballina
Greater
Taree Newcastle Narrabri
Coffs Harbour
0.02
LGAs
Figure 7.4 Order of regional LGA residuals for the specific model: New
South Wales
in the specific model. Table 7.2 provides a summary of those significant

explanatory factors.
Stimson, Robson and Shyy (2005, 2009a) plotted on a graph the order
of residual scores of LGAs from the line of best fit derived from the spe-
cific model (that is, the stepwise regression). And they also mapped the
spatial patterns of the residual scores for LGAs. Here we only reproduce
the plot for New South Wales (Figure 7.4), which shows there are no dis-
cernable outliers, with only two LGAs being above the +0.03 deviation.
The large majority of LGAs fall within the range +/0.01.
From this exploratory model application to the five mainland states of
Australia, it is evident that the specific model gives a relatively good fit,
especially in Victoria and South Australia and to a slightly lesser degree
in New South Wales, while in Queensland and Western Australia there is
a greater degree of deviation in from the line of best fit derived from the
model. The modelling conducted by Stimson, Robson and Shyy (2005,
2006, 2009a, 2009b) lends support to the importance of variables relating
to industrial structure and industry specialization, population size and to
population change, levels of human capital and aspects of job concentra-
tion in industry sectors and occupation categories that have been hypoth-
esized as being important in explaining differences in regional endogenous
growth performance. But there are considerable differences between the
five mainland states of Australia in the degree to which the individual
independent variables are significant independent (explanatory) factors
accounting for variations between the regional LGAs in their regional

endogenous growth performance over the decade 19912001. Further,

there are also some variations between those states in the direction of the
impact of some of the variables on regional endogenous growth.
From the specific model results, the following conclusions are drawn by
Stimson, Robson and Shyy (2006, 2009b):
1. Population growth emerges as a particular factor with a strong

positive impact in all five mainland states, and population size at the
beginning of the decade has an important positive impact in all states
except New South Wales.
2. The level of industry sector specialization, and the change in it over
time, both have an important positive effect in all states except New
South Wales, and the structural change index and the dynamics in
it over time are important positive factors in Queensland, South
Australia and Western Australia.
3. The level of income at the end of the decade 1991 to 2001 is a
significant factor only in Queensland.
4. Only in New South Wales do the unemployment variables have an
impact.
5. There is a mixed picture regarding the impacts of the degree of con-
centration of jobs in broad industry sectors, and the effects are most
apparent in New South Wales and in Western Australia. But there is
neither a consistent directional effect for these variables, nor for the
change over time in their location quotients.
6. There is a particularly marked impact from the incidence of people
with university and technical qualifications in New South Wales and
Queensland, but to a lesser degree in Victoria and the other states.
The incidence of people with university qualifications at the beginning
of the decade has a negative impact, but the change in that incidence
over the decade is a positive impact factor in New South Wales and
Western Australia, where the change in incidence of people with
technical qualifications also has a positive impact.
7. There is no really discernable pattern in the impact of the incidence
of jobs in the Reich (1991) broad occupation categories, except in
Victoria where the effects are negative, while in South Australia,
Western Australia and Queensland the effects are positive when such
a variable is significant.
8. Only in Western Australia and South Australia is it evident that the
proxy variables relating to location on or near the coast or proximity
to the metropolitan area have a significant effect, and that effect is
negative.

THE US APPLICATION
The purpose of the Stough et al. (2007) exploratory model was to empiri-
cally examine and model the sources of regional endogenous growth
across the US MSAs for the period 19992002. Specifically, the modelling
attempts to uncover or examine which factors assumed to be endogenous
to a region are actually affecting changes in the regional endogenous
employment growth of 245 MSAs between 1999 and 2002.7
Method
The US study followed the methodology employed in the Stimson,

Robson and Shyy (2005, 2006, 2009b) exploratory study in Australia dis-
cussed above. It uses the same set of variables relating to demographic and
economic characteristics of a region, but in this US application Stough et
al. (2007) added to the set of independent or explanatory data variables
a selection of variables that are surrogate measures (that is, indicators)
of factors relating to leadership, institutions and entrepreneurship. Thus,
this model application to MSAs in the US is an exploratory attempt to use
the full model framework for modelling regional endogenous growth pro-
posed by Stimson, Stough and Salazar (2003, 2005) and as set out earlier
in Figure 7.1.
The model uses secondary data available from multiple data sources,
including the US Census Bureau, Bureau of Economic Analysis (BEA)
and Bureau of Labor Statistics (BLS).
Spatial Patterns of Endogenous Employment Growth Performance
The map in Figure 7.5 shows the pattern of MSA scores on the depend-
ent variable measuring endogenous employment growth/decline for the
period 19992002. On the map the MSAs are differentiated according to
whether their score on the dependent variable was weak or strong, positive
or negative, using the cut-off points method of classification. MSAs are
differentiated according to their population size category.
It is evident that MSAs with different levels of population experienced
different levels of regional endogenous growth. Generally MSAs with
small population size experience relatively higher regional endogenous
economic growth than those with larger populations. However, the
variance for the small-size MSAs is much greater than for the larger-sized
MSAs. Specifically, the regional endogenous growth rate of the medium-
sized MSAs (population size between 0.2 million and 1 million) is lower
than that of the MSAs with a smaller population (less than 0.2 million) by

130
Source: The authors.
Figure 7.5 Pattern of MSA scores on the dependent variable standardized endogenous employment growth, 19992002,
by population size category
20/12/2010 15:12
9.53 per cent. Further, the rate for larger MSAs (population size between 1
million and 6 million) is even lower than that of the MSAs with population
size less than 0.2 million by 12.52 per cent.
It is interesting that the small-size MSAs dominate in the 20 top-ranking
MSAs with the highest positive scores on the endogenous employment
growth dependent variable, with only four medium-size MSAs with popu-
lations between 200 000 and 1 million being in the top 20 ranked MSAs.
Not one of the large MSAs with populations more than 1 million is in
the top 20 ranked MSAs. In contrast, nine of the large MSAs are in the
bottom 20 ranked MSAs with the greatest negative scores on the endogen-
ous employment growth dependent variable, and only one of the small
MSAs was in that list of 20 bottom-ranked MSAs.
General Model Results
For the application of the general model, the R-squared value for the
final general regression model was 0.57, with an adjusted R-squared value
of 0.55, meaning that about 55 per cent of the variance in the dependent
variable is explained by the final general OLS regression model, which
includes 11 significant explanatory variables. Table 7.3 lists the variables
used in the model, and Table 7.4 gives parameter estimation and statistical
fit information for the model.
From Table 7.4, some underlying factors capable of affecting regional
endogenous economic growth may be specified and explained. A summary
of the results follows.
Stough et al. (2007) show that four sets of variables measuring resource
endowments (POP_CH, BACH_00, DOCT_00), market fit (LQGOV_98,
INPERS_99, INPERS_CH), institutions (LGOVEM_CH) and entre-
preneurship (EM1_4_CH, EM5_9_CH) have differing levels of effect on
regional endogenous growth for the US MSAs over the study period. For
example, the following outcomes were identified:
1. Regional educational attainment level has different effects on regional

endogenous growth. A 1 per cent increase in the percentage of popu-
lation over 25 with a Bachelor Degree in 2000 is associated with a
0.88 per cent decrease in regional endogenous growth, while a 1 per
cent increase in the percentage of population over 25 with a Doctoral
Degree in 2000 is associated with an increase of 3.61 per cent in
regional endogenous growth.
2. Regional government plays an important role in metropolitan regional
economic growth. A 1 per cent increase in the regional location quo-
tient for government and government enterprises in 1998 is associated

Table 7.3 Definition of variables used in the US MSAs model application
Variable Definition Source

Dependent variable

REG_SHIFT Regional Shift (from 1999 to 2002)/employment (1999) BEA
Explanatory variables I: resource endowments

L_POP_99 Log (Population for all persons 1999) BEA
POP_CH Percentage change in Population from 1999 to 2002 BEA
L_INC_02 Log Per Capita Personal Income for 2002 BEA
UNEMP_00 Unemployment Rate in 2000 (for all persons) BLS
UNEMP_CH Unemployment Rate Change from 2000 to 2002 (for all persons) BLS
BACH_00 Percentage of Population over 25 with a Bachelor Degree in 2000 CENSUS
132
EDUMP_00 Percentage of Population over 25 with a Master Degree or Professional Degree in 2000 CENSUS
DOCT_00 Percentage of Population over 25 with a Doctoral Degree in 2000 CENSUS
Explanatory variables II: market fit

LQMAN_98 Location Quotient for Manufacturing Industry in 1998 BEA
LQSER_98 Location Quotient for Services Industry in 1998 BEA
LQGOV_98 Location Quotient for Government and Government Enterprises in 1998 BEA
LQMAN_CH Change in Location Quotient for Manufacturing Industry from 1998 to 2002 BEA
LQSER_CH Change in Location Quotient for Services Industry from 1998 to 2002 BEA
LQGOV_CH Change in Location Quotient for Government and Government Enterprises from 1998 to BEA
2002
ROUTW_99 Percentage of Total Occupations (all persons) as Routine Production Workers for 1999 BLS
INPERS_99 Percentage of Total Occupations (all persons) as In-person Service Workers for 1999 BLS
SYMBA_99 Percentage of Total Occupations (all persons) as Symbolic analysts for 1999 BLS
20/12/2010 15:12
ROUTW_CH Percentage Change of Total Occupations (all persons) as Routine Production Workers from BLS
1999 to 2002
INPERS_CH Percentage Change of Total Occupations (all persons) as In-person Service Workers from BLS
1999 to 2002
SYMBA_CH Percentage Change of Total Occupations (all persons) as Symbolic Analysts from 1999 to BLS

2002
Explanatory variables III: regional leadership

FORTU1000 Corporate Headquarters of Fortune 1000 located in an MSA, 2005 FORTUNE
Explanatory variables IV: regional institutions

LGOVEX_97 Log Local Government Expenditures in 1997 CENSUS
LGOVEM_97 Log Local Government Employment in 1997 CENSUS
LGOVEX_CH Percentage Change in Local Government Expenditures between 1997 and 2002 CENSUS
133
LGOVEM_CH Percentage Change in Local Government Employments between 1997 and 2002 CENSUS
SOCIAL_CAP Index of Social Capital Averaged during five years from 199398 DDB
Explanatory variables V: regional entrepreneurship

FIRM0_98 Percentage of Self-employed Firms (do not employ any employee) in the total number of SBA
firms in 1998
EM1_4_98 Percentage of the Employment in firms with 14 employees in the total employment of an SBA
MSA in 1998
MSA in 1998
MSA in 1998
FIRM0_CH Percentage Change of Self-employed firms (do not employ any employee) in the total SBA
number of firms between 1998 and 2002
20/12/2010 15:12
Variable Definition Source

EM1_4_CH Percentage Change of the Employment in firms with 14 employees in the total employment SBA
of an MSA between 1998 and 2002
EM5_9_CH Percentage Change of the Employment in firms with 59 employees in the total employment SBA
of an MSA between 1998 and 2002
EM10_19_CH Percentage Change of the Employment in firms with 1019 employees in the total SBA
134
employment of an MSA between 1998 and 2002
Explanatory variables VI: population dummies

POPDUM_1 Equals 1 if population<0.2 million and 0 otherwise
POPDUM_2 Equals 1 if 0.2 million<=population<1 million and 0 otherwise
POPDUM_3 Equals 1 if 1 million<=population<6 million and 0 otherwise
Source: Stough et al. (2007).
20/12/2010 15:12
Table 7.4 OLS general model results for the US MSAs model application
Coef. Value Std. Err. t-value P > |t|

Intercept 3.237 5.600 0.58 0.564
POP_CH 0.603 0.131 4.61 0.000
BACH_00 0.881 0.172 5.14 0.000
DOCT_00 3.612 0.763 4.74 0.000
LQGOV_98 2.031 1.225 1.66 0.099
INPERS_99 0.410 0.118 3.47 0.001
INPERS_CH 0.323 0.178 1.82 0.071
LGOVEM_CH 9.236 2.534 3.64 0.000
EM1_4_CH 7.043 1.900 3.71 0.000
EM5_9_CH 4.142 1.419 2.92 0.004
POPDUM_2 9.538 1.136 8.40 0.000
POPDUM_3 12.522 1.623 7.72 0.000
Source: Stough et al. (2007).
with an increase of 2.03 per cent in regional endogenous economic

growth. Meanwhile, a 1 per cent increase in local government employ-
ment is associated with an increase of 9.23 per cent in regional endog-
enous economic growth.
3. The new occupational structure of routine production workers, in-
person service workers and symbolic analysts created by Reich (1991)
has differential effects. Compared with the percentage of total number
of occupations for routine production workers, a 1 per cent increase
in the percentage of the total number of occupations for in-person
service workers can increase regional economic growth by 0.41 per
cent. However, a 1 per cent increase in the percentage change of
total number of occupations for in-person service workers decreases
regional endogenous growth by 0.32 per cent.
4. Regional entrepreneurship capital, measured by the percentage of
and percentage change of the employment in small businesses in the
total employment of a region, positively affects the level of regional
endogenous growth. Meanwhile, the pools of small businesses of dif-
ferent sizes in a region may have different effects on regional economic
growth. A 1 per cent increase in the percentage change of the employ-
ment in firms with 14 employees in the total employment is associ-
ated with a 7.04 per cent decrease in regional endogenous growth,
while a 1 per cent increase in the percentage change of employment in
firms with 59 employees in the total employment of an MSA during

the same period is associated with a 4.10 per cent increase in regional
endogenous growth.
Specific Model Results
In conducting the specific model, a number of issues were considered by

Stough et al. (2007):
1. First, since MSAs of different population sizes seem to behave dif-

ferently in terms of regional endogenous growth during the period of
1999 to 2002, it is necessary to examine how the regional endogen-
ous growth of each category of MSA segmented into population
size groups (small, medium and large) is endogenously determined
by regional factors and how these regional factors may differentiate
one another in terms of MSAs with different population size. Thus,
the original sample with 245 observations was divided into three
population size categories:
small-size MSAs with the population dummy POPDUM_1

equalling 1 (less than 0.2 million, 95 observations);
medium-size MSAs with the population dummy POPDUM_2
equalling 1 (between 0.2 and 1 million, 106 observations);
large-size MSAs with the dummy POPDUM_3 equalling 1
(between 1 and 6 million, 44 observations).
2. Variance Inflation Factors (VIFs) were examined for each sub-group

and explanatory variables with VIF values over ten were excluded
from each regression model.
3. Finally, ROUTW_99 and ROUTW_CH, EM5_9_98 were excluded
from the small-size MSAs model; EM5_9_98 and EDUMP_00 were
excluded from the medium-size MSAs model; and EM5_9_98, L_
POP_99, SYMBA_99, LQGOV_98, DOCT_00, FIRM0_98 and
INPERS_99 were excluded from the large-size MSAs model.
The specific model (backward removal stepwise approach) was then

applied to each of the three sub-groups with three specific models
obtained. The R-squared values for the final regression models for the
small-, medium- and large-size MSAs are 0.40, 0.60 and 0.82 respectively.
The adjusted R-squared values are 0.33, 0.57 and 0.77, indicating that
the endogenous growth in small MSAs from 1999 to 2002 are determined
more by such potential regional factors as resource endowments, market

fit, leadership, institutions and entrepreneurship than those in medium

and large MSAs.
The results of the three final specific models of the three size categories
of MSAs are not presented here, but they are both interesting, in some
cases provocative and in some cases even perplexing. For example, con-
sider the following:
1. While the population factor has significant effects on the regional eco-
nomic growth at all three levels of the US MSAs, the direction of that
effect varies. Population change has a greater impact on the regional
endogenous economic growth of medium-sized MSAs than for large-
size MSAs, while small population size negatively affects the regional
endogenous growth of small-size MSAs. This implies agglomeration
economies may be an important element in the generation of endogen-
ous growth, which is consistent with the literature.
2. Educational attainment level affects regional economic growth with
the highest level of education attainment (measured by the percentage
of population over 25 with a Doctoral Degree in an MSA) positively
affecting regional endogenous economic growth and the lower educa-
tion level (measured by the percentage of population over 25 with a
Bachelor Degree in an MSA) negatively affecting regional endogenous
economic growth.
3. Regional entrepreneurship capital affects regional endogenous growth
of MSAs, but is manifested differently depending on the measure of
entrepreneurship capital used. For small MSAs, the stock value of
the entrepreneurship capital measured by the percentage of employ-
ment in firms with 14 employees positively affects regional economic
growth, while the stock value of entrepreneurship capital measured by
the percentage of employment in firms with 1019 employees nega-
tively affects regional economic growth. Moreover, for small MSAs
the change in entrepreneurship capital also affects regional endogen-
ous growth: the percentage change of the employment in firms with
59 employees in the total employment has a positive impact on
regional endogenous growth over 1999 to 2002, while the percent-
age change of employment in firms with 1019 employees in the total
employment of an MSA between 1998 to 2002 has a negative impact
on regional endogenous growth. Change of regional entrepreneur-
ship capital negatively affects regional endogenous growth for both
medium and large MSAs compared with the small MSAs.

CONCLUSION
In this chapter we have outlined the development by the authors and
their collaborators of a new model framework to measure and investigate
the determinants of spatial variations in regional endogenous growth per-
formance, and in particular to identify the roles of leadership, institutional
factors and entrepreneurship. As exploratory applications of the model,
the sources of regional endogenous growth across the non-metropolitan
regions of the five mainland states of Australia over the decade 1991 to
2001 and across the US MSAs between 1999 and 2002 were empirically
modelled and then identified by a sequenced OLS regression and a step-
wise regression approach. By constructing two OLS endogenous growth
regression models with the regional share of the employment change for
a region as the dependent variable and some variables for measuring
regional resource endowments and market fit in the case of both explora-
tory model applications, plus the effects of variables relating to leadership,
institutions and entrepreneurship characteristics of the region in the case
of the latter exploratory model application, as explanatory variables, we
find that some of the regional endogenous factors do play important roles
in affecting the regional-share employment changes.
Of course much remains to be done in the development of these
operational models that seek to identify the roles of endogenous and non-
endogenous factors in explaining spatial variability of levels of regional
growth/decline performance across the spatial units of a space economy.
That includes the need to more effectively and directly address the issue
of spatial autocorrelation and incorporate weightings in the data matrix
to account for spatial spillover/proximity effects. Further, the problem of
directly measuring the mediating variables of entrepreneurship, leader-
ship and institutions leaves much to be desired and begs for significant
improvements in their measurement. It is interesting in the US case that
metropolitan regional size is found to be an important factor in regional
endogenous growth. The significance of this is that it strongly suggests the
importance of agglomeration forces in contemporary regional growth and
thus further contributes to the confirmation of the validity of the new eco-
nomic geography. This in turn casts additional concern on the importance
of the problem of stimulating and inducing economic growth in general
and endogenous economic growth in rural and non-urban areas.
Finally, the modelling approach might be enhanced by adopting a path-
analytic approach where the intervening/mediating effects of the leader-
ship, institutional and entrepreneurship variables on regional endogenous
growth performance are explicitly tested.

NOTES
* The authors wish to acknowledge the financial support of the Australian Research
Council Linkage International Scheme (grant #LX0346785) and the George Mason
University Foundation for the research project Regional Economic Development and
Performance: Roles of Leadership and Institutional Factors in Endogenous Growth
from which this chapter draws. We also acknowledge the contribution to that research by
Maria Salazar, Alistair Robson, Tung-Kai Shyy, Chunpu Song, Scott Jackson, Jiamin
Wang and Heifung Qian. Finally, we have gained considerably from the comments of
two anonymous reviewers.
1. It is important to note that the new model framework does not negate the importance of
exogenous factors and in fact recognizes these traditional sources of economic growth;
rather it emphasizes the largely missed importance of such mediating factors as entrepre-
neurship, leadership and more generally institutions in the development process.
2. It is important to note that OLS regression models need to be modified to incorporate the
effects of spatial spillover/proximity effects and to give consideration to spatial autocor-
relation as regional economic development and growth models use spatial data.
3. The interested reader is invited to review the work by Plummer and Sheppard (2006),
which illustrates both institutional and endogenous theory foundations for many of the
variables suggested here.
4. Some regions may have a high point-in-time value at the start of period of time but have
little change over that period; others may have a low point-in-time value at the start but
have a large change over time. The dynamic change-over-time variables are added to the
analysis to capture the effect of change during a period of time.
5. The decision to model the five mainland states separately does present a problem in that
the interaction effects between regions across state borders is lost. However, in policy
terms there is often a focus on state-level analysis in Australia.
6. Each step involved withdrawing one independent variable from the model. That deleted
variable had the highest probability that its absolute t-value was equal to or greater than
0.05. The model was then regressed, and new estimates of the model were obtained. This
process was repeated until the variable with the highest probability that its absolute
t-value was less than 0.05 was identified.
7. The period between 1999 and 2002 was selected for the availability of data; in particular,
the most recent data for some variables related to endogenous growth are only avail-
able in this period. Also, the US economy went into a recession in 2002, hence research
based on this time period may partly reveal the effect and importance of endogenous
growth factors when the national economy is in relative decline. However, this is likely
minimal because the recession that began in 2002 was by most standards a mild one and
its duration quite time limited.
REFERENCES
Arthur, W.B. (1994), Increasing Returns and Path Dependency in the Economy,
University of Michigan Press, Ann Arbor.
Barro, R.J. (1990), Endogenous Technological Change, Journal of Political
Economy, 98(5), S71S102.
Duranton, G. and Puga, D. (2000), Diversity and Specialization in Cities: Why,
Where and When Does it Matter?, Urban Studies, 37(3), 53355.
Erickson, R.A. (1994), Technology, Industrial Restructuring and Regional
Development, Growth and Change, 25(3), 35379.
Erickson, R.A. and Leinbach, T. (1979), Characteristics of Branch Plants

Attracted to Non Metropolitan Areas, in R. Lonsdale and H.L. Seyter (eds),

Non Metropolitan Industrialization, Winston, Washington, DC.
Fukuyama, F. (1995), Trust: The Social Virtues and Creation of Prosperity, Free
Press, New York.
Garlick, S., Taylor, M. and Plummer, P. (2006), An Enterprising Approach to
Regional Growth: The Role of VET in Regional Development, NCVER (National
Council for Vocational Education Research), Australian National Training
Authority, Adelaide.
Goetz, S.J. and Rapasingla, A. (2001), The Returns to Higher Education:
Estimates for the Contiguous States, paper to Regional Science Association
International, North American Annual Meeting, Charleston, November.
Gordon, I.R. and McCann, P. (2000), Industrial Clusters, Complexes,
Agglomeration and/or Social Networks, Urban Studies, 37(3), 51332.
Grossman, G.M. and Helpman, E. (1991), Innovation and Growth in the Global
Economy, MIT Press, Cambridge, MA.
Hanushek, E.A. and Kimko, D.D. (2000), Schooling, Labour-force Quality, and
the Growth of Nations, American Economic Review, 90(5), 1184208.
Henderson, J.V., Kuncoro, A. and Turner, M. (1995), Industrial Development in
Cities, Journal of Political Economy, 103(5), 106790.
Johansson, B., Karlsson, Ch. and Stough, R.R. (2001), Theories and Endogenous
Growth: Lessons for Regional Policy, Springer-Verlag, Heidelberg.
Kaufman, R. (1993), An Empirical Exploration of the Relation Among Diversity,
Stability and Performance in Economic Systems, Structural Change and
Economic Dynamics, 4(2), 299313.
Lande, P. (1994), Regional Industrial Structure and Economic Growth and
Stability, Journal of Regional Science, 34(3), 34360.
Lucas, R.E. (1985), Models of Business Cycles, Basil Blackwell, Oxford.
Markusen, A. (1985), Profit Cycles, Oligopoly and Regional Development, MIT
Norton, R.D. and Rees, J. (1979), The Product Cycle and the Decentralization of
North American Manufacturing, Regional Studies, 13(2), 14151.
Plummer, P. and Sheppard, E. (2006), Geography Matters: Agency, Structure
and Dynamics at the Intersection of Economics and Geography, Journal of
Economic Geography, 6(5), 61937.
Porter, M.E. (1990), The Competitive Advantage of Nations, Macmillan, New
York.
Productivity Commission (1998), Aspects of Structural Change in Australia,
Research Report, Ausinfo, Canberra.
Rebelo, S. (1991), Long Run Policy Analysis and Long Run Growth, Journal of
Political Economy, 98(5), S71S102.
Rees, J. (1979), State Technology Programs and Industry Experience in the USA,
Review of Urban and Regional Development Studies, 3, 3959.
Rees, J. (2001), Technology and Regional Development: Theory Revisited, in
B. Johansson, Ch. Karlsson and R.R. Stough (eds), Theories of Endogenous
Regional Growth, Springer-Verlag, Heidelberg, pp. 94110.
Reich, R. (1991), The Work of Nations: Preparing Ourselves for 21st Century
Capitalism, Vintage Books, New York.
Romer, P. (1986), Increasing Returns and Long Run Growth, Journal of Political
Economy, 94(5), 100237.

Romer, P.M. (1990), Endogenous Technological Change, Journal of Political

Economy, 98(5), S71S102.
Scott, A.J. (1988), New Industrial Spaces: Flexible Production Organization and
Regional Development in North America and Western Europe, Pion, London.
Solow, R.M. (1956), A Contribution to the Theory of Economic Growth,
Solow, R.M. (2000), Growth Theory: An Exposition, Oxford University Press, New
York.
Stimson, R.J., Robson, A. and Shyy, T.-K. (2005), Modeling Determinants
of Spatial Variations in Regional Endogenous Growth: Non-metropolitan
Regions in the Mainland States of Australia, 45th Congress of the European
Regional Science Association, Amsterdam, August.
Stimson, R.J., Robson, A. and Shyy, T.-K. (2006), Modeling Regional Endogenous
Growth: An Application to the Non Metropolitan Regions of Australia,
International Workshop on Creativity and Smart Policies as Signposts for
Innovative Development, Amsterdam: Tinbergen Institute and Free University,
May.
Stimson, R.J., Robson, A. and Shyy, T.-K. (2009a), Measuring Regional
Endogenous Growth, in R. Capello and P.J. Nijkamp (eds), Regional Growth
and Development Theories in the XXI Century: Theoretical Achievements and
Future Challenges, Edward Elgar, Cheltenham, UK and Northampton, MA,
USA, pp. 35472.
Stimson, R.J., Robson, A. and Shyy, T.-K. (2009b), Modeling Regional
Endogenous Growth: An Application to the Non-metropolitan Regions of
Australia, Annals of Regional Science, 43(2), 37998.
Stimson, R.J., Stough, R.R. and Roberts, B.H. (2006), Regional Economic
Development: Analysis and Planning Strategy, Springer, Berlin.
Stimson, R., Stough, R. and Salazar, M. (2003), Leadership and Institutional
Factors in Endogenous Regional Economic Development, paper presented
to The North American Regional Science Association 50th Annual Meeting,
Philadelphia, November.
Factors in Endogenous Regional Economic Development, Investigaciones
Regionales, 7, 2352.
Stimson, R., Robson, A.P., Stough, R. and Salazar, M. (2003), Leadership,
Institutions and Regional Economic Development: A New Conceptual
Framework, paper presented to The Pacific Regional Science Conference
Organization, Acapulco, Mexico, July.
Stough, R.R., Song, C., Wang, J. and Qian, H. (2007), Modeling Endogenous
Growth in U.S. Metropolitan Regions, paper presented at the Annual Meeting
of the Western Regional Science Association, Newport Beach, CA, February.
Taylor, P.J., Catalano, G. and Gane, N. (2002), A Geography of Global Change:
Services and Cities 200001, GaWC Research Bulletin, Globalisation and
World Cities Study Group and Network, 77, 19.
Thomas, M.D. (1975), Growth Pole Theory, Technological Change and Regional
Economic Growth, Papers of the Regional Science Association, 34(1), 325.

8. A theory of entrepreneurial rents in
endogenous growth: implications
for regional innovation policies
Zoltan Acs and Mark Sanders
INTRODUCTION
This chapter is about the role of rents and endogenous entry in economic
growth two issues we feel are not getting the attention they deserve in the
modern literature on endogenous growth. When Romer (1990) presented
his seminal article, his key contribution was to provide a private incen-
tive for doing R&D. A positive knowledge spillover (in the tradition of
Arrow, 1962) from current to future R&D then provided the mechanism
for endogenous growth. But in his model he attributed the full rents of new
entry to a specialized R&D sector that auctions off blueprints for innova-
tions to entrepreneurs. Competition among the latter leaves no rents for
them and their role is reduced to merely commercializing whatever the
R&D sector comes up with.
Aghion and Howitt (1992) on the other hand, put entry central in
their Schumpeterian model of growth. Entrepreneurs appropriate the
full rents to entry in their model, but Aghion and Howitt also have them
do the R&D that generates the knowledge in a tournament. There they
clearly deviate from Schumpeter who sees no role for the entrepreneur
in opportunity creation (Schumpeter [1911] 1934). In that way Aghion
and Howitt leave no rents to motivate R&D by firms who do not aim to
enter the market with new varieties or higher-quality products. In fact
they make a strong point to explain why incumbent firms do not engage
in R&D.
Both models are a huge improvement over earlier growth models. Yet
in the data we observe that the vast bulk of rents from innovation accrue
to the entrepreneur introducing the innovation in the market. And we can
see incumbent firms do the vast bulk of R&D in any market economy.1
It has long been recognized that rents are what motivates an entrepre-
neur to enter. Schumpeter (ibid.) claimed that entry was the key source
142

A theory of entrepreneurial rents 143
of innovation. In his model of economic growth there is no R&D and

knowledge creation and entrepreneurial opportunities are falling like
manna from heaven. A large body of evidence suggests that entrepreneur-
ial firms do in fact capture rents from entry. For example, the continued
entry of new firms and the going public of these entities suggest that the
entrepreneurial sector captures the lions share of rents in the economy.
This implies that these rents cannot be financing knowledge generation
because entry is not generating any knowledge from commercializing
ideas.
In this chapter we present a model in which incumbent firms finance
R&D for a clear profit motive but entrepreneurs capture the rents of com-
mercializing the opportunities that this R&D generates through regional
intra-temporal knowledge spillovers (Acs et al., 2009). In doing so we
answer the question why incumbent firms do not commercialize all the
knowledge their R&D generates and give entrepreneurs their proper func-
tion in generating economic growth, which is to recognize and exploit such
opportunities. It is also shown that these seemingly marginal changes have
some seriously different implications.
Our model predicts that in a decentralized equilibrium R&D will
be under-funded and there is room for welfare gains through policy.
However, a delicate balance now needs to be found when R&D and
entrepreneurship compete over the same resources. As more R&D implies
less entrepreneurship to pick up the opportunities that R&D generates, a
trade-off has to be made. The social optimum is to stimulate both R&D
and entrepreneurship and invest resources in facilitating the knowledge
spillover to entrepreneurs.
The purpose of this chapter is to develop a model that turns the origi-
nal Romer incentive structure on its head. Knowledge is generated inside
the R&D labs of incumbent firms. We develop a two-sector model where
incumbent firms pursue R&D investments to increase their flow of rents
through process innovations leading to efficiency gains. The incentive
structure for incumbent firms is such that they do not pursue new varieties
for fear of cannibalizing their own rents. While incumbent firms do not
fully exploit new ideas knowledge spills over to potential entrants. This is
both an intra-temporal spillover and a regional activity. The rents of com-
mercialization finance, motivate and endogenously attract the entrants.
Commercialization is driven by the pursuit of rents by entrepreneurs who
set up new ventures.
In what follows we review the literature, outline the model and present
its propositions, then draw some conclusions.

REVIEWING THE LITERATURE

Policy-makers are interested in promoting economic growth. This is true
at the national and the regional level. And politicians look to academia to
help them understand the process of economic development and inform
their decisions. Academics have, long ago, identified technical change
through innovation as a key process for generating long-term stable eco-
nomic growth. But that begs the question what causes innovation in a
region or economy?
Building on the Jaffe-Feldman-Varga model of regional knowledge
spillovers, Acs and Varga (2002) suggest a spatialized theoretical frame-
work of technology-led economic growth that explains three fundamental
issues.2
1. First, it provides an explanation of why knowledge-related economic

activities may concentrate in certain regions, leaving others relatively
underdeveloped.
2. Second, it answers the questions of how technological advance occurs
and what are the key processes and institutions involved, with a
particular focus on the geographic dimension.
3. Third, it presents an analytical framework where the role of techno-
logical change in regional and national economic growth is clearly
explained.
In order to explore these three issues Acs and Varga (ibid.) examine
three separate and distinct literatures:
the new economic geography;

the new growth theory;
the new economics of innovation.
The three literatures focus on different aspects but at the same time are
also complements of each other. The new theories of growth endogenize
technological change and as such link knowledge creation and spillovers
to technological change and macroeconomic growth at the aggregate level.
New growth theory has identified knowledge generation (Romer, 1986,
1990; Aghion and Howitt, 1992) through R&D and human capital accu-
mulation (Lucas, 1988) as the ultimate sources of economic growth. The
empirical evidence in support of this view is mounting (Temple, 1999). The
evidence clearly shows that more knowledge generation through R&D
and human capital accumulation is correlated with economic perform-
ance, both across countries and regions and over time, and there is hardly

any region that boasts periods of strong economic performance without

strong knowledge generation. But the opposite is not true.
There are several countries and regions in the world where R&D
expenditures and education are high by any standard and have been so
for quite some time and yet economic performance is low. The exemplary
case here is Sweden, which is leading the charts in educational attainment
and quality as well as in R&D expenditures per capita. This is one piece
of empirical evidence that does not fit well into the endogenous growth
framework and has been referred to as the Swedish or European Paradox.
After over a decade of poor economic performance, Japan may now also
be put in the list of countries where knowledge creation clearly turned out
to be no guarantee for growth. Therefore, despite the great improvement
that new growth models represent over the earlier neoclassical growth
model, new growth theory fails to capture the full complexity of the inno-
vation process. As we will argue below, the common assumption that
knowledge creation and commercialization are effectively a simultane-
ous process is particularly harmful to this line of research. Relaxing that
assumption may help explain the above-mentioned European Paradox as
well as provide useful points of entry for national and regional innovation
policy.
In contrast to new growth theory, the systems of innovation frame-
works are very detailed with respect to the different stages and agents
involved in the innovation process. The idea behind the innovation
systems approach is quite simple but extremely appealing. According to
this approach, innovation is a result of a complex process but this process
is shaped in a systemic manner. The elements of the system are innovat-
ing firms and firms in related and connected industries (suppliers, buyers),
private and public research laboratories, universities, supporting business
services (like legal or technical services), financial institutions (especially
venture capital) and the government. These elements are interconnected
by innovation-related linkages where these linkages represent knowl-
edge flows among them. Linkages can be informal in nature (occasional
meetings in conferences, social events etc.) or they can also be definitely
formal (contracted research, collaborative product development etc.).
The effectiveness (i.e., productivity in terms of number of innovations) of
the system is determined by both the knowledge already accumulated by
the actors and the level of their interconnectedness (i.e., the intensity of
knowledge flows). But behavior in these frameworks is typically assumed
to be predetermined and mechanical. The ability and motivations for
interactions are shaped largely by traditions, social norms, values and the
countries legal systems and do not respond much to economic incentives
and/or such behavioral responses are modeled as simple heuristics rather

than rational behavior. Moreover, the systems-of-innovation frameworks

have little to say about aggregate innovation rates and macroeconomic
growth.
Both new growth theory and the systems of innovation approach
typically also ignore the spatial dimension, although it arguably has been
introduced into the systems of innovation frameworks in recently devel-
oped regional innovation systems studies (Braczyk et al., 1998).
Finally, new economic geography models investigate general equilib-
rium in a spatial setting (Krugman, 1991). This means that they provide
explanations not only for the determination of equilibrium prices, incomes
and quantities in each market but also the development of the particular
geographical structure of the economy. In other words, new economic
geography derives economic and spatial equilibrium simultaneously
(Fujita et al., 1999; Fujita and Thisse, 2002). Spatial equilibrium arises
as an outcome of the balance between centripetal forces working towards
agglomeration (such as increasing returns to scale, industrial demand and
localized knowledge spillovers) and centrifugal forces promoting disper-
sion (such as transportation costs). Until recently, new economic geog-
raphy models did not consider the spatial aspects of economic growth.
However, even the recent models of technological change follow the
pattern as endogenous growth models and abstract from the complexity
inherent in innovation systems studies.
As emphasized by Acs and Varga (2002), each one of the above three
approaches has its strengths and weaknesses but they could serve to create
the building blocks of an explanatory framework of technology-led eco-
nomic growth at the regional level. In this chapter we outline a model in
the endogenous growth modeling tradition (in fact, an adaptation of the
Romer, 1990 model) that adds an essential element in the complexity of
the innovative process by separating the decision to create knowledge
from the decision to commercialize an opportunity. This is what we feel
provides the missing link between the three literatures described above:
entrepreneurship. We define entrepreneurship as the act of commercial-
izing new knowledge.3 And we show that separating knowledge creation
from knowledge commercialization implies that knowledge creation alone
is a necessary but not a sufficient condition to generate growth.
Our model gets a strong regional flavor when we realize that the
knowledge spillover from creators to commercializers necessarily involves
intense communication, as a lot of the knowledge is new, uncodifiable and
tacit. The costs of communicating such knowledge over large geographical
distances are large. In our model the knowledge spillover structure implies
that entrepreneurs will localize in regions with high levels of R&D and
knowledge creation in related industries. The vast empirical literature on

geographical clusters can be read to support the notion that such knowl-
edge spillovers are generally localized.
The combination of these two adaptations to the original Romer model
implies that regions with strong knowledge creation will experience high
growth only when there are enough entrepreneurial resources around to
get the new knowledge to the market. Our model then generates a testable
hypothesis. Growth should arise in those regions and countries that have
strong knowledge creation and entrepreneurship, whereas knowledge
creation without entrepreneurship and entrepreneurship without knowl-
edge creation generate less growth for the same levels of effort. Our model
predicts that both R&D and entrepreneurial activity will positively affect
growth in a region, but the marginal impact of either depends strongly on
the presence of the other.
Space and time do not permit us to test our hypotheses empirically in
this chapter, but the evidence in the literature is suggestive. The empiri-
cal evidence on the importance of entrepreneurship in explaining pat-
terns of regional growth and development has recently been surveyed in
Versloot and van Praag (2007) and Acs and Varga (2005) present results
for European country data and also find a positive and statistically sig-
nificant interaction effect between R&D and entrepreneurship. Audretsch
and Keilbach (2005) ran a test on German regions showing that indeed a
positive interaction exists at the regional level.
Based on the above-presented empirical results in the literature, however,
we feel it is worthwhile to dwell on the policy implications of our analysis.
There is a case to be made for R&D subsidies, as knowledge generation
has positive externalities, but given the current set of innovation policies in
place in most countries and regions, policy-makers should carefully identify
the bottleneck in the innovation chain from the original knowledge creation
to the final marketing of new products and services before committing more
public resources to the generation of more new knowledge. To assume that
new knowledge will create economic growth automatically and costlessly
is similar to dropping seeds on rock. Only a fraction of the thus developed
valuable knowledge will actually take root. In some regions more knowl-
edge investments may be needed, but we would argue that in most regions
more entrepreneurial resources would improve the effectiveness of existing
knowledge creation and generate more economic growth for the tax dollar.
We now proceed with a general outline of a growth model, where we
zoom in on the innovative process and present the formal economic deci-
sion problems for knowledge creators and commercializers. From that
analysis we can derive our main propositions in an a-spatial context.
After that we discuss the implications of taking into account spatial
considerations in the model.

OUTLINING A MODEL OF ENTREPRENEURIAL

RENTS AND GROWTH
Consider a three-sector, two-factor economy in which consumers consume,
save to accumulate raw capital and supply their labor exogenously. Final
goods producers produce consumption goods, using labor, intermediate
capital goods and invest in R&D (labor) to improve their productivity
and reduce production costs. And finally, intermediate goods producers
supply them with an expanding variety of intermediate goods that are
produced with raw capital, obtained in capital markets and introducing
a new variety requires the input of labor resources. The financial flows in
our model can then be illustrated in Figure 8.1, where the arrows represent
real money flows in terms of the final good that is the numeraire. Below
we introduce our notation and give the exact definition of the arrows.
Then we shortly discuss the actors and discuss what problem they solve
under what constraints. Finally we discuss how the markets in the model
equilibrate.
Consumers
With numbers referring to the arrows in the Figure 8.1, consumers have
two outgoing and two incoming flows:
(1) consumption of C (at price P=1);

(2) savings YC, which are invested in bonds, B, yielding interest rate r;
(1)
Consumers of final Producers of final
good C good C
(4) (8) (9)

(5) (6)
Labor Market (7)
(13)
(3) (10) Producers of n

Entry
Capital Market intermediate goods
(11)
(2)
(12)
Figure 8.1 Financial flows in the model

(3) interest income rB;

(4) labor income wL* where total labor is supplied inelastically and nor-
malized to 1.
Consumers in our model receive interest and labor income every period
and spend their income on consumption and the purchase of new bonds.
They maximize a standard log-linear utility function and face a standard
budget constraint:
max :U 5 e0`log [ Ct ] dt
Ct
#
s.t.: Bt 5 wtL* 1 rtBt 2 Ct
where U is the utility index and a dot over the variable signifies a time
derivative. Consumers then choose consumption (and implicitly savings)
in every period following the Ramsey rule, such that a constant fraction
of income is saved when the interest rate is constant and exceeds the rate
of time preference:
#
Ct/Ct 5 rt 2 r
Final Goods Producers
Total consumption equals the sales and production of final goods produ-
cers as we assume the market for final goods clears. The next four arrows
in Figure 8.1 then relate to the behavior of final goods producers, who are
price takers in factor and output markets:
(5) production wages wLP where LP is labor employed in production;

costs of n intermediate goods e0 c (i) x (i) di bought at price c(i) and in
n
(6)
quantity x(i);
(7) R&D wages wLR where LR is labor employed in R&D;
(8) investment in R&D (equal to labor costs) financed by issuing new
bonds;
(9) interest payments on issued bonds;
For the final goods producers to have an incentive to do R&D we intro-

duce the firm-specific factor knowledge, A, into their production function
and specify the process by which they can increase that knowledge stock.
Firms then choose the optimal levels of production and R&D employment
and their use of intermediate goods at every point in time. The knowledge
stock makes their problem an inter-temporal one that is given by:

` ` n
max : 3 e2rtP 5 3 e2rt aC 2 w (LP 1 LR) 2 3 c (i) x (i) dib

LP, x (i) , LR
0 0 0
n
s.t.: C 5 Y 5 AaLbP 3 x (i) 1 2a 2b

0
#
A 5 yAgn1 2gLR
where we have dropped the time arguments to save on notation and y is

a scaling parameter. Note that we have assumed that R&D in the final
goods sector receives a positive knowledge spillover from past R&D, A,
and the variety in intermediates, n. This reflects the assumption that it is
easier to do R&D from an already large knowledge base and it is easier
to increase productivity in the final goods sector when a lot of different
specialized intermediates are available. By the assumed symmetry and
constant returns to scale specification we can study the behavior of a rep-
resentative firm and solve the above problem.4 The demand for production
labor and individual intermediate variety i are given by:
bY
LDP 5
wP
21
c (i) a 1b
xD (i) 5 n 1 2a 2b
2 a 1b
(1 2 a 2 b) Y
a c (i)
i50
And the final goods sector will employ R&D workers as long as the wage
is below:5
aYy (A/n) 2g
wR 5
(r 2 w# R /wR 1 gn# /n)
Intermediate Goods Producers
(10) rental costs of the raw capital used in producing intermediate goods
and financed with bonds, rK;
(11) dividends on ownership shares and/or interest on loans equal to the
expected value of rents E0 [ e0 p (i) di ] at entry, to finance start-up
n
investments that equal the wages (forgone) by the entrepreneur.

(12) investment in entry (equal to labor costs) financed by issuing stocks

and/or bonds by entrants;
(13) labor costs of entry in intermediate sector financed by issuing stock
or bonds, wLE.
Intermediate producers are assumed to be monopolists and set prices to

maximize their profits. A simple production technology that converts one
unit of raw capital into a unit of the intermediate variety, completes the
problem for the intermediate producer:
max: p (i) 5 c (i) x (i) 2 rK (i)

c (i )
s.t.: x (i) 5 K (i) 5 xD (i)
The producer will set his price as a mark-up over marginal costs:
r
c (i) 5
12a2b
And as marginal costs are equal for all varieties, all varieties are priced and
consequently used at the same level. Given that producing a new variety
yields positive profits, new entrants have an incentive to commercialize
ideas for new varieties. We assume that this commercialization process is
costly in terms of labor and specify the entry process as:
#
n 5 ALE
where we have assumed that new variety creation is proportional to the

stock of accumulated R&D knowledge in final goods production. This
reflects our assumption that R&D in final goods production generates a
lot of knowledge spillovers in the form of new ideas and opportunities for
intermediate goods. Below we will discuss why such spillovers are gener-
ally local and will create the emergence of clusters of related economic
activity. In the model we have no geography and entry takes place as long
as the wage is below:
(a 1 b) (1 2 a 2 b) (A/n) Y
wE 5 # #
r 1 n/n 2 Y/Y
To finance the labor costs of entry, new entrants issue stock or bonds
and in equilibrium the profits from production are exactly equal to the
interest payments on bonds plus the returns on stock.

Equilibrium and Propositions
The model equilibrates when all flows into and out of the boxes add up
to 0 (which means agents solve their maximization problems given their
constraints and do not leave any resources idle) and prices equilibrate the
supply and demand on the two factor markets. It can be shown that:
Proposition 1: In equilibrium, growth in income can only come from

productivity gains in final goods production and variety expansion in
intermediates.
Proposition 1A: Optimal growth requires the stimulation of both R&D
and entrepreneurship.
Proposition 1B: The positive external effect that would justify a subsidy
on R&D, only materializes when entrepreneurs are present and able to
commercialize that knowledge, so any policy that helps R&D but hurts
entrepreneurship is therefore counterproductive.6
Proposition 1C: The model predicts that R&D and entrepreneurship both
contribute to economic growth. The latter mainly operates through a posi-
tive interaction effect between downstream R&D and upstream product
innovation through entry. The former operates through a direct effect of
R&D on process innovation.
Without going through the mathematical details we can derive these

main propositions by considering first the motives for innovative activity
of entrepreneurs and R&D workers in the model. Recall that intermediate
producers use a simple one-for-one technology to create their intermedi-
ates from raw, homogenous capital. As in Romer (1990) the Constant
Elasticity of Substitution (CES) production function at the final goods
production stage implies that the intermediate varieties are imperfect
substitutes in final goods production and thus a latent demand for all new
varieties exists. The monopolists in the intermediate sector earn monopoly
rents, creating an incentive for entry. Patent protection on existing inter-
mediates might be assumed to prevent entry in the existing intermedi-
ate markets and leave entry with a new intermediate variety as the only
alternative.7 Instead, one might also assume that the entrepreneur has and
retains exclusive knowledge regarding his venture and competitors can
never enter with perfect substitutes and drive profits to 0. 8 This implies
that entry can only take place with new varieties that are imperfect substi-
tutes. Romer (1990) then assumes that a specialized R&D sector generates
the blueprints for a new intermediate good and auctions them off to a
competitive fringe of potential entrants. The downstream rents thus moti-
vate and finance an R&D sector that generates new ideas. The only barrier

to entry in Romer is the possession of a blueprint and therefore the R&D

firm, the knowledge creator, appropriates the full discounted rents in equi-
librium. In Romer the entrepreneurial opportunity is created as a private
property and commercialization is costless and automatic.
We have assumed instead that new firm entry is costly and risky and
we follow Schumpeter ([1911] 1934)) in assuming that entrepreneurial
opportunities are pure and costless spillovers. Therefore, the associated
rents are fully appropriated by the entrepreneur, leaving the one that com-
mercializes knowledge as the residual claimant to the monopoly rents.9 It
implies that the rents from commercialization are no longer available to
finance knowledge creation and no independent R&D sector as in Romer
(1990) can exist.
However, knowledge needs to be created somewhere and for a clear
economic purpose if we wish to avoid a return to the neoclassical manna
from heaven growth models. Large amounts of investment in corporate
R&D also suggest that knowledge creation is somehow profitable to the
firms undertaking it. We would argue, however, that the improvement
and more efficient production of existing products is the stated aim of the
corporate R&D labs we see in the world today. Not the generation and
subsequent auctioning off of blueprints for new (intermediate) goods as
in Romer (1990). To make the generation of knowledge profitable to final
goods producers in our model, they cannot operate under perfect competi-
tion with constant returns to scale in intermediates and labor as in Romer
(1990). We assumed instead that the production function has constant
returns to three factors of production: labor; an aggregate of intermedi-
ates, that is, capital; and a stock of private production knowledge.
Price taking on the demand side in labor and intermediate markets then
implies that all firms have operating profits as wage and intermediates
costs do not exhaust sales. This profit is the return to the firm-specific
knowledge stock. We assumed that it needs to be accumulated prior to
production so a new entrant must first accumulate one for himself. Free
entry will therefore not eliminate the operating profits. The stock of
production knowledge can be augmented every period by doing R&D.
Profit-maximizing firms then choose a level of R&D labor that equates
the discounted future value of additional operating profits to the marginal
wage costs of their R&D workers.
It can be shown that in equilibrium all firms will have the same level
of production knowledge and R&D investment.10 Our structure makes
the intended outcome of R&D, efficiency gains to the firm, a pure private
good of which the returns can be fully appropriated.11 However, we also
assumed that the R&D generates an accidental by-product; knowledge
and opportunities that the final goods-producing firm does not find

profitable to commercialize. And we assume that firms cannot prevent the

spillover of that knowledge.12
The ideas for new intermediate goods are therefore a costless knowledge
spillover from incumbent firms R&D. Ongoing R&D in incumbent firms
generates a flow of ideas, some of which are shelved by the incumbents for
whatever reasons but then commercialized through new firm entry. The
markets for labor and capital connect all activities and close our model.
So far we have ignored the spatial dimension of our analysis and
provide a space-free aggregate macroeconomic model of innovation-
driven growth. Starting from this basic framework one can, however,
also develop the proposition that innovative activities will tend to cluster
spatially. To do so we consider the introduction of limited labor mobility,
transport costs and communication costs in our framework.
REGIONAL KNOWLEDGE SPILLOVERS
In our model, R&D generates an externality in the form of knowledge

spillovers to the entrepreneurial sector. For example, Acs and Varga
(2005) found empirical support for the claim that knowledge spillovers
between knowledge creators and knowledge commercializers are geo-
graphically localized. There are several ways in which our model might be
amended to explain that result.
A first and obvious one is the mobility of labor. If labor is more mobile
between R&D and entrepreneurship than it is among regions, then the
knowledge an individual has developed in the R&D lab presents an oppor-
tunity to leave the lab and start an intermediate firm. When geographic
labor mobility is low it is likely that such a venture is set up in the same
region. The mechanisms underlying this channel for knowledge spillovers,
the phenomena of spin-off and spin-out, have been studied intensively by
Klepper (2006), for example. He has found this conduit for knowledge
transfer to be highly relevant in, for example, the early automobile indus-
try and the tire industry around Detroit. We formulate Proposition 2:
Proposition 2: If we assume that knowledge spillovers require labor mobil-

ity between entrepreneurship and R&D, we can derive that knowledge
spillovers between R&D labs and entrants operate in particular within
regions with high labor mobility and are less important over longer geo-
graphical distances. Hence innovative activities will cluster geographically.
Second, we may introduce an agglomeration effect in the knowledge

commercialization function. In the presence of related activities, a regions

infrastructure is more likely to be adapted to the needs of a particular new

venture. It is also reasonable to assume that a new intermediate firm is
more likely to succeed when it is located in close geographical proximity to
its customers. This reduces transport costs and the costs of communicating
with the customer. In our model the customer is the final goods producer
that was the origin of the opportunity. This centripetal force is, of course,
stronger for more specialized intermediates that demand a more special-
ized and expensive infrastructure and for higher transport and transaction
costs.
Proposition 3: If we assume that infrastructural, transport and transac-

tion costs are large and increasing in the geographical distance between
intermediate firms and final goods producers, then intermediate firms will
locate close to each other and the final goods producer that generated the
opportunities on which the new ventures are founded. Hence innovative
activities will cluster geographically.
Third, we might introduce a parameter in our model to reflect the costs

of communicating the knowledge between creators and commercializ-
ers. This cost parameter would reflect language barriers, cultural differ-
ences, lack of appropriate expertise to appraise the opportunity and so
on, but would also capture institutional barriers to knowledge spillover.
Geographical proximity and agglomeration variables will then proxy in
empirical specifications for the importance of such knowledge transfer
costs, as, for example, in Varga (2006). This extension to the model, as
opposed to the ones above, also explains political border effects. Borders
between countries, ethnic communities, cultural identities but also scien-
tific disciplines will reduce the spillover and hence the introduction of such
communication costs explains the agglomeration and clustering of inno-
vative activities within countries, jurisdictions and cultural communities.
Proposition 4: If we assume that communication is costly and increasingly

so when we cross cultural boundaries, political borders, legal jurisdictions,
language barriers, scientific disciplines and common knowledge pools,
then the knowledge spillover theory of entrepreneurship predicts the
clustering of innovative activities. To the extent that these socio-cultural
distances are proxied by geographical distance, the model predicts geo-
graphical clustering of innovative activities.
The three propositions presented in this section all predict geographi-

cal clustering of innovative activity but identify very different channels
through which knowledge spillovers would cause that clustering. And this

list of channels is by no means claimed to be exhaustive. Understanding

the (relative) importance of the various channels is essential in formulat-
ing a suitable innovation policy at the aggregate, the regional and the local
level.
POLICY IMPLICATIONS AND CONCLUDING

REMARKS
Our model formalized the knowledge spillover theory of entrepreneur-
ship (Acs et al., 2009) and placed the entrepreneur as the missing link
between knowledge creation and economic growth (Branuerhjelm et
al., 2010) in an integrated general equilibrium framework and combined
modeling techniques developed in the growth literature with insights
from the more descriptive and empirical literature on innovation and
entrepreneurship.
Moreover, in accordance with the evidence, in our model the entrepre-
neur is the agent with whom the buck stops. The creation of the knowl-
edge he commercializes is not (necessarily) motivated by the rents that the
entrepreneur receives for commercialization. Such rents reward the act of
commercialization and as such should not be destroyed to enhance static
efficiency (a result we share with all innovation-driven growth models).
But the claim to these rents should also not be transferred to the genera-
tors of knowledge that may have had no intent of commercializing and/
or require no incentive to create such knowledge in the first place.13 Not
the generation but the implementation of new knowledge is what is valu-
able to society at large and what generates economic growth. Knowledge
creation, of course, is a necessary but insufficient condition for innovation
and growth, and creation without implementation is clearly a waste of
resources. We argue, therefore, that policy-makers should stop and think
about where the bottlenecks in the innovative process are, before commit-
ting large amounts of public money and/or entitlements to profits and rent
to the (formal) knowledge generation process.
These results also carry over to the regional level if we consider the
impact of limited geographical labor mobility, transport costs and com-
munication costs. As the knowledge spillovers that drive economic growth
are likely to be regionalized, regional policies should aim to smoothen
the spillover. A first requirement is that sufficient resources are available
for both knowledge creation and knowledge commercialization. And as
entrepreneurial talent is a key resource in the innovation chain, regional
policies should aim to develop it. Moreover, the impediments to knowl-
edge spillovers from creators to commercializers deserve attention. Legal

impediments such as non-competition clauses in labor contracts can easily

be abandoned. But by investing in physical and communication infrastruc-
tures and by stimulating or enabling the exchange of knowledge, local and
regional governments can support the entire innovation chain. Direct
support to new entrants or R&D should be given only as long as that does
not reduce the incentives to create or commercialize new knowledge.
The model outlined in this chapter has important policy implications
at the aggregate and regional level but also raises important questions to
be addressed. The presence of knowledge spillovers is well documented
in the literature. But the exact channels through which such spillovers
arise are a challenging area for further research. Our three proposi-
tions predicting regional clustering are empirically indistinguishable in
most studies, due to data availability issues. The detailed case studies by
Klepper (1996) provide support for the first channel that was identified.
The work by Florida (2002a and 2002b) presents evidence in support of
the third channel we have discussed, but to our knowledge studies that try
to distinguish between them have not yet been done. Also at the aggregate
level, our theory and its underlying assumptions require further empirical
scrutiny. We feel the available evidence supports our claim that knowledge
spillovers are important for (regional) economic growth but a lot more can
be done to test the model predictions. This empirical research agenda will
hopefully inspire other researchers to join us in pursuing it.
NOTES
1. The part that is not done by incumbent firms is done by publicly funded institutions
such as universities or specialized R&D firms as Romer (1990) envisioned them.
2. See Acs (2009) for a detailed and referenced exposition of this framework.
3 There are many definitions of entrepreneurship in the literature. We follow the
Schumpeterian tradition. See, for example, Braunerhjelm (2008) for a recent overview
of this literature.
4. Even if the knowledge stock is allowed to differ among final goods producers it can be
shown that only those that have A=Amax will employ R&D workers and increase their A
such that the firms with lower knowledge stocks will diminish.
5 There is a horizontal demand curve for R&D labor due to the assumed linearity in
R&D labor in the innovation function.
6. See Acs and Sanders (2008) for an application to the role of patent protection in this
context.
7. So far we copied the Romer (1990) structure. In Aghion and Howitt (1992) these
entrants drive out incumbents with a higher-quality version of existing varieties and
entry leads to average-quality improvement not to variety expansion. To keep the
model manageable we follow Romer (1990) here and focus on variety expansion but
we also feel that entrepreneurial entry may introduce improved versions of existing
products.
8. Patent protection is problematic in this model as we assume that the knowledge creator
is not the same agent as the knowledge commercializer. Patents are generally awarded

to the knowledge creator. There is a large literature (Acs, 2008) that stresses the
importance of the individual entrepreneur for the success of new ventures. His unique
combination of cultural background, skills, knowledge, access to finance and other key
resources and not least important, luck, makes it unlikely that any other entrant could
enter the same market and drive down profits to 0 by copying the incumbent.
9. Although in expectation terms the profits flow back to the consumers in Figure 8.1,
the entrepreneur, more often than not, is that specific consumer and the profits are his
expected returns on forgoing labor earnings during the entry stage. We have modeled
this as the entrepreneur issuing stock and bonds to finance his wage costs to reflect the
fact that they take such opportunity costs into account and desire a market-determined
return on their investment.
10. This follows intuitively from the assumption that all final goods-producing firms are
equal, face the same maximization problem, production possibilities, final demand
curve and set of input prices.
11. We discuss the precise set of assumptions we need to make for this result in Acs and
Sanders (2008).
12. And if there are such impediments, that would be a first target for policy. Braunerhjelm
et al. (2010) refer to the knowledge filter when they discuss the physical, cultural, politi-
cal and institutional barriers to such knowledge spillovers.
13. As, we have argued in Acs and Sanders (2008), may well be the effect of stronger patent
and IPR protection.
REFERENCES
Acs, Z.J. (2008), Foundations of High Impact Entrepreneurship, Foundations and

Trends, 4(6), 535620.
Acs, Z. (2009), Jaffe-Feldman-Varga: The Search for Knowledge Spillovers, in
A. Varga (ed.), Universities, Knowledge Transfer and Regional Development:
Geography, Entrepreneurship and Policy, Edward Elgar, Cheltenham, UK and
Northampton, MA, USA.
Acs, Z.J. and Varga, A. (2002), Geography, Endogenous Growth and Innovation,
International Regional Science Review, 25(1), 13248.
Acs, Z.J. and Varga, A. (2005), Entrepreneurship, Agglomeration and
Technological Change, Small Business Economics, 24(3), 32334.
Acs, Z.J. and Sanders, M. (2008), Intellectual Property Rights and the Knowledge
Spillover Theory of Entrepreneurship, Working Papers No. 08-23, Utrecht
School of Economics.
Acs, Z.J., Braunerhjelm, P., Audretsch, D. and Carlsson, B. (2009), The Knowledge
Spillover Theory of Entrepreneurship, Small Business Economics, 32(1), 1530.
Aghion, P. and Howitt, P. (1992), A Model of Growth through Creative
Destruction, Econometrica, 60(2), 32351.
Arrow, K.J. (1962), Economic Welfare and the Allocation of Resources for
Invention, in R.R. Nelson (ed.), The Rate and Direction of Inventive Activity,
Princeton, NJ: Princeton University Press, pp. 60926.
Audretsch, D.B. and Keilbach, H. (2005), Entrepreneurship capital and regional
growth, The Annals of Regional Science, 39(1), 45769.
Audretsch, D. and Lehmann, E. (2006), Does the Knowledge Spillover Theory of
Entrepreneurship hold for Regions, Research Policy, 34(8), 11912002.
Braczyk, H., Cooke, P. and Heidenreich, M. (1998), Regional Innovation Systems:
The Role of Governances in a Globalized World, London: UCL Press.

Braunerhjelm, P. (2008), Entrepreneurship, Knowledge, and Economic Growth,

Foundations and Trends in Entrepreneurship, 4(5), 451533.
Braunerhjelm, P., Acs, Z.J., Audretsch, D.B. and Carlsson, B. (2010), The
Missing Link: Knowledge Diffusion and Entrepreneurship in Endogenous
Growth, Small Business Economics, 34(2), 10525.
Florida, R. (2002a), Bohemia and Economic Geography, Journal of Economic
Geograhy, 2(1), 5571.
Florida, R. (2002b), The Economic Geography of Talent, Annals of the Association
of American Geographers, 92(4), 74355.
Fujita, M. and Thisse, J. (2002), Economics of Agglomeration. Cities, Industrial
Location, and Regional Growth, Cambridge, MA and London: Cambridge
University Press.
Fujita, M., Krugman, P. and Venables, A. (1999), The Spatial Economy,
Cambridge, MA: MIT Press.
Klepper, S. (1996), Entry, Exit, Growth and Innovation over the Product Life
Cycle, American Economic Review, 86(3), 56283.
Lucas, R. (1988). On the Mechanics of Economic Development, Journal of
Romer, P. (1986), Increasing Returns and Economic Growth, Journal of Political
Economy, 94(5), 100237.
Romer, P. (1990), Endogenous Technological Change, Journal of Political
Economy, 98(5), S71S102.
Schumpeter, J.A. [1911] (1934), The Theory of Economic Development, Cambridge,
MA: Harvard University Press.
Temple, J. (1999), The New Growth Evidence, Journal of Economic Literature,
37(1), 11256.
Varga, A. (2006), The Spatial Dimension of Innovation and Growth: Empirical
Research Methodology and Policy Analysis, European Planning Studies, 14(9),
117186.
Versloot, P.H. and Van Praag, C.M. (2007), What is the value of entrepreneur-
ship? A review of recent research, Small Business Economics, 29(4), 35182.

9. Foreign direct investment,
knowledge assets and the economic
geography of growth in the Asian
BRIICS countries
Tomokazu Arita, Chie Iguchi and
Philip McCann
INTRODUCTION
In the context of globalization, our objective in this chapter is to high-

light the role foreign direct investment plays in the growth of three of the
worlds largest newly-industrializing countries, and to examine the links
between these investment inflows and economic geography. The three
countries concerned are China, India and Indonesia.
The major features of the current phase of globalization that are
common to all regions of the world are:
increasing institutional integration between countries;

rapidly improving information and communication technologies;
rapidly increasing global investment activities (McCann, 2008).
The combination of these features produces complex interrelationships

between economic geography, economies of scale and global trade and
investment. These interrelationships have been discussed in detail else-
where (McCann, 2008, 2009; World Bank, 2009) so are not discussed
further here. However, one of the other major features of the current era of
globalization is the role played by the expansion of the global labour force.
This rapid expansion is dominated by six major newly-industrializing
economies, the so-called BRIICS countries, comprising Brazil, Russia,
India, Indonesia, China and South Africa (McCann, 2009). The open-
ing-up of these countries between 1989 and 1993 transformed the global
labour market in that they brought almost 1.5 billion new workers into the
global labour force.
160

Growth in the Asian BRIICS countries 161
Asia is often seen as the focus of the driving forces of this period of
globalization. The reason is that the three largest of these enormous newly
industrializing countries, namely China, India and Indonesia, are within
the South and East Asian economic arena. However, the performance of
these different countries in terms of global and regional output and the
role played by these economies is a complex picture. Figure 9.1 helps us to
visualize these differences. In order to understand the relative scale effects
of the Asian economies, we also include the two Australasian economies
of Australia and New Zealand as comparator cases.
If we consider the growth performance of these Asian countries during
the current phase of globalization, as we see from Figure 9.1 it is clear that
China and India are currently the two fastest-growing large economies
in the world, with 2005 growth rates of 10.2 per cent and 9.2 per cent
respectively. Chinas gross national income in 2005 was US$2269.7 billion,
which ranked it as the worlds fifth largest economy in 2005, and almost
identical in size to the UK economy; Indias economy in 2005 was some
US$804.4bn, and ranked as the tenth largest economy in the world, while
Indonesia, with an economy of some US$282.2bn in 2005 is ranked at
23rd. Yet scale and affluence are different things. If we consider per capita
income rather than the gross income, China, with a per capita income of
US$1740 is ranked 128th in the world, Indonesia with a per capita income
of US$1280 is ranked 139th, while India with a per capita income of
US$730 is ranked 158th in the world. As such, the potential for growth in
each of these countries is still enormous.
Economic growth in each of each of these countries has been associ-
ated with deregulation and liberalization, and the massive internal real-
location of production factors between sectors and regions. However, the
link between these internal growth processes and the broader processes
of globalization that are taking place is via the enormous inflows of
foreign direct investment (FDI) from the advanced economies into these
countries. Yet, inward FDI exhibits different levels of relative importance
in different host economies. Amongst developing countries, the trans-
nationality index of openness, which indicates the scale of inward mul-
tinational investment in terms of FDI inflows, stocks, value-added and
employment, relative to total GDP, ranks China 32nd, India 36th, and
Indonesia 38th. In general, across all developing economies, the overall
relative trans-nationality openness of countries to FDI tends to be higher
in small countries and lower in the larger economies, although Indonesia is
relatively more closed than its scale might suggest, while China is relatively
more open than its scale would suggest.
FDI will, however, continue to flow in very large quantities from devel-
oped countries into these countries over the foreseeable future. Recent

Figure 9.1 The size, wealth and growth of the Asian and Australasian
economies in 2006

UNCTAD survey evidence of multinational executives regarding the most

attractive locations for FDI over the coming years, finds that China is
ranked as the number 1 country in the world and India is ranked number
2. The investment inflows resulting from these perceptions will ensure that
these newly-industrializing countries will become increasingly open and
integrated into the global economic system. Yet, although the relative
levels of FDI currently appear to be small in these countries, multinational
inward investment plays an extremely important role in the economic
growth of these countries. In particular, as this chapter will demonstrate,
inward investment plays a crucial role in determining both the techno-
logical and geographical distribution of activities in each of these countries
during the current stages of economic growth.
The rest of the chapter is organized as follows. In the next section we
discuss the nature and scale of inward foreign direct investment in these
three countries, with a particular focus on the rise of knowledge-related
investments. In what follows we examine the spatial patterns of inward
investment in these countries and we relate these emerging patterns to
the economic geography of each country, and in particular, to the role of
agglomeration economies in each society. The final section provides some
brief conclusions. Unless otherwise stated, the empirical evidence in the
following sections comes from UNCTAD (2005, 2007) and the World
Bank (2009).
INWARD FOREIGN DIRECT INVESTMENT AND

R&D IN THE ASIAN BRIICS ECONOMIES
The dominant modes of FDI entry in developed economies are mergers

and acquisitions. In contrast, in newly-industrializing countries such as
China, India and Indonesia, the dominant modes of FDI entry are so-
called greenfield projects. Greenfield FDI projects are the foreign affiliate
investments whereby a brand new establishment is constructed on a new
site, and this mode of FDI obviously represents a very different form of
FDI from those undertaken by M&As. The number of such greenfield
FDI projects increased globally by 13 per cent to some 11 800 projects in
2005 (McCann, 2009). Manufacturing accounted for 54 per cent of these
projects, with the service sector accounting for 42 per cent and primary
industries accounting for 4 per cent. In terms of broad regions, South,
East and South East Asia accounted for 3515, or some 30 per cent of
these greenfield projects. Figure 9.2 gives us a sense of the relative order
of magnitude of these inward greenfield FDI projects in Asian economies
in 2005 in comparison to advanced economies. China alone accounted for

Figure 9.2 The number of greenfield inward FDI projects by country in

2005
1378 greenfield FDI projects, or 11.6 per cent of the global total. India
accounted for 981 greenfield projects, representing 8.3 per cent of the
global total, while Indonesia accounted for 93 projects (UNCTAD, 2007).
In order to give a sense of the relative scale of these numbers, in the same
year, as we see in Figure 9.2, the number of greenfield inward FDI projects
in the US was 723, UK 669, France 582. Asian economies, and in particular
China and India, are by far the most important locations in the world for
greenfield FDI projects (McCann, 2009). Yet as we see in Figure 9.2, apart
from China, the order of magnitude of these greenfield project numbers is
still roughly comparable to many of the large advanced economies. China
now has by far the largest number of domestically located multinational
foreign establishments with 42 753 foreign affiliates in 2004 with some 24
million employees. These affiliate establishments are heavily concentrated
in manufacturing. The 24 million employed in China in foreign affiliates
represents one-third of the global total workforce currently employed in
foreign affiliates. This number has increased fivefold from less than five
million in 1991, a number that is equivalent to the current total level of
domestic US employment in foreign affiliates (McCann, 2009).

In the case of Indonesia, a substantial phase of deregulation started

in 1986, and from the mid-1980s onward employment in foreign-owned
manufacturing increased rapidly. These increases were concentrated in the
machinery industries and in multinational firms with large foreign owner-
ship shares (Takii and Ramstetter, 2005). Inward foreign direct invest-
ment arrived primarily from North East Asia (ibid.), with Japan being
the dominant source of inward FDI (Syamwil and Tanimura, 2000). This
was just at the time that Indonesia needed to restructure its manufacturing
sector to be more labour-intensive and export-orientated (Timmer, 2004).
The initial 1986 deregulations regarding foreign investment were then
further bolstered by more fundamental reforms in 1989, 1992 and 1994.
Rapid growth and industrialization were both causes and results of even
larger increases in inflows of FDI during the economic boom of the late
1980s and early 1990s (Takii and Ramstetter, 2005), and then beyond the
economic crisis of 1997 (Sjoholm, 1997). The FDI policies in Indonesia are
now as liberal as they were in the late 1960s, and are largely in line with
most other countries. Although official estimates suggest that there was an
abrupt reversal of these flows in 199899 through to 2001, on the basis of
industrial and employment survey data, however, Takii and Ramstetter
(2005) find that the presence of multinational firms in Indonesia continued
through the crisis of 199798 and beyond at a modest rate both in terms
of employment and value-added. Aggregate employment in multinational
firms in Indonesia increased by some 20 per cent between 199597 and
200001, with the multinational firms shares accounted for by large plants
also increasing from 19 per cent to 21 per cent during the same period
(ibid.).
The knowledge base of the Asian BRIICS countries is also steadily
increasing, and the possibilities for upgrading R&D and innovation in
these countries over the long term are obvious given the sheer scale of the
countries. For example, in 2001, China and India together accounted for
close to one-third of the total global number of tertiary educated technical
people (McCann, 2009), while the Bangalore high-technology industries
alone have 35 000 people who are US educated or trained. Yet, of all
developing economies, it is Chinas growth in its R&D capacity that has
been the most remarkable (ibid.). Between 1996 and 2003 China increased
its domestic R&D expenditure by over 3.8 times, such that by 2002 China
became the only newly-industrializing economy that is amongst the
worlds top ten R&D expenditure countries. China is now ranked number
one for both total R&D and also business R&D expenditure, and India is
number four. After the US and UK, China is now third in the world for
the total number R&D-related foreign affiliates located there. In 1996 it
was outside of the top ten, but by 2002 it was ranked sixth in the world in

terms of total R&D expenditure and seventh in the world in terms of busi-
ness R&D expenditure (ibid.).
Amongst developing regions, it is the countries in South and East Asia
that are the major locations for multinational R&D investment. During
200204, of the 1773 inward FDI projects involving an R&D component,
1095 (62 per cent) were undertaken in developing or transition economies,
of which 861 projects (49 per cent) were undertaken in developing Asia
alone. In the case of developing Asia, the share of R&D accounted for by
the foreign affiliates of US multinational firms increased from 3 per cent
of their total foreign located R&D in 1994 to 10 per cent in 2002. Similar
trends are also observable for multinational firms from other developed
economies that are locating R&D-related investments in Asia (ibid.).
Within developing Asia itself, it is China in particular that dominates
inflows of multinational R&D investment, and the impact on China of
these inflows has also been the most marked. Between 1998 and 2002, the
share of total domestic business R&D in China accounted for by foreign
affiliates located there increased from 18 per cent to 22 per cent. The R&D
expenditure associated with this R&D-related inward FDI now accounts
for 13.5 per cent of Chinas total domestic public sector plus private sector
R&D expenditure (ibid.). To get a sense of how important these multi-
national R&D investments are to China, we can observe that the 42 000
foreign affiliates located in China are currently employing some 24 million
Chinese, and this still only represents 3.1 per cent of the total employed
workforce in China. As such, the relative importance of multinational
R&D expenditure to Chinas knowledge-related activities is four times
greater than the relative importance of multinational firms to Chinas
overall employment (ibid.). In contrast, in the case of India, the most
recent reliable estimates indicate that in 1999, multinational R&D expend-
iture accounted for just 3.4 per cent of domestic private sector R&D. The
types of R&D undertaken in different countries and regions therefore
appear to vary. In India, over three-quarters of R&D expenditure is on
services, and primarily on software development, whereas in China, most
multinational R&D focuses on adaptive innovations for the Chinese
market, but there is increasing evidence across Asia that innovative R&D
is growing.
It is clear that the scale of Chinas and Indias growth in both domestic
R&D and also inward FDI-related R&D growth is very significant, and
in both cases these are very much greater than the equivalent growth rates
for Indonesia. However, from Figure 9.3 we can get a sense of the relative
global and regional contribution of Chinas knowledge sectors by con-
sidering other indicators of innovation such as patents and trademarks.
From the perspective of such knowledge indicators, the relative role

Figure 9.3 The knowledge outputs of the Asian economies 200103

played by China and India is very much less than the output growth and
FDI figures imply.
The total 2002 R&D expenditure of developing countries in South, East
and South East Asia including both China and India is only 14 per cent
of the value for Japan and 6.7 per cent of US R&D expenditure (ibid.).
Similarly, if we consider the number of US patents and trademarks granted
to the residents of particular countries during 200103, we see that China
had 1543, while India had 1022, and Indonesia 108. For comparison, the
respective figures for Taiwan and Korea are 20 414 and 12 195. Similarly,
in terms of US patents and trademarks granted to firms or organizations
of particular countries during 200103 we see that India had 558, China
475 and Indonesia 31. Once again, for comparison purposes the equivalent
figures for Taiwan and Korea are 12 606 and 11 152, respectively (ibid.).
Therefore, although amongst the newly-industrializing countries China
and India are the leading knowledge generators, their relative contribu-
tions are not large in comparison to other advanced Asian economies.
In total, the share of global R&D expenditure accounted for by all the
BRIICS countries combined is only 4 per cent, of which China accounts
for more than half of this level. In contrast, the worlds developed econo-
mies account for 94.7 per cent of global R&D expenditure (ibid.). As
such, it will still take a long time for the knowledge base of the newly-
industrializing countries to catch up with that of the advanced economies.
THE CHANGING ECONOMIC GEOGRAPHY OF THE

ASIAN BRIICS COUNTRIES
The enormous transformations that are currently taking place within the
newly-industrializing economies also have a very explicit economic geog-
raphy logic to them that is more or less common across each of the coun-
tries. In each case, national economic growth is inherently about regional
restructuring and the role of agglomeration economies is central to these
growth processes.
China
Of all developing or transition economies, the most remarkable transfor-

mation has been that of China (McCann, 2009). Between 1980 and 2000
China increased its share of global exports from 0.9 per cent to 6 per cent,
its share of global imports from 1.1 per cent to 4.1 per cent and its share
of global GDP from 2.9 per cent to 3.4 per cent (Fujita, 2007). The result
of this increasing trade and openness was that between the 1980s and

2000, Chinas GDP per capita increased by ten-fold. However, the initial
impetus for Chinas restructuring and growth came from fairly modest
reforms. China began with the introduction of a rudimentary system of
property rights in order to create incentives and only recently gave con-
stitutional recognition to private property (World Bank, 2005). However,
recognizing the need to access global capital, technology and knowledge
assets via inward multinational investment, China has subsequently also
liberalized many rules regarding services and manufacturing industry
ownership. These changes now allow for greater levels of overseas owner-
ship in many advanced sectors and have been instituted because China
is aiming to attract both a broader range and a higher quality of inward
FDI. In particular, the National Economy and Social Development Plan
2005 emphasized the need to improve the quality of FDI by encouraging it
in high-technology industries, advanced manufacturing, modern services,
agriculture and environmental protection. The plan encourages the estab-
lishment of R&D centres, regional headquarters and bases of advanced
manufacturing. It also welcomed the role of FDI in the reform of state-
owned enterprises (McCann, 2009).
The growth in R&D-related FDI investments in China began in 1993
and reached some 700 projects by 2004 amounting to some $4bn in inward
FDI. Most projects were implemented after Chinas accession to WTO
in December 2001 (ibid.). These R&D investments are mainly focused on
technology-intensive industries such as ICT, automotive and chemicals,
and there is clear economic geography logic to these investments, which
are concentrated in a small number of locations. As seen in Figure 9.4, in
2004, Beijing had 189 foreign-owned R&D centres of which 60 per cent
were in ICT, Shanghai had 140 foreign-owned R&D centres of which 91
are in Pudong, and Guangdong and Jiangsu provinces in the south close
to Hong Kong are home to a combined number of over 100 R&D centres
(ibid.).
The fact that multinational R&D centres were being located in Shanghai,
Beijing, Guangdong and Jiangsu provinces displays a clear logic. These
locations are the core knowledge regions that are growing quickly,
and form the major locations for all types of international investment.
Although they cannot yet be described fully as global cities (McCann,
2008), both Beijing and Shanghai exhibit some world city characteristics.
At the same time, in 2003 the south eastern provinces of Guangdong and
Jiangsu individually accounted for 28 per cent and 19 per cent of FDI,
respectively. The reason that FDI in general, and knowledge-related FDI
in particular, was being located in these cities and regions, is because these
are the major growth regions of China (McCann, 2009).
During the 1970s and 1980s, inequality between provinces and also

Figure 9.4 The spatial distribution of R&D centres in China in 2004
between urban and rural areas in China fell consistently (Golley, 2007).
Until the mid-1980s, the growth in per capita productivity and expendi-
ture per capita was higher in rural than in urban areas, which suggested a
slow process of ruralurban convergence (Angang et al., 2005). Between
1978 and 1985, the ratio of per capita disposable income between urban
and rural residents had fallen from 2.57 to 1.85, and ratio of per capita
consumption had fallen to just over 2.1 (ibid.). However, from the mid-
1980s onwards this urbanrural ratio has been reversed. By 1990 the ratio
of both per capita disposable income and consumption had risen to above
2.0. Since the economic reforms started in earnest, as expected on the basis

Figure 9.5 Relative urban to rural incomes in China in 2004
of the earlier arguments, inequality between provinces in China has risen

continuously since 1991 (Golley, 2007). By 1997, the urbanrural ratios
of both per capita disposable income and consumption had increased to
approximately 2.5 (Angang et al., 2005). By 2001 the urbanrural ratio of
both per capita disposable income and consumption had risen to approxi-
mately 3:1, while the ratio of per capita income had risen to 2.1:1 (ibid.).
The result of this urban rural divergence was that, as we see in Figure
9.5, by 2004 the largest city-regions in China exhibit the highest per capita
incomes, with the ratio of per capita province GDP to national per capita
GDP being highest for Shanghai 5.0, then for Beijing 3.5, then Tianjin 2.7.

The next highest ratio areas are the regions of Zhejiang 2.1 and Jiangsu
1.8, which are the regions close to Shanghai, followed by the regions close
to Hong Kong or Guangdong 1.8, and then the coastal regions close to
the dominant cities of Fujian 1.6, Liaoning 1.6 and Shandong 1.4 (Golley,
2007). These ratios imply that the dominant cities currently exhibit GDP
per capita levels that are approximately 13 times that of the lowest regions
(Fujita, 2007).
If the three major city-regions of Shanghai, Beijing and Tianjin are
removed then the increase in regional inequality across China is notice-
ably reduced. On the other hand, however, if we group together all of the
coastal regions including the dominant city-regions, then regional inequal-
ity between the coastal and interior regions of China increases even more
dramatically (Golley, 2007). This demonstrates the role played by particu-
lar city-regions in the dramatic growth of China over the last two decades.
In 2000, the coastal region between Beijing and Hong Kong as a whole
produced 71 per cent of Chinas total industrial output. This enormous
output accounted for more than 60 per cent of output in all but two sectors
and at least 80 per cent of output in close to half of the industry sectors,
including 97 per cent of Chinas cultural, educational and sports outputs
(ibid.). The growth of China is a coastal phenomenon. However, even
within the coastal region of China there is a core region, which consists of
the south east regions adjacent or close to the major cities, and represents
broadly an arc of regions bounded by Shanghai and Hong Kong. These
core region provinces grew by more than the coastal region as a whole in
almost all of the sectors in which the coastal region grew (ibid.).
As expected on the basis of earlier arguments, increasing inter-regional
inequality is now a general phenomenon in China. However, in terms of
economic geography, the escalating growth and wealth of certain regions
is also highly associated with the increasing agglomeration of activities
in these regions. Once again, this is predicted by economic geography
arguments. The core regions of the south east are not only the fastest per
capita growth regions, but also they are the regions of the most rapidly
increasing agglomeration. Golley (2007) calculates that that between 1989
and 2000, 26 out of 28 major manufacturing and industrial sectors have
become more spatially concentrated, as reflected by increasing spatial
Gini coefficients. As such, the general trend towards increasing intra-
national inequality across many countries is clearly very evident in China
(McCann, 2009).
Regional economic restructuring in China has meant that poverty
reduction since the mid-1980s has been most dramatic in the eastern
regions, followed by the central regions, with poverty increasing in the
western regions (Angang et al., 2005). However, this is not just an urban

phenomenon. The ratio of per capita farming incomes in the east and
centre regions relative to the west region have also increased between 1980
and 2000, from 1.27 and 1.05, to 1.92 and 1.30, respectively (ibid.). Part
of the reason is human capital. The areas of highest growth are broadly
the regions with highest rates of literacy (ibid.). In addition, disparities in
Chinese income per capita are also exacerbated by a fiscal tax and transfer
system that significantly benefits urban residents (ibid.). More generally,
however, the competition and wealth effects associated with buoyant
regional growth across a range of local sectors tend to spill over to other
local sectors, and agriculture in such buoyant regions also benefits from
this.
India
Many of these same regional economic and economic geography phenom-

ena evident in China are apparent in countries such as India and Indonesia
(McCann, 2009). Like China, India began its economic restructuring with
initially modest reforms by reducing trade barriers and distortions within
the economy. In 1991 the average tariff was 83 per cent, and only 13 per
cent of goods were importable without a licence. By 1998 tariffs had been
reduced to 30 per cent, and the range of goods importable without a
licence was 57 per cent. Since then, its GDP per capita has increased four-
fold between 1980 and 2002 (World Bank, 2005). As in the case of China,
recognizing the need to access global capital, technology and knowledge
assets via inward multinational investment, India has also moved to
increase its attractiveness as a location for FDI. The Indian Investment
Commission aims at drawing FDI as well as domestic investment, while
the Foreign Investment Board is intended to act a one-stop service centre
and facilitator for FDI. In 2004 foreign equity ceilings in Indian aviation
services, private banks, non-news print publications, and the petroleum
industry were all adjusted upwards in order to attract more international
investment. As with China, the process of deregulation has also been
accompanied by enormous internal rural to urban migration flows, and
the expansion of many of the already immense cities. As we see in Figure
9.6, the major growth centres within India are the mega-cities of Mumbai,
Calcutta and Delhi, as well as the high-technology centre of Bangalore.
Yet, even though China and India are now often compared, there are
actually fundamental differences between the two economies and their
responses to globalization (McCann, 2009). First, one obvious difference
is simply sheer scale. The Chinese economy is almost three times the size
of the Indian economy, with per capita incomes of well over twice those of
India. Second, there are major trade performance differences between the

Figure 9.6 The major economic growth city-regions in India
two countries. In 1950, Chinas share of global trade was 1 per cent while
that of Indias was 2.2 per cent, whereas by 2002, Chinas share of global
trade had increased to 4.8 per cent while Indias had actually declined to
0.8 per cent (Lardy, 2005). In part, these trade performance differences are
because Chinas rapid growth began slightly earlier than Indias growth,
and also because it has been more dramatic than Indias, particularly in
opening trade. Third, there are also major differences between China and
India in terms of their structure of GDP composition (McCann, 2009).
The structure of GDP composition by industry in India is still rather
different from China; there is a much greater emphasis in India on serv-
ices than China, where manufacturing is still relatively more dominant
(Panagariya, 2005). In 1980 in China, agriculture accounted for 30.1 per
cent of GDP, industry for 48.5 per cent (of which manufacturing alone

accounted for 40.5 per cent) and services for 21.4 per cent of GDP. On
the other hand, for India, agriculture accounted for 38.6 per cent of GDP,
industry for 24.2 per cent (of which manufacturing accounted for only 16.3
per cent) and services for 37.2 per cent. Even though both countries have
undergone enormous changes during the last three decades, the legacy
of these inherited structures still remains. In 2002 in China, agriculture
accounted for 15.9 per cent of GDP, industry for 50.9 per cent (of which
manufacturing alone accounted for 34.5 per cent) and services accounted
for 33.2 per cent of GDP, whereas for India, agriculture accounted for
24.9 per cent of GDP, industry for 26.9 per cent (of which manufacturing
accounted for only 15.8 per cent) and services for 48.2 per cent of 2002
GDP (ibid.).
As a result of its different industrial structure and also its English
language advantages, the growth of FDI in India, and particularly the
growth of off-shoring FDI, has been dominated by a range of service
industries, rather than by manufacturing, which has been the case for
China (McCann, 2009). Yet, many aspects of the dynamic growth indus-
tries of India have similar features to China. First, most of the trade of the
Indian and Chinese economies is still in the form of re-exports of finished
or semi-finished products or services produced by multinational firms that
are based in Europe or the US. Second, many of the key growth centres are
dominated by external links with multinational companies. In the Indian
IT industry, which is dominated by the Bangalore region, two-thirds of all
sales are accounted for by foreign-owned multinational affiliates located
there (Scheve and Slaughter, 2007). Third, as India undergoes continuing
regional economic restructuring, firms located in the regions with large
home markets earn higher profits (Kambhampati and McCann, 2007).
This implies that in terms of economic geography, the large home market
effects associated with agglomeration are driving the internal economic
growth and restructuring within the Indian economy.
Indonesia
In terms of economic geography and the regional aspects of economic

growth, a similar picture emerges in the case of Indonesia. Indonesia has
a population of between 220 and 230 million, spread over 900 islands
spanning 5500km in an eastwest direction (Amiti and Cameron, 2006),
with almost two-thirds of the population living on the island of Java.
As with both India and China, deregulation has been followed by enor-
mous rural to urban migration flows, and the major economic growth
areas are the immense cities of Java. Excluding the oil and gas sectors,
the majority of overseas investments in Indonesia have occurred within

Figure 9.7 The major economic growth city-regions in Indonesia
the non-oil-related manufacturing sectors in these same regions. These

are the very same sectors that are not only key to the trade growth of
the Indonesian economy as a whole, but are also concentrated in the
expanding urban regions of Java (see Figure 9.7).
The reasons why FDI inflows exhibit such an economic geography
logic is partly due to the fact that there are major regional disparities
across Indonesia in terms of access to international transportation infra-
structure. Jakarta is the national hub for the international air-transport
system, and many parts of Indonesia are only accessible to the outside
world via Jakarta. Similarly, there is also great disparity across Indonesia
in terms of crucial communications infrastructure. While national fixed-
line density is only around 3.7 per cent, telecommunications infrastructure
density for metropolitan Jakarta is well above 25 per cent, with other
major cities such as Bandung and Surabaya having an average telecom-
munications infrastructure density of 11 to 20 per cent (Lee and Findlay,
2005). In contrast, average telecommunications infrastructure density for

rural Indonesia is only 0.2 per cent and more than 64 per cent of rural
villages are entirely without telecommunications (ibid.). Like China,
GDP per capita in Indonesia is also closely related to city size, with the
dominant city of Java exhibiting GDP per capital levels that are approxi-
mately 13 times that of the lowest regions (Fujita, 2007). The opening up
of the country is likely to accentuate these differences, except for the case
where multinationals are engaged in primary sector activities located in
other outlying regions.
These spatial disparities in terms of population and infrastructure are
also manifested in terms of firm investment behaviour. Since the emer-
gence of the Indonesian manufacturing industry from the 1970s onwards,
the role and importance of the major urban areas of Java has increased.
Moreover, as deregulation has continued apace, the relative advantage of
these urban regions has increased even further. Notwithstanding the one-
off shock of the 199798 crisis, over the long term more and more firms
have been investing ever-increasing amounts in the urban centres of Java,
the results of which will tend to be increasing regional inequality within
Indonesia. Yet, why firms do not generally relocate from urban Java to
other parts of Indonesia in order to take advantage of the very large varia-
tions in labour prices that exist across Indonesias regions, and even within
individual islands such as Java, is a moot point (Amiti and Cameron,
2007). Although trade protectionism probably played a role in the initial
concentration of production and investment in these regions, the reasons
why such an extreme spatial concentration of investment within Indonesia
continues appears to be related primarily to the existence of agglomeration
economies.
Indonesian manufacturing exhibits extreme spatial concentration in a
small number of regions within Java (Syamwil and Tanimura, 2000). Over
the last three decades these Javan regions have consistently accounted
for over 85 per cent of total national manufacturing output in large and
medium-sized industries (ibid.), and these geographical patterns of invest-
ment and employment are similarly exhibited for both domestic and
foreign-owned manufacturing firms. As such, increasing international
trade and investment openness appears to be associated with greater
disparities among Indonesias regions, as inward investing firms seek out
the pecuniary advantages of associated agglomeration. This is particu-
larly the case in the apparel and textiles industries, which is the sector in
Indonesia that has exhibited the greatest revealed comparative advantage
since the 1980s (James, 2007). If inward technology spillover effects also
operate in these sectors, these are likely to exacerbate the existing advan-
tages of agglomeration in the dominant urban regions of Java (Amiti and
Cameron, 2007).

On the basis of data from the 1980s and 1990s Amiti and Cameron
(ibid.) estimate the agglomeration advantages that arise from vertical
linkages between firms in Indonesia. Taking account of the location of
input suppliers to estimate cost linkages and the location of customers
to estimate demand linkages, Amiti and Cameron (ibid.) show that the
externalities that arise from demand and cost linkages in Indonesia are
quantitatively very important and highly localized. They find that an
increase in either cost or demand linkages from the 10th to the 90th per-
centile increases wages by more than 20 per cent. As such, firms benefit
greatly from proximity to a large supply of inputs and good market
access, and firms with the best supply or market access can afford to pay
more than 20 per cent higher wages than those with the poorest access
(Amiti and Cameron, ibid.). Moreover, the benefits of these vertical link-
ages are extremely highly localized. Only 10 per cent of the market access
benefit spreads beyond 108km and only 10 per cent of the supply access
benefit spreads beyond 262km (Amiti and Cameron, ibid.). At the same
time, firms located in the major urban areas tend to be more vertically
disintegrated and to exhibit strong local linkages with a variety of back-
ward input supply sources than firms in rural areas (Amiti and Cameron,
ibid.).
There are two crucial aspects to the findings of Amiti and Cameron
(ibid.). First, such extreme economic advantages associated with geo-
graphical localization imply that Indonesian agglomeration effects operate
almost entirely within Java. As agglomeration effects are not only associ-
ated with spatial clustering but also with economies of scale, market access
and infrastructure (Syamwil and Tanimura, 2000) these findings suggest
that Indonesian economic growth will continue to become systemati-
cally more oriented towards the urban areas of Java. For a country that
spans 5500km across 900 islands, the prospects for other regions within
Indonesia would appear to be very much more limited because firms
located in Indonesias outer islands are too far away to benefit from the
agglomeration advantages associated with industries on the main island of
Java (Amiti and Cameron, 2007). The logic of these agglomeration effects
therefore means that central government efforts to promote economic
growth in geographically peripheral and low wage regions face enormous
challenges.
The second crucial aspect of the findings of Amiti and Cameron (ibid.)
is regarding the nature of the agglomeration effects in Java. They find that
labour pooling is quantitatively much less important than the existence
of localized vertical linkages and there also appear to be little or no real
technology spillover effects. Regarding these local spillover effects, most
multinational firms in Indonesia are export-oriented and generally do not

supply Indonesian customers (Blalock and Gertler, 2007). As such, any

pecuniary or technology spillover effects that are found to operate are
likely to be upstream spillovers affecting the domestic supplier industries.
Even among Asian countries, however, Indonesia is a relative latecomer
in terms of its industrial sophistication, technological capabilities, R&D
effort and export orientation (Kumar et al., 1999). This could be due
to a lack of absorptive capacity, given the very low levels of domestic
Indonesian R&D. Although they are very limited, technology spillovers
from FDI are found primarily in sectors with a high degree of competition
(Sjoholm, 1997; Blalock and Veloso, 2007), and the degree of pecuniary
spillovers depends on the extent of the technology gap between domestic
and foreign firms, and the level of domestic market concentration. At the
same time, the very low observed levels of R&D underline the importance
of inward FDI as a potential external source of technology and knowledge
for Indonesias future growth.
CONCLUSIONS
The growth experience of China, India and Indonesia all share some
common elements. Most notably, the economic growth in each of these
countries is dominated by certain regions, in which agglomeration econo-
mies appear to be crucial. As such, major cities are the focus for economic
growth in all three societies. Both labour and capital are increasingly
mobile and the location of activity is shifting dramatically into the major
urban centres. The result of this is that each society is becoming more
unequal, and the gap between the urban and rural areas is increasing
rapidly. In addition, the gap between the dominant urban areas and other
smaller urban areas is also increasing. On the other hand, there is some-
thing of a difference between the countries in the role played by localized
technology spillovers. The positive effects of these local technology spillo-
vers appears to be more marked in China than in India, and in turn the
effect of these in India is far more marked than in Indonesia, where the
evidence for these is very limited indeed.
REFERENCES
Amiti, M. and Cameron, L. (2007), Economic Geography and Wages, Review of

Economics and Statistics, 89(1), 1529.
Angang, H., Linlin, H. and Zhixiao, C. (2005), Chinas Economic Growth
and Poverty Reduction, in W. Tseng and D. Cowen (eds), India and Chinas

Recent Experience with Reform and Growth, International Monetary Fund and
Palgrave, Basingstoke.
Blalock, G. and Gertler, P.J. (2007), Welfare Gains from Foreign Direct
Investment through Technology Transfer to Local Suppliers, Journal of
International Economics, 74(2), 40221.
Blalock, G. and Veloso, F.M. (2007), Imports, Productivity Growth, and Supply
Chain Learning, World Development, 35(7), 113451.
Fujita, M. (2007), Development of East Asian Regional Economies: A View from
Spatial Economics, in M. Fujita, Regional Integration in East Asia from the
Viewpoint of Spatial Economics, Palgrave, Basingstoke, UK.
Golley, J. (2007), The Dynamics of Chinese Regional Development, Edward Elgar,
Cheltenham, UK and Northampton, MA, USA.
James, W.E. (2007), Comparative Advantage in Thailand and Indonesia and
Potential Free Trade Agreements: Implications for Trade Diversion, in M.A.B.
Siddique (ed.), Regionalism, Trade and Economic Development in the Asia-
Pacific Region, Edward Elgar, Cheltenham, UK and Northampton, MA, USA.
Kambhampati, U.S. and McCann, P. (2007), The Regional Performance and
Characteristics of Indian Manufacturing Industry, Regional Studies, 41(3),
28194.
Kumar, V., Kumar, U. and Persaud, A. (1999), Building Technological Capability
Through Importing Technology: The Case of Indonesian Manufacturing
Industry, Journal of Technology Transfer, 24(1), 8196.
Lardy, N. (2005), Trade Liberalization and its Role in Chinese Economic Growth,
in W. Tseng and D. Cowen (eds), India and Chinas Recent Experience with
Reform and Growth, International Monetary Fund and Palgrave, Basingstoke.
Lee, R.C. and Findlay, C. (2005), Telecommunications Reform in Indonesia:
Achievements and Challenges, Bulletin of Indonesian Economic Studies, 41(3),
34165.
McCann, P. (2008), Globalization and Economic Geography: The World is
Curved, Not Flat, Cambridge Journal of Regions, Economy and Society, 1(3),
35170.
McCann, P. (2009), Globalisation, Multinationals and the BRIICS Countries, in
R. Lattimore and R. Safadi (eds) (2009), Globalisation and Emerging Economies,
OECD, Paris.
Panagariya, A. (2005), India in the 1980s and the 1990s: A Triumph of Reforms,
in W. Tseng and D. Cowen (eds), India and Chinas Recent Experience with
Reform and Growth, International Monetary Fund and Palgrave, Basingstoke.
Scheve, K.V. and Slaughter, M.J. (2007), A New Deal for Globalization, Foreign
Affairs, 3447, JulyAugust.
Sjoholm, F. (1997), Technology Gap, Competition and Spillovers from Foreign
Direct Investment: Evidence from Establishment Data, Working Paper No. 38,
WIJS The European Institute of Japanese Studies, Stockholm.
Syamwil, I.B. and Tanimura, P.H. (2000), The Spatial Distribution of Japanese
Manufacturing Industries in Indonesia, Review of Urban and Regional
Development Studies, 12(2), 12135.
Takii, S. and Ramstetter, E. (2005), Multinational Presence and Labour
Productivity Differentials in Indonesian Manufacturing, Bulletin of Indonesian
Timmer, C.P. (2004), The Road to Pro-poor Growth: The Indonesian Experience
in Regional Perspective, Bulletin of Indonesian Economic Studies, 37(1), 177207.

UNCTAD (2005), World Investment Report: Transnational Corporations and

the Internationalization of R&D, United Nations Conference on Trade and
Development, United Nations, New York and Geneva.
UNCTAD (2007), World Investment Report: Transnational Corporations,
Extractive Industries and Development, United Nations Conference on Trade
and Development, United Nations, New York and Geneva.
World Bank (2005), World Development Report 2005: A Better Investment Climate
for Everyone, Washington DC.
World Bank (2009), Reshaping Economic Geography: World Development Report,
Washington DC.

10. Implications of European
Union structural assistance to
new member states on regional
disparities: the question of
absorption capacity
Daniela Constantin, Zizi Goschin and
Gabriela Dragan
INTRODUCTION
In May 2004 the largest enlargement of the European Union (EU) took
place with ten new member states (NMS), mostly from Eastern and Central
Europe, joining it. They were then followed by Romania and Bulgaria in
January 2007. That expansion of the EU entailed an unprecedentedly high
amplitude of inter-regional disparities (Eurostat, 2007), and was of special
significance for the cohesion policy in the 200713 programming period.
In this chapter we discuss the implications of the EU structural assist-
ance to the NMS on regional disparities, special emphasis being placed on
the capacity of those states to absorb the allocated funds. First we provide
a brief outline of the EUs economic and social cohesion policy. We then
look at the influence of the EU enlargement on that cohesion policy in the
200713 programming period, revealing the close links between cohesion
and the EUs regional policy. Next the regional disparities in the NMS
within European context are discussed in connection with the contribu-
tion of the structural instruments to reducing these disparities. Finally
the absorption capacity of the EU funds in the NMS is examined as a
precondition for achieving this goal. A particular emphasis is placed on
the administrative absorption capacity, considering that the institutional
framework created in the NMS for the administration of the EU funds is
expected to play a decisive role for their successful and complete integra-
tion in the EU structures. That is, the focus is on considering the nature of
institutional issues as either enhancing or detracting endogenous factors.
182

Implications of European Union structural assistance 183
THE EU ECONOMIC AND SOCIAL COHESION

POLICY: ADDRESSING REGIONAL DISPARITIES
Addressing regional disparities through interventions that have the ulti-
mate objective of achieving convergence is a hallmark of regional policy
in the EU. The EUs economic and social cohesion may be considered in
terms of its two inter-related components, namely: (1) vertical cohesion,
referring to the alleviation of social disparities and the solidarity with
the disadvantaged social groups; (2) horizontal cohesion, which concen-
trates on the regional disparities decrease and the solidarity with the
lagging regions population. Based on this approach, the EUs regional
policy is closely related to the horizontal dimension of the cohesion
policy and is supported by the EU funds via allocations for convergence
competitiveness and employment European territorial cooperation
objectives of the current financial exercise. During the programming
period 200713, the cohesion policy ranks first in terms of expenditure
and coverage, with cohesion surpassing agriculture for the first time as
the largest area of expenditure undertaken by the EU (Leonardi, 2006).
In a total budget of approximately 862.4 billion euros, cohesion policy
accounts for 307.6 billion euros (35.6 per cent); that is an average annual
expenditure of 44 billion euros, compared with 41.8 billion euros allocated
to market-related expenditure and direct payments to agriculture.
As it was conceived from the very beginning as a necessary comple-
ment to the Single Market and Single Currency programmes, the cohe-
sion policy will be of a great importance to the objectives of economic
convergence real and nominal between the EU-15 and the NMS: (1)
real convergence refers to the narrowing of development gaps: similarity
of per capita GDP, nominal wage levels, equilibrium of real exchange
rates and related to this, price levels and tradable/non-tradable price
ratios; (2) nominal convergence aims at the narrowing and finally closing
of the gaps in macroeconomic stability currently existing between NMS
and incumbents. It focuses on the Maastricht criteria on inflation, inter-
est rates, fiscal variables and exchange rate stability (Kasman et al.,
2005).
Almost 82 per cent of the total budget for cohesion policy is allocated to
the objective of convergence of the member states and the regions. Its key
aim is to promote growth-enhancing conditions and factors leading to real
convergence within the EU. This objective covers the member states and
the regions whose development is lagging behind. The targeted regions are
those NUTS 2 regions with a GDP per capita less than 75 per cent of the
EU average. They are funded from the European Regional Development
Fund and the European Social Fund. At the same time the member states

whose GNI is less than 90 per cent of the EU average benefit from the
Cohesion Fund.
In this context it is obvious that the NMS are the main beneficiaries
of the renewed cohesion policy: all of them receive allocations from the
Cohesion Fund, while 51 regions out of 55 NUTS 2 regions in the NMS
are funded under the convergence objective. Hence, the big challenge these
countries have to face in the current financial circumstances of the EU is
whether they will be able to use the large amount of allocated funds. And
further on, the question is whether they will be able to promote adequate
economic policies and economic behaviour so as to generate high rates of
endogenous growth and, thus, ensure an effective use of these funds.
The answer to the first question is usually addressed in terms of the
so-called absorption capacity pertaining to the EU cohesion policy.
That defines the degree to which a country is able to effectively and effi-
ciently spend the financial resources allocated via European Funds. In
other words, it expresses the ability of an EU member state to digest and
consume the funds in order to foster its development and thus to improve
its economic and social performance (NEI, 2002; Horvat, 2004).
The absorption capacity can be addressed from the perspective of the
institutional system created in each member state in order to manage the
funds (the supply side) as well as from the perspective of the beneficiar-
ies of these funds (the demand side). The demand side mainly expresses
the ability of the potential beneficiaries public or private to generate
appropriate and acceptable projects (possible to be financed). The supply
side is determined by three main factors, leading to three components
of the absorption capacity, namely macroeconomic, administrative and
financial absorption capacity:
1. Macroeconomic absorption capacity indicates the rate of the EU

funding in terms of the GDP of the recipient member state. The
European Summit in Berlin (1999) and then the results of the
Copenhagen negotiations (2002) on financial chapters1 indicate an
upper limit for the Structural and Cohesion Funds set at 4 per cent of
the GDP of the respective member state. The capacity to absorb the
macroeconomic effects generated by the inflow of the supplementary
investments is also related to the macroeconomic absorption capacity.
2. Administrative absorption capacity represents the ability and skills
of central, regional and local authorities to prepare acceptable plans,
programmes and projects in due time, to decide on programmes and
projects, to arrange coordination among principal partners, to cope
with the vast amount of administrative and reporting work required
by the European Commission and to finance and supervise the

implementation properly, avoiding fraud as far as possible (Horvat,

2004).
3. Financial absorption capacity refers to the ability to co-finance
EU-supported programmes and projects, to plan and guarantee the
national contributions in multi-annual budgets and to collect these
contributions from several partners (public and private), interested in
a programme or project. The national co-financing is needed, since in
order to increase the incentive for using the funds efficiently the EU
structural assistance finances only a part of the costs of a programme
or project.
These components can be analysed by means of the EU programming

documents and various evaluation studies in both EU-15 member states
and NMS, although until now comprehensive studies regarding all three
components have not been carried out. Moreover, for the NMS only
macroeconomic and administrative absorption capacity can be evalu-
ated so far. The financial absorption capacity can be evaluated only ex
post, from 2009 onwards. According to the n+2 rule, each years tranche
that involves a programme co-financed by the EU funds must be used up
before the end of the second year following the commitment. Nevertheless,
useful information for the absorption capacity in the NMS can be found in
the studies focusing on the pre-accession funds as well as on the prepared-
ness of the (former now) candidate countries to absorb the EU funds after
accession.
IMPLICATIONS OF THE EU ENLARGEMENT FOR

THE COHESION POLICY 200713
Economic development within an integrated area might be considered

as the result of two complex components: (1) competitiveness, which
aims at the most efficient use of resources and factors; and (2) cohesion,
which addresses mainly the reduction of discrepancies among regions and
countries. However, for the EU, social and economic cohesion remains
the determinant element of this equation since the main goal of the entire
European construction namely the promotion of security, stability and
economic growth in the region cannot be attained as long as signifi-
cant disparities remain among different parts of this integrated area and
obstruct the competitiveness of the whole structure of the EU.
In terms of the number of newcomers, the latest round of enlargement
might be considered the EUs biggest ever, given that 12 NMS (ten in 2004
and two in 2007) joined the Union. The EU population has increased thus

Table 10.1 Impact of successive enlargements of the EU
Enlargement Population Surface GDP GDP/Capita

EU-9 (73) +33.4 +25.4 +32.2 0.9
EU-10 (81) +3.7 +7.9 +2.34 1.3
EU-12 (86) +17.5 +33.4 +11.3 5.5
EU-15 (95) +6.2 +34.9 +6.5 +0.2
EU-25 (04) +19.6 +18.0 +8.9 8.9
EU-27 (07) +6.5 +8.5 +2.0 4.0
Source: Compiled by the authors using data available from Eurostat, 2007.
by more than one-quarter, and its physical area by more than one-third.
Now with almost 500 million citizens, the EU generates approximately 31
per cent of the worlds nominal Gross Domestic Product (GDP) in 2007.
However, the combined GDP of all new member states has added only 11
per cent to the GDP of the EU-15, whereas the GDP per capita is 13 per
cent lower than before enlargement (see Table 10.1). Compared with the
EU average, the GDP per capita is 35 per cent per cent higher in the US
and 15 per cent higher in Japan.
If population and GDP per capita are considered, it is noticeable that
the NMS have brought about an important burden for the EU: their
populations represent approximately 21 per cent of the total population of
the EU, whereas in all cases the GDP per capita is below the EU average
(see Table 10.2). The worst situation is recorded by Bulgaria and Romania
with less than 40 per cent of the EU average. By contrast, GDP per capita
was 142.55 per cent in Ireland and 131.91 per cent in the Netherlands
(Luxembourg apart).
The changes of priorities in the new financial exercise for the period
200713 have been determined by changes in the overall background of
the EU under internal and external pressures. The new paradigm of the
200713 cohesion policy, as expressed by the European Commissioner,
Danuta Hubner (2007, p. 1), should be the creation of new opportunities
for the future ... rather than a compensation for the past. In fact this new
paradigm reflects the position of the most member states concerning the syn-
tagma competitiveness cohesion, which no longer represents an antinomy
competitiveness, versus cohesion, but a tandem of interdependent objectives.
Objectives
According to the Community Strategic Guidelines for 200713 and the

EU Budget, cohesion policy is considered as the main instrument at the

Table 10.2 Population and GDP per capita in the NMS, 2006a
Country Population (Million GDP Per Capita GDP Per Capita:

Inhabitants) (Euros, PPP) Percentage of the
EU Average
Bulgaria 7.70 8 700 37.02
Czech Republic 10.30 18 600 79.15
Cyprus 1.00 21 900 93.19
Estonia 1.30 15 900 67.66
Hungary 10.10 15 300 65.10
Latvia 2.3 13 100 55.74
Lithuania 3.40 13 500 57.45
Malta 0.40 17 700 75.31
Poland 38.10 12 400 52.76
Romania 21.58 8 800 37.45
Slovakia 5.40 14 700 62.55
Slovenia 2.00 20 800 88.51
EU-27 493.58 23 500 100.00
a
Note: EU-10 plus Romania and Bulgaria.
Source: Compiled by the authors using data available from Eurostat, 2008.
EU level for the accomplishment of the Lisbon Strategy. The three objec-
tives are as follows:
1. Convergence Objective. This objective concentrates on those regions

with a GDP per capita less than 75 per cent of the EU average. It
envisages 100 regions, which account for 35 per cent of the EU-27
population. The purpose is to accelerate the economic convergence
of less developed regions by improving conditions for growth and
employment as a result of investments in human and physical capital,
innovation and development of knowledge society, protection of
environment and/or improving the administrative capacity. From
the total amount of 264 billion euros allocated for this objective, the
distribution is as follows: 67.34 per cent go to regions whose GDP per
capita is below 75 per cent of the average; 8.38 per cent goes to regions
under statistical effect;2 23.86 per cent goes to cohesion countries;
and 0.42 per cent goes to the outermost regions3 (Euractiv, 2007).
For the NMS, the Convergence Objective will play a more significant
role given that the disparities among regions and states after the latest
enlargement are more important.
2. Regional Competitiveness and Employment Objective. This applies

to the rest of the EU, which means another 155 regions, including 61
per cent of the EU-27 population. It accounts for 15.8 per cent of the
funds allocated to the cohesion policy. These regions have relatively
high GDP levels, even if both growth and employment rates remain
weak in many regions. The regional development programmes will
strengthen regional competitiveness by supporting economic and
social innovation, knowledge society, entrepreneurship, protection of
environment and risk prevention.
3. European Territorial Cooperation Objective. This objective aims at
reinforcing cooperation at the cross-border, transnational and inter-
regional levels. It involves 2.44 per cent of funds. The objective is
complementary with the other two objectives, with it being possible
for eligible regions to be funded under both of those previous objec-
tives. The aim of this objective is to promote common solutions for
authorities of different countries in the domain of urban, rural and
coastal development, and to the development of economic relations
and the setting up of small and medium-sized enterprises (SMEs).
According to the EU, the total share of the convergence regions in

EU-27 GDP in 2002 was only 12.5 per cent, which compares with a 35
per cent population share (European Commission, 2006). GDP levels
also indicate widely differing regional situations. The GDP per inhabitant
in 2004 ranged from 24 per cent of the EU-27 average in the North-East
region of Romania to 303 per cent in Inner London (Table 10.3). Among
the 100 convergence regions, Romania and Bulgaria accounted for 12
of the 15 least prosperous regions (seven regions in Romania and five in
Bulgaria).
Structural Assistance
The structural assistance for 200713 allocated to all member states

represents 35 per cent of the EU budget (308 billion euros out of 862
billion euros total). For the EU-8 (the former communist countries) plus
Romania and Bulgaria, the total amount allocated is 175 billion euros,
representing more than one-half of the entire budget funds allocated for
the cohesion (Table 10.4).
Due to rule capping the structural assistance to a maximum of 4 per
cent of the GDP of each country,4 the less developed countries Romania
and Bulgaria have also the lowest allocation per capita. Assuming full
absorption of the ceiling for the payment appropriations, the EU consid-
ers that under the new financial framework for 200713, the NMS will be
net beneficiaries. However, the ability of the NMS to fully benefit from EU

Table 10.3 Regional GDP per capita in the EU-27 in 2004 (in PPS, EU-
27 = 100)
The 15 Highest Regions The 15 Lowest Regions

1 Inner London (UK) 303 1 Nord-Est (RO) 24
2 Luxembourg (LU) 251 2 Severozapaden (BG) 26
3 Bruxelles Cap (BE) 248 3 Yuzhen tsentralen (BG) 26
4 Hamburg (DE) 195 4 Severen tsentralen (BG) 26
5 Wien (AT) 180 5 Sud-Muntenia (RO) 28
6 Ile de France (FR) 175 6 Sud-Vest Oltenia (RO) 29
7 Berkshire, Buckinghamshire, 174 7 Severoiztochen (BG) 29
Oxfordshire (UK) 8 Yugoiztochen (BG) 30
8 Oberbayern (DE) 169 9 Sud-Est (RO) 31
9 Stockholm (SE) 166 10 Nord-Vest (RO) 33
10 Utrecht (NL) 158 11 Lubelskie (PL) 35
11 Darmstadt (DE) 157 12 Podkapackie (PL) 35
12 Prague (CZ) 157 13 Centru (RO) 35
13 Southern & Eastern (IE) 157 14 Podlaskie (PL) 38
14 Bremen (DE) 156 15 Vest (RO) 39
15 North Eastern Scotland (UK) 154
Source: Compiled by the authors using data available from Eurostat (news release,
23/2007, 19 February 2007).
transfers is limited due to certain agreed reforms and transitional arrange-

ments. As different authors have noted (see, e.g., Oprescu et al., 2005),
while the old member states had to create and adapt their administrative
structures to the requirements of Structural Funds gradually, for the NMS
all institutional and operational requirements had to be accomplished in a
very short period of time.
IMPLICATIONS OF REGIONAL DISPARITIES IN THE

NMS FOR STRUCTURAL ASSISTANCE
The Disparities
As already mentioned, the addition to the EU of the ten NMS in 2004

and a further two in 2007 has substantially increased economic regional
disparities. In EU-27, the GDP per capita in PPS is almost five times
higher in the top 10 per cent of regions than in the bottom 10 per cent of
regions, compared with a difference of less than three times in the EU-15
(Eurostat, 2007). In the enlarged EU, the ratio between GDP per capita in

Table 10.4 Total assistance allocated to the EU-10 for 200713
Country Total Conver- Competi- Territorial Assistance Percentage

Assistance gence tiveness Cooperation Per of GDP
(Billion (%) (%) (%) Capita
Euro), of (Euros)
Which:
Czech 26.69 97.0 1.6 1.4 2 627 3.5
Republic
Estonia 3.39 98.5 0 1.5 2 555 4.1
Hungary 25.31 90.0 8.0 2.0 2 561 3.9
Latvia 4.01 98.0 0 2.0 1 751 3.9
Lithuania 6.78 98.4 0 1.6 2 041 4.2
Poland 67.28 98.9 0 1.1 1 773 3.6
Slovakia 11.51 94.2 3.9 1.9 2 102 3.9
Slovenia 4.10 97.5 0 2.5 2 082 2.0
Bulgaria 6.67 97.3 0 2.7 901 4.0
Romania 19.67 97.7 0 2.3 911 3.2
Total 175.40 96.8 1.6 1.6 1 930 3.6
Source: Compiled by the authors using data available from The Economist Intelligence
Unit.
the top and bottom 25 per cent of regions also increased, from a ratio of
2:1 in EU-15 to a ratio of 3:1 in EU-27, and the average level of GDP per
capita was reduced by almost 12 per cent (2004 data). The regional distri-
bution of wealth among the 268 NUTS 2 regions of the EU-27 shows that
the regional GDP per capita (in PPS) relative to the EU-27 average ranges
from 23.58 per cent in North-East Romania to 302.9 per cent in the UK
capital region of Inner London.
GDP per capita is substantially lower in the NMS, where this indicator
is below 50 per cent of the EU-27 average in most regions (31 out of a total
number of 55 regions). A notable exception is Prague (Czech Republic),
which is the region with the highest GDP per capita in the NMS (157 per
cent of the EU-27 average).
Dynamics of Regional Development
In the NMS, the growth rate was especially high in the three Baltic States,
all having an average annual real GDP growth over 6 per cent (generating
an overall growth of 70 per cent in each country), as well as in Poland,
Slovakia, Hungary, Slovenia and Cyprus. The latest members Bulgaria
and Romania experienced long periods of economic decline during the

1990s, but recovered afterwards to have significant growth rates, especially

in the capital regions that concentrate most of their economic activities.
Many of the regions with a low GDP per capita in the NMS are catching
up fast. This trend is confirmed by a Eurostat statistical analysis, indicat-
ing regional convergence at the EU level based on the values of the Gini
coefficient and the coefficient of variation, both weighted by population
(Eurostat, 2007). But our own calculations of the Gini Inequality Index
clearly show that economic disparities are bigger within the 12 NMS for
which this indicator mounts to 0.2286, as compared with the regions in
the EU-15 countries, which have an overall Gini Index of only 0.1478.
Because of the higher differentials entailed by the last two enlargements
of the EU, the Gini Inequality Index has increased substantially reaching
a value of 0.2083 for EU-27 (2004 data). Smaller inequalities are found
within each individual country (see Table 10.5, column 3).
Following significant above-average growth rates in most of the NMS,
economic convergence between the regions of the EU-27, significantly
improved, with the difference in the ratio of GDP per capita in the most
developed region (Inner London) and the least prosperous region (North-
East Romania) declining from 13.9:1 in 2002 to 12.8:1 both in 2003 and
2004. This downward trend is encouraging although the differential is still
substantial. The number of regions with GDP per capita values below 40
per cent of the EU-27 average also decreased from 23 regions in 2002 to 21
regions in 2003 and to 17 in 2004.
The NMS are catching up with the EU-27 average at a rate of 0.8
percentage points every year (Eurostat, 2006). A closer look reveals that
not all regions have such good evolutions. Although many less developed
regions in NMS attained growth rates above the EU-27 average, there are
still 15 of the 55 regions in the NMS having disappointingly low dynamics,
of less than 2 per cent annually, which is the EU-27 mean growth. All 15
of these regions belong to three NMS, namely Romania, Czech Republic
and Bulgaria.
Within-country Disparities
Deep regional disparities exist even within the countries themselves, as

Table 10.5 clearly shows. In 2003, the highest value of regional GDP
per capita was more than double compared with its lowest value in 12
EU countries, the broadest inter-regional differences being in the United
Kingdom and Belgium, where the ratio between the two extreme values
was 3.7:1 and 3.1:1 respectively. The NMS also display comparatively
large regional disparities (see Table 10.5, column 3), although the range of
values slightly narrowed.

Table 10.5 Economic disparities and allocations for Convergence Objective in the new member states
Country GDP Per Highest/Lowest Population in Regional Gini Number of Structural and Indicative
Capita in 2004, Regional 2004, Inequality Convergence Cohesion Funds, Allocations

Euros PPS GDP Per Million Index for Regions 200406, for
Capita Inhabitants GDP Against Total Million Euros Convergence
in 2004 Per Capita Number of (Prices 2004) Objective,
in 2004 National 2007/13,
Regions Million Euros
(Current Prices)
Col 1 Col 2 Col 3 Col 4 Col 5 Col 6 Col 7 Col 8
Bulgaria 7 200 1.917 7.7 0.119 6/6 6 674
Czech 16 400 2.628 10.2 0.155 7/8 2 404 25 883
192
Republic
Estonia 12 300 1.4 1/1 615 3 404
Cyprus 19 700 0.7 108 213
Latvia 9 800 2.3 1/1 1 031 4 531
Lithuania 11 000 3.4 1/1 1 379 6 775
Hungary 13 800 2.426 10.1 0.173 6/7 2 837 22 890
Malta 16 400 0.4 1/1 81 840
Poland 11 000 2.184 38.2 0.109 16/16 11 202 66 553
Romania 7 200 2.734 21.6 0.159 8/8 19 213
Slovenia 18 300 2.0 1/1 423 4 101
Slovakia 12 200 3.054 5.2 0.246 3/4 1 544 10 912
Sources: Compiled by the authors using data available from Eurostat (2007).
20/12/2010 15:12
The dynamics of economic development between the regions in one

country can diverge almost as widely as it does between regions in different
countries. The greatest discrepancy is displayed by Romania, where the GDP
per capita in the most dynamic region (namely Bucharest-Ilfov) increased
six times more than in the least developed one (namely North-East).
The high regional inequalities of growth dynamics within the NMS are
largely determined by the strong economic dominance of their capital
regions. In all the NMS and in some of the EU-15 countries a sub-
stantial share of economic activity is concentrated in the capital regions,
which usually have the highest GDP per capita. The larger GDP per capita
of those capital city regions is mainly the effect of a substantially higher
productivity, and it is also due to in-commuting, which provides a larger
labour force relative to the inhabitants of the capital region.
Implications
The new financial perspective for the seven years between 2007 and 2013 is
thus for an EU comprising 27 member states that are displaying increased
economic inequalities. Disparities in the levels of development in the
enlarged EU imply the need for assistance to the least developed regions
and member states by means of an appropriate allocation of Structural
and Cohesion Funds. As previously mentioned, the new round of cohe-
sion policy will be focused on investment in a limited number of priorities
organized around the three main objectives; namely convergence, regional
competitiveness and employment and territorial cooperation.
The Convergence Objective is designed to diminish the amplitude of the
inter-regional disparities, focusing mainly on the least developed regions.
Eligible regions for funding under this objective are the current NUTS 2
regions whose GDP per capita (measured in purchasing power parities) is
below 75 per cent of the average GDP in EU-25 for the period 200002. In
EU-27 there are 84 regions in this category belonging to 17 member states.
From these, 51 regions (out of a total of 55 regions in NMS) belong to 11
NMS.
The overall level of allocations available under the Convergence
Objective amounts to 282.8 billion euros, representing 81.5 per cent of the
total budget for the EUs cohesion policy. Within these allocations, 199.3
billion euros are directed to the Convergence Objective aiming to speed up
the convergence of the least developed regions, preponderantly belonging
to the NMS (see Table 10.5, column 8).
The specific level of allocations to each member state is calculated on
the basis of relative regional and national prosperity and the unemploy-
ment rate of the eligible regions. The resulting indicative allocations of the

Convergence Objective for the current round of cohesion policy 200713

reflect the dimensions and development levels of the 51 eligible conver-
gence regions in 11 new member states (see Table 10.2 and Table 10.4).
Following the last two enlargements of the EU, the EU-25 average
GDP/capita decreased by 8 per cent, making 16 regions ineligible to
receive convergence funds and which previously received Objective 1
funding. Based on the fact that the economic development level in these
regions had not really improved, these so-called phasing-out regions
will still receive transitional funds that amount to 12.5 billion euros.
Another 13 regions (with a total of 19 million inhabitants) the so-called
phasing-in regions will receive special financial allocations (10.4 billion
euros) due to their former status as Objective 1 regions.
CAPACITY OF NMS TO ABSORB EU FUNDS
As demonstrated earlier in this chapter, the NMS will be by far the most
important net beneficiaries of the EU structural assistance funds. The
financial transfers are designed to increase the economic and social cohe-
sion among the member states, mostly via enhancing a faster catching-up
process of the less developed states and regions in terms of income
per capita. This issue is of a particular importance to the NMS, since
Structural Funds are more important when the economy is weak, the
marginal benefit of an efficient use of Structural Funds being higher in less
developed economies (Daianu, 2003).
However, there are experts who question the possibility of effective,
productive absorption of the substantial financial transfers by the former
centralized economies given all their structural, institutional and admin-
istrative problems (Kalman, 2002). Moreover, especially in the academic
debate, some authors doubt about the ability of fiscal transfers to bring
about economic convergence for the current net recipient member states,
or, in general, about whether convergence can be achieved and, even if so,
whether fiscal transfers are best tools for enhancing convergence (Boldrin
and Canova, 2001). In that debate, various convergence concepts (abso-
lute, conditional) have been discussed,5 but given the current options of
the EU cohesion policy, our objective here is to evaluate the capacity of
the NMS to absorb the large amount of allocated funds.
Methodology Used by the EU to Assess NMS Absorption Capacity
The question of absorption capacity started to concern the EC at the

beginning of the previous financial exercise (200006) when the current

NMS were preparing their accession to the EU. This concern is about
institutional effectiveness.
In 2002 the NEI Rotterdam developed a study commissioned by
the DG-Regio/DG-Enlargement, which proposed a methodology for
evaluating the capacity of the candidate countries to effectively manage
the Structural Funds (NEI, 2002, p. 1). The methodology analysed the
administrative absorption capacity only for the design (of Structural
Funds) phase, considering that it was premature to address the other two
phases, namely: (1) performance, or the extent to which the Structural
Funds have been managed efficiently and effectively; (2) functioning of
Structural Funds.
The design assessment focused on a series of indicators regarding:
management;
programming;
implementation;
evaluation and monitoring;
financial management and control.
These indicators were calculated for three main components, namely,

structure, human resources and systems and tools:
1. Structure, which refers to the clear assignment of responsibilities and

tasks to deal with the Structural Funds;
2. Human resources, which relate to detailed tasks and responsibilities at
the levels of preparing job description, the number and qualifications
of staff and fulfilling recruitment needs;
3. Systems and tools, which refer to the availability of instruments,
methods, guidelines, manuals, procedures, forms, and so forth.
The information provided by the candidate countries followed detailed

questionnaires sent out by the Commission in the spring of 2003. As a
result of this assessment the main message was that acceding countries
need to further strengthen their administrative capacity (Press Releases
Rapid, 2003).
Following that first exercise, experts and researchers employed the same
methodology for their own studies and evaluations Papadopoulos (2003)
for ten candidate countries; Horvat (2004) for five countries (Hungary,
Czech Republic, Slovakia, Estonia and Slovenia); and Oprescu et al.
(2005) for Romania. For example, the study carried out by Oprescu et al.
showed a weak capacity in the case of Romania, but at levels comparable
with the other five former candidate countries at approximately the same

Table 10.6 Administrative absorption capacity design phase: results of

the evaluation of the main indicators by country
Romania Hungary Czech Slovakia Estonia Slovenia

Rep.
Horizontal evaluation
Management C (72%) B (87%) B (75%) C (63%) B (87%) C (71%)
Programming C (52%) B (80%) B (80%) D (40%) B (87%) B (80%)
Implementation C (53%) C (72%) C (56%) C (52%) C (68%) C (52%)
Vertical evaluation
Structure B (76%) B (84%) B (79%) B (79%) A (95%) B (74%)
Human resources C (51%) C (74%) C (71%) D (41%) B (82%) C (59%)
Systems and D (45%) C (60%) C (50%) D (40%) C (60%) C (50%)
instruments
Notes:
A = Strong capacity: system ready for the Structural Funds (at least 90%).
B = Sufficient capacity, but weak points should be addressed (7590% from the maximum
score).
C = Capacity not sufficient yet, serious weaknesses must be addressed (5075%).
D = Insufficient capacity, there is no base for administrating the Structural Funds.
Source: Compiled by the authors using the evaluation by Oprescu et al. (2005) and
Horvat (2004).
time before accession, suggesting that the delays could be recovered and
its accession at the beginning of 2007 was still possible (see Table 10.6).
Subsequent Evaluations of Absorption Capacity
After the 2004 accession wave, further new studies have been carried out,
concentrating on the challenges that NMS had to face in implementing the
structural assistance allocated for 200406. What they reveal is discussed
in what follows.
A study undertaken by McMaster and Bachtler (2005) provides a
comparative analysis concerning how the NMS have accomplished three
essential functions:
programming and structural assistance;

institutional training;
the implementation of the funds.
In the programming process the difficulties were generated by the

decision-making with regard to the political choices of the strategic

development areas and the setting of clear targets and long-term objec-
tives. In most cases short- and medium-term objectives were preferred
over long-term ones and simple and direct interventions were preferred
to complex ones, able to combine several objectives simultaneously. For
example, direct support to enterprises was preferred to setting up services
for businesses, or modernizing the existing transport infrastructure to
developing combined or alternative transport modes.
Regarding the choices between the national and regional dimensions
of development, that is, between interventions at national level meant
to support general development and economic growth and those aiming
at stimulating the endogenous potential of regional and local develop-
ment, Baleanu (2007) shows that they were largely in favour of the
former. While in the EU-15 the regionalization of the Structural Funds
management has been carried on for more than one decade, in almost
all NMS that joined the EU in 2004 the governments have chosen to use
centralized management systems. Regionalization requires the transfer
of many programming and implementation responsibilities regarding
Structural Funds to the regional authorities, which is not to the advan-
tage of the countries with still weak regional and local administration.
For this reason Poland, Czech Republic and Slovakia even gave up the
design-specific regional programmes in the programming period 200406.
Instead they incorporated them into sectoral programmes or into a single
national programme (for example, Joint Operational Programme in the
Czech Republic).
Centralized management can have a negative effect on smooth imple-
mentation of the programmes and, as a consequence, it can diminish
the absorption speed of the funds, as shown by the low absorption rates
recorded. In the autumn of 2006, two years after the programmes had been
launched, an average of only one-third of the funds allocated to NMS-8
was used (see Table 10.7).
According to the n+2 rule, the EU funds could be spent by the end of
2008, based on a system of annual reallocation, so that the final results
regarding the absorption rate will depend to a great extent on the ability of
administration to strengthen its institutional capacity. So far the current
situation shows that after accession the NMS administrations did not
maintain the pace of reform, which resulted in a low ability to manage
public funds. The implementation of advanced human resource manage-
ment systems was in general limited and inconsistent. A survey conducted
by the World Bank (2006) has shown that innovation is still isolated in
public management: the administrative function of general coordination
is at much lower standards compared with advanced countries, being
unable to keep up with the requirements, politicization has reappeared in

Table 10.7 Absorption rate in NMS-8 between May 2004 and September
2006
Country Absorption rate (%)

Czech Republic 26
Estonia 29
Hungary 32.5
Latvia 25
Lithuania 25.5
Poland 24.5
Slovakia 27.5
Slovenia 34
Source: Compiled by the authors using data available from the European Commission, as
quoted by Baleanu (2007).
administration, new payment and incentive systems for civil servants are
not put into practice.
At the same time, the experience regarding partnership has been dif-
ficult to assess. This is a major principle in the management of Structural
Funds in order to be able to increase the effectiveness of the programmes
and to achieve the commitment of actors involved in various stages of the
programming cycle, as well as to create good practices in administration.
In many cases it has demonstrated that the participation is not authentic,
but mimic and formal, with negative consequences on the partners com-
mitment and on assuming the ownership of projects results (see Baleanu,
2007).
In order to reinforce the administrative reforms in the NMS and in
the old MS whose administrations still do not function at the required
level, the EC has introduced in the programming of Structural Funds
for 200713 a priority concentrating on the modernization of public
service. This is financed by the European Social Fund. The priority aims
to stimulate good governance practices and to strengthen the capacity of
administrations to meet the requirements for planning and implementing
development plans and for increasing the administrative effectiveness of
public service at national, regional and local level.
Supposing that all these measures will lead to a better administrative
capacity, and hence a high rate of absorption of the EU funds, a further
question relates to the demand side and impact of Structural Funds on
the economic and social welfare in the recipient state. In other words, what
are the effects of Structural Funds on economic growth and in achieving
real convergence? Considering the original idea behind regional policy

in which economic convergence was an ultimate goal, the input-oriented

approach based on administrative capability does not seem to be sufficient.
Impact of Structural Funds
Regarding the impact of Structural Funds, a series of so-called absorption

problems have been identified and require careful consideration on behalf
of policy-makers. They are pertaining to large-scale fiscal transfers that
can emerge for various reasons and can be significant in preventing the
economy from achieving its optimal growth path (Herv and Holzmann,
1998; Kalman, 2002). According to Kalman (2002, pp. 59) they can be
summarized as follows:
1. Administrative absorption problems. They result in a difference

between transfers and the increase in the productive capital. For
any given administrative capability there is a ceiling of absorption
capacity and therefore the authors suggest that transfers should
only be phased in gradually, starting from a low level and adjusted
upwards.
2. Rent-seeking. This phenomenon refers to the people who interfere for
the use of funds in purposes that afterwards lead to gaining personal
advantages by sharing the benefits resulting from the newly created
economic activities (Myrdal, 1972; Krueger, 1974). Rent-seeking
becomes manifest through external forms of corruption like bribery,
money laundering, traffic of influence and goods, black market, and
so on. In the EU context rent-seeking might appear at three levels:
between national governments and Brussels, between central and local
governments, between any government and private sector agents who
benefit from regional policy (e.g., consultants, supported SMEs, con-
struction companies, etc.). The forms of rent-seeking range from legal
lobbying activities to illegal forms such as bribery.
3. Use of funds for consumption instead of investment. As a result of
external funds injection, domestic investment financing may well be
reduced, which means that, unless effective constraints on the use of
external funds are imposed, the impact of transfers on capital accu-
mulation and growth will be lower than expected. Several empirical
studies (e.g., Herv and Holzmann, 1998) in development economics
confirm that a substantial part of foreign aid is in fact directed to
increasing consumption.
4. Timing-related problems. The long-term focused public investment
and infrastructure development-related decisions have significant
opportunity costs in the short run, such as delays in private investment

decisions owing to increased uncertainty or modified expectations, or

private investment even being crowded out by the public sector.
5. Information disadvantage of the transfer-generating authority/
principal agent problems. Agents (private or public, eligible for
structural funding) who propose projects have better information
on local conditions, thus on the expected private and social rates of
returns of the project than the allocating principals. Hence, they may
try to reap various advantages whereas the principal is not able to
correct the information disadvantage or may do it only at very high
costs.
6. Multiple priorities leading to sub-optimal choice. For example, when
economic growth is not the sole priority of the recipient country or
regional government, other considerations being followed as well,
such as equity or fiscal expansion for re-election purposes, they might
lead to non-optimal outcomes, such as selection of sub-optimal
investment projects either deliberately or not.
7. Problems resulting from relative price changes induced by transfers.
Two well-known examples are the Dutch disease6 and immiserizing
growth phenomena. The Dutch disease refers to the case when trans-
fers through excess demand effects in the non-tradable sector (e.g.,
construction) lead to upward pressure on the overall wage and price
level (inflation) and eventually determine a decline of the tradable
sectors. Immiserizing growth occurs when economic distortions influ-
ence the industrial structure in tradable sectors and transfers provide
benefits mostly to protected sub-sectors. An industrial restructuring
towards protected sectors might occur, which may be harmful for the
overall growth path of the economy and some backward regions may
become further disadvantaged.
CONCLUSION
In order to support the proper functioning of the single market and also
to ensure solidarity among its members, the EU cohesion policy has
an overall objective to stimulate the process of reducing the disparities
between states and between regions via the so-called convergence process.
Those disparities have significantly increased since the 2004 and 2007
accession waves. The structural financial assistance associated with the
cohesion policy plays a central role in this process, which has a special
significance to the NMS as the main net beneficiaries of the financial
transfers.

Without neglecting the importance of the amount of funds allocated,

it is even more important how these funds are used by the recipient state.
The absorption rate does matter, but at the same time the qualitative
aspects of the impact of structural assistance matter as well. Overall the
following conclusions may be drawn: (1) On the supply side, the success
of Structural Funds-based programmes is conditioned to a great extent
by the quality of public administration. The higher the quality, the higher
the impact of Structural Funds on the economic and social welfare in the
recipient state. (2) On the demand side, the scope of public benefits is con-
ditioned by the way the funds are employed: if they are invested in viable
projects, with big value-added and significant multiplier effects, the impact
of the funds will be also important.
The NMS should learn from the experience of the countries that have
successfully used the structural assistance that they should have open
economies, solid internal public policies and administrations able to
implement it.
NOTES
1. The financial chapters refer to Common Agriculture Policy, Structural and Cohesion
Funds and financial obligations of a particular country to the EU budget.
2. These are regions where GDP per capita would be below 75 per cent of the EU-15.
3. These regions are distinguished by their low population density and considerable dis-
tance from mainland Europe. There are seven outermost regions: Guadeloupe, French
Guiana, Martinique and Runion (the four French overseas departments), the Canaries
(Spain) and the Azores and Madeira (Portugal).
4. Even if according to the Berlin Summit (Council Regulation (EC) No. 1260/1999, Art.
7, 8) the upper limit for Structural and Cohesion Funds was set up at 4 per cent, for the
200713 period, the European Council (Dec. 2005) decided that the maximum level of
transfers towards individual member states had to be reduced and the upper limit has
been established between 3.71 and 3.2 per cent (and below) depending on the GNI per
head.
5. In this case the debate is around the possibility of market forces let alone to lead to
the convergence of income in the long run.
6. This refers to the effects of huge positive income shocks for the Netherlands in the 1970s.
REFERENCES
Baleanu, A. (2007), The Impact of Structural Funds Qualitative Aspects,

Working Papers Series, No. 20, September, European Institute of Romania,
Bucharest.
Berlin European Council (1999), Council Regulations, No. 1260/1999.
Boldrin, M. and Canova, F. (2001), Inequality and Convergence: Reconsidering
EU Regional Policies, Economic Policy, 16(32), 20753.

Copenhagen European Council (2002), Presidency Conclusions, December 2002.

Daianu, D. (2003), Absorption Capacity of EU Funds Plays a Critical Role for
Membership Seeking Countries, Southeast European Times, Bucharest, 26 May.
Euractiv (2007), The New EU Cohesion Policy, available at: http://www.euractiv.
com/en/regional-policy/new-eu-cohesion-policy-20072013-archived/article-
131988; accessed 26 May, 2010.
European Commision (2006), Fourth Progress Report on Cohesion: The Growth and
Jobs Strategy and the Reform of European Cohesion Policy Summary, available
at: http://europa.eu/legislation_summaries/regional_policy/review_and_future/
g24238_en.htm; accessed 15 June 2010.
Eurostat (2006, 2007, 2008), Regional Yearbook, 2006, 2007 and 2008.
Herv, Yves and Holzmann, Robert (1998), Fiscal Transfers and Economic
Convergence in the EU: An Analysis of Absorption Problems and Evaluation of
Literature, Nomos, Baden-Baden.
Horvat, A. (2004), Absorption Problems in the EU Structural Funds. Some Aspects
Regarding Administrative Absorption Capacity in the Czech Republic, Estonia,
Hungary, Slovakia and Slovenia, National Agency for Regional Development
of Slovenia.
Hubner, D. (2007), Future of Cohesion Policy: Towards Exploring Options,
Informal Ministerial Meeting on Territorial Cohesion and Regional Policy,
Ponta Delgada, Azores, Portugal, 24 November 2007, available at: http://
europa.eu/rapid/pressReleasesAction.do?reference=SPEECH/07/742&format=
HTML&aged=1&language=EN&guiLanguage=en; accessed 15 June 2010
Kalman, J. (2002), Possible Structural Funds Absorption Problems The Political
Economy View with Application to the Hungarian Regional Development
Institutions and Financial System, available at: http://lgi.osi.hu/publica-
tions/2002/105/Marcou-Hungary.pdf; accessed 26 May 2010.
Kasman, A., Kasman, S. and Turgutlu, E. (2005), Nominal and Real Convergence
Between the CEE Countries and the EU: A Fractional Cointegration Analysis,
Applied Economics, 37(21), 2487500.
Krueger, A.O. (1974), The Political Economy of the Rent-seeking Society, The
Leonardi, R. (2006), Cohesion in the European Union, Regional Studies, 40(2),
15566.
McMaster, I. and Bachtler, J. (2005), Implementing Structural Funds in the
New Member States: Ten Policy Challenges, available at: http://www.eprc.
strath.ac.uk/eprc/Documents/pdf_files/12A07_McMaster-Bachtler_paper.pdf,
European Policies Research Center; accessed 26 May 2010.
Myrdal, Gunar (1972), Asian Drama: An Inquiry Into the Poverty of Nations,
Random House, New York.
NEI (2002), Key Indicators for Candidate Countries to Effectively Manage the
Structural Funds: Final Report, available at: http://www.evaluace.cz/doku
menty/hodnot_zpr_eu/souhrnna_studie.pdf, prepared by the NEI Regional
and Urban Development for the EC-DG REGIO/DG-ENLARGEMENT,
Rotterdam; accessed 26 May 2010.
Oprescu, G., Constantin, D.L., Pislaru, D. and Ilie, F. (2005), Analysis of
Absorption Capacity of the EU Funds in Romania, The European Institute in
Romania, Bucharest.
Papadopoulos, A. (2003), Administrative Capacity Study Phare Region, Phase
Two. Country Reports, European Commission, March.

Press Releases Rapid (2003), Structural and Cohesion Funds: Acceding Countries
Need to Further Strengthen their Administrative Capacity, available at: http://
europa.eu/rapid/pressReleasesAction.do?reference=IP/03/103i&format=HTM
L&aged=0&language=EN&guiLanguage=en; accessed 26 May 2010. Brussels,
July.
World Bank (2006), EU-8. Administrative Capacity in the New Member States: The
Limits of Innovation, September.

11. Macroeconomic and
territorial policies for regional
competitiveness: theory and
empirical evidence from the EU
Roberto Camagni and Roberta Capello
INTRODUCTION
In this chapter we examine the new required policy targets and styles for
assuring regional performance and competitiveness for the European
Union (EU) and provide an empirical analysis of territorial policies for
regional competitiveness, with a focus on investigating territorial capital
and regional growth.
We first address the theoretical basis of the importance of regional
policies. The strongest argument in favour of regional policies lies in the
long-term persistence and even widening of inter-regional disparities. The
starting point in that theoretical reasoning is that regional policies are fun-
damental for the competitiveness of regional economies. Unlike countries,
regions compete on the basis of an absolute advantage principle, and,
whenever non-competitive, regions cannot rely on any automatic mecha-
nism in order to maintain some export specialization. Thus, in the case of
the EU, the fate of regions is mass unemployment and, in the case of insuf-
ficient public income transfers, emigration and possibly desertification.
In this chapter we stress the idea that the factors determining regional
performance do not lie just in each regions internal development capabil-
ity. In fact, among the causes of regional success and failure, one can find,
on the one hand, some pervasive characteristics of the national economy
and, on the other hand, its general performance. This aspect, linked to the
fact that national policies are not space-invariant within each country,
brings us to the conclusion that national policies also can explain regional
performance. That includes interest rate policies, monetary and fiscal poli-
cies driving movements of the exchange rate, and also such policies such
as transportation and TENs (Trans-European Network) policies, policies
204

Macroeconomic and territorial policies 205
for R&D and agricultural policies. All such policies bring selective effects
to different typologies of regions due to a wide array of transmission chan-
nels or preferential regional targets, and, therefore will strongly contribute
to the economic performance of regions.
Of course beyond the national component, the crucial component of
regional performance is each regions internal development capability.
These policies are discussed based on a new concept, that of territorial
capital, and that is analysed through a historical development perspective
of policy targets and styles.
Thanks to the existence of a regional growth forecasting model called
MASST built by the authors (Capello, 2007; Capello et al., 2008), we also
present simulations on the following:
The impact of macroeconomic policies at regional level in the EU.

As the simulation will demonstrate, they are not space-invariant, as
some regions gain more than others when, for example, a devalua-
tion policy is pursued.
The impact of European regional policies, like structural funds.
The role of specific elements of territorial capital on regional growth
patterns, on which some policies could be built.
PERSISTENCE OF REGIONAL DISPARITIES AS

THE STRONGEST RATIONALE FOR REGIONAL
POLICIES
The strongest argument in favour of regional policies lies in the long-

term persistence and even a widening of inter-regional disparities. In fact,
the history of the entire European integration process is characterized
apparently by slowly decreasing overall inter-regional disparities; but this
process, in reality, results generally from strong processes of catching-up
among nations in the presence of increasing disparities at the intra-
national level.
These last trends of increasing disequilibria inside single countries may
show a different intensity over time. The Williamson Law (Williamson,
1965) expects them to be naturally stronger in the early stage of the devel-
opment or integration process, while in a later stage they slow down or
may even reverse. This result was shown as empirically correct if inter-
preted inside the great macro-phases of European integration (Camagni
and Gibelli, 1996): (1) increasing disparities in the first development and
integration phase during the 1950s up to the mid-1960s, followed by a
reduction until 1980, mainly thanks to the catching-up process of a new

category of intermediate-income, third regions like the Third Italy,

Flanders, southern Germany, southern France and East Anglia; (2) an
upswing of disparities during the industrial restructuring processes of
the 1980s, the new integration phase of the Single Market and the path
towards the single currency, up to 19952000, and still no clear evidence
afterwards.
Explanations for the Difficulties Encountered by Peripheral Regions
The theoretical explanations of the causes of the difficulties encountered

by less advanced or peripheral regions in their path towards integration
and development may be classified into four main classes, which are dis-
cussed below:
1. Unlike countries, regions are not subject to the principle of com-

parative advantage governing international specialization and trade,
attributing each partner country some specialization sectors and a
condition of full employment. The reason for this resides in the fact
that the two equilibrating forces that in principle allow passing from
an absolute advantage to a comparative advantage regime namely
price flexibility and currency devaluation, fully active in the case of
countries either do not work properly or do not exist at the inter-
regional level.1 Therefore, regions are not granted some productive
specialization and some role within the spatial division of labour
whenever they prove less competitive in all production sectors with
respect to the external territories (Camagni, 1992, 2002). Regions in
fact compete on the basis of an absolute advantage principle, and,
whenever non-competitive, they cannot rely on any automatic mecha-
nism in order to maintain some export specialization; their fate is, in
this case, mass unemployment and, in the case of insufficient public
income transfers, emigration and possibly desertification (GREMI,
1995). For these territories, the possible strategy for development (or
survival) is threefold: complete autarchy (almost impossible); lobby-
ing for public income transfers (to be rejected flatly), or improving
competitiveness of some export sector; and attracting investments
from other regions and from abroad.2 As a consequence, in the case of
regions, cities and localities, caring about competitiveness and attrac-
tiveness is perfectly legitimate, and the criticisms that Paul Krugman
recently addressed to such concerns (Krugman, 1996, 1998), labelling
them as wrong and misleading, do not apply in their case.3
2. The second argument refers to the fact that, even in the case that
lagging regions could supply lower wages than advanced regions, their

advantage would be limited by the lower levels of labour productivity.

In fact, what matters to firms in choosing their location are unit labour
costs or efficiency wages (wages/productivity): in advanced regions,
a high productivity generated by a strong industrial culture, efficient
services and good infrastructure may well out-weigh the disadvantage
of higher salaries and generate a continuing external competitiveness,
as in the well-known Dixon-Thirwall model (1975).
3. All this leads to the further argument that, in absence of some
basic preconditions, namely in terms of infrastructure, accessibility,
general education and basic public services, growth could never start
in a region, given the strong locational advantage of the other com-
petitor sites (Rosenstein-Rodan, 1943; Armstrong and Taylor, 1993).
4. The fourth argument, dynamic in nature, regards the cumulative
nature of economic growth, due to:
inter-sectoral relationships and forward/backward linkages in

the general economy and along the main production filires
(Hirschman, 1957; Krugman and Venables, 1996);
increasing returns to scale, when the expanding size of the local
production fabric generates increasing productivity (once called
the Verdoorn Law, utilized by Kaldor, 1970, and more recently
rediscovered by Krugman, 1991);
demandsupply interaction on the goods and labour markets,
which generates a cumulative development process in core
regions (investment development in-migration new
local demand new investment new development) (Myrdal,
1957);
technological change, both embodied in new machinery or
endogenous in the sense of the cumulative expansion of local
know-how and knowledge through learning-by-doing processes
la Arrow.
These effects taking place in already developed regions are a measure

to overcome the trend towards decreasing capital productivity in
these regions (a consequence of the expanding capital stock), and to
counterbalance the dispersion of economic activities expected by the
simplified neoclassical model of regional growth of the early 1960s
(Borts and Stein, 1964).4
The above arguments look strong enough to disprove the optimistic view
about inter-regional convergence processes, mainly expressed by the neo-
classical school, both in its traditional and modern modelling approaches.

A Regional Convergence Process Might be Expected but Why Does it Not

Occur?
A convergence process is expected by the neoclassical school as a conse-

quence of many factors. These include the following:
An out-migration process of unemployed people from lagging

regions: this process takes place in fact, and statistically reduces per-
capita income disparities as it reduces the denominator of the per-
capita income indicator in these regions. One may question though
whether this can be considered a true convergence process, as it
adds human resources in already developed regions reducing the
inflationary effects of tight local labour markets, subtracting them to
lagging regions, and whether inter-regional and international mass
migrations could be still politically acceptable, at least in European
countries.
An outflow of capital from advanced regions, as a consequence of
the diminishing returns showing up along with the capital accumula-
tion process; this event may be easily counterbalanced by learning
processes and cumulative technological change.
An outflow of technology from advanced regions towards lagging
ones. This argument suffers from the consideration of single tech-
nologies or technological paradigms, which for sure, once developed
and introduced in some core localities, diffuse afterwards on the
territory. But what about the overall pace of technological creation
in the form of subsequent improvements to the single technology,
development of integrated technological systems and new inven-
tions leading to new technologies? And what about the capability
of developing new applications of existing technologies? The pace
of this technological creation process in core regions is likely to
be faster than the pace of technology diffusion and catching-up
by lagging regions, due to all sorts of learning processes, both
internal to the firm or taking place in a collective way on the ter-
ritory (Camagni and Capello, 2000; Keeble and Wilkinson, 2000;
Camagni, 2001).
Even modern neoclassical conditional convergence theory (Barro and

Sala-i-Martin, 1991, 1992), which still expects diffusion processes and
re-equilibrating processes (a sort of entropic trend towards spatial
homogeneity), acknowledges that these processes are conditioned
by other, opposite, endogenous or exogenous forces (industrial struc-
ture, propensity to save, institutional innovations like the creation of

custom unions, regionally asymmetric shocks, etc.), which diversify the

catching-up capability of the single regions and are often a measure to
postpone or overturn the process of convergence (Armstrong, 2002). But
a second aspect, regarding the social costs of non-intervention in a context
of increasing disparities and inter-regional competition (on the basis of an
absolute advantage principle), provides even clearer support to spatial
development policies. A strategy of non-intervention in fact presents the
following drawbacks:
huge social and political costs allowing the explosion of regional crises
and the cultural and environmental costs of regional desertification;
the risk of a super-concentration of population in the big urban
areas of lagging regions (aphenomenon that is typical of developing
countries), as a consequence of the crisis of the surrounding areas
and not of the attractiveness of these urban areas, of a push and not
of a pull factor;
the high opportunity cost of adding successful activities in already
successful areas: in a context of full employment, new workers for
new activities are found at the expense of existing activities, while in
weak areas they are drawn from the unemployment reservoir, and
their opportunity cost is close to zero;
the channelling of a wide share of national savings towards the
building and construction industry and real estate speculation in
advanced regions and cities, as a consequence of the migration
processes, subtracting it from more productive uses;
a lower exploitation of the creativity potential of all regional
communities, constrained by the presence of some basic locational
disadvantage (accessibility, services, infrastructure, and so on).
All the previous arguments may be considered, by and large, as vari-

ations on the theme of market failures. But other kinds of failures are
invoked by the neo-liberal opponents of any policy intervention, namely:
(1) government failures, or the unintentional drawbacks of public policies;
(2) the reduction of the competitive climate and the likely formation of
new social classes, specialized in incentive intermediation and lobbying,
can reduce the reaction capability of the local entrepreneurship and the
local economy, which is the well-known argument put forward by the
public choice school concerning the inability of bureaucracy to substitute
for the market signals and its orientation towards self-referential goals
instead of the public good.
All these arguments for which there are numerous confirmations in
the case of long-term lagging regions, supposedly call for a substantial

reduction in the role of governments regional, national, European

even in the presence of the social costs of continuing territorial disparities
(Buchanan and Tullock, 1982; Pennington, 1999). A possible answer in
this case concerns the need and the actual possibility of providing
the necessary countervailing powers and political control tools, coming
from transparent, democratic and participatory practices and procedures.
Many recent positive examples show that policy failures are not the neces-
sary outcome, while the failures of the market mechanism are in-built into
its constituency and operating rules.
Spatial Development Policies Derived from the Globalization Debate
An element that recently contributed to lowering the general attention to

re-equilibrium issues and spatial development policies derived from the
globalization debate. If, in the new integrated context, competitiveness is
the main issue, and if champion firms and territories act as driving forces
for the entire territorial system, a proper spatial policy it is argued
should care more about strong than weak territories, about winners rather
than losers, and should therefore concentrate investment and innovation
(e.g., public expenditure in advanced infrastructure and human capital) on
core regions and big global cities.
Two answers may be given to this argument, referring to two different
territorial levels, namely, the inter-regional and the intra-regional.
As far as the inter-regional level is concerned, a distinction has to be made
between the provision of advanced assets and infrastructure that has to
be secured to all regions and to advanced regions in particular and the
public support to the financing of this provision, which should be inversely
proportional to the capability of each region to provide the same assets
through a private financial procedure (e.g., through project financing).
Regarding the second level, the suggestion coming from the previous
argument can be acceptable: when intervening with public development
policies on limited territories like regions or sub-regions, an important
policy rule should be to select places with a maximum development poten-
tial (big cities, specialized medium-sized cities), in order to maximize
probability of success and save public money. At the intra-regional level,
spread effects expected in the long run from the intervention on champion
territories would be probably faster and stronger than at the inter-regional
level, and the social cost of letting some areas lag behind in the initial stage
more acceptable.
This issue presents itself very clearly in the case of the 12 new member
countries in the EU. There is in fact no doubt about the polarized
and regionally concentrated character of the new industrialization and

development phase, triggered by the accession to the EU, something that

will probably continue in the next decade, due to cumulative territorial
effects. Capital cities and territories located along the border belt with
the 15 old member countries are the preferred locations for new activities,
both industrial and tertiary, both in the production and in the commercial
spheres. Migratory processes from peripheral regions towards core ter-
ritories; investments linked directly to these migratory trends, in all kinds
of residential sectors; infrastructure provision that naturally follows the
priorities indicated by the demand for mobility and accessibility; all these
trends will generate an increase in internal disparities, very similar indeed
to what happened in Western European countries during the early phases
of integration. Furthermore, political attention is more and more directed
to core areas, as they are rightly seen as the true assets in international
territorial competition.
This situation is nonetheless full of risks. An excessive concentration
of economic growth in a few areas is likely to very quickly determine
tensions on the local labour markets and land markets, pushing wages
and land rents upwards in an unsustainable way, namely well beyond the
increase in productivity.5 All this, coupled with the usual indirect effects of
congestion, social costs and environmental decay, will soon generate a fall
in competitiveness and attractiveness of these areas, jeopardizing the eco-
nomic future of these countries. As usually happens, the potential trend is
not only costly in terms of territorial cohesion and equity, but particularly
on the grounds of economic efficiency itself.
A wise strategy in these cases will be, on the contrary, to widen the
potentially eligible areas for foreign and internal investments, strengthen-
ing second-rank cities and city-regions in the national urban hierarchy,
and linking them with the capital cities and the big foreign agglomera-
tions and markets through efficient transportation and communication
networks. A polycentric urban development, both at the inter-regional
and intra-metropolitan level, similar to what is suggested by the European
Spatial Development Perspective, could be a proper way for the con-
struction of a consistent spatial strategy, sustainable at the same time in
economic, social and environmental terms.
THE NATIONAL COMPONENT OF REGIONAL

GROWTH AND THE DIFFERENTIAL SPATIAL
EFFECT OF MACROECONOMIC POLICIES
The factors determining regional performance do not lie just in each

regions internal development capability. In fact, among the causes of

regional success and failure, one can find, on the one hand, some pervasive
characteristics of the national economy and, on the other hand, its general
performance.
The former elements refer to: (1) institutional factors like the perform-
ance of the high functions of the nation state legislative, justice and
government functions to organizational factors like the efficiency of
services of general interest like education, transportation, communication,
health and security services; (2) economic factors like general fiscal pres-
sure, effectiveness of public expenditure, pervasiveness of environmental
regulations and efficiency of contract enforcement procedures, general
price-competitiveness in case of less advanced countries.
The second element linking regional economies to the general perform-
ance of the national economy is represented by the high inter-regional,
within-countries integration, relative to international integration, in terms
of exchange of goods, services and production factors, due to proximity
effects and absence of any kind of institutional or linguistic barrier. All this
is reflected in the empirical evidence in the EU. By and large, the variance
of regional growth rates within countries is lower than the variance of
international growth rates, all regions benefiting from a good short-term
and long-term performance of their national economy.6
This second element looks even more relevant if interpreted in the light
of the goal of reducing inter-regional disequilibria. As illustrated in Figure
11.1, the results of the MASST simulations and foresights on European
regions (see later in the chapter) show that, between now and 2015, dis-
parities in the EU-27 will decrease only thanks to the national growth
component, namely to fast processes on international catching-up (mainly
by new member states), with a slight increase in intra-national disparities
(see Figure 11.1a).7
In conclusion, if aggregate, national development represents an impor-
tant part of regional development, then it follows that first class territorial
development policies reside in sound and consistent policies internal to
each country, addressed towards a pervasive effectiveness of the public
administration and provision of public goods and externalities enhancing
the development capability of all local economies.
A further element, linked to the previous one, that is worth an in-
depth inspection is the impact of macroeconomic and structural policies
managed at the national level in determining regional performance. In
this case, going beyond the open question of the general effectiveness of
these policies, it looks sound to expect highly differentiated effects on
the different regions. In fact, interest rate policies, monetary and fiscal
policies driving movements of the exchange rate, but also such policies as
transportation and TENs policies, excellence policies in R&D, agricultural

0.25 (a) EU-27 countries
0.2
0.15
0.1
0.05
0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Total (EU-27) Between countries (EU-27) Within countries (EU-27)
0.25 (b) 15 old and 12 new countries
0.2
0.15
0.1
0.05
0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Total (old 15) Total (new 12)

Between countries (old 15) Between countries (new 12)
Within countries (old 15) Within countries (new 12)
Source: The authors using simulations through the MASST model.
Figure 11.1 The evolution of per-capita income disparities in the EU,

200215
policies, bring selective effects to different typologies of regions, due to a

wide array of transmission channels or preferential regional targets.
The case of agricultural policies is easy to understand, bringing direct
support to rural regions that, at present, are often among the richest, at
least in many of the 15 old member countries (Arkleton Institute, 2005).
But also other policies have selective effects: TENs and excellence policies,

for example, are naturally biased in favour of stronger regions, where most
of the favourable preconditions for successful competitiveness policies are
present and where the demand for new transportation infrastructure is
higher and guarantees the highest economic return to investment.
Another typology of national policies, concerning the macroeconomic
sphere, is usually considered as neutral in terms of inter-regional dispari-
ties. On the contrary, selective effects are visible also in this case, due to
inter-sectoral reasons. In fact, a rise in interest rates is likely to hit most
regions specialized in manufacturing and building and construction rather
than tertiary regions; a similar effect may be expected as a consequence of
a real revaluation of the currency, generating wider tensions in industries
and regions characterized by a wider degree of openness to international
trade. Furthermore, a revaluation process is likely to hit more, ceteris
paribus, those regions specialized in labour-intensive industries, as it
raises labour costs expressed in international currency (while the cost of
capital will remain unchanged, at the level determined internationally,
especially if the revaluation process is interpreted as the effect of a strong
and potentially fast-growing economy).
THE ENDOGENOUS COMPONENT OF REGIONAL

GROWTH: THE CONCEPT OF TERRITORIAL
CAPITAL
Beyond the national component, of course the crucial component of

regional performance is each regions internal development capability. In
the interpretation of this endogenous capability, spatial economic theory
has shown a triple paradigm shift in the last few decades, namely:
from development (or even location) factors to innovation factors

(Nijkamp, 1986);
from hard to soft factors, residing either in intangible, atmosphere-
type, local synergy and governance factors (Becattini, 1990;
Camagni, 1991), or in human capital and knowledge assets (Foray,
2000);
from a functional approach to a cognitive approach.
The last and more recent trend deserves some inspection. A cognitive
approach is increasingly superseding the traditional functional approach
to show that causeeffect, deterministic relationships should give way to
other kinds of complex, inter-subjective relationships that impinge on the
way economic agents perceive economic reality, are receptive to external

Functional approach
(substantive rationality)
Infrastructure Human capital

Factor endowment Competences
Industrial structure Education and training
Financial incentives Quality of life
Development Innovation
factors factors
Connectivity Knowledge, identity

Interconnection Receptivity
Openness Milieu effect
Go-global Collective learning
Cognitive approach
(procedural rationality)
Figure 11.2 Paradigm shifts in regional development factors (and

policies)
stimuli, can react creatively and are able to cooperate and work syner-
getically. Local competitiveness is interpreted as residing in cooperation,
trust and sense of belonging rather than in pure availability of capital; in
creativity rather than in the pure presence of skilled labour; in receptivity
to new business ideas and organizational styles more than in the presence
of SMEs per se; in connectivity and relationality more than in pure acces-
sibility; in local identity as well as local efficiency and quality of life (Figure
11.2).
The theoretical elements that support the new methodological approach
may be found in the following:
The theory of bounded rationality and decision-making under

conditions of uncertainty, from the seminal contributions of Simon
(1972) to their application to industrial innovation (Dosi, 1982;
Nelson and Winter, 1982).
The institutional approach to economic theory based on a theory
of contracts, which emphasizes the importance of rules and behav-
ioural codes, and of institutions that embed transactions in more
protective governance structures (Williamson, 2002, p. 439),

reducing conflicts and allowing mutual advantages to be gained

from exchange.
The cognitive approach to district economies and synergies, which
comprises the Italian school (Becattini, 1990), the French proxim-
ity approach (Gilly and Torre, 2000), the GREMI approach to
local innovative environments (Camagni and Maillat, 2006), and
Michael Storpers concept of untraded interdependencies (Storper,
1995). The GREMI group conceives proximity space or the local
milieu as an uncertainty-reducing operator that works through
socialized transcoding of information, cooperation enhancing and
the supply of the cognitive substrate represented mainly by the
local labour market in which processes of collective learning are
embedded (Camagni, 1991; Capello, 2001).
All the above elements which add to, and do not substitute for, more
traditional, material and functional approaches may be encompassed
and summarized by a concept that, strangely enough, has only recently
made its appearance, and has done so outside a strictly scientific context:
the concept of territorial capital. This was first proposed in a regional
policy context by the OECD in its Territorial Outlook (OECD, 2001),
and it has been recently reiterated by DG Regio of the Commission of the
European Union:
Each Region has a specific territorial capital that is distinct from that of other
areas and generates a higher return for specific kinds of investments than for
others, since these are better suited to the area and use its assets and potential
more effectively. Territorial development policies (policies with a territorial
approach to development) should first and foremost help areas to develop their
territorial capital. (CEC, 2005a, p. 1)
In our view territorial capital may be seen as the set of localized assets
natural, human, artificial, organizational, relational and cognitive
that constitute the competitive potential of a given territory. In this
very large sense it encompasses (Camagni, 2008; Camagni and Capello,
forthcoming):
natural resources and social overhead capital;

impure public goods or mixed public/private (p/p) goods (landscape,
cultural heritage);
agglomeration and district externalities;
club goods such as proprietary networks;
private fixed capital stock and relational private services;
social and relational capital;

human capital, entrepreneurship, creativity and leadership;

cooperation networks and strategic p/p partnerships in knowledge
creation;
governance structures.
It is evident that the concept of territorial capital encompasses not

just immaterial and relational assets, but also material ones, in the
form of natural and cultural resources, public goods and general urban
structure. In this second respect, it helps us understand the role of a
well-shaped geographical structure and form of settlements for both the
efficiency-competitiveness of territories and the general welfare condi-
tions of populations. In this sense, integrated spatial/urban development
policies were recently indicated by OECD (2001) and the EU (CEC,
2005b; Ministers for Spatial Planning, 2007) as the consistent new policy
approach, integrating economy and territory and targeting the best utili-
zation of the economic potential and the territorial capital of each local
economy. The concept comes close to the French tradition of amnage-
ment du territoire, with its integrated and wide-area approach to spatial
development.
RENEWED REGIONAL POLICIES FOR REGIONAL

COMPETITIVENESS
As stated before, the new scenario in which territorial systems compete

and act nowadays is a scenario of increasing complexity and uncertainty.
Decision-making centres have to manage increasingly interdependent
and integrated systems, globally stable but locally fragile; logical causal
chains present wide and embarrassing holes; the behavioural logics of
the different actors show increasing instability, limiting forecasting pos-
sibilities; transformations in production and consumption habits appear
faster and faster; available information grows exponentially, but our
capability of selection, interpretation, transcoding and evaluation does
not grow in a parallel way. As a consequence, our cognitive models
evolve in new directions, requiring new rationalities in the decision-
making process.
The logics founded upon a substantive rationality, widely utilized in
economic thinking, appear inadequate, as do the pretentions of the so-
called rational-comprehensive planning models of the recent past. But all
this does not question the very existence and necessity of spatial policies,
as some modern approaches seem to imply; it only requires a new type of
rationality, acknowledging its limitations and fully assuming uncertainty

as a normal condition in cognitive processes. Decision-making procedures

shall be characterized by flexibility, partnership, iterative processes of
decision-monitoring-assessment-decision, precautionary and no regret
principles as far as environmental matter is concerned and search for
consensus by the concerned communities. Even if a strong rationality is
useless, we still need some form of rationality in spatial policies: we still
need to find good solutions, to evaluate the effects of past decisions, to
provide consistency and compatibility among different decisions (about
projects and schemes that are frequently managed separately), to assess
the likely impact of single projects not just in terms of their direct effect on
employment, income, mobility, environmental quality, but on the general
long-term competitiveness and liveability of territories.8
During the last half-century, both theoretical reflections and the evolu-
tion itself of spatial systems have determined a substantial enrichment of
policy strategies and policy tools in spatial development practices, shifting
the emphasis successively towards new, more suitable goals and keywords.
We can summarize these evolutions as follows:
1950s: infrastructure, as precondition for growth;

1960s: attraction of external activities, development poles, export
industry;
1970s: endogenous development: SMEs, local competencies;
1980s: innovation, technological diffusion, innovative milieux;
1990s: knowledge base, intangible factors, local culture;
2000s: relational capital, collective learning, interconnection,
territorial capital.
The general approaches followed in the earliest period, in Europe in

particular, fully recognized the balanced nature of the economic develop-
ment process; all preconditions had to be present, or provided, at the same
time. Subsequent reflections, typical of the 1960s, were mainly oriented by
the sense of urgency and the need to find short-cut solutions to develop-
ment needs of territories: a big push and an unbalanced growth strategy
was felt to be economically and politically superior. Development poles,
cathedrals or white elephants forced into the desert were supposed to
bring faster, even if less certain, effects (Parr, 1999).
During the 1970s, taking advantage of the results achieved mainly
through the diffusion of public infrastructure and services throughout
the countries (but also trying to counterbalance the negative effects that
in many cases derived from the location of big external plants, separated
from the local economy and society), development policies addressed
themselves towards the strengthening of the local fabric of SMEs, in a first

stage through generalized incentives and subsequently trying to enhance

local synergies and local specializations (the so called development from
below model: Sthr, 1990).
Subsequently, during the 1980s, both the theorization on the importance
of innovation processes for spatial development (Ewers and Wettmann,
1980; Camagni and Cappellin, 1985; Nijkamp, 1986) and that on indus-
trial districts (Bagnasco, 1977; Brusco, 1982; Becattini, 1990) and milieux
innovateurs developed by the GREMI (Camagni, 1991) helped the under-
standing of the necessity of valorization of local entrepreneurship, even
if weak and of a mainly handicraft nature; of collaboration between the
public administrations and the organizations of entrepreneurs, devising
common projects and schemes addressed towards the strengthening of
the local productive vocation; the necessity of concentrating in single
sites different policy tools, from economic and fiscal incentives to educa-
tion and training services and infrastructure provision; the necessity of
stimulating innovation more than simply efficiency (Camagni,1995). But
the reflection on local milieux also helped in drawing attention once again
to the characteristics of the territory: not only to its physical configuration
(infrastructure, accessibility, environmental amenities) but especially to
its social and relational aspects. Incentives to innovation through inter-
firm cooperation; provision of local schemes developed in privatepublic
partnership; orientation of public funding towards specialized territorial
districts; all these measures represented policy innovations during the
1980s and 1990s, recently revisited through the new tool represented by
the territorial pacts.
Since the mid-1980s, the knowledge base as a source of continuous
innovation was highlighted as one of the major factors for long-term
growth (Knight, 1984; Castells, 1985; Tatsuno, 1986). Also in this case,
different stages were experimented with in policy strategies: interventions
addressed directly to the R&D functions of firms; interventions establish-
ing facilities and places devoted to knowledge construction outside firms
(technology poles, science parks, etc.), and more recently interventions
addressed towards creating a local atmosphere conducive to innovation,
through wide training programmes dealing with general education and
extensive use of new technologies, inside and outside firms, or building
advanced schemes in public/private partnership (Ewers and Allesch,
1990; Cooke and Morgan, 1998). The core issue is increasingly placed
on enhancing local creativity (Andersson, 1985), recreating the particu-
lar atmosphere felt (in particular periods) in such innovative milieux as
Silicon Valley, Orange County, the Swiss Jura and the Italian industrial
districts. Along the same logical trajectory, trust (Knack and Keefer,
1997), relational capital (Camagni, 1999; Camagni and Capello, 2000)

and collective learning (Camagni, 1991; Lundvall and Johnson, 1994;

Capello, 1999a and 1999b; Keeble and Wilkinson, 1999) are more recently
theorized as the main factors or preconditions for a creative environment
that calls for completely new policy styles, still to be fully understood and
developed.
We have spoken about strategy enrichment, in the sense that previ-
ous goals and keywords were not really abandoned subsequently: rather,
the policy vision was enriched and widened, the main limits of previous
policies corrected, emphasis shifted to new issues, on a path towards com-
plexity and integration. Spatial development is increasingly understood
as a complex, multi-dimensional phenomenon, and the illusion about the
existence of simple, short-cut strategies progressively abandoned.
As stated before, a general long-term trend is apparent, in the direction
of some new target elements for a renewed and modern territorial develop-
ment policy:
intangible factors, like human capital and knowledge, and the

operators that could translate their potential in actual growth
projects;
relational factors, creating synergies, promoting cooperation and
partnership, exploiting the richness of local relationships that define
a productive vocation, a local know-how and a local culture: social
and relational capital;
advanced communication networks and communication services, in
order to get a global reach on markets, information, business oppor-
tunities: public goods (but also toll goods provided by private
bodies) addressed towards the efficiency of territory.
But also a change in policy styles is needed, residing in the goal of:
preparing territories for innovation, enhancing their adaptability to

a changing external context, promoting their openness and recep-
tivity to new business ideas and organizational styles, rather than
forcing the locational decisions of single firms;
negotiating the terms for a fruitful cooperation between territories
and firms, rather than just supplying favourable location factors;
reinterpreting a bottom-up, generative approach to development
rather than a top-down, competitive one where regions and cities
fight among each other for the attraction of a given (and increas-
ingly scarce) amount of public resources and private investments, in
a zero-sum game.

EMPIRICAL EVIDENCE FROM THE MASST MODEL:

MACROECONOMIC AND REGIONAL POLICIES IN
THE EU
A Brief Introduction to the MASST Model
The econometric model estimating the system of causeeffect relationships

termed the MASST (MAcroeconomic, Sectoral, Social and Territorial)
model is a new one, conceptually defined for the purpose of investigat-
ing regional growth, its determinants and its territorial evolution. It draws
on the most advanced theories of regional growth, without denying the
importance of the achievements accomplished by the traditional theories.
MASST explains relative regional growth through territorial and spatial
factors such as agglomeration economies, territorial capital and spatial
spillovers (i.e., the influence of each region on the growth trajectories of
neighbouring regions). These factors determine the cumulative nature of
regional growth patterns, as widely emphasized by the new endogenous
growth theories and the new economic geography rooted in Myrdals
(1957) and Kaldors (1970) cumulative causation theory.9
Social elements (demographic change due to natural population change
or migration flows) are included in MASST and have a role in explaining
regional growth patterns together with the widely recognized factors of
local competitiveness, namely accessibility, presence of human capital and
local resource endowment.
The MASST model is structured as follows. It comprises two blocks
of equations, one explaining national growth, and the other explaining
regional differential growth. The sum of the two provides, by definition,
total or absolute regional growth. This structure differs substantially from
the existing econometric regional growth models, which in general move
towards a direct interpretation of absolute regional growth either by
replicating national macroeconomic models, or by constructing complex
systems of equations for each region linking the region to both the
national aggregate economy and to the other regional economies through
inputoutput technical coefficients.
The advantage of the MASST models structure is that a strong inter-
connection between regional and national growth is established: national
macroeconomic trends and policies generate an effect on both national
and regional growth, but at the same time regional structures and policies
affect both regional and national performance in an interactive national
regional manner. This structure allows account to be taken of complex
vertical feedbacks between the regional and national economy without
imposing a complex system of interlinked equations.

The most innovative aspect of MASST is that, thanks to its simulation

algorithm, it can be considered a generative model of regional growth
in the sense defined by Richardson (1969), even if it also encompasses
macroeconomic and institutional aspects, which are typically national and
top-down. In MASST, regional growth has a role in determining national
performance. The model thus supersedes the limiting and erroneous role
given in general to the regional side of growth models: that of simply dis-
tributing national growth among regions in a typical top-down approach.
The scenario-building methodology on which simulations lie is based
on identification of the institutional, socio-demographic and economic
driving forces of change, and their possible alternative trajectories, which
give rise to different opportunities for growth and patterns of territorial
distribution. Once the driving forces have been identified, they are trans-
lated into quantitative assumptions concerning the independent causal
variables of the forecasting model. Consequently, growth rates of GDP
and population as well as their levels for each year up to 2015 are simu-
lated. The aim is not to achieve precise quantitative values of economic
elements, nor, on the other hand, is it merely to provide a qualitative
image of what the economic system will look like; the aim is to show the
main trends and relative behavioural paths that will be at work under
specific assumptions on how the main driving forces of change will evolve.
Both the values assigned to the target variables and the regional values
emerging from the final results indicate an order of magnitude and some
relative behavioural classes (high-medium-low increase or decrease),
rather than precise quantitative values. The results of the simulations pre-
sented here are based on a baseline scenario built on the assumption that
the present trends affecting growth and the associated policies put in place
will continue in the future.
The Regional Impact of Macroeconomic Policies
Macroeconomic policies are usually considered to be neutral in terms of

inter-regional disparities. Yet, for inter-sectoral reasons, selective regional
effects may be expected: capital-intensive vs. labour-intensive, exposed vs.
non-exposed sectors to international trade, are differently hit by policies
having effects on real interest rates or exchange rate movements. These
differentiated effects have been demonstrated by means of targeted simula-
tions with the MASST model.10 In fact, by focusing on the spatial effects of
specific national policies and changes in macroeconomic contexts, it was
possible to disentangle the effects of single policy trends and, in particu-
lar, of a revaluation of the EU currencies generated by monetary policy
or macroeconomic conditions. In general terms, a revaluation would

negatively affect growth in all regions of Europe, but less intensively so in

the east, where cost competitiveness is higher for historical reasons (Figure
11.3).
The Impact of EU Policies: the Case of Structural Funds
Figure 11.4 concerns the impact of structural funds on regional growth.

The map shows the regional average growth rate between 2002 and 2015 if
structural funds are set to 0. The impact is easy to envisage, for structural
funds provide direct support to peripheral regions. The figure shows that
the Objective 1 regions suffer the most from the lack of structural funds.
All Objective 1 regions lose growing capacity: all Greece, most of Spain,
Ireland, north of Great Britain, the north of Scandinavian countries, and
the south of Italy retain a negative growth rate. Eastern countries are also
penalized by the lack for structural funds that in the baseline simulation
are also devoted to them.
What is less immediate to envisage is that the lack of structural funds
has an indirect negative impact on most European regions, since in the
regional integration perspective like the one modelled by MASST, the
negative growth rates of peripheral regions pervade the national eco-
nomic systems (via regional spillovers and national growth rates); nega-
tive growth rates are registered in most central regions, like most of the
French regions, German and Dutch regions, in the UK, in Scandinavian
countries. The very few exceptions regard the regions with capital cities or
with a dynamic city located within them; it is the case of London and some
of its surrounding regions, of Paris, Helsinki, Milan, Rome, Bologna,
Amsterdam and Rotterdam, just to quote the most relevant ones.
Territorial Capital and Regional Growth
Figure 11.5 contains the results of the contribution of territorial capital

to regional growth. Territorial capital is presented here in four catego-
ries, in particular social overhead capital, receptivity,11 entrepreneurship,
creativity all elements on which appropriate and renewed territorial
policies may impinge. The first important result is that each component
produces a rather different picture. The map in Figure 11.5a shows the
contribution of social overhead capital to regional growth. Peripheral
areas in Northern Europe (the Scandinavian countries and Scotland)
together with most areas in Eastern countries receive the highest contri-
bution to regional growth from transport infrastructure, thanks to efforts
to overcome remoteness. However, also, central regions in the Pentagon
(London-Paris-Milan-Munich-Hamburg), some advanced regions in

Politecnico di Milano - MASST model - December 06

Effects of a revaluation of the exchange rate
0.199 0.184
0.184 0.174
0.174 0.159
0.159 0.134
0.134 0.108
0.108 0.086
0.086 0.057
0.057 0.015
EU mean = 0.133
Figure 11.3 Spatial effects of a revaluation of the exchange rates across

EU regions

Politecnico di Milano - MASST model - June 2007

Difference in average annual % GDP growth rate 2002-2015 in Baseline Scenario if Structural Funds are set to 0
< 0.2418
0.2418 0.1493
0.1493 0.0905
0.0905 0.0505
0.0505 0.0318
0.0318 0.0178
0.0178 0
>0
Figure 11.4 The impact of structural funds on regional growth rates

across EU regions
northern Italy, some agglomerated areas like Barcelona and Madrid in

Spain, and Porto and Lisbon in Portugal perform relatively well thanks
to the infrastructure increase, probably thanks to the efforts to overcome
congestion.

Social overhead capital

0.523 0.334
0.333 0.202
0.201 0.148
0.147 0.106
0.105 0.066
0.065 0.019
0.018 0.031
0.32 0.092
(a)
Source: The authors, derived from MASST model.
Figure 11.5 Contribution of territorial capital to regional growth across

EU regions: (a) social overhead capital
A different picture is presented by the map in Figure 11.5b, where the

role of receptivity in regional growth (i.e., that part of regional growth
that is dependent on the performance of neighbouring regions, like a
sort of growth spillover) is highlighted. The map shows that receptivity
has a positive effect on growth in central, Pentagon Europe, spreading
around major capital and mega regions (London, Paris, Milan, Munich,

Receptivity
0.418 0.127
0.126 0.001
0.002 0.267
0.268 0.523
0.524 0.713
0.714 0.904
0.905 1.049
1.050 1.203
(b)
Figure 11.5 (continued) (b) receptivity
Brussels, etc.), but, interestingly enough, also scattered towards more

peripheral territories.
The map in Figure 11.5c is devoted to the contribution provided by
entrepreneurship to regional growth. It is evidently shown that the role
of entrepreneurship in eastern regions growth is rather limited, and it
is instead more important in peripheral countries of the 15 old member
states, like Italy, Spain and Greece. A relatively important role is also
played by entrepreneurship in some regions of the Pentagon area. As
expected, entrepreneurship plays a limited role in the dynamics of capital

Entrepreneurship
0.140 0.244
0.245 0.330
0.331 0.429
0.430 0.535
0.536 0.676
0.677 0.866
0.867 1.157
1.158 2.097
(c)
Figure 11.5 (continued) (c) entrepreneurship
regions, the latter being probably more influenced by the presence of

value-added functions, not represented by a variable of self-employment.
Interestingly enough, the map in Figure 11.5d presents a rather unex-
pected result for what concerns the contribution of creativity to regional
growth: a decisive contribution of creativity to regional growth is shown
in eastern regions, while in the 15 old member states the highest values are
registered in some regions of the Pentagon.
If all effects are counted together, the outcome is the one presented in
the map in Figure 11.5e. Territorial capital provides an important contri-
bution to regional growth in most Pentagon regions, in some peripheral

Creativity
0.276 0.238
0.237 0.196
0.195 0.008
0.009 0.024
0.025 0.056
0.057 0.137
0.138 0.233
0.234 0.424
(d)
Figure 11.5 (continued) (d) creativity
areas, like Greece and part of Spain, and France. It is evidently lacking
in eastern countries, in peripheral countries in the north (Scandinavian
countries, UK and Italy). Interestingly enough, capital regions, like
Madrid, Lisbon, Paris, Athens, London, Copenhagen, Oslo, Helsinki,
and important agglomerated regions like the regions in Northern Italy,
Barcelona, Cte dAzur, receive a small contribution from territo-
rial capital to regional growth. We are inclined to say that, as with all
productive factors, territorial capital also shows decreasing marginal
productivity.

Total effect
0.261 0.128
0.129 0.413
0.414 0.655
0.656 0.869
0.870 1.044
1.045 1.236
1.237 1.482
1.483 2.514
(e)
Figure 11.5 (continued) (e) total effect
CONCLUSIONS
The objective of this chapter was twofold: on the one hand to present the
rationale for regional competitiveness policies, highlighting the theoreti-
cally based reasons for their existence; and on the other to present a his-
torical development perspective of regional policy targets and styles.
The main messages stemming from the analysis can be summarized as
follows. The persistence of regional disparities justifies the relevance of

regional policies; regions compete on the basis of absolute advantages,

and therefore need solid policy tools to enhance their competitiveness.
The policies for regional competitiveness are both national and regional
policies: the former are generally felt to be space-invariant policies; what
we claim in this chapter is that, on the contrary, they have a different
impact according to the structural characteristics of regions; and the latter
require a new approach, widening their scope in the direction of soft and
knowledge components of territorial capital.
Based on simulation exercises through the MASST model, the chapter
presents empirical evidence on the impact of:
macroeconomic policies, and in particular a devaluation policy;

the impact of a European policy, and in particular the impact of
structural funds;
the role of territorial capital elements on regional growth patterns.
The results support and highlight important elements. Macroeconomic

policies are not neutral in terms of inter-regional disparities; for inter-
sectoral reasons, selective effects are visible. In fact, a real revaluation of
the currency generates wider tensions in industries and regions character-
ized by a greater degree of openness to international trade. Furthermore,
revaluation is likely to hit more, ceteris paribus, those regions specialized
in labour-intensive industries, because it raises labour costs expressed in
international currency (while the cost of capital will remain unchanged at
the level determined internationally, especially if the revaluation is indica-
tive of a strong and potentially fast-growing economy). Moreover struc-
tural funds turn out to provide direct support to peripheral regions, and
to have an indirect positive impact on most European regions, due to the
inter-regional integration that characterizes the economy.
The simulation exercise clearly shows that in those regions where ter-
ritorial capital assets play an important role on regional growth, the
overall performance of the regions is higher. Moreover, it clearly demon-
strates that territorial capital, as with all production factors, is subject to
strong decreasing returns to scale: in fact, in those regions (agglomerated
and mega-regions) in which the level of territorial capital is higher, its
effects on regional growth are more contained. A last interesting aspect
emerges: the different factors of territorial capital analysed (receptivity,
social overhead capital, entrepreneurship and creativity) play a differ-
ent role on local growth according to the settlement structure and rela-
tive location of regions. All this allows us to say that territorial capital
appears as a new, fruitful concept on which modern territorial policies
have to be built.

NOTES
1. The classical equilibrating process relies on downward flexibility of prices and wages,
which is hampered by the existence of national wage contracts in both private and
public structures and by the homogeneity of import prices (remembering that regions
are very open economies). The second, modern process relies on the devaluation of the
currency, and it is automatically excluded in an inter-regional context.
2. The importance of this last reflection is magnified by the consideration that, with the
creation of exchange rate agreements and large monetary unions, states will increas-
ingly lose control over the external value of their currency, and will increasingly behave
inside the unions like regions inside a nation.
3. Krugman speaks of a general competitiveness obsession, mainly referring to the case
of countries.
4. From a modelling point of view, already 15 years ago it was shown that, if the linearity
assumptions of the traditional neoclassical model were abandoned and some non-
linearities and increasing returns introduced, the same neoclassical model could well
accommodate varied possible outcomes both spatial diffusion and concentration
(Miyao, 1987); and also endogenous growth models, which include cumulative pro-
cesses into a neoclassical production function (Romer, 1986), mainly end up with an
inter-regional divergence process.
5. The Italian experience of the first economic boom, 195764, totally concentrated in
only three areas Milan, Turin and Genoa, the north-western triangle, is telling in this
respect: wage and cost inflation, massive migrations from the south, social and political
tensions, deterioration of environmental quality due to excessive speed in economic
transformation, generated the anticipated end of the expansion phase, an early vanish-
ing of previous competitiveness and 15 years of slow growth in these areas up to the end
of the following decade. History might repeat itself.
6. In the simulation experiment of regional growth in the EU up to 2015, realized through
the MASST model (see later in the chapter) this fact is evident, almost all countries
showing a standard deviation in inter-regional growth rates lower than the interna-
tional standard deviation of growth rates of EU-27 countries. This is particularly true
for the 15 old member countries (with the exception of Ireland and the partial exception
of Spain and Portugal, showing an internal standard deviation similar to the interna-
tional one). On the other hand, all 12 new member countries show a higher variability
of internal regional growth rates, bringing support to expectations la Williamson
about the increasing inter-regional disparities in the first phase of a development or
integration process.
7. This general trend is generated by two differentiated evolutions inside the two blocks
of EU countries. In fact, in the 12 new member countries, inter-regional disparities will
increase, due to a strongly widening intra-national dualism between core and periphery
regions, insufficiently counterbalanced by slowly decreasing international disparities.
On the other hand, a partly different trend is likely to characterize the 15 old member
countries: here too intra-national disparities will increase, though at a much slower pace
than in the 12 new member countries, while international disparities will also increase,
for the first time, though only slightly. Some countries in fact, like Ireland, have already
more than completed their catching-up process, and their persisting miracle will widen
international disparities; other countries, relatively wealthy, like Holland and Italy, are
losing momentum, while some others, less advanced, like Greece or partially Portugal,
are not expected to show convincing performances (see Figure 11.1a and 11.1b).
8. An operational methodology for Territorial Impact Assessment of EU policies, along
the lines suggested by the Commission, enphasizing the three dimensions of territo-
rial cohesion, namely territorial efficiency, quality and identity, was recently devel-
oped with the TEQUILA model Territorial Efficiency, Quality, Identity Layered
Assessment; see Camagni (2006).
9. For a wide explanation of MASST, see Capello (2007); Capello et al. (2008).

10. These simulations were performed as part of the ESPON 3.4.2 Project exploring the
Territorial Impact of EU Economic Policies. For the complete report, see the ESPON
website, www.espon.eu; accessed 27 May 2010.
11. Receptivity is here defined as that part of regional growth that is dependent on the
performance of neighbouring regions, like a sort of growth spillover.
REFERENCES
Andersson, A. (1985), Creativity and Regional Development, Papers of the
Regional Science Association, 56(1), 520.
Arkleton Institute, University of Aberdeen (2006), The Territorial Impact of CAP
and Rural Development Policy, ESPON Project 2.1.3., available at: www.espon.
eu accessed 27 May 2000, Luxembourg.
Armstrong, H.W. (2002), Regional Policy: Reconciling the Convergence and
Evaluation Evidence, in J.R. Cuadrado and M. Parellada (eds), European
Union Regional Policy: Reconciling the Convergence and Evaluation Evidence,
Springer Verlag, Berlin, pp. 23172.
Armstrong, H.W. and Taylor, J. (1993), Regional Economics and Policy, Harvester
Wheatsheaf London.
Bagnasco, A. (1977), Le Tre Italie, Il Mulino, Bologna.
Barro, R.J. and Sala-i-Martin, X. (1991), Convergence Across States and
Regions, Brookings Papers in Economic Activity, 22(1), 10782.
Barro, R.J. and Sala-i-Martin, X. (1992), Convergence, Journal of Political
Economy, 100(2), 22351.
Becattini, G. (1990), The Marshallian Industrial District as a Socio-economic
Notion, in F. Pyke, G. Becattini and W. Sengenberger (eds), Industrial Districts
and Inter-firm Cooperation in Italy, ILO, Geneva.
Borts, G.H. and Stein, J.L. (1964), Economic Growth in a Free Market, Columbia
University Press, New York.
Brusco, S. (1982), The Emilian Model: Productive Decentralisation and Social
Integration, The Cambridge Journal of Economics, 6(2), 16784.
Buchanan, J.M. and Tullock, G. (eds) (1982), Towards a Theory of the Rent
Seeking Society, A&M Press, Austin.
Camagni, R. (1991), Technological Change, Uncertainty and Innovation
Networks: Towards a Dynamic Theory of Economic Space, in R. Camagni
(1991), Innovation Networks: Spatial Perspectives, Belhaven-Pinter, London.
Camagni, R. (1992), Development Scenarios and Policy Guidelines for the
Lagging Regions in the 1990s, Regional Studies, 26(4), 36174.
Camagni, R. (1995), The Concept of Innovative Milieu and its Relevance for
Public Policies in European Lagging Regions, Papers in Regional Science, 74(4),
31740.
Camagni, R. (1999), The City as a Milieu: Applying GREMIs Approach to
Urban Evolution, Revue dEconomie Rgionale et Urbaine, 3, 591606.
Camagni, R. (2001), Policies for Spatial Development, in OECD (2001),
pp. 14769.
Camagni, R. (2002), On the Concept of Territorial Competitiveness: Sound or
Misleading?, Urban Studies, 39(13), 2395412.
Camagni, R. (2006), Territorial Impact Assessment TIA: A Methodological

Proposal, in Scienze Regionali Italian Journal of Regional Science, 5(2),

13546.
Camagni, R. (2008), Territorial Capital and Regional Development, in R. Capello
and P. Nijkamp (eds), Handbook of Regional Dynamics and Growth: Advances in
Regional Economics, Edward Elgar, Cheltenhan, UK and Northampton, MA,
USA, pp. 11832.
Camagni, R. and Capello, R. (2000), The Role of Inter-SME Networking
and Links in Innovative High-tech Milieux, in D. Keeble and F. Wilkinson
(eds), High-technology Clusters, Networking and Collective Learning in Europe,
Ashgate, Avebury, pp. 11855.
Camagni, R. and Capello, R. (forthcoming), Territorial Capital, Entrepreneurship
and Regional Development in the EU, Regional Studies.
Camagni, R. and Cappellin, R. (1985), Sectoral Productivity and Regional Policy,
CEC, Brussels.
Camagni, R. and Gibelli, M.C. (1996), Cities in Europe: Globalisation, Sustain-
ability and Cohesion, in Presidenza del Consiglio dei Ministri Dipartimento
Politiche Comunitarie, European Spatial Planning, Rome, Poligrafico dello Stato.
Camagni, R. and Maillat, D. (eds) (2006), Milieux Innovateurs: Thorie et Politiques,
Economica, Paris.
Capello R. (1999a), A Measurement of Collective Learning Effects in Italian
High-tech Milieux, Revue dEconomie Rgionale et Urbaine, 3, 44968.
Capello, R. (1999b), Spatial Transfer of Knowledge in High-technology Milieux:
Learning vs. Collective Learning Processes, Regional Studies, 33(4), 35365.
Capello, R. (2001), Urban Innovation and Collective Learning: Theory and
Evidence from Five Metropolitan Cities in Europe, in M.M. Fischer and
Froehlich J. (eds), Knowledge, Complexity and Innovation Systems, Springer,
Berlin, Heidelberg, New York, pp. 181208.
Capello, R. (2007), A Forecasting Territorial Model of Regional Growth: The
MASST Model, Annals of Regional Science, 41(4), 75387.
Capello, R., Camagni, R., Chizzolini, B. and Fratesi, U. (2008), Modelling
Regional Scenarios for the Enlarged Europe: European Competitiveness and
Global Strategies, Springer Verlag, Berlin.
Castells, M. (1985), High Technology, Space and Society, Sage, Beverly Hills.
CEC Commission of the European Communities (2005a), Territorial State and
Perspectives of the European Union, Scoping Document and Summary of Political
Messages, Brussels, May.
CEC Commission of the European Communities (2005b), Cohesion Policy
in Support of Growth and Jobs: Community Strategic Guidelines, 20072013
(Communication from the Commission), COM(2005).
Cooke, P. and Morgan, K. (1998), The Associational Economy. Firms, Regions and
Innovation, Oxford University Press, Oxford.
Dixon, R.J. and Thirwall, A.P. (1975), A Model of Regional Growth Rate
Differences on Kaldorian Lines, Oxford Economic Papers, 27(2), 20114.
Dosi, E. (1982), Technological Paradigms and Technological Trajectories: A
Suggested Interpretation of the Determinants and Directions of Technical
Change, Research Policy, 11(3), 14762.
Ewers, H.J. and Allesch, J. (eds) (1990), Innovation and Regional Development, De
Gruyter, Berlin.
Ewers, H. and Wettman, R. (1980), Innovation Oriented Regional Policy,
Regional Studies, 14, 16179.

Foray, D. (2000), LEconomie de la Connaissance, La Dcouverte, Paris.

Gilly, J.P. and Torre, A. (eds) (2000), Dynamiques de Proximit, LHarmattan,
Paris.
GREMI (1995), Cohesion and the Development Challenge Facing the Lagging
Regions, Regional Development Studies, No. 24, directed by R. Camagni and
M. Quevit, European Commission, Bruxelles.
Hirschman, A.O. (1957), Investment Policies and Dualism in Underdeveloped
Countries, The American Economic Review, 47(5), 55070.
Kaldor, N. (1970), The Case for Regional Policies, Scottish Journal of Political
Economy, 17(3), 33748.
Keeble, D. and Wilkinson, F. (1999), Collective Learning and Knowledge
Keeble, D. and Wilkinson, F. (eds) (2000), High-technology Clusters, Networking
and Collective Learning in Europe, Ashgate, Aldershot.
Knack, S. and Keefer, P. (1997), Does Social Capital have an Economic Payoff? A
Cross-country Investigation, Quarterly Journal of Economics, 112(4), 125188.
Knight, R. (1984), The Advanced Industrial Metropolis: A New Type of World
City, paper presented at the seminar The Future of the Metropolis, Berlin,
October.
Krugman, P. (1996), Making Sense of the Competitiveness Debate, Oxford
Review of Economic Policy, 12(3), 1725.
Krugman, P. (1998), Pop Internationalism, MIT Press, Cambridge, MA.
Krugman P. and Venables (1996), Integration, Specialisation and Adjustment,
European Economic Review, 40, 85780.
Lundvall, B.A. and Johnson, B. (1994), The Learning Economy, Journal of
Industrial Studies, 1(2), 2342.
Ministers for Spatial Planning (2007), Leipzig Charter on Sustainable European
Cities, Informal Meeting of Ministers, Leipzig, May.
Miyao, T. (1987), Dynamic Urban Models, in E. Mills (ed.), Urban Economics;
Handbook of Regional and Urban Economics, North-Holland, Amsterdam, vol.
2, pp. 877925.
Myrdal, G. (1957), Economic Theory and Underdeveloped Regions, Gerald
Duckworth & Co., London.
Nelson, R. and Winter, S. (1982), An Evolutionary Theory of Economic Changes,
Harvard University Press, Cambridge, MA.
Nijkamp, P. (ed.) (1986), Technological Change, Employment and Spatial
Dynamics, Springer Verlag, Berlin.
OECD (2001), Territorial Outlook, Paris.
Pennington, M. (1999), Free Market Environmentalism and the Limits of Land
Use Planning, Journal of Environmental Policy & Planning, 1(1), 4359.
Parr, J. (1999), Growth Pole Strategies in Regional Economic Planning: a
Retrospective View. Part 2. Implementation and Outcome, Urban Studies,
36(8), 124768.
Richardson, H.W. (1969), Regional Economics, World University, Redwood Press
Limited, Trowbridge, Wiltshire.
Romer, P. (1986), Increasing Returns and Long-run Growth, Journal of Political
Economy, 94(5), 100237.

Rosenstein-Rodan, P.N. (1943), Problems of Industrialisation of Eastern and

South-Eastern Europe, The Economic Journal, 53(210/211), 20211.
Simon, H. (1972), From Sustantive to Procedural Rationality, in C.B. McGuire
and R. Radner (eds), Decision and Organization, North Holland, Amsterdam,
pp. 11676.
Stohr, W. (ed.) (1990), Global Challenge and Local Responses, United Nations
University Press/Mansell, London.
Storper, M. (1995), The Resurgence of Regional Economies Ten Years Later: The
Region of Untraded Interdependencies, European Urban and Regional Studies,
2, 191221.
Tatsuno, S. (1986),The Technopolis Strategy, Prentice Hall, New York.
Williamson, J.G. (1965), Regional Inequality and the Process of National
Development: A Description of the Patterns, Economic Development and
Cultural Change, 13(4), 345.
Williamson, O. (2002), The Lens of Contract: Private Ordering, American
Economic Review, Papers and Proceedings, 92(2), 43853.

12. Endogenous employment growth
and decline in Australian capital
city statistical divisions*
Alistair Robson
INTRODUCTION
The Australian economy has experienced significant changes in its indus-

trial structure in recent decades (Productivity Commission, 1998, p. xiv).
Broadly, there has been a large shift in employment from the manufac-
turing industry to the services industries. Such change is similar in many
other OECD countries. Research on regional structural economic change
is copious, and the Productivity Commission report of 1998 provides a
good overview albeit 12 years old on the topic. There are, however,
considerable regional differences in the nature and magnitude of changes
in employment structures across industry sectors. In this chapter the
reasons for these changes are considered, with the focus being explicitly
on the nations state capital city metropolitan areas with populations
exceeding 1 million (hereafter referred to as the five-capital cities). Those
five-capital cities in descending order of population size are: Sydney
(the capital city of the state of New South Wales); Melbourne (Victoria);
Brisbane (Queensland); Perth (Western Australia); and Adelaide (South
Australia).
The chapter begins with an overview of the economic structure of the
state capital cities. That is followed by a shift-share analysis of change in
employment in industry sectors over the decade 1996 to 2006 to explain
the divergent employment growth between the cities. The emphasis is
on analysing the regional shift component as a surrogate measure of
endogenous performance of the industry sectors within each of the capital
cities. The chapter concludes with a summary of the differences between
the cities in their endogenous employment performance for the decade
1996 to 2006.
237

OVERVIEW OF THE ECONOMIC STRUCTURE OF

THE CAPITAL CITIES
Capital City Dominance
In Australia the capital city regions1 are dominant in terms of popula-

tion, employment and production. They are effectively primate cities in
their respective states according to the Jefferson (1939) criteria. The latest
available data from the 2006 Census of Population and Housing shows
these regions accounted for 63 per cent of the nations population and
65 per cent of total employment.2 However, historically the capital city
regions have not always been so dominant. For example, the capital city
regions accounted for only 36 per cent of Australias population in 1901
(Australian Bureau of Statistics, 1997, p. 18). But since that time they have
reached their current level of dominance (see Figure 12.1).
The dominant pattern of the five-capital cities is evident in all states, but
it is less so Queensland, which is a comparatively special case where the
regions to the south, west and north3 of Brisbane have strong economic
linkages to each other that is not as evident in other states capital city
regions (such as in Sydney with Newcastle and Wollongong). Referred to
as the South East Queensland region, it accounts for about two-thirds of
both the population and the employment in the state, a level broadly on a
par with the other capital city regions.
The capital cities have a high employment to working age population.
And they also tend to have relatively higher income per capita compared
with other non-metropolitan regions. However, in an analysis of regional
economic change in Australia over the period 1991 to 1996, Harding et
al. (2000) found a divergence in economic performance as measured by
incomes between the capital cities.
Changing National Shares
Ideally, production data for the five-capital cities would be used in an

analysis of their economic performance. However, this data does not
exist for the capital city regions. But state production data does exist
and it provides a context for capital city employment changes. Over the
period 1996 to 2006, falls in the share of national real Gross Domestic
Product were recorded in New South Wales (down 1.4 percentage points
to 34 per cent), South Australia (down 0.5 of a percentage point to 7
per cent), Victoria (down 0.3 of a percentage point to 24 per cent) and
Tasmania (down 0.2 of a percentage point to 2 per cent). Increases in
the share were recorded in Queensland (up 1.9 percentage points to 18

Endogenous growth and decline in Australian capital cities 239
2.0
1.5
1.0
Change in share (p.p:s)
0.5
1 2 3 4 6 10 11 19 20
0.0
5 7 8 9 12 13 14 15 16 17 18
0.5
1.0
2.0
Industry
1 Agriculture/forestry & fishing 11 Financial & insurance services
2 Mining 12 Rental/hiring & real estate services
3 Manufacturing 13 Professional/scientific & technical services
4 Electricity/gas/water & waste services 14 Administrative & support services
5 Construction 15 Public adminstration & safety
6 Wholesale trade 16 Education & training
7 Retail trade 17 Healthcare & social assistance
8 Accommodation & food services 18 Arts & recreation services
9 Transport/postal & warehousing 19 Other services
10 Information media & telecommunications 20 Inadequately described/Not stated
Source: Compiled by the author using data available from the Australian Bureau of
Statistics (2007).
Figure 12.1 Percentage point change in capital Cities share of national

employment by industry, 19962006
per cent) and Western Australia (up 0.6 of a percentage point to 12 per
cent).
The shifts in production over this period are largely replicated in the
shifts in employment. The states losing national share of employment were
New South Wales (1.6 percentage points), South Australia and Tasmania
(both 0.2 of a percentage point). The states recording an increase in their
share of national employment were Queensland (1.6 percentage points),
Western Australia (0.3 of a percentage point) and Victoria (0.2 of a
percentage point).
Capital City Shares of State Population and Employment
Over the decade from 1996 to 2006 there was little change in the rela-
tive dominance of the capital city regions compared with their states as a

Table 12.1 Capital city shares of state employment and population, 2006
Employment Working Age Employment

Population to Working
(1564) Age Population
(1564)
Ratio
Level in Change Level in Change Level in Change
2006 19962006 2006 19962006 2006 19962006
(% of (pps of (% of (pps of (% of (pps of
state) state) state) state) state) state)
Sydney 65 0.2 65 1.2 101 2.2
Melbourne 74 0.3 74 1.2 100 1.2
Brisbane 47 0.4 45 0.6 103 0.4
Adelaide 74 74 0.1 100 0.1
Perth 74 1.4 73 0.6 101 1.1
Five-capital city 65 65 0.6 101 1.0
weighted average
Source: Compiled by the author using data available from the Australian Bureau of
Statistics (2007).
whole in terms of population and employment (see Table 12.1). In 2006,

the five-capital cities aggregate share of state employment was 65 per cent
(unchanged from 1996), the same as their share of state population (up 0.6
of a percentage point from 1996). Overall, the ratio of employment to the
working age population (1564) was 1.0 per cent higher in the five-capital
cities than their states, a fall of 1.0 percentage points from 1996, hence the
patterns detected in the state data can be inferred as a good guide to the
performance of the capital city regions.
Indeed, over this time no capital city, or rest of state area, recorded a fall
in either population or working age population (the age groups 15 to 64
years). Very prominent was the relatively large increase of 30 per cent in
employment and 20 per cent in the working age population in Queensland
over the decade to 2006, which was considerably above the national
figures of a 19 per cent and 13 per cent increase respectively. The stronger
growth in Queenslands employment compared with working age popula-
tion was reflected in the 6.9 per cent rise in its employment to working age
population ratio. That compared with 5.3 per cent for Australia as a whole
and 4.2 per cent for the five-capital cities aggregate average. That stronger
growth in the employment to working age population ratio in Queensland
occurred even though it was equal to the national average in 1996.

Divergent Rates of Growth and Change in the Capital Cities
One explanation for the divergent growth rates of employment is the

different industrial structures of the capital city economies. Some of the
five-capital cities may have relatively more employment in industries that
are growing. Comparatively, other capital cities may have relatively more
employment in industries that are declining.
As shown in Table 12.2, the industry sector with the most variation in
employment share in 1996 was the manufacturing sector. The lowest share
of employment in the manufacturing industry was in Perth (10 per cent
of total employment), followed by Brisbane (11 per cent of total employ-
ment). Comparatively, Melbourne had the highest proportion of employ-
ment in the manufacturing industry (17 per cent), followed by Adelaide
(15 per cent) and Sydney (12 per cent). Employment in the manufacturing
industry in the rest of Australia represented 10 per cent of total employ-
ment. This represents one of the major structural differences between
capital city regions and the rest of Australia.
By 2006, the manufacturing industry sector recorded some relatively
large falls in share of employment (see Table 12.3). In Melbourne, its share
fell by 3.7 percentage points over the decade to 2006 the largest percent-
age share fall of any industry by capital city region. The share of employ-
ment in the manufacturing industries sector also fell in Sydney (by 2.2
percentage points), in Adelaide (by 2.1 percentage points), Brisbane (by
0.5 percentage points) and Perth (by 0.4 percentage points). Over this same
period, the share of employment in the five-capital cities increased the
most in the construction industry (up 1.4 percentage points). The largest
share increases were in Perth (up 2.0 percentage points) and Adelaide (up
1.9 percentage points).
Nationally, overall the most prominent changes in the share of national
employment over the decade 1996 to 2006 were in the manufacturing
industries (down 1.6percentage points), and in both the agriculture, for-
estry and fishing industries and the wholesale trade industries (down 1.2
percentage points). Comparatively, there were increases in the share of
national employment over the same period in the construction industry
(up 1.6 percentage points), the retail trade industry (up 1.2 percentage
points), and the health care and social assistance industries (up 1.1 per-
centage points).

Table 12.2 Capital city industry structure share of employment, 1996
Sydney Melbourne Brisbane Adelaide Perth Five-capital Rest of

(%) (%) (%) (%) (%) cities Australia

(%) (%)
Agriculture\ forestry & fishing 1 1 1 1 1 1 10
Mining 0 0 0 0 2 0 2
Manufacturing 12 17 11 15 10 13 10
Electricity\ gas\ water & waste 1 1 1 1 1 1 1
services
Construction 6 6 6 5 7 6 6
Wholesale trade 7 6 6 6 6 6 4
Retail trade 10 10 10 10 11 10 10
242
Accommodation & food services 6 5 6 6 6 6 7
Transport\ postal & warehousing 6 5 5 4 4 5 4
Information media & 4 3 2 2 2 3 2
telecommunications
Financial & insurance services 6 5 4 4 4 5 2
Rental\ hiring & real estate 2 1 2 2 2 2 1
services
Professional\ scientific & technical 8 8 7 6 7 7 4
services
Administrative & support services 3 3 3 3 3 3 2
Public administration & safety 5 5 7 6 6 5 7
Education & training 7 7 8 8 8 7 8
Health care & social assistance 9 9 10 12 10 9 9
20/12/2010 15:12
Arts & recreation services 1 2 1 2 2 2 1
Other services 4 5 5 5 5 5 4
Inadequately described/Not 3 3 3 4 4 3 3
stated
243
Total employment (%) 100 100 100 100 100 100 100
Note: Figures rounded to nearest whole percentage point.
Source: Compiled by the author using data available from the Australian Bureau of Statistics (2007).
20/12/2010 15:12
Table 12.3 Capital city change in industry structure share of employment over the decade 1996 to 2006
Sydney Melbourne Brisbane Adelaide Perth Five-capital Rest of

(pps) (pps) (pps) (pps) (pps) cities (pps) Australia

(pps)
Agriculture\ forestry & fishing 0.2 0.3 0.4 0.1 0.5 0.3 2.6
Mining 0.0 0.1 0.1 0.2 0.9 0.1 0.1
Manufacturing 2.2 3.7 0.5 2.1 0.4 2.2 0.7
Electricity\ gas\ water & waste 0.0 0.1 0.3 0.2 0.1 0.1 0.0
services
Construction 0.9 1.6 1.5 1.9 2.0 1.4 1.9
Wholesale trade 1.1 0.7 1.7 1.6 1.6 1.2 1.2
Retail trade 0.9 1.3 0.9 1.6 0.8 1.1 1.3
244
Accommodation & food services 0.2 0.4 0.1 0.0 0.2 0.1 0.0
Transport\ postal & warehousing 0.2 0.1 0.1 0.0 0.2 0.0 0.0
Information media & 0.5 0.6 0.8 0.5 0.7 0.6 0.5
telecommunications
Financial & insurance services 0.4 0.0 0.2 0.0 0.5 0.0 0.2
Rental\ hiring & real estate 0.2 0.2 0.2 0.0 0.0 0.1 0.1
services
Professional\ scientific & 0.7 0.6 0.1 0.4 0.5 0.5 0.4
technical services
Administrative & support services 0.1 0.6 0.4 0.4 0.1 0.3 0.5
Public administration & safety 0.3 0.2 0.3 1.2 1.0 0.5 0.8
Education & training 0.6 0.6 0.2 0.0 0.1 0.4 0.2
Health care & social assistance 1.0 1.1 1.0 1.2 1.0 1.0 1.3
20/12/2010 15:12
Arts & recreation services 0.0 0.1 0.1 0.2 0.2 0.0 0.0
Other services 0.7 0.9 0.8 0.9 0.9 0.8 0.4
Inadequately described/Not 0.3 0.5 0.5 1.6 1.2 0.6 0.9
stated
245
Total employment 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Note: Figures rounded to nearest whole percentage point (0 refers to less than 0.5).
Source: Compiled by the author using data available from the Australian Bureau of Statistics (2007).
20/12/2010 15:12
A SHIFT-SHARE ANALYSIS OF EMPLOYMENT

CHANGE OVER THE DECADE 1996 TO 2006
As we will see later in the chapter, the changes in industry employment
structure in the five-capital cities discussed above explain little of the actual
employment growth performance of the capital city regions. One method
that allows us to explain the employment growth performance of these
regions is shift-share analysis. This technique allows for a decomposition
of change in employment that is attributable to:
the share of national growth (national shift effect, NS);

the mix of industries (industry mix effect, IM);
the shift change in activities toward (or away from) a capital city
region (differential or regional shift effect, RS).
The regional shift effect provides us with the net impact of factors affect-
ing employment from in the region; or alternatively the endogenous
factors. Hence the technique can provide the new insight that is central to
the concern of this chapter. These factors can be useful to planners and
researchers as they can be influenced more so than factors from outside
the region (national shift effect and industry mix effect that are exogenous
to the region).
Background and Method
Shift-share analysis has a long history of use in regional economic analy-

sis. It was developed to project future employment in a particular region
and is an off-shoot of the Constant Share Approach (for a discussion see
Wadley and Smith, 2003). Like most methods there has been much debate
on the efficacy and accuracy of the approach. There is also considerable
debate about how much of the regional shift effect, which is a residual,
truly reflects endogenous factors.
There have been numerous refinements to the shift-share technique,
particularly a version by Stilwell (1969), which provided a traditional
shift-share technique with a working tableau and relevant equations
(see Wadley and Smith, 2003, p. 1). In response to this, Chalmers (1971),
Edwards et al. (1978), Markusen et al. (1991) and Noponen et al. (1997)
have provided refinements to his method. While Wadley and Smith (2003,
p. 1) argue that the method still has a way to go, the benefits of the method
are its simplicity and low level of required data. Nonetheless, shift-share
analysis is a powerful technique and is widely used in regional econom-
ics. Recent uses in Australia include a study by Mitchell (2008) who used

labour force survey employment data to determine regional patterns of

employment change adjusting for part-time and full-time status. Labour
force survey data is not used in the research reported in this chapter as
data from the Census of Population and Housing is considered by the
author to be more robust and reliable (with the trade-off being fewer
observations). The Australian Bureau of Statistics makes very clear that
while it has attempted to improve the quality of employment data at geo-
graphic levels below the state and territory level, care should be taken in
the interpretation of regional estimates due to relatively high standard
errors (Australian Bureau of Statistics, 2009). Nonetheless, broad findings
from that study are similar to those shown here.
In the shift-share analysis in this chapter, the national shift, industry mix
and regional shift effects are calculated by using a modified shiftshare tech-
nique developed by Haynes and Dinc (1997). There are more advanced
methods such as using efficiency measures using production data but
the data needed to use these methods are unavailable at the capital city
level. In the Haynes and Dinc methodology the total change in employ-
ment over a time period is referred to as the total shift (TS). As with all
shift-share techniques, this methodology decomposes the total shift in
employment into its exogenous and endogenous shift effects. The exoge-
nous shift effects in this methodology are those effects attributable to both
the increase in national employment (known as the national shift effect,
NS) and the impact of growth in certain industries nationally (known as
the industry mix effect IM). Comparatively, the endogenous shift effect
(known as regional or differential shift effect, RS) is due to the impact of
factors within the local area. This can be shown mathematically as:
TS = NS + IM + RS
The NS effect is the change in employment in a certain industry that

would have taken place in employment had it grown at the same rate of
that for all industries nationally. It is calculated by multiplying the level of
employment in the local area by the growth rate of employment nation-
ally: thus:
NS = Ei gn
where:
NS = the national shift effect;

Ei = the employment in industry i at the initial year;
gn = the growth rate of employment nationally.

The IM effect is the change in employment in a certain industry that

would have take place if it had grown at the same rate as that industry
nationally: thus:
IM = Ei gi, n
where:
IM = the industry mix effect;

gi,n = the growth in employment in industry i nationally.
The RS effect (also known as the differential shift effect) is the residual
between the total shift and the combination of the national shift effect and
the industry mix effect. It can be calculated as either the difference between
those three, or more properly as:
RS = Ei (gi, r gi, n)
where:
RS = the regional shift effect;

gi, r = the regional growth rate of employment in that industry;
gi, n = the national growth rate of employment in that industry.
The RS effect can be derived in aggregate for a region, or its individual

industry components. The RS effects for each industry in a region are dif-
ferent from the IM effects.
Attempts to separate exogenous and endogenous shift effects on employ-
ment for Australian regions have been few (Mitchell, 2008, p. 3). Recent
work, apart from that by Mitchell, has been performed by Stimson et al.
(2005, 2009). The technique used here is the same as those publications,
albeit looking at capital city regions rather than non-metropolitan regions.
The analysis reported here uses data based on the place of enumera-
tion count methodology. This was because the Australian Bureau of
Statistics releases time series data on CD, only using this count method
for all Statistical Divisions.4 For consistency purposes this is the count
method used throughout this chapter. The implications of using this count
method is that employment data refer to people who are employed based
on where they were counted. As such, a higher level of employment does
not necessarily mean an area has more jobs in that area. This is not an

issue for large Statistical Divisions, such as the capital city metropolitan
areas, as they are largely self-contained. All data is sourced from the
2006 Census of Population and Housing Time Series Profile (Australian
Bureau of Statistics, 2007). There are particular issues with this data set.
These include the facts that over time, geographies and industry clas-
sifications change. The Australian Bureau of Statistics adjusts (called
concordance) for these changes by particular methods. Another issue is
that people filling out the industry of employment on the census form
may not accurately identify the industry of their employer. An example
the Australian Bureau of Statistics (2006, p. 198) provides is a person who
works for a coal mining company as a driver of the companys coal trucks.
The individuals occupation is truck driver. However, the industry of the
individuals employer is coal mining and not transport.
To provide some measure of standardization, the regional shift effect
throughout this chapter is compared with the change in working age popu-
lation. There are issues with this measure, particularly where the level of
employment self-containment varies between capital cities. Nonetheless,
standardizing the regional shift effect provides a relatively good comparative
tool. This is the first time such a standardization method has been applied to
the results of the Haynes and Dinc method to the knowledge of the author.
Results
Using the shift-share method discussed above, the NS, IM and RS effects
were obtained for the five-capital city regions for the period 1996 to 2006.
The results are given in Table 12.4. Two distinct patterns are evident: In
aggregate the five-capital cities recorded negative RS effects in employment
of 15 668 or 1 job per 100 change in working age population. But most
of this was attributable to Sydney, which recorded a negative RS effect
of 113 034 (or 37 jobs per 100 increase in working age population).
Adelaide recorded a much smaller negative RS effect of 17 193 jobs (or
37 per 100 increase in working age population). In contrast, positive RS
effects were recorded in Brisbane (+66 832 or 29 jobs per 100 increase in
working age population), Melbourne (+22 948 or +7 jobs per 100 increase
in working age population) and Perth (+24 714 or +17 jobs per 100
increase in working age population).
When the decade 1996 to 2006 is segmented into the two inter-census
periods 19962001 and 200106 we get the results shown in Table 12.5.
The following conclusions may be drawn:
Most of Sydneys negative RS effects over the decade 1996 to 2006

occurred in the five-year period 2001 to 2006. There was a noticeable

Table 12.4 Shift-share results for capital city Statistical Divisions, 19962006

Total Shift National Shift Effect Industry Mix Effect Regional Shift Effect
(No.) Per 100 (No.) Per 100 (No.) Per 100 (No.) Per 100
increase in increase in increase in increase in
working working working working
age age age age
population population population population
(no.) (no.) (no.) (no.)
250
Sydney 219 197 72 322 080 106 10 126 3 113 034 37
Melbourne 286 066 87 267 520 82 4 412 1 22 948 7
Brisbane 202 382 89 126 212 56 9 366 1 66 832 29
Adelaide 68 374 146 83 925 180 1 638 4 17 193 37
Perth 139 700 96 106 380 73 8,598 6 24 714 17
Australia 1 467 862 94 1 467 941 94
Five-capital cities 915 719 87 906 064 86 25 323 2 15 668 1
Rest of Australia 552 143 107 561 798 109 25 323 5 15 668 3
Source: Compiled by the author using data available from the ABS 2006 Census of Population and Housing data.
20/12/2010 15:12
Table 12.5 Capital city regional shift effects: 19962001, 200106 and 19962006

19962006 19962001 200106
(No.) Per 100 increase (No.) Per 100 increase (No.) Per 100 increase
in working age in working age in working age
population (no.) population (no.) population (no.)
Sydney 113 034 37 13 336 7 98 021 85
Melbourne 22 948 7 27 475 17 8 532 5
251
Brisbane 66 832 29 14 133 14 51 436 40
Adelaide 17 193 37 8 604 39 7 945 32
Perth 24 714 17 3 665 5 19 419 26
Australia
Five-capital cities 15 668 1 23 333 4 43 643 9
Rest of Australia +15 668 3 23 333 10 43 643 16
Note: Data is not additive.
Source: Compiled by the author using data available from the ABS 2006 Census of Population and Housing data.
20/12/2010 15:12
acceleration in the negative RS effect in Sydney from 7 jobs per 100

increase in working age population to 85 jobs.
Melbourne recorded a positive RS effect in the 19962001 period of
+17 jobs per 100 increase in working age population, which was fol-
lowed by a negative RS effect of 5 jobs per 100 increase in working
age population.
Adelaides RS effect over these two periods was negative, but margin-
ally less negative in the later period at 32 jobs per 100 increase in
working age population compared with 39 jobs per 100 increase in
population in the period 19962001.
Both Brisbane and Perth recorded accelerating positive RS effects over
the two periods (Brisbane with +14 jobs per 100 increase in working
age population in the 19962001 period to +40 jobs in the 200106
period; and Perth with +5 jobs per 100 increase in working age popu-
lation in the 19962001 period to +26 jobs in the 200106 period).
INDUSTRIAL EXPLANATIONS FOR REGIONAL

SHIFT EFFECTS IN CAPITAL CITIES
While in aggregate there have been some significant movements in RS

effects for Australias five-capital cities, by looking at the RS effects by
industry it is possible to see which particular industries have experienced
strong endogenous employment growth. This is, of course, different from
the IM effect described previously.
The industry sectors mainly affecting aggregate RS effects as opposed
to the IM effect described earlier were: manufacturing (particularly for
Sydney and Melbourne) and construction (particularly for Sydney).The
large negative RS effect of manufacturing employment in Sydney and
Melbourne, and to a lesser degree Adelaide, was countered by positive RS
effects in the rest of Australia, Brisbane and Perth.
Proportionate to their 1996 employment levels, the mining industry
recorded large RS effects in the five-capital cities in aggregate. But Sydney
and Melbourne recorded strong negative proportionate RS effects, while
Perth, Adelaide and Brisbane recorded strong positive proportionate RS
effects. Other proportionately large RS effects were recorded in the electri-
city gas and water supply industries for Brisbane (positive), the construc-
tion industry for Sydney (negative) and the arts and recreation services
industries for Adelaide (negative).
Overall, most of the negative RS effects for all the five-capital cities in
aggregate over the period 1996 to 2006 were concentrated in these industry
sectors:

the manufacturing industries (20 838 jobs or more than the total
negative RS effect);
the construction industry (7468 jobs or 48 per cent of the total
negative RS effect);
the administrative and support services industries (6078 jobs or 39
per cent of total negative RS effects jobs).
The highest positive RS effects were recorded in:
the wholesale trade industry (+8629 jobs);

the mining industry (+6372 jobs);
the financial and insurance services industries (+6080 jobs).
Proportionately, compared with 1996 employment levels over this

period, the following patterns are evident: (1) The highest positive RS
effect was in the mining industry at 32 per cent, followed by the electri-
city, gas and water supply industries at 6 per cent, 3 per cent for whole-
sale trade industry and for information, media and telecommunications
industries, and 2 per cent for education and training. (2) The highest
negative proportionate RS effects over this period were recorded in the
agriculture, forestry and fishing industry at 8 per cent, followed by the
administrative and support services industries at 4 per cent, and then
at 3 per cent manufacturing, construction and other services industry
sectors.
The focus now turns to an examination of some individual industries
that have had relatively large contributions to RS effects in the five-capital
cities. These are the manufacturing industries (large negative RS effect),
the wholesale trade industry (also a large negative RS effect) and the
construction industry (a relatively large positive RS effect).
Manufacturing Industry Employment
Large RS effects were recorded in the manufacturing industries sector

in the five-capital cities. Overall, there was a negative regional shift of
20 838 jobs in the manufacturing employment over the decade 1996 to
2006 (see Table 12.6). This represented 3 per cent of the 1996 level of
employment in the manufacturing industry sector.
But there were marked differences in the RS effect between the cities:
The negative RS effects took place in Sydney (22 609 jobs represent-
ing 11 per cent of the 1996 employment level), Melbourne (21 824
jobs representing 9 per cent of the 1996 employment level) and

Table 12.6 Regional shifts in the manufacturing industry, 19962006
Total Shift Regional Contribution Regional

(no.) Shift Effect to Total Shift c.f. 1996
(no.) Regional Employment
Shift (%) Level (%)
Sydney 16 270 22 609 20 11
Melbourne 14 537 21 824 95 9
Brisbane 18 766 16 387 25 22
Adelaide 362 2 434 14 4
Perth 11 440 9 642 39 17
Five-capital 963 20 838 133 3
cities
Rest of 30 080 20 838 39 8
Australia
Source: Compiled by the author using data available from the ABS 2006 Census of
Population and Housing data.
Adelaide (2434 jobs representing 4 per cent of the 1996 employment

level).
In contrast, there was a positive RS effect in Brisbane (+16 387 jobs
representing 22 per cent of the 1996 employment level) and Perth
(+9642 jobs representing 17 per cent of the 1996 employment level).
The rest of Australia recorded a positive RS in the manufactur-
ing industries of 20 838 jobs or 8 per cent of the 1996 employment
level. Most of this took place in the Gold Coast Statistical Division
(+7266 jobs) and the Sunshine Coast Statistical Division (+2197
jobs).
In essence, employment in the manufacturing industry has shifted from

the two largest capital cities in Australia (and to a minor degree Adelaide)
to the rest of the country. This has occurred due to endogenous factors
within these places.
Wholesale Trade Industry Employment
One of the offsetting factors for the negative RS effect in the five-capital
cities in aggregate was the wholesale trade industry (see Table 12.7). The
wholesale trade industry recorded a positive RS effect in employment of
+8629 jobs for the period from 1996 to 2006.
Again the performance varied between the cities:

Table 12.7 Regional shifts in wholesale trade industry, 19962006

Shift (%) Level (%)
Sydney 6 421 503 0
Melbourne 5 514 10 960 48 13
Brisbane 1 827 7 07 1 2
Adelaide 4 381 2 867 17 12
Perth 2 725 674 3 2
Five-capital 9 840 8 629 55 3
cities
Rest of 16 473 8 629 55 7
Australia
Most of this positive RS performance was attributable to the strong

positive RS effect in Melbourne (+10 960 jobs or 13 per cent com-
pared with its 1996 employment level). There was a smaller positive
RS effect recorded in Brisbane (+707 jobs or 2 per cent compared
with its 1996 employment level) and Sydney (+503 jobs or less than
1 per cent compared with its 1996 employment level).
Falls were recorded in Adelaide (2867 jobs or 12 per cent com-
pared with its 1996 employment level) and Perth (674 jobs or 2 per
cent compared with its 1996 employment level).
In the rest of Australia there was a negative RS effect on employ-
ment, with much of it in New South Wales (with 27 per cent of
the rest of Australian negative RS effect located in the Hunter and
Illawarra Statistical Divisions).
Construction Industry Employment
The next highest RS effect for employment impact in the five-capital cities
was in the construction industry (see Table 12.8). On average, the five-
capital cities recorded a negative RS effect of 7468 jobs or 3 per cent of the
1996 level of employment in the industry. This was entirely due to Sydney:
For Sydney, over the period 1996 to 2006, it was equal to +22 395
jobs, which contributed 20 per cent of its total negative RS. It

Table 12.8 Regional shifts in the construction industry, 19962006

Shift (%) Level (%)
Sydney 30 059 22 395 20 22
Melbourne 43 507 3 687 16 5
Brisbane 26 134 4 566 7 11
Adelaide 12 856 2 503 15 12
Perth 23 359 4 172 17 11
Five-capital 3 039 7 468 48 3
cities
Rest of 3 7 468 48 7
Australia
represented 22 per cent of the 1996 level of employment in the con-

struction industry.
All other capital cities recorded positive RS effects in the construction
industry, with Brisbane recording the highest (+4566 jobs or 11 per
cent of 1996 levels of employment), followed by Perth (+4172 jobs
or 11 per cent of 1996 levels of employment), Melbourne (+3687
jobs or 5 per cent of 1996 levels of employment) and Adelaide
(+2503 jobs or 12 per cent of 1996 levels of employment).
There was a positive RS effect of +7468 jobs in the rest of Australia,
which is equal to 7 per cent of this industrys level of employment
in 1996.
FOCUSING ON EACH CAPITAL CITY
Each individual capital is now examined to identify the impact of each

industry sector on the total RS effect.
Sydney
Sydney is the capital city with the largest negative RS effect on employment
of 122 993 over the decade 1996 to 2006, which compares with a total
shift of +219 197 jobs (see Table 12.9). A little under two-thirds of the
negative RS effect is attributable to four industries:

Table 12.9 Regional shift effects in Sydney, 19962006
Total Regional Contribution Regional

Shift Shift Effect to Total Shift
(No.) (No.) Regional c.f. 1996
Shift (%) Employment
Level (%)
Agriculture\ forestry 2 623 1 113 1 16
& fishing
Mining 323 1 147 1 33
Manufacturing 16 270 22 609 20 11
Electricity\ gas\ water 2 194 1 754 2 13
& waste services
Construction 30 059 22 395 20 22
Wholesale trade 6 421 503 0 0
Retail trade 38 505 14 346 13 9
Accommodation & 15 961 4 434 4 4
food services
Transport\ postal & 9 137 8 636 8 9
warehousing
Information media & 2 135 2 757 2 5
telecommunications
Financial & insurance 21 093 3 455 3 3
services
Rental\ hiring & real 6 796 1 600 1 6
estate services
Professional\ scientific 31 770 7 539 7 6
& technical services
Administrative & 9 819 9 599 8 17
support services
Public administration 17 769 8 916 8 10
& safety
Education & training 25 391 2 123 2 2
Health care & social 37 938 12 465 11 8
assistance
Arts & recreation 2 686 1 821 2 7
services
Other services 3 196 4 034 4 5
Inadequately 1 047 4 823 4 9
described/Not stated
Total employment 219 197 112 993 100 7

manufacturing industry (22 609 jobs or 20 per cent of total RS

effect);
construction industry (22 395 jobs or 20 per cent of total RS effect);
retail trade industry (14 346 jobs or 13 per cent of total RS effect);
health and community services and social assistance industries
(12 465 jobs or 11 per cent of total RS effect).
Despite this there were three industry sectors that recorded positive RS
effects, namely:
financial and insurance services industries (+3455 jobs or 3 per cent

of total RS effect);
information, media and telecommunications industries (+2757 jobs
or 2 per cent of total RS effect);
wholesale trade industry (+503 jobs or less than 1 per cent of total
RS effect).
Proportionately, and compared with 1996 employment levels, the highest

negative RS effect in Sydney was in the mining industry (33 per cent),
followed by the construction industry (22 per cent) and the administra-
tive and support services industries (17 per cent). The highest positive RS
effect in Sydney was in the information, media and telecommunications
industries (+5 per cent), the financial and insurance services industry (3 per
cent) and the wholesale trade industry (less than 0 per cent).
Melbourne
Compared with Sydney, Melbourne recorded a modest positive RS effect

of +22 976 jobs (compared with a total shift of +286 066 jobs) over the
decade 1996 to 2006 (see Table 12.10). Most of the positive RS effect was
attributable to these industry sectors:
wholesale trade industry (+10 960 jobs or 48 per cent of the total RS
effect);
accommodation and food services industries (+6481 jobs or 28 per
cent of total RS effect);
health care and social assistance industries (+5351 jobs or 23 per
cent of total RS effect).
As with Sydney, in Melbourne there was a very large negative RS effect

in manufacturing industries (21 824 jobs or 95 per cent of the total RS
effect). Other industries that recorded a negative regional RS were:

Table 12.10 Regional shift effects in Melbourne, 19962006

Level (%)
Agriculture\ forestry & 2 415 823 4 11
fishing
Mining 272 988 4 33
Manufacturing 14 537 21 824 95 9
Electricity\ gas\ water & 3 378 880 4 10
waste services
Construction 43 507 3 687 16 5
Wholesale trade 5 514 10 960 48 13
Retail trade 50 417 3 997 17 3
Accommodation & 21 558 6 481 28 9
food services
Transport\ postal & 14 931 2 808 12 4
warehousing
Information media & 1 526 2 097 9 5
telecommunications
Financial & insurance 14 261 2 715 12 4
services
Rental\ hiring & real 7 283 2 193 10 13
estate services
Professional\ scientific 32 185 1 490 6 1
Administrative & 17 899 3 209 14 8
support services
Public administration 16 263 3 946 17 6
& safety
Education & training 29 609 5 351 23 5
Health care & social 43 280 1 464 6 1
assistance
Arts & recreation 6 188 2 153 9 10
services
Other services 1 997 2 708 12 4
Inadequately described/ 540 3 781 16 8
Not stated

public administration and safety industries (3946 jobs or 17 per

other services industries (2708 jobs or 12 per cent of total RS effect);
mining industry (988 jobs or 4 per cent of total RS effect);
agriculture, forestry and fishing industries (823 jobs or 4 per cent
total RS effect).
Proportionately, when compared with 1996 employment levels, the whole-

sale trade and the rental, hiring and real estate services recorded the highest
positive RS effect at +13 per cent, followed by the electricity, gas and water
supply and the arts and recreation services industries at +10 per cent. The
highest negative proportionate RS effects were recorded in the mining
industry at 33 per cent, followed by the agriculture, forestry and fishing
industries at 11 per cent and the manufacturing industries at 9 per cent.
Brisbane
Brisbane recorded a strong positive RS effect of +66 805 jobs over the
decade 1996 to 2006 compared with a total shift of +202 825 jobs (see
Table 12.11). Most of this strong positive RS effect was attributable to
these industry sectors:
manufacturing industries (+16 387 jobs or 25 per cent of total RS

effect);
health care and social assistance industries (+6634 jobs or 10 per
retail trade industry (+6222 jobs or 9 per cent of total RS effect).
There was a negative RS effect recorded in the:
information, media and telecommunications industries (364 jobs

or 1 per cent of total RS effect);
agriculture, forestry and fishing industries (110 jobs or less than 1
per cent of total RS effect).
Proportionately, when compared with 1996 employment levels, the elec-

tricity, gas and water supply industries had the highest positive RS effect at
+48 per cent, and was followed by the mining industry at +38 per cent and
the manufacturing industries at +22 per cent. There were negative propor-
tionate RS effects in the agriculture, forestry and fishing industries at 3
per cent and the information, media and telecommunications industries at
2per cent.

Table 12.11 Regional shift effects in Brisbane, 19962006

Level (%)
Agriculture\ forestry & 1 042 110 0 3
fishing
Mining 1842 1129 2 38
Manufacturing 18 766 16 387 25 22
Electricity\ gas\ water & 3753 2 345 4 48
waste services
Construction 26 134 4566 7 11
Wholesale trade 1827 707 1 2
Retail trade 28 475 6222 9 9
Accommodation & 11 267 3144 5 8
food services
Transport\ postal & 12 057 5337 8 15
warehousing
Information media & 1729 364 1 2
telecommunications
Financial & insurance 6377 1931 3 8
services
Rental\ hiring & real 5191 1515 2 12
estate services
Professional\ scientific 15 790 2111 3 4
Administrative & 9180 2459 4 13
support services
Public administration & 16 461 2792 4 6
safety
Education & training 17 204 4541 7 9
Health care & social 28 454 6634 10 10
assistance
Arts & recreation 1857 172 0 2
services
Other services 2366 2028 3 7
Inadequately described/ 1806 3259 5 16
Not stated

Perth
Perth recorded a positive RS of +24 728 jobs over the decade 1996 to 2006
compared with a total shift of +139 700 jobs (see Table 12.12). This was
mainly attributable to these industry sectors:
manufacturing industries (+9642 jobs or 39 per cent of total RS

effect);
mining industry (+6589 jobs or 27 per cent of total RS effect);
spublic administration and safety industries (+5356 jobs or 22 per
cent of total RS effect).
The highest negative RS effects were in:
financial and insurance services industries (1657 jobs or 7 per cent

of total RS effect;
administrative and support services industries (1252 jobs or 5 per
arts and recreation services industries (774 jobs or 3 per cent of
total RS effect).
Proportionately, when compared with 1996 employment levels, the mining

industries recorded the highest positive RS effect at +74 per cent, followed
by both the manufacturing industries and the public administration and
safety industries at +17 per cent. The highest negative proportionate RS
effect was recorded in the agriculture, forestry and fishing industries at
13 per cent, followed by the arts and recreation services industries at 9
per cent, and the financial and insurance services industries at 8 per cent.
Adelaide
In Adelaide over the decade 1996 to 2006 there was a negative RS effect of
17 184 jobs compared with a total shift of +68 374 jobs (see Table 12.13).
Marginally over half of this negative RS effect is attributable to these
industry sectors:
health care and social assistance industries (3366 jobs or 20 per

education and training industries (2889 jobs or 17 per cent of total
RS effect);
wholesale trade industry (2867 jobs or 17 per cent of total RS
effect).

Table 12.12 Regional shift effects in Perth, 19962006

Level (%)
Agriculture\ forestry & 1517 563 2 13
fishing
Mining 8731 6589 27 74
Manufacturing 11 440 9642 39 17
Electricity\ gas\ water & 1738 257 1 5
waste services
Retail trade 20 309 899 4 2
Accommodation & 6912 67 0 0
food services
Transport\ postal & 4757 6 0 0
warehousing
Information media & 1723 696 3 6
telecommunications
services
estate services
Professional\ scientific 13 175 2102 8 5
support services
Public administration & 14 745 5356 22 17
safety
Education & training 11 435 704 3 2
assistance
services
Not stated

Table 12.13 Regional shift effects in Adelaide, 19962006

Level (%)
Agriculture\ forestry & 33 648 4 22
fishing
Mining 1098 788 5 61
Manufacturing 362 2434 14 4
Electricity\ gas\ water & 1438 327 2 8
waste services
Retail trade 15 413 527 3 1
Accommodation & 4017 1170 7 5
food services
Transport\ postal & 2968 510 3 3
warehousing
Information media & 1 022 179 1 2
telecommunications
services
estate services
Professional\ scientific 5966 1297 8 5
support services
Public administration 10 163 2395 14 9
& safety
Education & training 5586 2889 17 8
assistance
services
Not stated

The industry sectors with positive RS effects were:
construction industry (+2503 jobs or 15 per cent of total RS effect);

public administration and safety industries (+2395 jobs or 14 per
mining industry (+788 jobs or 5 per cent of total RS effect).
Proportionately, when compared with 1996 employment levels, the arts

and recreation services industries recorded the highest negative RS effect at
18 per cent, followed by the rental, hiring and real estate services indus-
tries at 17 per cent and the wholesale trade industry at 12 per cent. In
contrast, the mining industry had the highest proportionate positive RS
effect at +61 per cent, followed by the agriculture, forestry and fishing
industries at +22 per cent, and the construction industry at +12 per cent.
CONCLUSION
The research reported in this chapter found that negative regional shift
effects in employment in Sydney and Adelaide have been the drivers of
relatively poor employment growth in those cities over the decade 1996 to
2006. In Sydney, Australias largest city, this was particularly so in the final
half of that period after 2001. For Brisbane and Perth, the regional shift
effects were positive over this period. These findings fit within the broader
context of contemporary regional economic trends during this period.
Given its size, the impact of Sydneys relatively poor regional employ-
ment shift effects are important to the employment performance of the
nation as a whole. Most of the negative regional shift effect in Sydney
over the period was attributable to the manufacturing, construction and
retail trade industry sectors. Comparatively, most of the positive regional
shift effect in Brisbane the nations rapidly growing sub-belt metropo-
lis has been attributable to the manufacturing, health care and social
assistance industry and the retail trade industry sectors.
The national economy has experienced significant changes in its indus-
trial structure in recent decades. This is also true for employment in
the five-capital cities (those with populations in excess of 1 million) of
Australia. One of the key determinants of change in employment for
capital cities has been regional shift effects (i.e., due to endogenous
factors). The research reported here has examined the regional shift effects
on employment in the capital city regions for the period from 1996 to
2006, which was the last decade or so of the long boom experienced by
Australia from the early 1990s.

Future research may examine these regional shift effects for different
count methods of employment. The optimal method would be to use
place of employment (i.e., where the jobs are). Additionally, techniques to
measure production at the capital city level may allow future research into
regional shift effects for production in the capital cities.
NOTES
* The research reported in this chapter is part of a wider project analysing endogenous
regional growth performance across Australia funded in a grant to Robert J. Stimson
under the Australian Research Council Discovery Scheme Grant # DP0879819.
1. As defined by Statistical Divisions.
2. By place of enumeration count method.
3. The southern region refers to the Gold Coast Statistical Division, the western region
refers to the West Moreton Statistical Division, and the northern region refers to the
Sunshine Coast Statistical Division.
4. Place of usual residence method is available for most statistical divisions. This data needs
to currently be downloaded one Statistical Division at a time consuming a significant
amount of time to download all the required data. For most Statistical Divisions there is
little difference between the two count methodologies. On balance, I determined that the
place of enumeration count method was the optimal count method to use at this point in
time.
REFERENCES
Australian Bureau of Statistics (1997), Australian Demographic Trends, Catalogue

No. 3102.0, Canberra.
Australian Bureau of Statistics (2006), 2901.0 Census Dictionary, 2006 (Reissue),
Canberra.
Australian Bureau of Statistics (2007), 2006 Census of Population and Housing,
Catalogue No. 2069.0.30.003 Census of Population and Housing: Time Series
Profile Datapack, 2006, Canberra.
Australian Bureau of Statistics (2009), 6262.0 - Information Paper: Regional
Labour Force Statistics, February 2009, Canberra.
Chalmers, J.A. (1971), Measuring Changes in Regional Industrial Structure: A
Comment on Stilwell and Ashby, Urban Studies, 8(3), 28992.
Edwards, J.A., Harniman, K.F. and Morgan, J.S. (1978), Regional Growth and
Structural Adaption: A Correction to the Stilwell Modification, Urban Studies,
15(1), 97100.
Haynes, K.E. and Dinc, M. (1997), Productivity Change in Manufacturing
Regions: A Multifactor/Shift-Share Approach, Growth and Change, 28(8),
20121.
Harding, A., Hellwig, O. and Lloyd, R. (2000), Regional Divide? A Study of
Incomes in Regional Australia, 29th Conference of Economists, Gold Coast,
36 July 2000.
Jefferson (1939), The Law of the Primate City, Geographical Review, 29, 22632.

Markusen, A.R., Noponen, H. and Driessen, K. (1991), International Trade,

Productivity, and US Job Growth: A Shift-Share Interpretation, International
Regional Science Review, 14(1), 1539.
Mitchell, B. (2008), Exploring Regional Disparities in Employment Growth,
Working Paper, No. 08-12, Centre of Full Employment and Equity, University
of Newcastle, Australia.
Noponen, H., Markusen, A. and Driessen, K. (1997), Trade and American Cities:
Who Has the Comparative Advantage?, Economic Development Quarterly, 1(1),
6787.
Productivity Commission (1998), Aspects of Structural Change in Australia,
Canberra, Australia.
Stilwell, F.J.B. (1969), Regional Growth and Structural Adaptions, Urban
Studies, 6(2), 16278.
Stimson, R.R., Robson, A.P.L. and Shyy, T. (2005), Modeling Determinants of
Spatial Variation in Regional Endogenous Growth: Non-metropolitan Regions
in the Mainland States of Australia, in Proceedings of 29th Annual Conference
of the Australian and New Zealand Regional Science Association International,
Sydney, Australia.
Stimson, R.R., Robson, A.P.L. and Shyy, T. (2009), Modeling Regional
Endogenous Growth: An Application to the Non-metropolitan Regions of
Australia, Annals of Regional Science, 43(2), 37998.
Wadley, D. and Smith, P. (2003), Straightening up Shift-Share Analysis, Annals
of Regional Science, 37(2), 25961.

13. A case study approach to
investigating local development
initiatives in rural small towns in
Victoria
John Martin*
INTRODUCTION
An interest in endogenous development comes out of the research under-

taken on Australian small towns in recent years, in particular research on
small towns in the state of Victoria.1 Studying small towns is problematic
because they are, paradoxically, similar in many ways, but cannot be treated
as such (Powers, 1991). Across Victoria there are some 300 small towns with
populations ranging from 250 to 10 000. The future of these places has been
the basis of much contentious discussion in recent years (see Sher and Sher,
1994; Forth, 2000; OConnor et al., 2001; Stimson, 2002) and the continu-
ing public policy response seems to be to leave them to the competing forces
of globalization and their own capacity to prosper and survive. Of course,
there are also many smaller communities, communities with fewer than 250
people, simply categorized by the demographers as rural other.
A key policy question facing Australian state governments in recent
decades has been how to assist small towns with the considerable change
impacting them (Garlick, 2005). That includes change in the scale of agri-
culture, more extreme climatic regimes, innovations and improvements in
transportation and communications technologies and the ageing of people
living in these places as the flight of youth takes this cohort to education
and employment opportunities and the excitement of larger cities. A clear
policy preference by state governments for these places has come out of the
social capital literature, especially that reflecting Robert Putnam, a regular
speaker at state and local government forums in Australia, where he
advocates intervention strategies based on participation and engagement
(see Eversole and Martin, 2005 for recent Australian cases). Yet when
one looks closely at these programmes one cannot find clear measures as
268

Initiatives in rural small towns in Victoria 269
to their impact in the places they are applied. Most evaluations are based
on opinion surveys with central players with no real tangible evidence of
improved or sustainable development relating to investment, production,
employment and the overall standard of living of people in these commu-
nities. The initial idea for this chapter grew out of the idea of being able to
identify how intra-regional endogenous growth is measured, thus influen-
cing the focus on local policy development.
In his keynote address at the annual conference of the Australian and
New Zealand Regional Science Association International (ANZRSA) held
in Manukau in September 2007, Roger Stough referred to the doctoral
research of Mark De Santis (1993) on intra-regional endogenous growth.
De Santis research was on the relationship between leadership, resource
endowment and exogenous and endogenous drivers of regional economic
development. The author became interested in how it might apply to better
understanding of small town development. The focus was on inter-regional
endogenous growth at a scale several times greater than our interest in small
rural towns. Herein lay the complication I was to discover when research-
ing small town development strategies in two rural shires in Victoria. The
question that arose was: Can the factors informing inter-regional success
be applied to small rural towns on an intra-region comparison?
Local government policy-makers, who are the place managers responsi-
ble for ongoing support for these small towns, would like to know whether
these factors also apply. Stimson et al. (2005) have proposed their virtu-
ous circle for sustainable regional development, which also reflects De
Santis findings. Thus, the question is: Do they apply to intra-regional
assessment of small towns in large rural shires?
After a brief discussion of methodology, there is a discussion of some
of the literature relating to community economic development. That is
followed by a review of two rural shires, and several towns in each shire,
which are used as case studies. There is then a discussion of both the rel-
evance of the social capital concept and measurement and the applicability
of the virtuous circle for sustainable regional development model as it
applies to the four case study small towns. The focus in the chapter reflects
a wider research interest in indicators for small town development, which
represents a much smaller spatial scale of analysis than is usual in the
conventional studies on regional development.
PURPOSE AND METHOD
In the De Santis study referred to above, the sample was 35 metropolitan

areas in the US, each with an urban centre of at least 50 000 people and

a total population of at least 100 000. While the conceptual framework

driving his research was interesting, the regional scale was much greater
than that faced in dealing with small towns in Victoria.
The regional shires in Victoria range in size from 15 to 25 000 people in
total, with maybe one urban centre approaching 5 000 to 10 000 people,
and all-up these shires might be lucky to have a total of 15 000 people.
The overall population density is low. These are rural places with overall
an ageing and mostly static population, with perhaps some growth on
the back of city dwellers making lifestyle purchases with their rural block
or holiday house. In focusing on small rural towns, the question arises
as to why does one town do so much better than another of similar size
and apparent resources? The Australian inter-regional comparisons
have been ably done by others (see, for instance, OConnor et al., 2001).
Comparing small towns within Australian regions (local government
authorities) is not something that has been systematically researched in
the way that the De Santis study in the US did by analysing change over
a decade.
For the study reported here the author was curious to see if it was
possible to identify the criteria by which one could identify and evaluate
change in order to compare inter-town performance and success within
shires. As part of the research, secondary sources on small town develop-
ment were obtained for two Victorian shires identified as Shire A and
Shire B for purposes of confidentiality for reasons that will become
apparent when I report on how challenging it is to get information from
local governments, even when one of them has won a national award for
its programme and the other has presented itself at regional conferences
as being an innovator and leader in small town engagement and develop-
ment. It is clear that there is a managerial and political overlay at the
micro-level that comes into play, therefore it is necessary to protect the
people in these small towns who could easily be identified. With the focus
on local government processes, what is being sought is to identify suc-
cesses and failures in the eyes of local government officials responsible for
these programmes aimed at enhancing local development. It would have
been possible to have gone to a local government shire that did not have
a comprehensive small town engagement strategy across its jurisdiction,
but that would have taken more time and resources to get information on
different town performance. So the primary focus was to begin to identify
those factors that are critical to the success of small rural towns (in the
state of Victoria), and to measure those factors and determine how to use
them in any assessment of small town performance within a local govern-
ment jurisdiction.

THEORETICAL VIEWS ON ENDOGENOUS

DEVELOPMENT
De Santis (1993) conclusion to his intra-regional comparison of US cities
referred to above is summarized in Table 13.1. High economic perform-
ance was found in those places where there were both high levels of leader-
ship and resource endowments.
De Santis findings thus support the sometimes hypothesized causal
relationship between social capital and regional economic performance.
Places rich in social capital in this case leadership are more likely to
have higher economic performance when they are also endowed with
appropriate material resources, than places that have less amounts of
these two factors. In the social capital literature it is typically presented
as a metaphor for advantage (Burt, 2001). The more social capital a place
has the more likely it will succeed on a number of grounds, including
economic performance. More recent research by Eastwood (2007) has
shown that the combination of creative entrepreneurs networking in an
economic free environment (p. 157) is the basis for sustainable economic
development. The question is: Does this align with De Santis conclusion
that effective leadership in resource-endowed cities leads to high economic
performance?
I would suggest that these two empirical studies are not at odds with
each other. A more general assessment is that in places endowed with eco-
nomic resources, however widely defined, and where individual creativity
is valued and fostered, economic growth and development is more likely to
occur than in places that do not have these characteristics.
The research focus in the study of small towns in Victoria discussed
Table 13.1 Relationship between leadership, resource endowments, and

economic performance
Leadership
High Low
Resource High High Moderate
Endowments economic economic
performance performance
Low Moderate Low
economic economic
performance performance
Source: Compiled by the author and derived from De Santis (1993, p. 91).

here was to identify the factors that contribute to small town success
including social and economic success with the aim being to inform
policy-makers as to how best they might work with such places for their
continuing sustainability. Defining success is thus critical. The criteria will
determine what to measure, and begin to value.
Stimson et al. (2005) presented their virtuous circle for sustainable
regional development with ten factors that constitute the basis for iden-
tifying the success of local community approaches to sustainable devel-
opment. They condense them into leadership, institutions and resource
endowments and market fit, and see how they might interact within what
is termed a regional competitiveness performance cube. Those over-
arching factors are not, I believe, inconsistent with the conclusions drawn
by De Santis (1993) and Eastwood (2007). My view is that the ten factors
that constitute the conceptual model provide the basis for finding out the
local development potential of a small town. For example the Stimson
et al. (2005) premise is that strong proactive leadership that helps build
a vision for future development is backed up by strategy, plans and pro-
cesses and facilitates institutional change to enhance regional capacity and
capability, then the local development process is enhanced. That requires
effective institutions and regional infrastructure, and needs processes and
mechanisms for using resource endowments so as to maximize the poten-
tial to tap market conditions. That may lead to sustainable development
(after Stimson et al., 2005).
It was clear in the research on small towns outlined below that many
of these factors are evident in their small town development projects. The
question is: Why dont the local shires use them as part of their ongoing
evaluation?
The literature on social capital is now extensive and central in efforts by
researchers to understand social structures and processes (Putnam, 1993,
2000; Woolcock, 1998; Lin, Cook and Burt, 2001). Putnams argument
that the degree of association or participation by individuals in society is
related to the success of communities is central to the interests reflected
in this chapter. There is a need to know what these processes are in small
rural towns of several hundred people and how it impacts local economic
development: What are the indicators? How do we measure them, and
what does it tell us about what governments can do to facilitate local eco-
nomic success?
Burt (2001) also raises the idea of brokerage across structural holes as
the source of value-added, and closure can realize value buried in holes.
He contrasts processes of external and internal lack of constraint (see
Figure 13.1). The sociometry of these relationships is presented diagram-
matically in Figure 13.2. As will be seen from the discussion below, it is the

D A
High
Disintegrated group of
Maximum
diverse perspectives,
performance
skills, resources
External lack of
constraint Minimum Cohesive group
performance containing only one
Low perspective, skill,
resource
B
Low Internal lack of constraint High
Source: Compiled by the author and derived from Burt (2001, p. 48).
Figure 13.1 Social capital matters
balance between external and internal attention that seems to be a critical

factor in a communitys success in garnering development support.
In his extensive theoretical discussion of the relationship between social
capital and economic development, Woolcock (1998, p. 153) has com-
mented on both the epistemology underpinning research and the scale at
which it is focused. He asserts that:
The subsequent discrediting of modernization theory in both its sociologi-

cal and economic guise, however, and its replacement, respectively, by world
systems and neo-classical growth theories have led to a situation where the
units of analysis in contemporary studies of development are either nation-
states and transnational corporations or rational individuals and firms. The
contribution of civil society and other institutional arrangements mediating
the space between states and markets has been lost, incorporating themselves
neither into a coherent body of knowledge nor sensitive and sensible policy
prescriptions.
Woolcock (1998, p. 187) also provides a cautionary conclusion to the rela-

tionship between social structures and processes and development success.
He asserts that:
the mediating variable is the extent to which a mutually beneficial interaction

coordinates specific levels, dimensions, and combinations of social relation-
ships. The structure of the state, the nature and extent of its involvement in civic
and corporate life, and the organization of society together constitute the key
factors determining whether a country succeeds or fails in development.
Sociological studies that attempt to identify those associational factors

that influence development in small communities also reflect Woolcocks

D B
Source: Compiled by the author and derived from Burt (2001, p. 48).
Figure 13.2 The sociometry of the relationships social capital matters
caution. In their excellent research on the structure of access to embedded

resources in Canadian First Nations communities, Enns et al. (2008) apply
Lin, Fu and Hsung (2001) position generator to these small coastal com-
munities. Position generators:

use a sample of ordered structural positions salient in a society and ask

respondents to indicate contacts, if any, in each of these positions [thus] it
becomes possible to construct measures of (1) range of accessibility to different
hierarchical positions; (2) extensity or heterogeneity of accessibility to different
positions; and (3) upper reachability of accessed social capital. (Lin, Fu and
Hsung, 2001, p. 63; italics original)
Enns et al.s (2008) findings support Putnams (1993, 2000) view that:
civic participation is related to the development of critical networks that
are crucial in gaining access to embedded resources and developing social
capital (quoted in Enns et al., 2008, p. 37). They also report a limitation to
this social research technique of interest to the research question being
asked in this chapter that: the position generator is not able to explicitly
capture the social processes whereby people utilize their network ties for
social benefit (ibid., p. 36)
Nevertheless, Enns et al.s (2008) research is worthy of greater consid-
eration in investigating the relationship between intra-region, or shire
council, jurisdictions, as it tells us something about the relationship
between, for example, gender, education and the strength of ties people
have both within and outside their communities. Enns et al. support both
Leonard and Onyx (2003) and Woolcock (1998) who also
advocate for balanced networks with no particular emphasis on one type of tie
in a given location ... involvement in groups that span communities leads to
both weak and strong ties located both inside and outside the community, and
thus allows residents to form diverse, stable networks to resource-rich persons.
(Enns et al., 2008, p. 38)
For Woolcock (1998) it is the intra-community (bonding) extra-

community (bridging) ties working together that differentiates success
between communities. Importantly Woolcock asserts the important link
between social capital and economic success: you cant have the latter
without the former. Thus, the question is: Can we assume that economic
development within small towns will follow if there is strong social
capital?
THE CASE STUDY SHIRE COUNCILS2
Two rural/regional local government areas (called shires) in the state

of Victoria, each containing numerous small towns, were chosen as the
case studies for the research discussed here. They have these common
characteristics:

there is only one large town of approximately 8000 population;

numerous other small towns with a population above 250 (as well as
many locales with fewer than 250 people);
significant agricultural activity;
an ageing population.
Both have active community development strategies.3

Victoria is the most closely settled state of mainland Australia, and the
state government has had a deliberate small towns development strat-
egy working through its dependent local government councils. Of the
two councils one (Shire B) is a national award winner for its community
development programme. In 2006 it was committing more than A$200 000
per year (for three years) to support Local Community Planning Groups
representing ten communities and 100 localities. The other (Shire A) is
also regarded across the state as a leader in this field. Statistical data and
documents were reviewed, and the research team members spoke by phone
to key players in the councils.
Shire A
Shire A has labelled its small towns programme as The Community

Connect Model. The shire identified several characteristics as underpin-
ning this programme. That includes the following:
open and honest communication between council and communities

(trust?);
plans reflect community vision (a shared view?);
community control of the process (involvement);
collaborative opportunities for council, communities and other
agencies to work together (engagement).
The focus of the strategy was to engage citizens in a discussion about facil-
ities and local government service improvement in each town. A deliberate
planning process was set in place, which resulted in specific plans for each
town. The processes used in Shire A were to:
conduct workshops to explore vision;

learn how to plan and work together;
support and learn from each other;
identify community priorities and projects;
apply for funding from state government.

It is clear from Shire As plans and publicity information that it places a

heavy emphasis on community engagement. The councils first goal in its
200610 plan is: Vibrant Communities: To create healthy, informed and
proud communities. Indicators of success identified are:
civic centre redevelopment;

attendance at a community event in past six months;
volunteer rate;
community satisfaction with health and human services;
immunization rates;
reportable incidences of communicable diseases;
percentage of population who feel they have opportunities to par-
ticipate in affordable local community events;
membership of an organized group;
community satisfaction rating with recreation facilities;
community satisfaction rating for local roads and footpaths.
These are outcome, output and process measures, taken in large part from
state government indicators linked to funded programmes. They give no
clear view of the overarching conceptual framework in which they are
located. It is suggested that this is why the attempt to define and use a wide
range of social indicators in local governments in Victoria largely falls on
deaf ears at the local government level as there is no apparent utilization
in the local decision-making.
Shire As community plan development process is the primary source of
information used for these two case study towns in the shire. Each town
was engaged by the shire in a planning process that established a commu-
nity contact team. Other key stakeholders for the planning process were
identified by this group. The shire then developed a community profile for
each town. Those two towns are now discussed.
Town A1
In the case of Town A1 the profile lists population over time, age cohorts,
employment, buildings, birth rate and a half-page profile, which reads like
a tourism brochure. The community future recognizes that the town is on
the edge of the growth area that is the southeast corridor of the Melbourne
metropolitan region. There are and will continue to be a significant pro-
portion of residents who commute elsewhere for employment. Economic
activity reflects the changing land use in small towns around two hours
drive from metropolitan Melbourne. While dairying and beef produc-
tion remain prominent agricultural industries the town now has many

self-employed tradespeople as well as small home-based businesses with

people who live and work in the area.
The community plan development process follows a strict formula:
revisit our vision statement (who do we want to be) followed by a work-
shop process that identifies priority areas, projects, scoring or polling
participants to determine priorities. At a follow-up workshop after the
shire officers processed the initial discussions, the participants reviewed
and consolidated projects and priorities. The third and final workshop
developed action plans and agreed the top eight projects. This completed
the series of workshops for input into the development of the [Shires]
Community Plan (Town A1 Community Plan).
The eight Community Plan priority areas and projects are:
sport and recreation;

public infrastructure;
public facilities and services;
town beautification;
business development;
walking and cycling tracks;
roads;
education.
The projects all concern the provision of infrastructure in one form or

another; for example, skateboard park, community centre, street scaping
and parks, walking and cycling tracks.
Importantly this Town A1 has also organized itself independently of
the shire council and its development efforts to undertake two major
local economic development projects. One is the establishment of a
Bendigo Bank Community Bank, which is a regional banking initiative
taken more than a decade ago in Bendigo, a regional town in Victoria,
and which has spread widely across rural areas of Australia. The other
is setting up a town historical society (recognized by the shire council in
its community plan, reflecting its involvement, but just how much is not
clear). Both of these cases demonstrate that in many small towns, and we
suspect this will be the majority,4 that individuals do cooperate to develop
local social and economic institutions and facilities in spite of what gov-
ernments do or do not do. The challenge for local policy-makers is to work
with people when such initiative occurs.
Town A2
This town is similarly located to Town A1 with respect to its proximity
to the metropolitan region, predominant agricultural production around

dairy and beef production and a significant number of commuters enjoy-

ing a rural lifestyle on a few acres.
The community plan development process, priority areas and projects
are also remarkably similar to Town A1. This is not surprising as the shire
council developed the approach and applied it across a number of small
towns in their jurisdiction. The eight priority areas were:
sport and recreation;

public facilities and services;
walking and cycling track and trails;
environment;
signage;
events;
town beautification and improvements;
strategic planning.
The focus is also on infrastructure such as the improvement of play-

grounds, provision of a mower, car parking, walking tracks, improved
signage, and retaining walls. The last priority area strategic planning
has a one-line project Controlled Development in [Town].
A defining characteristic of the outcomes of these two case study
towns in Shire A is the provision of local infrastructure, which, at first
glance, appears minor. But in these small towns such local improve-
ments to amenity impact the lives of the local citizens making their town
a better place in which to live as well as providing greater amenity for
visitors.
Shire B
The overarching strategy for this Shire B is found in their award-winning
Community Development Programme.5 From the shires website they
outline the programme as follows:
a programme that sets out to fully engage people in shaping the

future of their community and local economy;
a public community meeting develops the plan for each participat-
ing district;
public appointment of a community group steers action on the
district plan;
the district plan takes into account the strengths of each
community;
council works in partnership with the various groups, working on
projects arising from the district plan.

The community plan is developed as follows:
participate in the annual public community meeting to set the plan

for the district;
join a community planning group to help steer action on your dis-
tricts plan;
the members of community groups seek out interested community
members, to join inthe development of projects arising from their
districts plans;
actively seek and listen to views;
exchange ideas and opinions;
exchange information;
progress is reviewed at a public meeting each year.
There is a clear process set out by the shire council for community plan-
ning. This process provides indicators of success and can be measured.
They are largely institutional and should not be overlooked as important
factors contributing to endogenous development.
Town B1
Located within an irrigated agricultural region, Town B1 has been very
successful in leveraging capital works funds from the shire council to
attract additional monies to upgrade infrastructure in the town including
its business centre and street scaping. The town has also established its own
local fuel outlet that operates out of the business centre being staffed by
some 70 volunteers to ensure the locals do not have to drive many kilome-
tres to the next largest town for fuel. Significantly many of these volunteers
worked together to upgrade the fuel outlet to a modern service station.
One of the local town leaders attributes the success to people with
ideas, motivations and leadership. He asserted that its all about getting
excited about the little things. Many of these volunteers also help out at
the Bush Nursing Centre driving patients to other centres for their particu-
lar health needs.
In 1992 the community formed a progress association, and one of their
first projects was community housing for retirees. They now have 12
units completed, with land for a total of 20 units. Another project was to
convert an irrigation drain through the middle of town into an aquatic
centre playground for the community and visitors. They now have an
ornamental lake, an island, rotunda boardwalk and fountain, which make
their town attractive for both residents and visitors. A more recent project
is their living heritage complex. This is an historical centre showcasing the
development of agriculture in the region. Volunteers have together spent

thousands of hours and also raised tens of thousands of dollars for the
centre.
This town is clearly one that has been more successful than most in the
shire in obtaining resources from external sources as well as from amongst
the residents. It is, in fact, held up as a model of small town success and the
stories presented in the media typically show a range of citizens taking an
active role in their community. As the shire organization applied its com-
munity development model to the town it responded readily and was able
to capitalize on the processes put in place.
Town B2
Town B2 is also located within the irrigation schemes off the Murray
River but has been impacted by water restrictions over the last decade.
It contains a major milk manufacturing facility, now with international
owners who control this factory from the states capital city, Melbourne.
Like Town B1, Town B2 has also engaged in a series of planning meetings
to improve local infrastructure and services, ranging from improvements
to the public swimming pool, construction of a multi-purpose stadium
adjacent to the local school, road intersection realignments and signage. It
has also developed several projects independently from the shire-initiated
community planning process. One was to establish a banking service
with the Bendigo Bank as part of a longer-term development to establish
a community bank. Another recent success was the establishment of a
Rural Transaction Centre with support from the Australian government.
The leaders of these initiatives, who were marginal to the council-initiated
process, also developed a major national campaign Sponsor a Cow. The
district development committee, we are referring to here, created this pro-
gramme to give city folk and farmers the opportunity to Build Bridges
between the city and country and for sponsors to learn first hand the extent
of this drought and the impact and affect it is having on the whole popu-
lation (from Sponsor a Cow website). This project has support from the
Victorian Farmers Federation and Dairy Australia. The sponsorship has
a real impact enabling dairy farmers to save some of their breeding stock
and maintain some dignity during this time, which is the worst drought in
living memory. While the programme ran for two years and raised more
than A$100 000 for participating farmers, the stories of families coming up
from Melbourne to visit their cow and meet the dairy farming family tell
of a much more powerful outcome: of urban Australians feeling what it is
to be a dairy farmer in trying environmental and economic times.
Both Town B1 and Town B2 have benefited from the initiatives of the
shire council with its award-winning community development planning
programme, so much so that it has been put on hold because of funding

difficulties across the whole shire programme. This suggests that for most
small towns their expectations are for funding support from elsewhere for
their local projects. Our two case study towns have also shown they are
able to work together beyond the (exogenous) shire support to develop
their own (endogenous) development strategies.
CONCLUSION
The common characteristics of small town development processes in the
two case study shires were the structuring of relationships to address the
provision of basic infrastructure and local leadership through community
planning processes. It is not known which priorities have been agreed by
each council and subsequently developed. But what was learned through
the studies enquiring into four small towns across two shires in rural
Victoria was that local economic development projects have been initiated
in some towns, but are not as evident in the others. It cannot be denied
that such local initiative does not occur in those other places. It was only in
two where a community-wide approach to such development was found.
Coincidentally it was in the shire that won a national award for its com-
munity development programme a few years earlier. In the other shire, the
community planning process is still relatively new. It will be interesting
to learn if local economic development initiatives spin off from the com-
munity planning processes established by the shire and the communities
themselves attend to local social and economic development.
What intra-regional endogenous development factors can be effectively
evaluated? From the discussion of the social capital literature and a brief
review of intra-regional endogenous development in four small towns
across two rural shires there appear to be several characteristics that are
key to local development:
1. The application of the position generator to these places will tell local
government policy-makers who talks to whom about what and how
often they do this. Quite simply is there a sufficient level of engagement
in the community such that they can have a dialogue, a conversation
about what needs to happen and how they can make it work together?
2. The balance of internal community and external networks needs to
be right, and local leaders can influence these networking opportuni-
ties once they appreciate the importance of striking the right balance,
which will be different for different leaders, at different times working
on different issues.
3. Leadership pervades all of these processes. If appropriate individuals

with the skills and motivation to lead are in sufficient number then
De Santis findings for inter-regional success will also apply for intra-
regional success. But just how many leaders are required is a vexed
question. Of course there is another view that leadership is what is
required, something that many people in a community can display.
The community planning processes developed by both the case study

shires would have clearer conceptual direction if they followed, for
example, the virtuous circle for sustainable regional development pro-
posed by Stimson et al. (2005) outlined earlier in this chapter. It may be
argued that a comprehensive, over-arching measurement strategy for local
development should start with an assessment of social capital using Lin,
Fu and Hsungs (2001) position generator, as demonstrated by Enns et al.
(2008). This should be a baseline for future surveys to learn how the exter-
nal and internal links had changed as a result of an intervention strategy
by either the shire council, or when local circumstances change through
the actions of locals, regardless of the involvement of government.
The Stimson et al. (2005) model for sustainable development is reflected
in the small town development strategies implemented by the two shires
and four towns reported in this chapter. While such a concept was not
revealed by the informants in the scoping study undertaken in the case
studies, it is believed that it would certainly resonate with them. The big
question is: Would these shires be prepared to enter into a pre- and post-
evaluation strategy as suggested here? Or alternatively: Does local politi-
cal control dictate such processes in that other agendas, including being
seen to take action are as important as any genuine attempt at achieving
an outcome for the sustainability of these places? It may be that the local-
ized evaluation research may prove all too challenging, for political and
managerial reasons, like: Do we have the staff and resources to do this
analysis? and If we do, then what would we do with this information?
The concepts and measures are available to monitor local social and
economic development. Their use by local government councils, especially
in small and under-resourced councils, will be dependent on their under-
standing of the concepts and tools involved in this type of research and the
preparedness of local councillors and professional staff to face both good
and bad news that will come from such an assessment.
NOTES
* The author thanks Ms Barbara Beattie who provided excellent research assistance in
obtaining information from both shires and the four towns for the research on which this
chapter is based.

1. See The Study of Small Towns in Victoria Revisited, available at: http://www.dse.vic.
gov.au/CA256F310024B628/0/E352E96216614FDDCA2572CF0021B941/$File/07_08_
Small_Towns_Revisited.pdf; accessed 28 May 2010.
2. These are referred to as Shire A and Shire B.
3. Although one shire has just signalled to its community leaders that their small towns
strategy is currently on hold due to a lack of funding. This notification came about as
we started to interview key players and the purpose of our research became clear. This
is unfortunate as our intentions are to identify how best to measure intra-regional small
town development success to assist local governments in their future planning for these
places.
4. See The Study of small towns in Victoria Revisited, available at: http://www.dse.vic.
gov.au/CA256F310024B628/0/E352E96216614FDDCA2572CF0021B941/$File/07_08_
Small_Towns_Revisited.pdf; accessed 28 May 2010.
5. From the shires website.
REFERENCES
Burt, R. (2001), Structural Holes versus Network Closure as Social Capital, in N.

Lin, K. Cook and R.S. Burt (eds), Social Capital: Theory and Research, Aldine
De Gruyter, New York, pp. 3156.
De Santis, M. (1993), Leadership, Resource Endowments, and Regional Economic
Development, DPA, George Mason University, Fairfax, Virginia.
Eastwood, B.M. (2007), Explaining U.S. Urban Economic Growth 20002004:
The Role of the Creative Class, Social Capital, Economic Freedom, Distributive
Politics and Defense Spending, PhD thesis, West Virginia University,
Morgantown.
Enns, S., Malinick, T. and Matthews, R. (2008), Its Not Only Who You Know,
Its Also Where They Are: Using the Position Generator to Investigate the
Structure of Access to Embedded Resources, in N. Lin and B.H. Erickson (eds),
Social Capital: An International Research Program, Oxford University Press,
New York, pp. 25581.
Eversole, R. and Martin, J. (eds) (2005), Participation and Governance in
RegionalDevelopment: Global Trends in an Australian Context, Aldershot, UK:
Ashgate.
Forth, G. (2000), Following the Yellow Brick Road and the Future of Australias
Declining Country Towns, paper presented at the First National Conference
on the Future of Australias Country Towns, La Trobe University, Bendigo,
Australia.
Garlick, S. (2005), Regional Growth, Enterprising Human Capital and
Community Engagement, Monograph, Institute for Sustainability, Health
and Regional Engagement, University of the Sunshine Coast, Queensland,
Australia.
Leonard, R. and Onyx, J. (2003), Networking Through Loose and Strong Ties:
An Australian Qualitative Study, Voluntas: International Journal of Voluntary
and Non-profit Organizations, 14(2), 189203.
Lin, N., Fu, X.-C. and Hsung, R.-M. (2001), The Position Generator:
Measurement Techniques for Investigations of Social Capital, in N. Lin, K.
Cook and R.S. Burt. (eds), Social Capital: Theory and Research, Aldine De
Gruyter, New York, pp. 5781.

Lin, N., Cook, K. and Burt, R.S. (eds) (2001), Social Capital: Theory and Research,
Aldine De Gruyter, New York.
OConnor, K., Stimson, R.J. and Daly, M. (2001), Australias Changing Economic
Geography: A Society Dividing, Oxford University Press, Melbourne.
Powers, R. (1991), Far From Home: Life and Loss in Two American Towns,
Doubleday, New York.
Putnam, R.D. (1993), Making Democracy Work: Civic Traditions in Modern Italy,
Princeton University Press, Princeton, NJ.
Putnam, R.D. (2000), Bowling Alone: The Collapse and Revival of American
Community, Simon and Schuster, New York.
Sher, J.P. and Sher, K.R. (1994), Beyond the Conventional Wisdom: Rural
Development as if Australias Rural People and Communities Really Mattered,
Journal of Research in Rural Education, 10(1), 243.
Stimson, R. (2002), Planning for Fewer People, available at www.uq.edu.au/
news/?article=3069, accessed 28 May 2010.
Stimson, R.J., Stough, R.R. and Salazar, M. (2005), Leadership and Institutional
Woolcock, M. (1998), Social Capital and Economic Development: Toward
a Theoretical Synthesis and Policy Framework, Theory and Society, 27(2),
151208.

14. Economic development incentives
and the measurement of local
endogenous growth: is there a need
for modeling adjustment?
Terry Clower
INTRODUCTION
Over the past quarter-century there has been much addition to and
improvement in theoretical constructs of endogenous growth models.
Those improvements include addressing human capital characteristics
and investments (Romer, 1986), the role of firms in regional economies
(Markusen, 1985), innovation diffusion (McCombie, 1982), entrepreneur-
ship (Acs and Armington, 2004), local leadership (Stough et al., 2007) and
many others. Of course, this sampling of contributions does not include
notable works by many researchers, nor does it cover the extensive theo-
retical debates centered on regional convergence/divergence. While the
theoretical debates and theorem proofs are rich, relatively little of this
research has employed empirical techniques to test these theories.
The first, and in many respects foremost, challenge in assessing and
quantitatively modeling endogenous growth at the local level is determin-
ing the best measure of growth to serve as a dependent variable. Unlike
national and state-level analyses, which can rely on well-tested measures
of gross product output, there are few demonstrably valid measures
of regional product available from public sector sources for sub-state
regions. Nascent efforts by the US Bureau of Economic Analysis are
just beginning to offer a measure of regional product for Metropolitan
Statistical Areas, but these data are only available for a few areas/years.
The author is not aware of any public sources for local/regional output
estimates for Australia, the UK or Canada. While there are estimates of
US local area output available from several private vendors, these data
can be cost-prohibitive for research requiring hundreds of cases to support
complex endogenous growth regression equations.
286

Economic development incentives and measurement of growth 287
In response to these data limitations, the majority of local endogenous

growth modeling efforts have relied on employment change as the best-
available measure of growth. The chapter starts with a consideration
of the strengths and weaknesses of employment change as a measure of
local endogenous growth, highlighting a particular concern regarding the
confounding influence of large non-local government incentives targeted
to specific firms on employment change measures. This concern is pre-
sented first on a conceptual level illustrated by some infamous examples of
state governments buying jobs and then offered as a simulation exercise
focused on a particular site location case in north central Texas. Then
follows a discussion on how the influence of non-local incentives could be
addressed through alternative or new measures on either the left- or right-
hand side of endogenous growth model equations.
MEASURES OF LOCAL ENDOGENOUS GROWTH
As noted above, the traditional measure for growth has been one or more
calculations of gross output. This works well for most national-level
studies and is fairly well supported by gross product measures at the state/
provincial level. However, these data are not readily available at the local
level. Some researchers have used measures of personal income for meas-
uring local-level growth (see, for example, Gkritza et al., 2007). From a
data availability standpoint, employment is the most consistently avail-
able measure of local growth, though some caution must be taken based
on these measures. For example, in the US, the Department of Commerce
offers annual estimates of firm employment by detailed North American
Industry Classification System (NAICS) code. These estimates are based
on place of work. In contrast, the US Bureau of Labor Statistics employ-
ment data are based on place of residence. This potential difference is not
great if one defines the subject area based on labor markets.1 However, the
main concern with using employment data as a measure of endogenous
growth is the effect of non-local economic influences.
One of the most useful and intellectually satisfying contributions to the
literature on the practice of endogenous growth research is the methodol-
ogy proposed by Stimson et al. (2002) using the regional share component
of a shift-share equation to represent endogenous growth essentially the
dependent variable in any number of regression models accounting for
regional development. More recently, Stimson et al. (2006) have adapted
an enhanced shift-share model suggested by Haynes and Dinc (1997).
Starting with Rigby and Andersons efforts to include output and produc-
tivity change to account for ambiguity in the source of employment shifts

in the regional component, Haynes and Dinc (1997) offer an approach for
separating the influences of labor and capital in the productivity change
portion of the regional component.
The intention of using the regional shift component is to remove any
portion of employment change that is not attributable to local character-
istics. In most cases, this approach is likely to work very well. However,
there is a possibility that the regional shift component may not sufficiently
partial out the influence of development incentives that are provided by
non-local government entities. Specifically, the problem is caused when
local economic development is advanced through one-off (non-recurring)
grants or other incentives funded through non-local sources such as state
government. Since the incentives are non-recurring, does the incentive
represent an endogenous or exogenous characteristic?
THE INFLUENCE OF INCENTIVES IN LOCAL

ECONOMIC GROWTH
For practitioners and applied researchers, economic development is based

on three sources, often called the legs of the economic development stool:
1. Attraction refers to bringing existing businesses and organizations

to the local market. This includes relocated plant sites or corporate
headquarters and possibly large-scale government enterprises such as
a military base or prison.
2. Retention strategies focus on keeping whatever firms and activities
that are located in the local community from relocating elsewhere or,
in the current economic climate, helping companies stay in business.
3. Business development is the creation of new local firms or the
expansion of existing local businesses.
Business retention and development are clearly based on local factors of

production. An argument can be made that business attraction can be
exogenous growth. The business was created and built based on factors
that may not have had anything to do with local factors of production.2
Site location decisions are typically made based on local characteristics.
The Area Development annual survey of corporate executives ranks site
location factors based on level of perceived importance. As shown in
Table 14.1, traditional endogenous factors such as convenient infrastruc-
ture access, labor costs and availability, tax rates and proximity to markets
are all important in site location decisions (Gambale, 2008). Government-
funded incentives and tax exemptions make the top ten in important local

Table 14.1 Site selection factors according to corporate executives
Rank Factor 2007 2006

1 Highway accessibility 96.9 90.9
2 Labor costs 92.3 95.0
3 Energy availability and costs 89.0 82.4
4 Availability of skilled labor 88.7 85.1
5 Occupancy or construction costs 88.2 85.5
6 Available land 85.4 73.3
7 Corporate tax rate 83.8 90.8
8 State and local incentives 83.4 88.6
9 Environmental regulations 83.2 68.9
10 Tax exemptions 82.8 86.7
10t Proximity to major markets 82.8 76.9
11 Availability of advanced ICT services 82.2 n/a
12 Low union profile 80.6 78.4
13 Availability of buildings 79.3 n/a
14 Right-to-work state 72.1 67.1
15 Proximity to suppliers 71.8 49.3
16 Expedited or fast-track permitting 71.5 n/a
17 Availability of unskilled labor 65.2 65.3
18 Availability of long-term financing 63.0 64.1
19 Raw materials availability 62.5 64.1
20 Training programs 56.6 56.0
21 Accessibility to major airport 54.4 61.4
22 Railroad service 38.1 20.8
23 Proximity to technical university 32.7 30.0
24 Waterway or ocean port accessibility 15.2 17.0
Source: Compiled by the author using information available from Area Development
(www.areadevelopment.com).
characteristics for companies ahead of buildings, availability of financing,

training programs and other factors. The findings are similar for Area
Developments survey of site location consultants with incentives and tax
exemptions ranking number 6 and 7 in importance, respectively (ibid.).
Compared with most other developed nations, even those that share a
history in English common law, the US federal system pushes more gov-
ernmental functions and taxing authority to state and local governments.
With the advent of first wave (Blakely and Bradshaw, 2002) industrial
recruitment strategies that included direct incentives, states and the locales
began competing for firm relocations. Some have even termed this compe-
tition the economic war among the states (Burnstein and Rolnick, 1995).

One reason that states and local government have increased their use of
incentives has been effective convergence in state and local tax burdens
driven by interstate competition. According to the Tax Foundation, the
average combined state and local tax burden in the US is 10.10 percent of
per capita income (CNN/Money, 2008). Thirty-five of the 50 states have a
combined average state and local tax burden within 1.0 percent (ibid.).
Based on an ongoing survey conducted by the Council for Community
and Economic Research, 25 of the 50 US states have non-targeted3 grants
as a part of their incentive programs (C2ER, 2008).
Bartik (2005) concludes that lowering taxes has a statistically significant
but small effect on business activity levels. This effect is larger for subur-
ban communities in intra-regional competition for employers. Peters and
Fisher (2004) conducted a meta-review (their terminology) of the litera-
ture, reporting that incentives were found in most studies to have ambigu-
ous or very minor impacts. Gabe and Kraybill (2002) found a negative
relationship between employment change and state incentives in Ohio.
There are, of course, many other studies that could be cited, but the results
are consistently unclear. There is no consistent evidence that incentives
lead to growth, especially employment growth. Why then would there be
an issue with the measurement of endogenous growth?
Simply stated, the magnitude of incentives is tracking upwards and
in many cases has become large enough that dispassionate practitioners
strongly feel that they play a role in site location decisions. Admittedly,
this observation is based largely on casual empiricism, but one can find
compelling evidence to support the notion that incentives can trump other
business location factors.
Alabama
The state of Alabama has taken a very aggressive approach to attract-

ing auto assembly plants. The first to arrive was Mercedes Benz in 1993.
Mercedes Benz artfully played North Carolina, South Carolina and
Alabamas state negotiating teams off against each other to drive up
the value of the incentives being offered (Buchholz, 2009). In the end,
Mercedes Benz, arguably one of the premier auto manufacturers in the
world, chose Vance, Alabama, population 480, as the site for a new assem-
bly plant. Vance, Alabama, is located between the cities of Birmingham
(45km) and Tuscaloosa (20km). In exchange, the company received incen-
tives, including direct payments for workers salaries in the early years of
the agreement, that were valued at US$300 million, or somewhere between
US$153 000 and US$220 000 per job created. Demonstrating further that
these incentives were beyond the capacity of local government, the state of

Alabama eventually had to borrow money from its own state employees
pension fund to pay for the promised incentives (ibid.).
Cautioned to a small extent by the near financial disaster of the
Mercedes Benz deal, the state of Alabama and local governments offered a
comparatively paltry US$158 million incentive package to Honda in 1999
or about $105 000 per job created (Anonymous, 1999). The states largess
grew once more in the early part of this decade with a state and local
incentive package for Hyundai totaling about US$253 million, or about
US$117 000 per promised job (Beyerle, 2002).
Texas
In Texas, the state legislature set up funding for the Texas Enterprise Fund
to be used by the governors Office of Economic Development as a deal
closer. These funds are effectively grants to attract or retain businesses,
or to promote expansion in a targeted industry. In Fiscal Years 200407
(the state of Texas operates on a biennium budget), the governors office
granted over US$360 million to 39 companies, of which US$324.1 million
has actually been distributed. Tables 14.3 and 14.4 provide information
on the grants. All grants are tied to job creation and capital investment
criteria. So far about US$300 000 has been recovered through clawbacks
that are included in the grant contracts for failure to create the promised
number of jobs.
Included in the Texas Enterprise Fund recipients is Ruiz Foods, a maker
of frozen Mexican food. Ruiz foods relocated from California to the
cityof Denison located in Grayson County, Texas. This will be the subject
of the case study/simulation discussed later in the chapter. In exchange
for locating in Denison and bringing 423 jobs to the local economy, Ruiz
Foods received an Enterprise Fund (EF) grant of US$1.5 million in June
of 2005. The EF grant is in addition to any local incentives such as tax
abatements, grants, permit fee waivers and the like. There have been nine
EF grants of at least US$10 million with two at the US$50 million level
(Texas Instruments and Sematech).
THE PROBLEM FOR ENDOGENOUS GROWTH

MEASURES
As noted earlier, using gross regional employment data as the dependent

variable in an endogenous growth model would potentially bias the result-
ing regression coefficients due to overall economic and industry-specific
trends. The regional component of a shift-share analysis is intended to

capture the local competitive effect of the sum of all local characteristics.
These local effects are further refined by the Haynes and Dinc (1997) mod-
eling approach. Still, even with these model improvements over traditional
shift-share adjustments, employment change is still supposed to be related
to local characteristics in the regional component. However, the shift-
share methodology has no recognized method for adjusting the regional
component for exogenous incentives.
The presence of a state-level incentive is, in and of itself, not a problem
when assessing local economic development characteristics, as long as the
incentive is available on a recurring basis. Otherwise, the assessment of
any endogenous growth model would be unreliable because the assumed
competitive advantage would not be present in another time period.
Therefore, if the time period covered by the model coincides with a local
firm receiving a large, one-off subsidy, the assessment of regional competi-
tive strengths is potentially rendered invalid. This is especially true if the
incentivized firm represents a substantial change in area employment, such
as the aforementioned Mercedes Benz assembly plant, which could bias
the coefficients of the growth model at unknown magnitudes. To illus-
trate this point I will use the traditional shift-share calculation for food
manufacturing firms (NAICS 311) to assess the economy of the Sherman-
Denison Metropolitan Area (MA) of Texas.
The Sherman-Denison Example
Sherman-Denison MA is composed of Grayson County. In recent years

two major employers in Sector 311, Pillsbury and Oscar Mayer, closed
plant operations in Grayson County. This resulted in substantial employ-
ment declines in Grayson County while this industry was comparatively
stable at the national level, potentially suggesting that this areas endogen-
ous characteristics were somehow not competitive (see Table 14.2).
As noted earlier, Ruiz Foods moved their food processing plant to
Denison, Texas in June 20054 occupying the building abandoned by Oscar
Mayer (a meat processing plant) and employing 423 individuals. The
availability of a building contributes to Grayson Countys positive endo-
genous characteristics. On the other hand, we have been informed by local
economic development professionals that very few of the former employ-
ees of either Pillsbury or Oscar Mayer are currently employed by Ruiz
negating the potential advantage of an available, specialized labor force.
For illustrative purposes, the simulation focuses on the period subse-
quent to the last national recession. From 2003 through March of 2005,
employment in Sector 311 continued to decline in Grayson County but
rose after Ruiz Foods relocated to the area. As shown in Table 14.3, the

Table 14.2 Employment and establishments in Grayson County and the

United States Sector 311 food manufacturing and totals
Year Grayson County Total Grayson US 311 US Total

311 Employment County Employment Employment
Employment
1998 1 445 38 179 1 464 419 108 117 731
1999 914 38 147 1 464 354 110 705 661
2000 749 39 519 1 468 254 114 064 976
2001 728 39 252 1 470 146 115 061 184
2002 387 36 693 1 443 766 115 061 184
2003 139 37 337 1 495 998 113 398 043
2004 125 36 644 1 482 261 115 074 924
2005 111 37 036 1 469 730 116 317 003
2006 404 37 459 1 458 738 119 917 165
Source: Compiled by the author using data available from the US Department of
Commerce, Census Bureau and County Business Patterns.
national share component of the industry-specific shift share totaled eight

jobs. The industrial share component, reflecting declining employment
at the national level, is 11.5. That leaves an employment gain of 268.5
for the regional share, indicating that the entire gain in employment is
attributable to Ruiz Foods. It appears that sector employment would
have declined further from 2005 levels if Ruiz Foods had not located to
the area, but local economic development practitioners suggest that the
relocation would not have happened without the state incentive funds. It
is reasonable to assume that other site location factors influenced Ruizs
decision; however, the growth in the regional share for this sector in
Grayson County is, at least in part, not attributable to endogenous growth
factors.
In broader assessments of endogenous growth, employment change is
measured across all employment sectors of the local economy simultane-
ously, often at a two-digit data disaggregation (major industry sectors).
Is the potential bias observed when examining a single industry a problem
for assessments that look more broadly at local employment change? To
examine this possibility, separate shift-share calculations are made for 19
major industry sectors present in Grayson County.
As shown in Table 14.4, Grayson County shows substantial competi-
tive weakness as indicated by the sum of the regional shares for the major
industry categories at negative 1266 jobs on a base of 37 334 jobs in 2003.
The strongest performing sectors were lodging and food services (349.0

Table 14.3 Shift-share calculations for Sector 311 food manufacturing

Grayson County, Texas 200306
Component % Change Share

National 119 917 165113 398 043 0.0575 139* 0.0575 8.0
shift
Industry 1 458 7381 495 998 0.0249 139* (0.02490.0575) 11.5
shift
Regional 404139 1.9065 139* (1.90650.0249) 268.5
shift
Total shift 265.0
Source: The author.
regional share), manufacturing (324.7), and retail trade (152.2). If you

remove the effect of the Ruiz Foods relocation, traditional basic industries
shrank substantially in Grayson County. The strongest performing sectors
were in low-value-added services. Perhaps this indicates that Grayson
County really requires exogenous intervention to promote basic industry
development.
It is noteworthy that without Ruiz Foods, the regional share for the
manufacturing major industry category would have likely been negative.
For Grayson County, the introduction of an incentivized firm distorts the
regional component of the shift-share analysis at the detailed industry
level (311) and the major industry group level of aggregation (31). It could
also impact an assessment of the regional strength of basic industries
when using a shift-share approach. However, even though the removal
of the jobs associated with Ruiz foods would decrease the sum of the
regional share components by an additional 30 percent, the presence of
Ruiz Foods does not greatly alter the overall assessment that the Grayson
County economy was not competitive during the 2003 to 2006 period.
Nonetheless, for this case study the impact of adding a firm to the region
through exogenous incentives could impact a statistical analysis of the
relative impacts of endogenous growth characteristics on local economic
performance.
Addressing the Potential Problem
The degree to which the potential problem of biased coefficients in an

endogenous growth model needs to be specifically addressed is first a
function of the sample used for the model. If the sample is large and the
number of subject areas having a firm that was influenced in their location

Table 14.4 Shift-share calculations Grayson County, Texas 200306
Industry 2003 2006 National Industry Regional

Employment Employment Share Share Share
11: Agriculture 15 10 0.9 2.1 3.8
21: Mining 80 85 4.6 13.0 12.6
22: Utilities 198 225 11.4 29.4 45.0
23: Construction 2 322 2 347 133.5 214.9 323.4
31: Manufacturing 6 855 6 937 394.1 636.8 324.7
42: Wholesale Trade 995 1 085 57.2 28.9 61.7
44: Retail Trade 5 861 6 368 336.9 17.9 152.2
48: Transportation 693 764 39.8 0.8 30.4
51: Information 486 486 27.9 55.4 27.5
52: Finance 2 465 2 322 141.7 71.8 212.9
53: Real Estate 418 412 24.0 11.1 41.2
54: Professional 1 107 877 63.6 44.0 337.7
Services
55: Management 54 60 3.1 2.4 5.3
56: Administrative 2 073 1 720 119.2 244.3 716.5
Support
61: Education 435 455 25.0 6.8 11.8
62: Health Care 7 935 7 892 456.2 46.0 545.2
71: Arts & 377 308 21.7 7.3 97.9
Entertainment
72: Lodging & Food 3 459 4 120 198.9 113.1 349.0
Services
81: Other Services 1 506 1 573 86.6 60.9 41.4
Totals 37 334 38 046 2 146.3 168.5 1 265.8
Source: Compiled by the author using data available from US Department of Commerce,
Census Bureau, and County Business Patterns.
decision by exogenous incentives is small, then the potential bias is likely

inconsequential. For example, Stough et al. (2007) examined 245 metro-
politan statistical areas in their endogenous growth model. In contrast,
Stimson et al. (2008) studied local government areas by state in Australia
with as few as 47 subjects in their model of the Victoria economy. It is
beyond the scope of this analysis to perform sensitivity analyses for the
described effect, but one could imagine that a few instances of exogenous
incentives would have little measurable influence on the models, but the
likelihood of such influence is greater with a small sample.
Unfortunately, large samples do not necessarily shield studies of US
sub-state economic areas from the effect described in this chapter. As

noted earlier, the use of state-level, one-off incentives is hardly a rare

occurrence. Therefore, one would need to at least suspect that many of the
metropolitan areas studied by Stough et al. (2007) are home to firms influ-
enced by exogenous incentives. Still, if there is reason to conclude that the
number of industry share components impacted by the presence of firms
receiving exogenous incentives is small, a note offering reader caution is
probably sufficient.5
If regression modeling operated perfectly, the influence of any con-
founding factor would be partialed to the error term and left as unex-
plained variance. One would have to test for any correlations between
the error term and the independent variables to make this assumption. Of
course, cases of very large incentives targeted to a small economic area,
such as Mercedes Benz in Alabama, are likely to appear as outliers in the
model and would either be dropped from the model or have a statistical
treatment to mute the influence of the case on model coefficients.
Another argument for the do nothing approach is to recognize that
over the course of several years, the effect of the exogenous incentive would
diminish. For example, recipients of funds from the Texas Enterprise
Fund may guarantee that they will sustain their promised employment
level for a certain number of years. But, having the firm choose to stay in
the area, or in the best case where the incentivized firm grows, most likely
reflects usual endogenous economic conditions in the local area.
Adjustments to the dependent variable (employment) are problematic.
While there are methods and techniques that could be applied to assessing
the relative impact of a particular incentive on site location decisions, the
incidence of strategic behavior on the part of business leaders makes the
finding suspect. The author is not aware of any research that has gained
acceptance among scholars and practitioners on the assessment of the con-
tributions of incentives by type and magnitude on site location decisions
that are scientifically rigorous.
A more practical approach to dealing with the issue of partialing out
the effects of exogenous incentives from endogenous growth models on the
left-hand side of the equation is to choose a variable that measures growth
that is not greatly sensitive to job counts. Per capita measures of income
or perhaps local tax revenues could be tested, although this would address
local/regional economic development, not just economic growth an
important difference. Still, it would be interesting to test existing endogen-
ous growth models using per capita income or per capita government
revenue to see how the explanatory variables relate to these alternative
measures of economic performance.
Theoretically, the issue raised here could be addressed in the right-hand
side of the regression equation by including a variable(s) describing the

competitive effect of the exogenous incentive. However, in practice this

would be very difficult because the researcher would need to know what
other offers the incentivized firm received before making their site location
decision, which is impossible in practice. More importantly, the type of
incentives discussed here are often highly political in nature. Their applica-
tion may not follow any economically rational model.
Another right-side-of-the-equation approach would be to consider
variables that account for why a particular firm is chosen to receive non-
recurring exogenous incentives. As noted above, there are often political
considerations in granting firms large state-level grants and other incen-
tives. These political considerations could reside with the firm itself call
it political connectedness or influence. However, the political dimension
to selecting incentive recipients could be based on the political influence of
the host community.
Stimson et al. (2005) introduced leadership and institutional factors
to endogenous growth modeling. Though their variables do not specifi-
cally address political leadership, this provides a basis for the inclusion of
variables that may explain additional variance in the endogenous growth
model. Simply put, the ability of an economic area to attract directly, or
through the firm, an exogenous incentive is effectively an endogenous
local trait. A variable testing of this trait could be developed as an index
such as a local area government equivalent of the Power Advantage Index
(Mielcova and Cemerkova, 2008) or the Hirschmann-Herfindahl index of
market share adapted for political representation.
CONCLUSIONS
The influence of government incentives on site location decisions remains

problematic for researchers and economic development practitioners,
especially in the US. This chapter has sought to draw attention to the
particular problem of potential bias caused by large, usually one-off
exogenous incentives on local endogenous growth models when the
dependent variable is job change from the regional share component of
a shift-share calculation. The case study of a food manufacturing firm
relocating to Grayson County, Texas, after receiving a large grant from
the Texas Enterprise Fund illustrates the concern. It was shown that if
one assumes that the incentive affects the firms location decision, then
the regional share component is heavily biased if the analysis focuses on
a single detailed industry. This case also shows the bias would exist to a
lesser extent when the share component represents an aggregated major
industry sector (manufacturing) and suggests bias if the model considered

only traditional basic sectors. However, the case study showed that the
influence was greatly muted when major industry regional shares are
summed for the subject area.
Several alternatives are offered for addressing the potential problem
including doing nothing, choosing an alternative dependent variable, or
addressing the concern by including appropriate explanatory variables
in the model to partial out the influence of exogenous incentives. It is
suggested that testing alternative dependent variables and new explana-
tory variables may contribute to endogenous growth modeling exercises
beyond addressing the particular concern highlighted in this chapter.
The choice among the offered alternatives, or preferably the develop-
ment of better alternatives, is left to the discretion of the research teams.
However, even if a do nothing alternative is chosen, researchers should
be able to identify within the cases examined in their models instances of
very large exogenous incentives occurring during their study period with
appropriate cautions for the potential of biased coefficients.
There is one additional perspective that could prove important. The
argument presented in this chapter focuses on the potential for assigning
a better assessment of economically competitive endogenous characteris-
tics than is deserved by the subject area. However, another area is equally
affected by the influence of exogenous incentives but in the opposite
direction. The economic area that loses employment because a firm is
lured away by large incentives may indeed have a competitive disadvan-
tage but is it an endogenous weakness? In the US, where the frequency
and magnitude of economic development incentives increasingly distort
rational economic (site location) decision-making, one has to wonder if
endogenous models at a local/regional level can really capture economic
reality.
NOTES
1. Metropolitan Statistical Areas in the US are aggregations of counties based, in part, on

worker commuting patterns.
2. This assumes that the local market is not a prime market area for the firms products and
is not a source of scarce raw materials.
3. Non-targeted meaning that they are not specifically designated for a particular use, such
as training, or to a particular area, such as enterprise zones.
4. Annual employment data presented in Table 14.2 is for the week of 12 March for the
specified year; therefore Grayson Countys Sector 311 totals for 2005 do not reflect the
presence of Ruiz Foods.
5. This would be similar to the caution offered by Stimson et al. (2008) on the potentially
biasing impact of their data reflecting employee place of residence rather than place of
work. It is a potential weakness in their analysis, but one for which there is little that can
reasonably be done to correct.

REFERENCES
Acs, Z. and Armington, C. (2004), Employment Growth and Entrepreneurial
Activity in Cities, Discussion Papers on Entrepreneurship, Growth, and Public
Policy, No. 1304, Max Planck Institute for Research into Economic Systems,
Jena, Germany.
Anonymous (1999), Incentive Deal of the Month, The Site Selection Online
Insider, available at: http://www.siteselection.com/ssinsider/incentive/ti9906.
htm; accessed 28 May 2010.
Bartik, T. (2005), Solving the Problems of Economic Development Incentives,
Growth and Change, 36 (2), 13966.
Beyerle, D. (5 April, 2002), State Unveils Hyundai Incentives, Times Daily
(Florence, Alabama), 1B, 2B, available at http://news.google.com/newspapers
?Nid=1842&dat=20020405&id=DWweAAAAIBAJ&sjid=VskEAAAAIBAJ&
pg=5017,478953; accessed 10 June 2010.
Blakely, E. and Bradshaw, T. (2002), Planning Local Economic Development:
Theory and Practice, 3rd edition, Sage, California.
Buchholz, D. (2009), Business Incentives Reform. Case Study: Mercedes Benz:
Deal of the Century, CFED, available at http://www.cfed.org/assets/pdfs/
business-incentives-reform.pdf; accessed 10 June 2010.
Burnstein, M. and Rolnick, A. (1995), Congress Should Stop the Economic War
Among the States, The Region, 9, 320.
C2ER (2008), State Development Incentive Database, see www.c2er.org for
access.
CNN/Money (2008), How Tax-friendly is Your State?, available at http://money.
com/pf/ features/lists/taxesbystate2005/index.html; accessed 28 May 2010.
Gabe, T. and Kraybill, D. (2002), The Effect of State Economic Development
Incentives on Employment Growth of Establishments, Journal of Regional
Science, 42(4), 70330.
Gambale, G. (ed.) (2008), The 22nd Annual Corporate Survey, available at: www.
areadevelopment.com; accessed 28 May 2010.
Gkritza, K., Sinha, K., Labi, S. and Mannering, F. (2007), Influence of Highway
Construction Projects on Economic Development. An Empirical Assessment,
Annals of Reginal Science, 42(3), 54563.
Haynes, K. and Dinc, M. (1997), Productivity Change in Manufacturing Regions:
A Multi-factor/Shift-Share Approach, Growth and Change, 28(2), 20121.
Markusen, A. (1985), Profit Cycles, Oligopoly, and Regional Development, MIT
McCombie, J. (1982), How Important is the Spatial Diffusion of Innovations in
Explaining Regional Growth Rate Disparities?, Urban Studies, 19(4), 37782.
Mielcova, E. and Cemerkova, S. (2008), Political Power of the Czech
Representatives in the European Parliament, Munich Personal RePEc Archive,
paper no. 12258, available at: http://mpra.ub.uni-muenchen.de/12258/; accessed
28 May 2010.
Peters, A. and Fisher, P. (2004), The Failures of Economic Development
Incentives, Journal of the American Planning Association, 70(1), 2738.
Romer, P. (1986), Increasing Returns and Long Run Growth, Journal of Political
Economy, 94(5), 100237.
Stimson, R., Robson, A. and Shyy, T. (2006), Modeling Regional Endogenous

Growth: An Application to the Non-Metropolitan Regions of Australia,

Annals of Regional Science, online edition, 16 April, 120.
Stimson, R., Robson, A. and Shyy, T. (2008), Modeling Determinants of Spatial
Variation in Regional Endogenous Growth: Non-metropolitan Regions in the
Mainland States of Australia, paper presented at the Annual Conference of the
Australia-New Zealand Regional Science Association, Adelaide, December.
Stimson, R., Stough, R. and Roberts, B. (2002), Regional Economic Development:
Analysis and Planning Strategy, Springer, Berlin.
Stough, R., Song, C., Qian, H. and Wang, J. (2007), Modeling Endogenous
Growth in US Metropolitan Regions, paper presented to the Western Regional
Science Association Annual Meeting, Newport Beach, February.

15. Regional growth and development
theories revisited
Roberta Capello and Peter Nijkamp
TRENDS IN REGIONAL ECONOMIC GROWTH

THEORY
Regional development covers a wide range of economic policy issues

related to the need to exploit appropriate productive resources that may
contribute, or form an impediment, to the welfare of a region, in either
an absolute or a relative sense. Consequently, regional development is
associated with both efficiency objectives, such as the optimal use of scarce
factor inputs, and equity objectives (such as social cohesion and distribu-
tion of wealth), issues in modern jargon sometimes referred to as territo-
rial cohesion. Clearly such elements are also prominent in conventional
macroeconomic growth analysis (e.g., at a national scale), but a special
reason for giving explicit attention to regional economic growth lies in the
relatively small and open character of a region (or a system of regions).
Since a region forms an intermediate or hybrid spatial unit between a
nation and its citizens, regional economic growth theory uses elements
from both macroeconomic growth policy and individual welfare theory.
Against this background it comes as no surprise that in recent years
endogenous growth theory has gained much popularity in regional eco-
nomics; it is essentially a blend of microeconomics and macroeconomic
growth theory, in which smart use of the indigenous resources of a region
plays a critical role. It covers, inter alia, the linkages between income,
employment, investments, infrastructure and suprastructure. In particu-
lar, much emphasis is placed on the study of spatial socio-economic dis-
parities or convergence (including labour migration), with a particular
view to the way spatial disparities can be influenced by the deliberate
actions of stakeholders (e.g., industry, government).
It is thus increasingly recognized that regional development is not only
a spatial efficiency issue in economic policy. It is also an equity issue,
because economic development normally exhibits a significant degree of
spatial variability, and, furthermore, it is increasingly conceived of as a
301

spatial sustainability issue with strong regional and urban dimensions.

Over the past few decades, the continuous concern about unbalanced
regional development has prompted various strands of important research
literature, in particular: the measurement of inter-regional disparity; the
causal explanation for the emergence or persistent presence of spatial
variability in economic development; and the impact assessment of policy
measures aimed at coping with undesirable spatial inequity conditions.
The study of socio-economic processes and inequalities at the meso-scale
and regional levels positions regions as the core places of policy action,
and hence warrants intensive conceptual and applied research efforts.
Economic analysis of regional growth and its distribution already has a
long history and dates back to classical economists such as Adam Smith
and Alfred Marshall. From an analytical perspective, the foundations of
modern economic growth theory can be found in the early work of Solow
(1956), in which he argues that, in a neoclassical economic world, the
growth rate of a region (measured in per capita income) is inversely related
to its initial per capita income, a thesis that offers an optimistic perspective
for poor regions. Interesting regional growth models have been extensively
developed in the 1960s, in particular in a neoclassical framework (Borts,
1960; Borts and Stein, 1964 and 1968). The spatial-economic convergence
idea has attracted considerable attention over the years and has generated
interesting applied research on evolving convergence versus persistent dis-
parities (see, e.g., Barro and Sala-i-Martin, 1992). In addition to statistical
analysis, this research and policy issue has also led to new insights into the
contextual drivers of disparities, such as mobility, product diversification,
monopolistic competition, institutional impediments, and so on.
It is thus clear that, over the past few decades, a persistent unequal
distribution of welfare among regions and/or cities has been a source of
concern and inspiration for both policy-makers and researchers. Regional
development is at the heart of this concern, as it is about the geography of
welfare and its evolution. It has played a central role in such disciplines as
economic geography, regional economics, regional science and economic
growth theory. The concept is not static in nature, but refers to complex
spacetime dynamics of regions (or an interdependent set of regions). In
addition, the actual measurement is also dependent on the geographical
scale used. Changing regional welfare positions are often hard to measure,
especially in a comparative multi-regional context. In practice, we often
use Gross Domestic Product (GDP) per capita (or growth therein) as a
statistical approximation (see Stimson et al., 2006). Sometimes alternative
or complementary measures are also used, such as per capita consump-
tion, poverty rates, unemployment rates, labour force participation rates,
or access to public services. These indicators are more social in nature

Regional growth and development theories revisited 303
and are often used in United Nations welfare comparisons. An example

of a rather popular index in this framework is the Human Development
Index, which represents the welfare position of regions or nations on a
01 scale using quantifiable standardized social data (such as employ-
ment, life expectancy, or adult literacy) (see, e.g., Cameron, 2005). In all
cases, however, spatial socio-economic disparity indicators show much
variability.
The history of economic research has witnessed an ongoing debate on
income convergence among countries or regions, both theoretically and
empirically, often with due emphasis on effective and efficient policy meas-
ures and strategies. This continues to be an important research topic, since
regional disparities may have significant negative socio-economic cost
consequences because of, for instance, social welfare transfers, inefficient
production systems (e.g., due to the inefficient allocation of resources), and
undesirable social conditions (see Gilles, 1998). In a neoclassical frame-
work of analysis, these disparities (e.g., in terms of per capita income)
are assumed to vanish in the long run, because of the spatial mobility of
production factors, which ultimately results in an equalization of factor
productivity in all regions. Clearly, long-range factors such as education,
R&D and technology play a critical structural role in this context. In the
short run, however, regional disparities may show rather persistent trends
(see also Patuelli, 2007).
Inter-regional disparities can be as mentioned above measured using
various relevant categories, such as (un)employment, income, investment,
growth, and so on. Clearly, such indicators are not entirely independent,
as, for instance, is illustrated in Okuns Law, which assumes a relationship
between economic output and unemployment (see Okun, 1970; Paldam,
1987). The empirical research on convergence has often been based on
cross-sectional analysis, for example, on the basis of concepts related to
beta-convergence and sigma-convergence (see, e.g., Baumol, 1986; Barro,
1991). More recently, time-series analysis has also been used extensively,
based on notions from stationary time processes (see, e.g., Bernard and
Durlauf, 1995). The findings from these different strands of the literature
are not always identical, however, and in recent years this has stimulated
new research efforts inspired by endogenous growth theory. The conver-
gence of regional disparities is clearly a complex phenomenon, as it refers
to a variety of mechanisms through which differences in welfare between
regions may vanish (see Armstrong, 1995). In the modern convergence
debate, we observe increasingly more attention to the openness of spatial
systems, reflected, inter alia, in trade, labour mobility, commuting and
so on. (see, e.g., Magrini, 2004). In a comparative static sense, conver-
gence may have various meanings in a discussion on a possible reduction

in spatial disparities among regions (see also Baumol, 1986; Barro and
Sala-i-Martin, 1992; Bernard and Durlauf, 1996; Boldrin and Canova,
2001). In particular, there is: b-convergence: a negative relationship
between per capita income growth and the level of per capita income in
the initial period (e.g., poor regions grow faster than initially rich regions);
s-convergence: a decline in the dispersion of per capita income between
regions over time.
The convergence hypothesis in neoclassical economics has been widely
accepted and discussed in the literature, but is critically dependent on two
hypotheses (see Cheshire and Carbonaro, 1995; Dewhurst and Mutis-
Gaitan, 1995): (1) diminishing returns to scale in capital should prevail,
which means that output growth will be less than proportional with
respect to capital; (2) technological progress will generate benefits that also
decrease with its accumulation (i.e., diminishing returns).
A wealth of studies has been carried out to estimate the degree of
b-convergence and s-convergence (see, e.g., Barro and Sala-i-Martin
1991, 1992). The general findings are that the rate of b-convergence is
in the order of magnitude of 2 per cent annually, while the degree of
s-convergence tends to decline over time, for both US states and European
regions. Clearly there is still an ongoing debate worldwide on the type of
socio-economic convergence, its speed, its multidimensional conceptuali-
zation and its causal significance in the context of regional policy measures
(see, e.g., Fagerberg and Verspagen, 1996; Galor, 1996; Fingleton, 1999).
Important research topics in the current literature appear to be: the role
of knowledge and entrepreneurship; spatial heterogeneity in locational
or socio-cultural conditions; and institutional and physical barriers. An
important new topic in the field is group convergence (or club conver-
gence), which means a convergence of sets of regions towards a more
homogeneous cluster.1
We may conclude that the research field of spatial disparities is still
developing and is prompting fascinating policy issues that deserve in-
depth attention in the years ahead. Besides the concern on policy issues,
also in the academic arena too much interest has arisen over the last
decade in spatial development. The degree of scientific consensus as well
as the degree of cross-fertilization of ideas between regional economists
and mainstream economists has led to interesting debates in regional
science circles.
In addition, in a period of globalization (including financial crises)
and of the creation of broad single-currency areas, regions (and also
nations) have to keep an eye on the competitiveness of their production
systems, because no spontaneous or automatic adjustment mechanism
is yet at work to counterbalance a lack (or an insufficient growth rate)

of productivity. Local specificities and local material and non-material

assets (including knowledge and learning mechanisms) become strategic
elements upon which the competitiveness of regions is based. More than
ever, theories of regional growth and development need to be able to inter-
pret the way in which regions achieve a role in the international division of
labour and, more important, the way in which regions can maintain this
role over time.
Finally, it should be recognized that regional growth trajectories are
often confronted with sustainability issues and scarcity of environmental
resources, such as environmental quality, natural resource security, urban
quality of life, or spatial biodiversity. Such important issues are increas-
ingly a critical element of a regional welfare function. Some sustainability
elements may be interpreted in terms of scarce input factors, such as bio-
ethanol, solar, or wind energy, while others are seen as welfare constitu-
ents, such as the quality of the living environment. This new scarcity has
important spatial components and has to be analysed from the perspective
of regional development as well. Consequently, regional development is
not only a matter of regional competition for the use of scarce resources
but also a matter of creating the environmental conditions for sustainable
wealth creation and enjoyment. This calls increasingly for a broader welfare
interpretation of regional growth, in which the triangle of efficiency, equity
and sustainability form the driving forces of regional development.
Our aim in this chapter is to present a selection of recent contributions
that explain regional growth and local development, with the aim to
highlight:
recent advances in theories;

policy implications of these theories;
cross-fertilization of ideas among regional economists and main-
stream economists.
In the sections that follow we first discuss the theoretical advances on

regional growth achieved in recent years in different parts of the world.
Next we address sustainability issues from a regional perspective and then
we provide an overview of future challenges in this field. That is followed
with some concluding remarks.
DEVELOPMENTS IN REGIONAL GROWTH THEORY
As mentioned above, the complex issue of regional development and

growth has been a focal point of interest in recent decades. The great

Table 15.1 Main tendencies in theories of regional economics
New Regional Growth Regional Development

Theoretical Theories Theories
Tendencies
Realism in Endogenous growth Intangible assets as sources of
theoretical determinants regional competitiveness
approaches A role in growth models for Determinants of success and
the complex non-linear and failure of clusters of SMEs,
interactive behaviour and local districts, milieux
processes that take place in An active role of space in
space knowledge creation
Imperfect market conditions
in growth models
Growth as a long-term
competitiveness issue
Technological progress as an
endogenous factor of growth
Dynamic Evolutionary trajectories of Dynamic rather than static
approaches non-linear interdependencies agglomeration economies
of complex systems Dynamics of local districts and
Non-linear, dynamic growth local milieux
models
Source: Elaboration on Capello (2008).
number of relatively new and advanced contributions in the area of

regional development/growth theories does not allow us to offer a detailed
review of all individual achievements made; moreover, a detailed presenta-
tion of all new ideas would probably not be very stimulating. Our view is
that selective and focused attempts to highlight general theoretical trends
offers a more fruitful basis for a debate on current weaknesses and on pos-
sible future directions of regional economics. Inevitably, the set of mega-
trends in regional growth analysis offered in this section is both selective
and incomplete (Capello, 2008).
Table 15.1 summarizes the two main mega-trends that, in our view,
largely characterize the theoretical developments over the last two decades
in regional economics, and that are common to urban economics and to
many other disciplines (Capello and Nijkamp, 2004):2 the need for more
realism and the move towards dynamic rather than static approaches.
These theoretical perspectives are presented in Table 15.1 for both
regional growth and regional development theories, the former aiming
at explaining the aggregate growth rate of income and employment in a

formalized and quantitative way, the latter oriented towards the identi-
fication of all tangible and intangible qualitative elements of the growth
process of regions.
The Need for More Realism
The first tendency that has accompanied the theoretical development in

the field is the need for more realism in sometimes rather abstract con-
ceptual approaches, by relaxing most of the glaringly unrealistic assump-
tions of the basic theoretical models. This tendency is justified by the need
to broaden the interpretative capacity of the theoretical toolbox in this
research field by searching for theories that are better able to reflect issues
and policy strategies for the real world.
In recent years, more realism has been required to insert into growth
models the complex non-linear and interactive behaviour and processes
that take place in space, and to understand regional competitiveness in
terms of endogenous factors. The question of whether a region is intrinsi-
cally capable of growing as a result of endogenous forces has been a source
of debate for decades. Industrial specialization, infrastructure endow-
ment, central location, production factor endowment, or agglomeration
economies have all been emphasized in the academic arena as driving
forces of local economic success. The theory of industrial organization has
also provided important contributions here.
One important step forward in this field has been the focus on econo-
mies of scale in production, which, together with non-linear transpor-
tation costs, have been introduced into a (quantitative) inter-regional
growth model; the final spatial distribution of activities critically depends
on initial conditions, including the starting distribution of activities and
the nature of the non-linearities embedded in the activitytransportation
interactions, which give rise to multiple equilibria (Krugman, 1991). The
additional value of this approach sometimes referred to as the new
economic geography rests in skilfully modelling the interaction between
transportation costs and economies of scale in production, although the
determinants of endogenous growth have already long been emphasized,
starting with the Myrdal-Kaldor model (increasing returns, cumulative
self-reinforcing growth patterns). The aim to incorporate agglomeration
economies in the form of increasing returns into conventional models
of a strictly macroeconomic nature was made possible by advances in
more sophisticated mathematical tools for the analysis of the qualita-
tive behaviour of dynamic non-linear systems (bifurcation, catastrophe
and chaos theory), together with the advent of formalized economic
models that abandoned the hypotheses of constant returns and perfect

competition (Fujita and Thisse, 1996, 2002). Further links to complexity

theory are also plausible in this context.
In the spirit of Krugmans efforts, in the field of endogenous determi-
nants great emphasis has recently been put on knowledge as a driving
force for development, and, what is really new, on the endogenous self-
reinforcing mechanisms of knowledge creation. Macroeconomic models
of endogenous growth, where knowledge is generally embedded in human
capital (Romer, 1986; Lucas, 1988), have widely dominated the academic
arena in the last decade. Their main aim was to insert more realism into
growth models by relaxing the unrealistic assumption that technologi-
cal progress is an exogenous process in an economic system. In the new
growth theories, instead technological progress is regarded as an endogen-
ous response of economic actors in a competitive environment. More
specifically, increasing returns in factor productivity stemming from
endogenous factors such as innovation, scale economies and learning
processes are included in a neoclassical production function, where they
offset the effect of the marginal productivity of the individual factors,
which the traditional neoclassical approach assumes to be decreasing.
These adjusted assumptions have led not only to interesting analytical
findings but also to new far-reaching policy implications.
Endogenous growth theory is already two decades old and has played
a central role in the growth debate since the 1990s. The main idea of these
new contributions is that technological progress is not exogenously given,
but a self-organized response of individual agents in an entrepreneurial
business environment. Consequently, in contrast to earlier macroeconomic
explanatory frameworks, the emphasis is much more on firms individual
economic behaviour (see, e.g., Barro and Sala-i-Martin, 1997; Aghion and
Howitt, 1998). In this way, it can be demonstrated that regional growth
is not the result of exogenous productivity-enhancing factors but rather
is the result of deliberate choices of individual actors (firms and policy-
makers). Individual smart behaviour is thus at the centre of this debate.
The strategic relevance of knowledge for innovation and entrepreneur-
ship is thus increasingly recognized. The spatial distribution of knowledge
and its spillovers are considered as important success factors for regional
development in an open competitive economic system. Thus, the geo-
graphical patterns of knowledge diffusion, as well as the barriers to access
to knowledge, are decisive for regional development in a modern global
and open space economy. Consequently, knowledge policy often insti-
gated by ICT advances is a critical success factor for regional welfare
creation (see, e.g., Keeble and Wilkinson, 1999; Acs et al., 2002; Dring
and Schnellenback, 2006), where knowledge refers to education, learning,
training, creativeness and R&D.

It is noteworthy that the identification of endogenous determinants of

growth was the crucial scientific issue that explained the birth of regional
development theories. Development is in fact by definition endogenous. It
is fundamentally dependent on the concentrated organization of the ter-
ritory, embedded in which is a socio-economic and cultural system whose
components determine the success of the local economy:
entrepreneurial ability;
local production factors (labour and capital);
the relational skills of local actors.
All these generate cumulative knowledge acquisition and, moreover, a

decision-making capacity that enables local economic and social actors to
guide the development process, support it when it is undergoing change
and innovation and enrich it with the external information and knowledge
required to harness it to the general process of growth, and to the social,
technological and cultural transformation of the world economy. The
micro-behavioural nature of these approaches allowed a deep understand-
ing of the sources of territorial externalities, and of increasing returns in
the form of agglomeration economies, which are at the basis of industrial
cluster formation. Within this approach, much emphasis is given to the role
of entrepreneurship in regional development (Nijkamp and Stough, 2004).
More realism in the study of clusters and their determinants has called
for a better understanding of the success and failures of local productive
systems, hardly explained in the first theories proposed. Dynamic agglom-
eration economies defined as territorial advantages that act on the
capacity of firms and regions to innovate have become the centre of most
recent theoretical reflections in this field, giving rise to neo-Schumpeterian
approaches in regional development. A major debate dominates the aca-
demic arena, with the aim of identifying the role of space in innovative
processes. This has prompted a new strand of growth literature on regions
or cities.
In the vast regional and urban literature created in this field,3 the
endogenous determinants of innovation are increasing returns in the form
of dynamic locational advantages deriving from the following:
1. The spatial or geographical proximity of firms, which facilitates the

exchange of tacit knowledge. This characterizes reflection by eco-
nomic geographers concerned to explain the concentration of innova-
tive activities.
2. The relational proximity of firms, defined as interaction and coop-
erativeness among local agents, the source of collective learning

processes and socialization to the risk of innovation (i.e., territorial-

ized relations among subjects operating in geographical and social
proximity). This was the approach taken by territorial economists in
explaining the dynamics of local systems in terms of local innovative
capacity.
3. Institutional proximity, which takes the form of rules, codes and
norms of behaviour that facilitate cooperation among actors and
therefore the socialization of knowledge and assist economic actors
(individual people, firms and local institutions) to develop organiza-
tional forms that support interactive learning processes. This aspect
was emphasized by more systemic approaches that sought to under-
stand the evolution of complex systems like the innovative system in a
broad global competitive network.
Dynamic Approaches
A second clear mega-trend in theoretical developments typical only of

regional development/growth theories has been the attempts to move
towards dynamic approaches. Time matters, as well as space, in regional
science, and this also holds in regional economics. The effort to encapsu-
late time in spatial analyses has been accomplished in two different ways,
according to two different meanings of time applied in the two fields of
analysis: a more traditional chronological time; and time as the rhythm of
innovative phenomena that occur in the territory that has been applied in
regional growth models all over the world.
The introduction of chronological time within spatial analysis is not
at all a simple task, since it requires a mathematical and methodological
toolbox, only recently available to regional scientists. Theories on non-
linear regional dynamics framed in the context of chaos theory, syn-
ergetics theory, or predatorprey analysis may be mentioned here (see
Nijkamp and Reggiani, 1999). In growth models, until a few years ago, the
vast majority of experiments and applications took for granted the exist-
ence of linear and thus regular growth processes. Linear models are cer-
tainly able to generate unstable solutions, but the solutions of such models
are restricted to certain regular standard types. Such models may provide
approximate replications of short-run and medium-run changes, but fail
to encapsulate long-term developments characterized by structural shifts
of an irregular nature. This limitation has recently been overcome with the
adoption of non-linear models, which allow for a change in the dynamics
of a system generated even by small perturbations in structural forms;
structural instability means the possible existence of significant qualitative
changes in the behaviour of the system (i.e., in the state variables) that are

closely connected with bifurcation and catastrophe phenomena, which

occur if the parameter values (i.e., the control variables) are changing (see
Fujita and Thisse, 1996, 2002). The application of non-linear models to
the well-known neoclassical and Keynesian models has shown that the
deterministic and unique results achieved by the dynamic linear models
are no longer guaranteed: inter-regional income convergence determined
by the traditional neoclassical model collapses and opens the way to alter-
native possible trajectories, and equilibria solutions; non-linear Keynesian
Myrdal-Kaldor models substitute the deterministic result of continuous
growth or decline with new and opposite development trajectories, after
catastrophe phenomena occur (Miyao, 1984, 1987a, 1987b). We thus
observe a wealth of new theoretical contributions on the dynamics of
spatial development.
Such a theoretical improvement has also been useful in achieving a
greater realism of these models, which are now able to incorporate the
dynamic interactions between the components of a spatial system repre-
sented in a network constellation. The latter network approaches are func-
tionally determined by interdependencies between the behaviour of actors
and distance frictions. Such spatial interactions may be stable in nature
(i.e., operating under fixed external conditions) or subject to change as a
result of dissipative evolutionary processes in the external world. In the
latter case, model parameters become time-dependent, so that non-linear
complex dynamics may emerge (see Puu, 1991; Nijkamp and Reggiani,
1993 and 1999; Nijkamp, 2006).
Next, we also note that, in the field of regional development, conceptu-
ally speaking a different concept of time has been developed and applied.
Time la Bergson-Heidegger is interpreted as duration and a continuous
process of creation, characterized by discontinuity, irreversibility, sequen-
tiality and cumulativity. Time has thus been conceived by an important
part of regional studies as the pace of learning, innovation and creation
processes. Local clusters (and industrial districts) are, by definition, the
loci where learning and cumulative learning processes take place. The
identification of the sources and the endogenous determinants of such
processes, apart from simple physical proximity, represent a great chal-
lenge for regional economists. Knowledge spillovers, collective learning,
learning regions (or learning space) and knowledge-based regions are all
theories that embrace the most advanced perspectives in this direction. In
these theoretical approaches, therefore, innovation has become the critical
survival factor in a competitive space economy and determines the direc-
tion and pace of regional development (Nijkamp and Abreu, forthcoming).
We may thus conclude that the introduction of micro-based endogenous
growth components has led to an enrichment of regional growth theories,

in which the role of public policy is no longer seen as a top-down control

and command approach, but as a partnership model, in which strategies
are to be developed in cooperation with all stakeholders in space.
THE CHALLENGE OF REGIONAL SUSTAINABILITY
Concern for the local environment is not a recent policy issue. Already in
Ancient Rome we find examples of policy measures to mitigate the noise
nuisance of horse-drawn carriages during the night. And several medi-
eval cities in Europe had guidelines on where and when to park carriages
(including priority rules for wealthy citizens). In the welfare economics
literature, such non-market phenomena were called externalities. To
guarantee again an equalization of marginal costs and benefits, a system
of taxes (or subsidies) was foreseen, so that an optimal market equilibrium
could be re-established. These Pigouvian welfare rules were seen as inter-
esting exceptions to an otherwise perfectly working market mechanism.
However, in the age of mass industrialization and large-scale mobility,
such externalities became a dominant phenomenon rather than an excep-
tion. Already in the 1950s and 1960s, the first voices on environmental
decay were heard, often inspired by concerns about declining water quality,
noise nuisance, local health and air pollution. The real breakthrough and
awareness took place in the first wave of environmental consciousness, viz.
after the publication of the First Report to the Club of Rome in 1973. In
that global study, much concern was expressed about environmental pol-
lution, scarcity of natural resources, the threat of the population bomb,
lack of food and the position of the developing countries. The next wave of
increased interest emerged in 1987, with the publication of the Brundtland
Report. This document was greatly concerned about the issue of sustain-
able development, in particular from the viewpoint of the developing
world and from the viewpoint of the next generations. Since then, sus-
tainability has become a fashionable word but, unfortunately, it lacked
an operational definition, so that in practice this concept was uncritically
used for sectoral development, for regional or local developments, or for
development with regard to new generations. However, this concept has
prompted an avalanche of research including applied modelling research
on various elements of local, urban or regional sustainability, from the
perspective that our world calls for a holistic view on future development.
And, finally, we witness the third wave, where the interest in global devel-
opments such as climate change or the rise in sea level is anchored in
local or regional settings. This embeddedness of the global change debate
in local and regional development has prompted new types of research in

the area of regional science (see Batabyal and Nijkamp, 2008). From the
wide range of issues, we now select five important focal points that serve
to illustrate the importance of the recognition of sustainability issues in
Regional Economic Development
Regional economic development comprises more elements than regional

economic growth: for example, access to social facilities, a healthy living
environment, a high quality of education, a sustainable living environ-
ment, and so on. Such conditions manifest themselves in many facets of
the local or regional environment. We mention here three examples:
1. First, business location is nowadays not a simple cost-minimization

question, but a multidimensional trade-off of many factors, in which
the (image of the) quality of the local environment or access to recrea-
tional or sports facilities plays a critical role.
2. Second, public environmental policies may increase the locational
costs of new or incumbent firms, and hence influence their locational
decisions. Thus, the way externalities are treated by the public sector
will impact on the willingness to invest by business firms.
3. Third, the local institutional setting for example, the presence of
publicprivate partnerships may create a sense of common respon-
sibility, also for the local quality of life, so that the institutional
structure may be a driving force for local or regional sustainable
development.
The welfare of regions and the use of the physical resource base of these
regions are clearly mutually interwoven phenomena. It is thus clear that
regional economic development and sustainability strategies may be seen
as mutually complementary forces that may reinforce each other.
Natural Resources
Natural resources may have a productive meaning (e.g., mines, oil

reserves) or a consumption meaning (e.g., recreation parks or lakes for
sailing or fishing). They may be decisive for balanced regional develop-
ment, as will now be exemplified by means of three selective illustrations
on natural resource use:
1. First, deforestation may be a wise strategy from a short-term produc-

tive perspective, but calls for a careful and balanced policy in view of

long-term sustainability. An equilibrium policy that attains a balance

between long-term wealth and short-term revenues may therefore be a
wise strategy.
2. A second example concerns water provision and use. The situation
regarding the quality of water in many regions and cities is rather
critical nowadays, and requires intensified policy initiatives. Proper
management of scarce water reserves in terms of both quality and
quantity is a condition sine qua non for healthy and sustainable local
and regional development.
3. A final example concerns waste management. Waste is a necessary
by-product of any production and consumption process; it is often
a cost factor, but it may be turned into a revenue factor (e.g., in the
case of district heating, waste incineration, and so on.). Thus, the
challenge will be to turn environmental bads into environmental
goods through the proper usage of sticks and carrots. And this calls
for operational environmental research at a local scale.
Smart local resource use may improve local welfare conditions. We may
conclude, therefore, that proper waste management at local or regional
level will create favourable conditions that stimulate balanced local or
Environmental Regulation
Local or regional economic policy is not only a matter of taxes, subsidies

or prohibitions but also calls for transparent environmental regulations.
We will give a few examples here.
1. Environmental regulations have often been criticized because they

might cost jobs. But this calls for a careful empirical analysis, as there
are also many counterfactual findings based on several experiences all
over the world. There is an increasing body of literature that demon-
strates that strict environmental regulations by no means have a nega-
tive effect on economic activity, but most likely a negligible or even
positive effect in the long run.
2. Pigouvian taxation is seen as a proper response to the existence of
externalities. This may certainly be true from an economic perspective,
but the way such taxes are then allocated is decisive for their political
support and acceptability. This is also of great importance for local
taxation schemes (e.g., for parking, waste treatment, and so forth.).
3. Finally, the local or regional dimension of environmental policy is of
great importance. This has a particular relevance for the local tax base

and the distribution of public money in a system of fiscal federalism.

Such spatial demarcation issues may greatly impact on local sustain-
able development.
Environmental regulations may act as competitiveness vehicles in regional

development. All in all, we may conclude that spatial heterogeneity in
environmental regulation systems will significantly influence the economic
and environmental aspects of any sustainable development policy.
Regional Climate Change
The relationship between global change and local development has often
been neglected, but is receiving increased attention. Global warming will
certainly have a far-reaching impact on local or regional development
conditions, as illustrated by the following examples.
Changes in our ecosystems with more extremes and outliers will call
for adjusted ecosystems management at local or regional level so that these
systems can be better protected against floods or other nature disasters.
This may call for other types of land use policy.
Another example concerns the interrelationship between local and global
climatic conditions (e.g., in terms of building regulations, or water or energy
supply). This prompts increasingly proactive policies (e.g., in the EU).
Many cities are to be found along the coast or river flood plains. The rise
in sea level calls for new design and security principles in urban planning
in the decades to come. We may clearly draw the conclusion that sustain-
ability policy will call for drastic adjustments of local and regional land
use, environmental and resource policy.
Modelling Local and Regional Environments
Much quantitative research has been undertaken in the past few decades
to map out the complex policy dynamics of the modern space economy.
Computable General Equilibrium modelling and simulation modelling
have proven to be appropriate tools.
More recently, we have seen the emergence of geographic information
systems (GIS) as an example of operational model applications. A great
many useful applications can be found in the literature. For instance: (1)
Several studies have been published on housing planning and demand,
with respect not only to the quality of the dwellings in a narrow sense,
but also to the neighbourhood and accessibility effects in a broader sense
(Kain and Quigley, 1970). (2) Another example concerns location analysis
in the context of land use and land values.

GIS has been very instrumental in dealing with large spatial data sets,
for example, in evaluating commercial buildings and densities, in making
spatial assessment of land values, or in identifying risk-prone areas
(Geoghegan et al., 1997; Weng, 2002). GIS has indeed become an impor-
tant toolbox for regional science research, but its potential for local or
regional socio-economic development policy is still underutilized. There is
still a long way to go before geo-science modelling will be an integral part
of local or regional development policy.
The interwoven nature of the ecology of our planet at differ-
ent geographic-scale levels with local and regional development thus
prompts many new research endeavours. There is yet one final question:
Has sustainability had an impact on regional development policy, and is
there a link to endogenous growth policy? The answer is in the affirma-
tive. Environmental and resource conditions at local or regional scales are
no longer permanently given factors. They can be influenced and adjusted
by smart and appropriate decisions and investments in both the public and
the private sector. In other words, regional sustainable development from
an endogenous perspective is the result of deliberate choices and actions
of local and regional stakeholders, including public policy-making bodies.
NEW RESEARCH CHALLENGES IN REGIONAL

GROWTH AND DEVELOPMENT
The previous observations have clearly demonstrated that, nowadays,

fascinating new theoretical challenges are being faced by regional scien-
tists, and have to be addressed. A first challenge emerged from the attempt
to obtain advantage from a future convergence in different theoretical
approaches, a convergence only partially obtained by the new regional
growth theories. New growth theories make a commendable effort to
include space in strictly economic models. Also to be commended is the
implicit merging in their theoretical structure of the various concep-
tions of space put forward over the years: that is, the merging of the
physical-metric space represented by transport costs with the diversified
space, which assumes the existence of certain territorial polarities where
growth cumulates. However, the new economic geography is still unable
to combine the economic laws and mechanisms that explain growth
with territorial factors that spring from the intrinsic relationality present
at local level. Any approach that did achieve this would represent the
maximum cross-fertilization between location theory, development theory
and macroeconomic growth theory a synthesis that would bring out the
territorial micro-foundations of macroeconomic growth models.

What is needed, therefore, is a convincing umbrella model that

comprises the micro-territorial, micro-behavioural and intangible elements
of the development process. What is required for this purpose is the defini-
tion of patterns, indicators and analytical solutions to be incorporated into
formalized models necessarily more abstract and synthetic in terms of their
explanatory variables, variables besides the cost of transport, which remove
the role of territory in the development process. A move in this direction is
the quantitative sociology that embraces the paradigm of methodological
individualism and seeks to measure the social capital of local communi-
ties. It is obviously necessary to bring out territorial specificities within a
macroeconomic model. Or, in other words, it is necessary to demonstrate
the territorial micro-foundations of macroeconomic growth models.
Serious risks both from disciplinary barriers and from adherence to
interdisciplinary perspectives on strategic problems can be mentioned
here. They are the result not only of regional scientists narrow perspec-
tive, as mentioned by Bailly and Coffey (1994), but also of some rather
idiosyncratic attitudes of mainstream disciplines towards a clearly multi-
disciplinary science like regional science.
In this respect, the following examples may be illuminating. The first
concerns the theory on social capital developed by quantitative sociology:
the concept could take advantage of, and provide advantage to, all reflec-
tions on local synergies and milieu effects developed by regional and urban
economists, and by the strategic planning studies in the field of urban
planning. The reflections in the field of knowledge spillovers developed by
industrial economists could take advantage of regional science concepts
of collective learning and relational proximity, in which the endogenous
spatial development patterns of knowledge are not left to simple probabil-
istic contacts, but are explained through territorial processes (Camagni and
Capello, 2002). Last but not least, the theoretical reflections that character-
ize the new economic geography seem to be the result of a skilful effort of
a group of mainstream economists, driven, however, by a somewhat inex-
plicable attitude that denies the importance of well-known spatial concepts
(i.e., technological spatial externalities), or (re-)invents important spatial
concepts (i.e., cumulative self-reinforcing processes of growth; transporta-
tion costs vs. agglomeration economies in location choices). The inevitable
consequence of such an attitude is to mix the important and undeniable
steps forward made by the new economic geography school with already
well-known knowledge in the field of regional science.
Especially in the case of economics, we hope that, after the (re-)
discovered interest by mainstream economists in space, and in spatial phe-
nomena, the attitude towards regional science will change in favour of a
more cooperative and pronounced interest.

Related to the interdisciplinary challenge, a last important point is

worth mentioning. An interdisciplinary approach should lead scientists
to explore new frontiers and achieve new interpretative analytical frame-
works. However, the tendency shown in this respect is quite different,
being merely inclined to exploit passively the new ideas suggested by com-
plementary disciplines. A case worth mentioning is the enthusiastic way in
which regional scientists accepted the spatial spillover theory as a theory
that could provide a new interpretation of the role of space as a channel
for knowledge transfer.
Instead, a more critical approach to this theory shows that to some
extent it has gone some steps backwards in the interpretation of space
in spatial knowledge creation. Space is purely geographical, a physical
distance among actors, a pure physical container of spillover effects that
come about according to the epidemiological logic adopted simply as
a result of physical contact among actors. Important consequences ensue
from this interpretation of space. First, this view is unable to explain the
processes by which knowledge spreads at local level, given that it only
envisages the probability of contact among potential innovators as the
source of spatial diffusion. Second, it concerns itself only with the dif-
fusion of innovation, not with the processes of knowledge creation. It
thus imposes the same limitations as did Hgerstrands (1967) pioneering
model in regard to the spatial diffusion of innovation: the diffusion of
knowledge means adoption, and adoption means more innovation and
better performance. This ignores, however, the most crucial aspect of the
innovation process: how people (or the context) actually learn. This calls
for a more thorough and innovative investigation of cognitive processes in
a regional context (Capello, 2009). This is an aspect of overriding interest
not only for researchers but also, and especially, for policy-makers should
they wish to explore the possibilities of normative action to promote local
development.
RETROSPECT AND PROSPECT
In the globalization process of creating an open economy, local factors

and local specificities are fundamental elements on which the competitive-
ness of countries depends. They are, therefore, important areas where
practitioners and policy-makers require a sophisticated and advanced
toolbox to intervene.
Regional economics has been subject to varied and creative advances
in theoretical economic contributions. Various core tendencies in the
development trajectories of the discipline have been stressed in this

review, in particular the attempt to introduce more realism into the

theoretical approaches, combining rigorous theoretical reflections with an
understanding of the reality of place.
The force field of regional welfare is varied and sometimes unpredict-
able, as the region is part of a complex global system. An important
contributor to regional development is technological progress, an exten-
sively studied topic in the recent economic growth literature. From a
geographic (regional, urban, or local) perspective, in recent years much
attention has been paid to the spatial conditions that induce technologi-
cal progress (e.g., entrepreneurial climate, availability of venture capital,
incubator facilities, institutional transformations, etc.) (see also Longhi
and Musolesi, 2007). Furthermore, the spatial diffusion of technology
has also received considerable attention, in particular in the geography
literature.
The development of regions in an open world is also strongly influ-
enced by their environmental and resource base. And, therefore, sus-
tainability is a critical task for regional policy as well. This concept is,
however, not only a burden but opens the door for new and creative
regional growth strategies. In fact, in the last few years, spatial develop-
ment has been put vigorously on the agenda of policy-makers who foresee
economic competitiveness as highly dependent on an efficient territorial
system of regions and cities. At the European level, the concept of ter-
ritorial sustainability has come to the fore, meaning the normative aim
of complementing economic equity aims with social, environmental and
territorial ones.
It is noteworthy that regional scientists increasingly address, in their
scientific agenda policy, issues that have a strong societal interest; con-
vergence problems, on the one hand, and endogenous determinants of
regional growth (like knowledge creation), on the other. These two themes
both have a practical interest and reflect a need for a multidisciplinary
approach, providing regional scientists with all the challenges to identify
new pathways. Whether this will actually happen is a matter of willing-
ness to grasp the opportunities that are provided in the present era, and
to respond to the plea of policy-makers for a more locally-oriented under-
standing of the real world.
The history of regional growth theory covers almost half a century. We
have seen a deepening of our understanding of the complex backgrounds
of, and impediments to, balanced regional development. But regions are
not islands that can be isolated from the rest of the world. They are sub-
jected to the same forces as nations and continents, and, consequently
they have to be positioned in a similar conceptual explanatory frame-
work, complemented with regional socio-economic specificities. Regional

growth is thus a race without a finish; and that also applies to regional
growth theory.
NOTES
1. For further discussion on club convergence, we refer to Baumont et al. (2003); Chatterji
(1992); Chatterji and Dewhurst (1996); Fischer and Stirbock (2006); Islam (2003);
Lpez-Bazo et al. (1999); Quah (1996); Rey and Montouri (1999); Sala-i-Martin (1996).
2. For an extensive review of regional growth theories, see, among others, Johansson et al.
(2001).
3. For the literature on spatial spillovers, see, amongst others, Anselin et al. (1997 and 2000);
Audretsch and Feldman (1996); Aydalot (1986); Feldman (1994); Feldman and Audretsch
(1999); de Groot et al. (2001); Jaffe (1989); Jaffe et al. (1993); Maier and Sedlacek (2005);
on collective learning, see Bellet et al. (1993); Camagni (1991); Capello (1999) and (2001);
Crevoisier and Camagni (2000); Maillat et al. (1993); Rallet (1993); Rallet and Torre
(1995); Ratti et al. (1997); on learning regions, see Lundvall (1992); Lundvall and Johnson
(1994); Maskell and Malmberg (1999); on knowledge-based regions, see Florida (1995);
Malecki (2000); Nijkamp and Stough (2004); Simmie (1997).
REFERENCES
Acs, Z.J., Anselin, L. and Varga, A. (2002), Patents and Innovation Counts as
Measures of Regional Production of New Knowledge, Research Policy, 31(7),
106985.
Aghion, P. and Howitt, P. (1998), Endogenous Growth Theory, MIT Press,
Cambridge, MA.
Anselin, L., Varga, A. and Acs, Z. (1997), Local Geographic Spillovers between
University Research and High Technology Innovations, Journal of Urban
Economics, 42(3), 42248.
Anselin, L., Varga, A. and Acs, Z. (2000), Geographic and Sectoral Characteristics
of Academic Knowledge Externalities, Papers in Regional Science, 79(4), 43543.
Armstrong, H.W. (1995), Convergence among the Regions of the European
Union, 19501990, Papers in Regional Science, 74(2), 14352.
Audretsch, D. and Feldman, M. (1996), R&D Spillovers and the Geography of
Innovation and Production, American Economic Review, 86(3), 63040.
Aydalot, Ph. (ed.) (1986), Milieux Innovateurs en Europe, GREMI, Paris.
Bailly, A. and Coffey, W. (1994), Regional Science in Crisis: A Plea for a More
Open and Relevant Approach, Papers in Regional Science, 73(1), 314.
Barro, R. (1991), Economic Growth in a Cross-section of Countries, Quarterly
Barro, R.J. and Sala-i-Martin, X. (1991), Convergence across States and Regions,
Brookings Papers on Economic Activity, 1, 10782.
Barro, R.J. and Sala-i-Martin, X. (1992), Convergence, Journal of Political
Economics, 100 (2), 22351.
Barro, R. and Sala-i-Martin, X. (1997), Technology Diffusion, Convergence and
Growth, Journal of Economic Growth, 2(1), 127.
Batabyal, A.A. and Nijkamp, P. (2008), Sustainable Development and Regional

Growth, in R. Capello and P. Nijkamp (eds), Handbook of Regional Growth

and Development Theories, Edward Elgar, Cheltenham, UK and Northampton,
MA, USA.
Baumol, W.J. (1986), Productivity Growth, Convergence, and Welfare: What the
Long-run Data Show, American Economic Review, 76(5), 107285.
Baumont, C., Ertur, C. and LeGallo, J. (2003), Spatial Convergence Clubs and
the European Regional Growth Process, 19801995, in B. Fingleton (ed.),
European Regional Growth, Springer-Verlag, Berlin.
Bellet, M., Colletis, G. and Lung, Y. (1993), Introduction au Numro Spcial sur
Economie et Proximit, Revue dEconomie Rgionale et Urbaine, 3, 35761.
Bernard, A. and Durlauf, S. (1995), Convergence in International Output,
Journal of Applied Econometrics, 10(2), 97108.
Bernard, A. and Durlauf, S. (1996), Interpreting Tests of the Convergence
Hypothesis, Journal of Economics, 71(12), 16174.
Boldrin, M. and Canova, F. (2001), Europes Regions: Income Disparities and
Regional Policy, Economic Policy, 16, 20753.
Borts, G.H. (1960), The Equalisation of Returns and Regional Economic
Growth, The American Economic Review, 50, 31947; reprinted in D. McKee,
R. Dean and W. Leahy (eds) (1970), Regional Economics: Theory and Practice,
The Free Press, New York, pp. 14776.
Borts, G.H. and Stein, J.L. (1964), Economic Growth in a Free Market, Columbia
University Press, New York.
Borts, G.H. and Stein, J.L. (1968), Regional Growth and Maturity in the United
States: A Study of Regional Structural Change, in L. Needleman (ed.), Regional
Analysis, Penguin Books Ltd., Harmondsworth, pp. 15997, original edition
(1962), in Schweizerische Zeitschrift fr Volkswirtscahft und Statistik, 98,
290321.
Camagni, R. (1991), Local Milieu, Uncertainty and Innovation Networks:
Towards a New Dynamic Theory of Economic Space, in R. Camagni
(ed.), Innovation Networks: Spatial Perspectives, Belhaven-Pinter, London,
pp. 12144.
Camagni, R. and Capello, R. (2002), Apprendimento Collettivo, Innovazione
e Contesto Locale, in R. Camagni and R. Capello (eds), Apprendimento
Collettivo e Competitivit Territoriale, Franco Angeli, Milano, pp. 1126.
Cameron, R. (2005), Spatial Economic Analysis, Journal of Development
Capello, R. (1999), Spatial Transfer of Knowledge in High-technology Milieux:
Learning vs. Collective Learning Processes, Regional Studies, 33(4), 35365.
Capello, R. (2001), Urban Innovation and Collective Learning: Theory and
Evidence from Five Metropolitan Cities in Europe, in M.M. Fischer and J.
Froehlich (eds), Knowledge, Complexity and Innovation Systems, Springer,
Berlin, Heidelberg, New York, pp. 181208.
Capello, R. (2008), Regional Economics in its Fifties: Recent Theoretical
Directions and Future Challenges, Annals of Regional Science, 42(1), 74767.
Capello, R. (2009), Spatial Spillovers and Regional Growth: A Cognitive
Approach, European Planning Studies, 17(5), 63958.
Capello, R. and Nijkamp, P. (2004), The Theoretical and Methodological
Toolbox of Urban Economics: From and Towards Where?, in R. Capello and
P. Nijkamp (eds), Urban Dynamics and Growth: Advances in Urban Economics,
Elsevier, Amsterdam, pp. 130.

Chatterji, M. (1992), Convergence Clubs and Endogenous Growth, Oxford

Review of Economic Policy, 8(4), 5769.
Chatterji, M. and Dewhurst, J.H.L. (1996), Convergence Clubs and Relative
Economic Performance in Great Britain: 19771991, Regional Studies, 30(1),
3140.
Cheshire, P. and Carbonaro, G. (1995), Convergence-Divergence in Regional
Growth Rates: An Empty Black Box?, in H.W. Armstrong and R.W. Vickerman
(eds), Convergence and Divergence among European Regions, Pion, London,
pp. 89111.
Crevoisier, O. and Camagni, R. (eds) (2000), Les Milieux Urbains: Innovation,
Systmes de Production et Ancrage, EDES, Neuchtel.
de Groot, H., Nijkamp, P. and Acs, Z. (2001), Knowledge Spillovers, Innovation
and Regional Development, Papers in Regional Science, 80(3), special issue.
Dewhurst, J.H.L. and Mutis-Gaitan, H. (1995), Varying Speeds of Regional GDP
Per Capita Convergence in the European Union, 198191, in H.W. Armstong
and R.W Vickerman (eds), Convergence and Divergence among European
Regions, Pion, London, pp. 2239.
Dring, Th. and Schnellenback, J. (2006), What Do We Know About Geographical
Knowledge Spillover and Economic Growth?, Regional Studies, 40(3), 37595.
Fagerberg, J. and Verspagen, B. (1996), Heading for Convergence? Regional
Growth in Europe Reconsidered, Journal of Common Market Studies, 34(3),
43148.
Feldman, M. (1994), The Geography of Innovation, Kluwer Academic, Boston, MA.
Feldman, M. and Audretsch, D. (1999), Innovation in Cities: Science-based
Diversity, Specialisation and Localised Competition, European Economic
Review, 43(2), 40929.
Fingleton, B. (1999), Estimates of Time to Economic Convergence: An Analysis
of Regions of the European Union, International Regional Science Review,
22(1), 534.
Fischer, M. and Stirbock, C. (2006), Pan-European Regional Income Growth and
Club-convergence, Annals of Regional Science, 40(4), 693721.
Florida, R. (1995), Toward the Learning Region, Futures, 27(5), 52736.
Fujita, M. and Thisse, J.-F. (1996), Economics of Agglomeration, Journal of the
Japanese and International Economies, 10(4), 33978.
Fujita, M. and Thisse, J.-F. (2002), Economics of Agglomeration: Cities, Industrial
Location and Regional Growth, Cambridge University Press, New York.
Galor, O. (1996), Convergence? Inferences from Theoretical Models, Economic
Journal, 106(437), 105669.
Geoghegan J., Wainger, L.A. and Bockstael, N.E. (1997), Spatial Landscape
Indices in a Hedonic Framework: An Ecological Economics Analysis Using
GIS, Ecological Economics, 23(3), 25164.
Gilles, S.-P. (1998), The Political Consequence of Unemployment, Working
Paper, No. 343, Department of Economics, Universitat Pompeu, Fabra.
Hgerstrand T. (1967), Innovation Diffusion as a Spatial Process, University of
Chicago Press, Chicago.
Islam, N. (2003), What Have We Learnt from the Convergence Debate?, Journal
of Economic Surveys, 17(3), 30962.
Jaffe A. (1989), Real Effects of Academic Research, American Economic Review,
79(5), 95770.
Jaffe A., Trajtenberg, M. and Henderson, R. (1993), Geographic Localisation of

Knowledge Spillovers as Evidenced by Patent Citation, Quarterly Journal of

Economics, 63(4), 57798.
Johansson, B., Karlsson, C. and Stough, R. (eds) (2001), Theories of Endogenous
Regional Growth, Springer-Verlag, Berlin.
Kain, J.F. and Quigley, J. (1970), Measuring the Value of Housing Quality,
Journal of the American Statistical Association, 65(330), 53248.
Kaldor, N. (1970), The Case for Regional Policies, Scottish Journal of Political
Economy, 17(3), 33748.
Keeble, D. and Wilkinson, F. (1999), Collective Learning and Knowledge
Longhi, S. and Musolesi, A. (2007), European Cities in the Process of Economic
Integration: Towards Structural Convergence, Annals of Regional Science,
41(2), 33551.
Lpez-Bazo, E., Vay, E., Mora, A. and Suriach, J. (1999), Regional Economic
Dynamics and Convergence in the European Union, Annals of Regional Science,
33(3), 34370.
Lucas, R. (1988), On the Mechanics of Economic Development, Journal of
Lundvall, B.A. (1992), Introduction, in B.A. Lundvall (ed.), National Systems
of Innovation: Towards a Theory of Innovation and Interactive Learning, Pinter
Publishers, London, pp. 119.
Lundvall, B.-A. and Johnson, B. (1994), The Learning Economy, Journal of
Industry Studies, 1(2), 2342.
Magrini, S. (2004), Regional (Di) Convergence, in V.J. Henderson and J.-F.
Thisse (eds), Handbook of Regional and Urban Economics, vol. IV, Elsevier,
Amsterdam, pp. 274296.
Maier, G. and Sedlacek, S. (eds) (2005), Spillovers and Innovations Space,
Environment and the Economy, Springer Verlag, Wien.
Maillat, D., Quvit, M. and Senn, L. (eds) (1993), Rseaux dInnovation et Milieux
Innovateurs: un Pari pour le Dveloppement Rgional, EDES, Neuchtel.
Malecki, E. (2000), Creating and Sustaining Competitiveness, in J.R. Bryson,
P.W. Daniels, N. Henry and J. Pollard (eds), Knowledge, Space, Economy,
Routledge, London, pp. 10319.
Maskell, P. and Malmberg, A. (1999), Localised Learning and Industrial
Competitiveness, Cambridge Journal of Economics, 23(2), 16785.
Miyao, T. (1984), Dynamic Models of Urban Growth and Decay: A Survey and
Extensions, paper presented to the Second World Conference of Arts and
Sciences, Rotterdam, 415 June.
Miyao, T. (1987a), Dynamic Urban Models, in E. Mills (ed.), Urban Economics;
Handbook of Regional and Urban Economics, vol. 2, North-Holland, Amsterdam,
pp. 877925.
Miyao, T. (1987b), Urban Growth and Dynamics, in E. Mills (ed.), Urban
Economics. Handbook of Regional and Urban Economics, North-Holland,
Amsterdam, pp. 141.
Myrdal, G. (1957), Economic Theory of Under-developed Regions, Gerald
Duckworth & Co, London.
Nijkamp, P. (2006), Ceteris Paribus: Spatial Complexity and Spatial Equilibrium.
An Interpretative Perspective, mimeo, available from the author.

Nijkamp, P. and Abreu, M. (forthcoming), Regional Development Theory,

Encyclopedia of Geography.
Nijkamp, P. and Reggiani, A. (eds) (1993), Nonlinear Evolution of Spatial
Economic Systems, Springer-Verlag, Berlin.
Nijkamp, P. and Reggiani, A. (1999), The Economics of Complex Spatial Systems,
Elsevier, Amsterdam.
Nijkamp, P. and Stough, R. (eds) (2004), Entrepreneurship and Regional Economic
Development, Edward Elgar, Cheltenham, UK and Northampton, MA, USA.
Okun, A.M. (1970), The Political Economic of Prosperity, Brookings Institution,
Washington, DC.
Paldam, M. (1987), How Much Does One Percent of Growth Change the
Unemployment Rate?, European Economic Review, 31(12), 30613.
Patuelli, R. (2007), Regional Labour Markets in Germany, PhD Dissertation,
Free University, Amsterdam.
Puu, T. (1991), Non-linear Economic Dynamics, Springer-Verlag, Berlin.
Quah, D.T. (1996), Empirics for Economic Growth and Convergence, European
Economic Review, 40(6), 135375.
Rallet, A. (1993), Choix de Proximit et Processus dInnovation Technologique,
Revue dEconomie Rgionale et Urbaine, 3, 36586.
Rallet, A. and Torre, A. (eds) (1995), Economie Industrielle et Economie Spatiale,
Economica, Paris.
Ratti, R., Bramanti, A. and Gordon, R. (eds) (1997), The Dynamics of Innovative
Regions, Ashgate, Aldershot.
Rey, S.J. and Montouri, B.D. (1999), US Regional Income Convergence: A
Spatial Econometric Perspective, Regional Studies, 33(2), 14356.
Romer, P. (1986), Increasing Returns and Long-run Growth, Journal of Political
Economy, 94(5), 100237.
Sala-i-Martin, X. (1996), The Classical Approach to Convergence Analysis,
Economic Journal, 106(437), 34370.
Simmie, J. (ed.) (1997), Innovation, Networks and Learning Regions, Jessica
Kingsley, London.
Solow, R. (1956), Contribution to the Theory of Economic Growth, Quarterly
Stimson, R.J., Stough, R.R. and Roberts, H.B. (2006), Regional Economic
Development, Springer-Verlag, Berlin.
Weng, Q. (2002), Land Use Change Analysis in the Zhujiang Delta of China
Using Satellite Remote Sensing, GIS and Stochastic Modeling, Journal of
Environmental Management, 64(3), 27384.

Index
absolute advantage principle 206, 209 new regionalism and embededdness
absorption capacity see European 447
Union structural assistance and Australia:
regional disparities: absorption national economy perspective 161,
capacity 162
agglomeration(s) 5, 6, 434, 11415, planners perspective 77
1545, 179, 309 vocational education and training
Indonesia 1778 47, 489, 523, 54
see also Silicon Valley see also employment growth and
agriculture 65, 213, 277, 2789, 280 decline in Australian capital
see also employment growth and city statistical divisions; model
decline in Australian capital determinants; Victoria
city statistical divisions Australian and New Zealand Regional
amnagement du territoire 217 Science Association International
appreciative intelligence see Silicon (ANZRSA) 269
Valley Autocracy (human resource practice
Area Development annual survey blueprint) 93
2889
area-based approach 64 backward removal stepwise
Arrow, K. 267, 28, 30 (regression) approach 136
Arrow-Romer model 28 Bangladesh: Grameen Bank 81
Asian BRIICS countries see foreign Basque Madrigon region 81
direct investment, knowledge behavioural/management science
assets and economic geography of perspective see Silicon Valley:
growth appreciative intelligence
attraction (economic development) 288 Belgium 191, 206
Australia 14, 4950, 138 Blakely, E.J. 2
Australia: geographers perspective 12, bottom-up approach 64
3954, 578 Brazil 160
analytical capabilities and policy Brundtland Report 312
advice 4753 Bulgaria 182, 186, 187, 188, 190, 191,
enterprising human capital and 192
regional engagement 512 Bureaucracy (human resource practice
qualitative information and blueprint) 93
facilitated workshops 4951 business development (economic
theoretically informed empiricism development) 288
489
case study regions and their Canada: First Nations 274
characteristics 5051 capital outflow from advanced regions
data set 578 208
endogenous regional theory 414 causation, circular and cumulative 30
325

centralized management 197 objective 184, 187, 1934

China see foreign direct investment, real 183
knowledge assets and economic sigma-convergence 3034
geography of growth in Asian spatial socio-economic 301
BRIICS countries cooperation between areas 65
cities, scale of see employment growth cost parameter 155
and decline in Australian capital creative capital 11
city statistical divisions creative destruction 42
civic spirit 80 creativity 9, 209, 21920, 223, 2289,
see also Victoria 231, 308
climate change 315 critical success factors 10
coalitions of people 53 cross-fertilization 30
Cobb-Douglas production function 24, cultural assets, potential 7072
267 cultural attributes 80
cognitive approach to district cultural factors 5, 114
economies and synergies 21415, culture economy approach 623
216 currency devaluation 206
cohesion 185 Cyprus 187, 190, 192
horizontal and vertical 183 Czech Republic 187, 190, 191, 192,
Cohesion Funds (EU) 184, 193 196, 197, 198
commercialization process 151 Joint Operational Programme 197
Commitment (human resource practice
blueprint) 934, 97, 101 decision-making under conditions of
Common Agricultural Policy 65 uncertainty 215
comparative statics 223 deforestation 31314
competency building 52 Delbecq, A.L. 91
competitiveness 185, 1878, 272 demand-supply interaction 207
and regional policies 21720 demography 120
see also European Union: development:
macroeconomic and territorial dependent 61
policies for regional destructive 61
competitiveness dictated 61
Computable General Equilibrium distorted 61
modelling 315 from below model 219
Constant Elasticity of Substitution initiatives in small rural towns see
production function 152 Victoria
Constant Share Approach 246 in regional growth theory 30512
consumers 1489 dynamic approaches 306, 31012
contemporary challenges 69 realism 306, 30710
contracts theory 21516 see also economic development
convergence 20810 differential growth, regional 221
beta-convergence 312, 3034 differential/regional shift effect 117,
conventional gap (mean reversion) 2478, 249
econometric model 49 see also model determinants:
economic 312 Australia and United States
group/club 304 diminishing returns to scale 304
hypothesis 304 diversity and endogeny in explanatory
income 303 paradigms 847
neoclassical conditional 208 Dixon-Thirwall model 207
nominal 183 do nothing approach 296, 298

Index 327
Dutch disease 200 Brisbane 261

dynamic approaches 306, 31012 Melbourne 258, 259
Perth 263
Eastern and Central Europe 182, 223, Sydney 257
229 Adelaide 237, 240, 262, 264, 265
ecological capital 11 divergent rates of growth and
economic cohesion 185 change 2415
economic development 273, 313 employment and population 240
planning, endogenous 779 industrial explanations for
regional 11315, 116, 119 regional shift effects 252,
see also United States: economic 2546
development incentives and shift-share analysis 24952
measurement of endogenous administrative and support and
growth safety 239, 242, 244, 253
economic factors 212 Adelaide 264
economic geography see foreign direct Brisbane 261
investment, knowledge assets and Melbourne 259
economic geography of growth; Perth 262, 263
new economic geography Sydney 257, 258
economic performance 271 agriculture/forestry and fishing 239,
Economic and Social Cohesion Policy 241, 242, 244, 253
see European Union Adelaide 264, 265
economic success 275 Brisbane 261
economies of scale 160, 307 Melbourne 259, 260
economists perspective 1112, 2036 Perth 262, 263
comparative statics 223 Sydney 257
economic convergence, empirical arts and recreation services 243, 245,
testing for 312 252
exogenous growth models 234 Adelaide 264, 265
policy implications 323 Brisbane 261
precursors to modern endogenous Melbourne 259, 260
growth theory 257 Perth 262, 263
recent models 2830 Sydney 257
specific regional context 3031 Brisbane 237, 238, 26061, 265
economy 120 divergent rates of growth and
economy-wide aggregate production change 2415
function 27 employment and population
education 8, 9, 29, 3031, 303, 308 240
efficiency objectives 301 industrial explanations for
efficiency wages 207 regional shift effects 252,
embeddedness 12, 447 2546
institutional 7, 9, 114 shift-share analysis 24952
Empathic Knowledge Management Bureau of Statistics (Australian)
104 2489
employment growth and decline in capital city dominance 238
Australian capital city/cities Census of Population and Housing
statistical divisions 15, 23766 Time Series Profile (2006)
accommodation and food services 249
239, 242, 244 city shares of state population and
Adelaide 264 employment 23940

construction 239, 241, 242, 244, Perth 262, 263

2523, 2556 Sydney 257, 258
Adelaide 264, 265 Melbourne 237, 25860
Brisbane 261 divergent rates of growth and
Melbourne 259 change 2415
Perth 263 employment and population 240
Sydney 257, 258 industrial explanations for
education and training 242, 244, regional shift effects 2526
253 shift-share analysis 24952
Adelaide 262, 264 mining 239, 242, 244, 2523
Brisbane 261 Adelaide 264, 265
Melbourne 259 Brisbane 261
Perth 263 Melbourne 259, 260
Sydney 257 Perth 262, 263
electricity/gas/water and waste Sydney 257, 258
services 239, 242, 244, 2523 national shares, changing 2389
Adelaide 264 New South Wales 238, 239
Brisbane 261 Perth 237, 262, 263, 265
Melbourne 259, 260 divergent rates of growth and
Perth 263 change 2415
Sydney 257 employment and population 240
financial and insurance services 239, industrial explanations for
242, 244, 253 regional shift effects 252,
Adelaide 264 2546
Brisbane 261 shift-share analysis 24952
Melbourne 259 professional/scientific and technical
Perth 262, 263 services 239, 242, 244
Sydney 257, 258 Adelaide 264
healthcare and social assistance 241, Brisbane 261
242, 244, 265 Melbourne 259
Adelaide 262, 264 Perth 263
Brisbane 260, 261 Sydney 257
Melbourne 259 public administration and safety
Perth 263 242, 244
Sydney 257, 258 Adelaide 264, 265
industry structure of employment Brisbane 261
2425 Melbourne 259, 260
information, media and Perth 262, 263
telecommunications 239, 242, Sydney 257
244, 258 Queensland 238, 239
Adelaide 264 regional shift effects 247, 251, 265
Brisbane 260, 261 Adelaide 264
Melbourne 259 Brisbane 261
Perth 263 Melbourne 259
Sydney 257, 258 Perth 263
manufacturing 239, 242, 244, 2524, Sydney 257
265 retail/hiring and real estate services
Adelaide 264 239, 241, 242, 244, 265
Brisbane 261 Adelaide 264, 265
Melbourne 259, 260 Brisbane 260, 261

Index 329
Melbourne 259, 260 new growth theory 1446

Perth 263 policy implications 1567
Sydney 257, 258 systems of innovation framework
share of national employment by 1457
industry 239 entrepreneurial talent 156
shift-share results of city Statistical entrepreneurship 7, 114, 116, 119, 120,
Divisions 250 129, 131, 1378, 309
South Australia 238, 239 capital 1356, 137
Sydney 237, 2568, 265 European Union 223, 2278, 231
divergent rates of growth and United States 129, 131, 137
change 2415 environmental policies and regulation
employment and population 240 313, 31415
industrial explanations for equity issue 301
regional shift effects 2526 Estonia 187, 190, 192, 196, 198
shift-share analysis 24952 Europe 147
Tasmania 238, 239 European Commission 184
transport/postal and warehousing European Community Strategic
239, 242, 244 Guidelines (200713) 186
Adelaide 264 European Paradox 145
Brisbane 261 European Regional Development
Melbourne 259 Fund 183
Perth 263 European Social Fund 183, 198
Sydney 257 European Territorial Cooperation
Victoria 238, 239 Objective 188
Western Australia 239 European Union 28, 32, 304
wholesale/trade 239, 241, 242, 244, European Union: macroeconomic and
253, 2545 territorial policies for regional
Adelaide 262, 264, 265 competitiveness 15, 20433
Brisbane 261 competitiveness and regional
Melbourne 259, 260 policies 21720
Perth 263 impact of structural funds on
Sydney 257, 258 regional growth rates 225
employment scale weighted location per-capita income disparities 213
quotient 117 persistence of regional disparities as
endogenous shift effect see differential/ rationale for policies 20511
regional shift effect convergence process 20810
Engineering (human resource practice peripheral regions, difficulties
blueprint) 93 encountered by 2067
Enns, S. 275 spatial development policies
enterprising skills 52 derived from globalization
entrepreneurial rents 14, 14258 debate 21011
consumers 1489 regional growth and differential
equilibrium and propositions 1524 spatial effect of macroeconomic
final goods producers 14950 policies 21114
intermediate goods producers spatial effects of revaluation of
15051 exchange rates 224
knowledge creation/spillovers 1447, see also MASST model
1546 European Union:
new economic geography models sociological perspective 60
146 territorial capital 22630

territorial sustainability 319 phasing-out regions 194

see also LEADER programme Poland 187, 190, 192, 197, 198
European Union Budget 186 Regional Competitiveness and
European Union structural assistance Employment Objective 1878
and regional disparities: regional development dynamics
absorption capacity 15, 182201 19091
absorptional capacity methodology Romania 182, 186, 187, 188, 190,
1946 191, 192, 193, 195, 196
administrative absorption capacity Single Market and Single Currency
1845, 199 programmes 183
Belgium 191 Slovakia 187, 190, 192, 196, 197, 198
Bulgaria 182, 186, 187, 188, 190, Slovenia 187, 190, 192, 196, 198
191, 192 social cohesion 185
Cohesion Funds 184, 193 Structural Funds 184, 189, 193,
Cohesion Policy 1839 1945, 1978, 199200
Community Strategic Guidelines structure 195
(200713) 186 subsequent evaluations of
Convergence Objective 187, 1934 absorption capacity 1969
Cyprus 187, 190, 192 systems and tools 195
Czech Republic 187, 190, 191, 192, Territorial Cooperation Objective
196, 197, 198 188
disparities 18990 United Kingdom 190, 191
Eastern and Central Europe 182 vertical cohesion 183
economic cohesion 185 within-country disparities 1913
Estonia 187, 190, 192, 196, 198 excellence policies 213
EU Budget 186 exogenous development:
EU-8 188 and endogenous development,
EU-9 196 comparison of 612
EU-10 186, 190 sociological perspective 5960
EU-12 186 exogenous factors 2, 4
EU-15 185, 186, 189, 190, 191, 193, exogenous forces 4, 7
197 exogenous growth models 234, 25
EU-25 186, 193, 194 exogenous incentives 294, 2957
EU-27 186, 187, 188, 189, 190, 191, exogenous shift effect 2478
193, 212 exploratory model 117
financial absorption capacity 185 external funds injection 199
horizontal cohesion 183 external networks 282
human resources 195 externalities 312, 313
Hungary 187, 190, 192, 196, 198
implications 1934 falsification 856
Ireland 186 final goods producers 14950
Latvia 187, 190, 192, 198 First Report to the Club of Rome (1973)
Lithuania 187, 190, 192, 198 312
macroeconomic absorption capacity Florida, R. 90
184 foreign direct investment, knowledge
Malta 187, 192 assets and economic geography of
n+2 rule 185 growth in Asian BRIICS
Netherlands 186 countries 1415, 16079
NUTS 2 regions 190, 193 China 16061, 1636, 16873, 1745,
phasing-in regions 194 179

Index 331
inter-regional inequality 16970, growth econometrics model 489

172 growth stage 113
National Economy and Social growth theory see new growth theory
Development Plan (2005) 169
urbanrural ratio 17072 Hamel, G. 1067
greenfield projects 1634 Harrod-Domar model 234
India 16061, 163, 1646, 168, Hirschmann-Herfindahl index 297
1735, 179 historical society 278
economic growth city-regions 174 Honda 291
Foreign Investment Board 173 housing planning and demand 315
GDP composition by industry Hsung, R.-M. 275
174 human capital 5, 1011, 29, 42, 49
Investment Commission 173 Australia 512
Indonesia 16061, 1646, 168, 1759 planners perspective 80
agglomeration economies 1778 regional shift component 114
economic growth city-regions 176 Human Development Index 303
trade protectionism 177 human resources (indicator) 195
inward FDI and R&D 1638 Hungary 187, 190, 192, 196, 198
manufacturing 163 Hyundai 291
primary industries 163
service sector 163 immiserizing growth phenomena 200
forward/backward linkages 207 increasing returns to scale 207
France 164, 206, 223, 229 India see foreign direct investment,
proximity approach 216 knowledge assets and economic
Fu, X.-C. 275 geography of growth in Asian
BRIICS countries
G-factor 94 individual agency 40
geographers perspective see Australia Indonesia see foreign direct
geographic information systems investment, knowledge assets and
31516 economic geography of growth in
geographical clusters 147, 155, 309, Asian BRIICS countries
311 industrial structure 114
see also agglomeration industry mix effect 247, 248
Germany 147, 206, 223 industry specialization 49
Gini Inequality Index 191 infrastructure:
global labour force expansion 160 provision 278, 279, 281
global trade 160 support and institutional thickness
globalization 56, 7, 39, 83, 115, 160, 58
161 innovation 9, 64, 113, 165
Google 956, 99 innovative milieu 5, 114
government industry assistance see institutional arrangements 8
United States: economic institutional embeddedness 7, 9, 114
development incentives and institutional factors 7, 116, 138, 212,
measurement of endogenous 297
growth institutional proximity of firms 310
government intervention 49 institutionalist perspective 48, 215
government policy 245 institutions 5, 40, 119, 120, 129, 131,
Greece 223, 227, 229 137
greenfield projects 1634 integrated approach 65
GREMI 216, 219 inter-regional level 210, 212

inter-regional performance see leadership 5, 7, 271, 272, 2823, 297

European Union structural regional shift component 114, 116,
assistance and regional disparities: 11819, 120, 129, 137, 138
absorption capacity technological 49, 57
inter-regional trade 58 learning 26, 308
inter-sectoral relationships 207 -by-doing 25, 27, 2830, 42
intermediate markets 15051, 152, 1534 collective 311
internal community networks 282 region 9
intra-regional level 210 Lin, N. 275
investment 160 linear models 310
see also foreign direct investment Lisbon Strategy 187
Ireland 186, 223 Lithuania 187, 190, 192, 198
Italian school 216 local action group 64
Italy 206, 223, 225, 227, 229 local control 61
local environment quality 313
Japan 168 local financing and management 65
local human resource base 58
Kaldor, N. 26 local institutional setting 313
Kaldor regional models 31 local integration of small firms 578
Keynesian approaches 3, 4 local level of scale see Victoria
Keynesian economics 21 local sectoral specialization 58
Keynesian models 311 lodging and food services 293
knowledge 5, 6, 29, 44, 143, 149, 3089 Lucas models 33
accumulation 27 Lucas, R. 2930
acquisition 309
Asian BRIICS 165, 167 macroeconomic long-run growth
assets see foreign direct investment, theory 22
knowledge assets and economic macroeconomic policies see European
geography of growth Union
base 219 Malta 187, 192
commercialization 146, 1567 Malthusian effect 28
creation 57, 1447, 153, 1567 manufacturing 294
disembodied 28 see also employment growth and
production function 42 decline
spillovers 8, 42, 1447, 150, 1547, market fit 118, 119, 131, 1367, 138
311, 317 market imperfections 6, 115
Korea 168 MASST Model (macroeconomic,
sectoral, social and territorial)
labour 288 205, 212, 22130
force survey employment data 247 regional impact of macroeconomic
mobility 154 policies 2223, 224
productivity 207 structural funds 223, 225
land use: territorial capital and regional
as endogenous planning base 767 growth 223, 22630
and land values location analysis 315 measurement of endogenous growth
Latvia 187, 190, 192, 198 see United States: economic
LEADER programme 60, 636 development incentives and
components 645 measurement of endogenous
phases 634 growth
significance 656 Mercedes Benz 29091, 292, 296

Index 333
Metzker, C. 98 population size 129, 131, 136, 137

migration, simplified theories of 23 spatial patterns of endogenous
mixed-method research design 47 employment growth
model determinants: Australia and performance 12931
United States 14, 11139 specific model results 1367
Australia 12028 modelling local and regional
concentration of jobs in broad environments 31516
industry sectors 128 modern endogenous growth theory,
income level 128 precursors to 257
industry sector specialization 128 monetarism 3, 4
jobs in broad occupation Morans I statistic 124
categories 128 Myrdal-Kaldor model 307, 311
Local Government Areas 12021,
123, 127 n+2 rule 185, 197
location on or near coast or to national economy perspective see
metropolitan area 128 foreign direct investment,
model results 1238 knowledge assets and economic
model variables 121 geography in Asian BRIICS
patterns of endogenous regional countries
growth performance 1213 national growth 221
population growth 128 national and regional dimensions of
unemployment 128 development 197
university and technical national savings 209
qualifications 128 national shift effect 2478
evolution of regional economic natural resources 31314
development theory 11315 nature of regional development 24
measurement and modelling 11520 neo-endogenous development 62
exploratory model applications neoclassical conditional convergence
11920 theory 208
new model framework 11516 neoclassical economics 234, 26, 32,
potential independent and 40, 207, 208, 304
intervening variables 11819 neoclassical framework 302, 303
proxy measure for dependent neoclassical models 234, 25, 30, 31,
variable 11617 33, 311
United States 12937 neoclassical production function 308
educational attainment level 131, Netherlands 186, 223
137 networking 65
employment growth/decline new economic geography 39, 41, 146,
129 307, 316, 317
entrepreneurship capital 1356, new growth theory 1, 46, 30, 42,
137 1446, 316
exploratory model 129 regional shift component 11315
general model results 1316 new regionalism 3940, 41, 447, 54
government, regional 131, 135 new research challenges 31618
method 129 New Zealand 161, 162
Metropolitan Statistical Areas non-competition clauses 157
120, 129, 130, 131, 1324, non-linear models 31011
136, 137, 138 North American Industry
occupational structure of routine Classification System code 287
production 135 NUTS 2 regions 190, 193

Okuns Law 303 product cycle trap 789

oligopoly 114 production functions 24, 267, 152,
on-the-job training 29, 31 308
opportunity costs 209 productive capital 10
ordinary least squares regression 117, productivity change 2878
120, 121, 123, 131, 135, 138 profit cycles 114
Organizational Learning Laboratory programming process 1967
104 progress association 280
organizing metaphors 84 proxy variables 43
out-migration process of unemployed public services modernization 198
people from lagging regions 208
outcome measures 277 qualitative approach 54
output 277, 2878 qualitative information 47, 4951
qualitative measures 2
paradigms, changing 34 qualitative methodologies 4041
partnership approach 64, 198 quantitative element 48
PayPal 97, 98100 quantitative information 47
performance differentials see European quantitative measures 2
Union: macroeconomic and quantitative methodologies 4041
territorial policies for regional quantitative sociology 317
competitiveness
Pigouvian taxation 314 R&D 8, 9, 1423, 145, 303, 308
Pigouvian welfare rules 312 Asian BRIICS countries 1638
planning profession and rural, urban China 169, 170
or regional endogenous costs 29
development 13, 7382 Indonesia 179
economic development planning R-squared values 131, 136
779 racial segregation 76
evolution of planning as indigenous/ random shocks 49
endogenous assessment 746 rationalism 3, 4, 7
land and its uses as endogenous rationality 21718
planning base 767 bounded 215
social space, endogenous plan as realism 306, 30710
8081 receptivity 223, 226, 227, 231
pluralism 47 recreational/sports facilities 313
Poland 187, 190, 192, 197, 198 Regional Competitiveness and
political border effects 155 Employment Objective 1878
political costs 209 regional competitiveness performance
politicization 1978 cube 272
population super-concentration in relational proximity of firms 30910
urban areas 209 rent-seeking 199
Portugal 225 resource endowments 118, 119, 131,
position generators 2745, 283 136, 138, 271
post-Fordism 59 retail industry 294
post-modernism 59 see also employment growth and
Power Advantage Index 297 decline in Australian capital
power of large corporations 58 city statistical divisions
price flexibility 206 retention strategies (economic
process measures 277 development stool) 288
product cycle theory 45, 11314 revaluation 231

Index 335
reverse simulation 1034 social space, endogenous plan as

Romania 182, 190, 191, 192, 193, 195, 8081
196 sociological perspective 1213, 5967,
population and GDP per capita 186, 7072
187, 188 characteristics and values of ERD
Romer model 12 6061
Route 128 8794, 100, 106 cultural assets, potential 7072
rules of the game 85 culture economy approach 623
Rural Transaction Centre 281 definition 5960
Russia 160 European Union LEADER
programme 636
Saxenian, A. 87 exogenous development and
Schumpeter, J. 20, 256 endogenous development,
self-determination 61 comparison of 612
shift-share analysis 117, 2878, 2912, Solow model 1, 4, 11, 30, 32, 113
294, 295 Solow residual 26
Silicon Valley 5, 114 Solow-Swan formulation 42
Silicon Valley: appreciative intelligence South Africa 160
13, 83107 Spain 223, 225, 227, 229
Asian immigrants 912 spatial autocorrelation 138
components and qualities of spatial concepts 317
appreciative intelligence 968 spatial development policies derived
discrimination 92 from globalization debate 21011
diversity and endogeny in spatial efficiency issue 301
explanatory paradigms 847 spatial equilibrium 146
ecology of Silicon Valley versus spatial patterns of endogenous
Route 128 8794 employment growth performance
employee attachment 8991 (United States) 12931
human resource practices 934 spatial proximity of firms 309
old boys network 92 spatial socio-economic disparities or
small worlds phenomenon 91 convergence 301
social networks 93 spatial spillover/proximity effects 138,
transferable lessons 100107 318
simulation modelling 315 spatial sustainability issue 302
Single Currency programme 183 spatial-economic convergence idea 302
Single Market programme 183 spatial/urban development policies,
Slovakia 187, 190, 192, 196, 197, 198 integrated 217
Slovenia 187, 190, 192, 196, 198 specific regional context 3031
social attributes 80 spin-off and spin-out phenomena 154
social capital 5, 11, 42, 8081, 114, 317 stability and growth 114
local level of scale in Victoria, standardization stage 113
Australia 271, 2723, 274, 275, Star (human resource practice
282 blueprint) 93
social cohesion 185 step-wise regression 120, 136, 138
social costs 209 Stimson, R.J. 34
social elements 221 strategic planning 279
social networks 93 structural assistance see European
social overhead capital 223, 226, 231 Union
social planners 289 structural change and adjustment 115
social regulation 40 structural characteristics 49

Structural Funds 184, 189, 193, 1945, triple helix scenario 9

1978, 199200, 223, 225 trust 5, 114
structure (indicator) 195
stylized facts using abstract United Kingdom 9, 190, 191, 206, 223,
mathematical reasoning 412, 229
434, 467 greenfield projects 164
subsidiarity 61 United States 14, 26, 32, 26970, 304
sustainability 61, 302, 305, 31216, United States: economic development
319 incentives and measurement of
sustainable development 6, 1011, 312 endogenous growth 16, 28698
see also virtuous circle Bureau of Economic Analysis 286
Sweden 145 Bureau of Labor Statistics 287
systems of innovation framework Department of Commerce 287
1457 employment and establishments in
systems and tools (indicator) 195 Grayson County and Sector 311
food manufacturing and totals
Taiwan 168 293
tax burdens 290 federal system 28990
technical augmentation factor 27 Grayson County (Texas) 293, 297
technological change 4, 24, 42, 113, incentives, influence of in local
207 economic growth 28891
technological leadership 49, 57 Alabama 29091
technological progress 21, 24, 26, 113, Texas 291
304, 308, 319 lodging and food services 293
technology 8, 303 manufacturing 294
-based theories 5 Mercedes Benz 29091, 292, 296
high 49 potential problems 2947
-led economic growth 144 retail trade 294
outflow from advanced regions 208 Ruiz Foods 291, 2924, 297
TENs 213 Sherman-Denison Metropolitan
territorial capital 21417, 22630, 231 Area (Texas) 2924
territorial policies see European Union shift-share calculations Grayson
territorial sustainability 319 County (Texas) 294, 295
Thatchenkery, T. 98 site selection factors according to
Third Wave Theory 77 corporate executives 289
third-way growth econometrics 423 Texas Enterprise Fund 291, 296, 297
time, concept of 311 United States:
time-series analysis 303 Emeryville (California) 79
timing-related problems 199200 federal deficit 33
total shift effects 247 greenfield projects 164
trade analysis 25 Jim Crow Laws 76
trade protectionism 177 New Orleans 80
traditional management theories 101 planners perspective 77
trans-nationality openness 161 R&D 168
transfer-generating authority/principal Route 128 86, 8794, 97, 100, 106
agent problems 200 San Jose (California) 80
transition dynamics 49 Youngstown (Ohio):
transportation costs 307 deindustrialization process 79
trends in economic growth theory see also model determinants; Silicon
3015 Valley

Index 337
untraded interdependencies 216 community engagement 277

urban scale 114 community plan development
process 277, 278, 279, 283
value 5, 114 purpose and method 26970
Variance Inflation Factors 136 Sponsor a Cow campaign 281
vector-autoregressive 30 theoretical views on endogenous
verification 856 development 2715
Vernon, R. 79 virtuous circle for sustainable
Victoria: local development initiatives regionaldevelopment 11112,
in small rural towns 1516, 269, 272
26884, 295
Bendigo Bank Community Bank waste management 314
278, 281 water provision and use 314
case study shire councils 27582 Weiss, J. 91
shire A 2769 Williamson Law 205
shire B 27982 Woolcock, M. 273
Community Connect Model 276 World Trade Organization 66
community contact team 277
Community Development Program Yelp 99
27980, 281, 282 YouTube 99


Endogenous Regional Development

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Endogenous Regional Development

Uploaded by

Copyright:

Available Formats

Endogenous Regional Development

STIMSON PAGINATION (M2469).indd i 20/12/2010 15:12

Entrepreneurship, Industrial Location and Economic Growth

STIMSON PAGINATION (M2469).indd ii 20/12/2010 15:12

NEW HORIZONS IN REGIONAL SCIENCE

STIMSON PAGINATION (M2469).indd iii 20/12/2010 15:12

All rights reserved. No part of this publication may be reproduced, stored in a

Edward Elgar Publishing, Inc.

A catalogue record for this book

Library of Congress Control Number: 2010927669

ISBN 978 1 84980 456 1

Typeset by Servis Filmsetting Ltd, Stockport, Cheshire

STIMSON PAGINATION (M2469).indd iv 20/12/2010 15:12

1 Endogenous regional development 1

STIMSON PAGINATION (M2469).indd v 20/12/2010 15:12

11 Macroeconomic and territorial policies for regional

STIMSON PAGINATION (M2469).indd vi 20/12/2010 15:12

STIMSON PAGINATION (M2469).indd vii 20/12/2010 15:12

Roberta Capello is Professor of Regional Economics at Politecnico

Terry Clower is Director of the Center for Economic Development and

Daniela Constantin is a Professor of Regional Economics at the Academy

Gabriela Dragan is a Professor of European Integration at the Bucharest

Zizi Goschin is a Professor of Statistics, Econometrics and Regional

Jessica Heineman-Pieper is an Assistant Professor in the Master of Science

Chie Iguchi is Assistant Professor of Global Strategic Management at

STIMSON PAGINATION (M2469).indd viii 20/12/2010 15:12

University, Kyoto, and was also a Visiting Research Fellow at Rutgers

STIMSON PAGINATION (M2469).indd ix 20/12/2010 15:12

STIMSON PAGINATION (M2469).indd x 20/12/2010 15:12

STIMSON PAGINATION (M2469).indd xi 20/12/2010 15:12

This book is, we believe, a unique contribution to the growing literature

Robert Stimson, Roger R. Stough, Peter Nijkamp

STIMSON PAGINATION (M2469).indd xii 20/12/2010 15:12

STIMSON PAGINATION (M2469).indd 1 20/12/2010 15:12

THE NATURE OF REGIONAL DEVELOPMENT

Stimson et al. (2006, p. 4) observe that it is often difficult in regional eco-

STIMSON PAGINATION (M2469).indd 2 20/12/2010 15:12

This multidimensional aspect of economic development led Stimson

Regional economic development is the application of economic processes and

Policy for economic development and regional planning strategy has

STIMSON PAGINATION (M2469).indd 3 20/12/2010 15:12

businesses need to understand the implications of the paradigm shifts occur-

We now turn to a more detailed consideration of the emergence of the

THE EVOLUTION OF THE NEW GROWTH THEORY

STIMSON PAGINATION (M2469).indd 4 20/12/2010 15:12

STIMSON PAGINATION (M2469).indd 5 20/12/2010 15:12

in a globalized economy and a highly competitive and rapidly changing

THE CONTEMPORARY CHALLENGES

We certainly live in a rapidly changing and increasingly competitive world

STIMSON PAGINATION (M2469).indd 6 20/12/2010 15:12

At one time regions were protected from outside competition, and

STIMSON PAGINATION (M2469).indd 7 20/12/2010 15:12

interesting to see whether these much-changed macro-circumstances will

STIMSON PAGINATION (M2469).indd 8 20/12/2010 15:12

In the context of regional development, Taylor (2009) has referred to

STIMSON PAGINATION (M2469).indd 9 20/12/2010 15:12

TOWARDS A SUSTAINABLE DEVELOPMENT

1. The availability of productive capital (PC): this corresponds to neo-

Source: Compiled by author Peter Nijkamp.

Figure 1.1 A pentagon model of creative forces for sustainable regional

STIMSON PAGINATION (M2469).indd 10 20/12/2010 15:12

factor. Clearly a balanced distribution of human capital over people

The pentagon factors can in principle be measured and quantified,

THE ORGANIZATION OF THIS BOOK